Copyright (c) 2025 Advanced Micro Devices, Inc., or its affiliates. AMD Public Use. MIT 2025-02-19 1.1.0 AMD RDNA 4 10 ENC_SOP1 32 11111111100000001111111100000000 10111110100000000000000000000000 10111110100000000000000100000000 10111110100000000000001000000000 10111110100000000000001100000000 10111110100000000000010000000000 10111110100000000000010100000000 10111110100000000000100000000000 10111110100000000000100100000000 10111110100000000000101000000000 10111110100000000000101100000000 10111110100000000000110000000000 10111110100000000000110100000000 10111110100000000000111000000000 10111110100000000000111100000000 10111110100000000001000000000000 10111110100000000001000100000000 10111110100000000001001000000000 10111110100000000001001100000000 10111110100000000001010000000000 10111110100000000001010100000000 10111110100000000001011000000000 10111110100000000001011100000000 10111110100000000001100000000000 10111110100000000001100100000000 10111110100000000001101000000000 10111110100000000001101100000000 10111110100000000001110000000000 10111110100000000001110100000000 10111110100000000001111000000000 10111110100000000001111100000000 10111110100000000010000000000000 10111110100000000010000100000000 10111110100000000010001000000000 10111110100000000010001100000000 10111110100000000010010000000000 10111110100000000010010100000000 10111110100000000010011000000000 10111110100000000010011100000000 10111110100000000010100000000000 10111110100000000010100100000000 10111110100000000010101000000000 10111110100000000010101100000000 10111110100000000010110000000000 10111110100000000010110100000000 10111110100000000010111000000000 10111110100000000010111100000000 10111110100000000011000000000000 10111110100000000011000100000000 10111110100000000011001000000000 10111110100000000011001100000000 10111110100000000011010000000000 10111110100000000011010100000000 10111110100000000011011000000000 10111110100000000011011100000000 10111110100000000100000000000000 10111110100000000100000100000000 10111110100000000100001000000000 10111110100000000100001100000000 10111110100000000100010000000000 10111110100000000100011100000000 10111110100000000100100000000000 10111110100000000100100100000000 10111110100000000100101000000000 10111110100000000100110000000000 10111110100000000100110100000000 10111110100000000100111000000000 10111110100000000100111100000000 10111110100000000101001100000000 10111110100000000101100000000000 10111110100000000110000000000000 10111110100000000110000100000000 10111110100000000110001000000000 10111110100000000110001100000000 10111110100000000110010000000000 10111110100000000110010100000000 10111110100000000110011000000000 10111110100000000110011100000000 10111110100000000110100000000000 10111110100000000110100100000000 10111110100000000110101000000000 10111110100000000110101100000000 10111110100000000110110000000000 10111110100000000110110100000000 10111110100000000110111000000000 default != .fieldderef .fieldderef INST SOP1 SSRC0 255 I 32 Scalar ALU operations with one destination and one source. Allowed patterns: SOP1 (32 bits) SOP1 + LITERAL (64 bits) ENCODING Encoding. 9 23 OP Opcode. 8 8 SDST Destination operand for instruction. 7 16 SSRC0 Source operand for instruction. 8 0 ENC_SOPC 32 11111111111111110000000000000000 10111111000000000000000000000000 10111111000000010000000000000000 10111111000000100000000000000000 10111111000000110000000000000000 10111111000001000000000000000000 10111111000001010000000000000000 10111111000001100000000000000000 10111111000001110000000000000000 10111111000010000000000000000000 10111111000010010000000000000000 10111111000010100000000000000000 10111111000010110000000000000000 10111111000011000000000000000000 10111111000011010000000000000000 10111111000011100000000000000000 10111111000011110000000000000000 10111111000100000000000000000000 10111111000100010000000000000000 10111111010000010000000000000000 10111111010100010000000000000000 10111111010000100000000000000000 10111111010100100000000000000000 10111111010000110000000000000000 10111111010100110000000000000000 10111111010001000000000000000000 10111111010101000000000000000000 10111111010001010000000000000000 10111111010101010000000000000000 10111111010001100000000000000000 10111111010101100000000000000000 10111111010001110000000000000000 10111111010101110000000000000000 10111111010010000000000000000000 10111111010110000000000000000000 10111111010010010000000000000000 10111111010110010000000000000000 10111111010010100000000000000000 10111111010110100000000000000000 10111111010010110000000000000000 10111111010110110000000000000000 10111111010011000000000000000000 10111111010111000000000000000000 10111111010011010000000000000000 10111111010111010000000000000000 10111111010011100000000000000000 10111111010111100000000000000000 default && != .fieldderef .fieldderef INST SOPC SSRC0 255 I 32 != .fieldderef .fieldderef INST SOPC SSRC1 255 I 32 Scalar ALU comparison operations with two sources. Allowed patterns: SOPC (32 bits) SOPC + LITERAL (64 bits) ENCODING Encoding. 9 23 OP Opcode. 7 16 SSRC0 First source operand for instruction. 8 0 SSRC1 Second source operand for instruction. 8 8 ENC_SOPP 32 11111111111111110000000000000000 10111111100000000000000000000000 10111111100000010000000000000000 10111111100000100000000000000000 10111111100000110000000000000000 10111111100001010000000000000000 10111111100001110000000000000000 10111111100010000000000000000000 10111111100010010000000000000000 10111111100010100000000000000000 10111111100010110000000000000000 10111111100100000000000000000000 10111111100100010000000000000000 10111111100100100000000000000000 10111111100101000000000000000000 10111111100111110000000000000000 10111111101000000000000000000000 10111111101000010000000000000000 10111111101000100000000000000000 10111111101000110000000000000000 10111111101001000000000000000000 10111111101001010000000000000000 10111111101001100000000000000000 10111111101100000000000000000000 10111111101100010000000000000000 10111111101101000000000000000000 10111111101101010000000000000000 10111111101101100000000000000000 10111111101101110000000000000000 10111111101110000000000000000000 10111111101110010000000000000000 10111111101110100000000000000000 10111111101110110000000000000000 10111111101111000000000000000000 10111111110000000000000000000000 10111111110000010000000000000000 10111111110000100000000000000000 10111111110000110000000000000000 10111111110001000000000000000000 10111111110001100000000000000000 10111111110001110000000000000000 10111111110010000000000000000000 10111111110010010000000000000000 default 1 U 1 Scalar ALU operations for control flow. Allowed patterns: SOPP (32 bits) SOPP + LITERAL (64 bits) ENCODING Encoding. 9 23 OP Opcode. 7 16 SIMM16 16-bit integer input for opcode. Signedness is determined by opcode. 16 0 ENC_SOPK 32 11111111100000000000000000000000 10110000000000000000000000000000 10110000100000000000000000000000 10110001000000000000000000000000 10110111100000000000000000000000 10111000000000000000000000000000 10111000100000000000000000000000 10111001000000000000000000000000 10111010000000000000000000000000 default 1 U 1 Scalar ALU operations with one destination and one 16-bit constant source. Allowed patterns: SOPK (32 bits) SOPK + LITERAL (64 bits) ENCODING Encoding. 4 28 OP Opcode. 5 23 SDST Destination operand for instruction. 7 16 SIMM16 16-bit integer input for opcode. Signedness is determined by opcode. 16 0 ENC_SOP2 32 11111111100000000000000000000000 10000000000000000000000000000000 10000000100000000000000000000000 10000001000000000000000000000000 10000001100000000000000000000000 10000010000000000000000000000000 10000010100000000000000000000000 10000011000000000000000000000000 10000100000000000000000000000000 10000100100000000000000000000000 10000101000000000000000000000000 10000101100000000000000000000000 10000110000000000000000000000000 10000110100000000000000000000000 10000111000000000000000000000000 10000111100000000000000000000000 10001000000000000000000000000000 10001000100000000000000000000000 10001001000000000000000000000000 10001001100000000000000000000000 10001010000000000000000000000000 10001010100000000000000000000000 10001011000000000000000000000000 10001011100000000000000000000000 10001100000000000000000000000000 10001100100000000000000000000000 10001101000000000000000000000000 10001101100000000000000000000000 10001110000000000000000000000000 10001110100000000000000000000000 10001111000000000000000000000000 10001111100000000000000000000000 10010000000000000000000000000000 10010000100000000000000000000000 10010001000000000000000000000000 10010001100000000000000000000000 10010010000000000000000000000000 10010010100000000000000000000000 10010011000000000000000000000000 10010011100000000000000000000000 10010100000000000000000000000000 10010100100000000000000000000000 10010101000000000000000000000000 10010101100000000000000000000000 10010110000000000000000000000000 10010110100000000000000000000000 10010111000000000000000000000000 10011000000000000000000000000000 10011000100000000000000000000000 10011001000000000000000000000000 10011001100000000000000000000000 10011010000000000000000000000000 10011010100000000000000000000000 10100000000000000000000000000000 10100000100000000000000000000000 10100001000000000000000000000000 10100001100000000000000000000000 10100010000000000000000000000000 10100011100000000000000000000000 10100100000000000000000000000000 10100100100000000000000000000000 10100101000000000000000000000000 10100101100000000000000000000000 10100110000000000000000000000000 10100110100000000000000000000000 10100111000000000000000000000000 10100111100000000000000000000000 10101000000000000000000000000000 10101000100000000000000000000000 10101001000000000000000000000000 10101001100000000000000000000000 10101010000000000000000000000000 10101010100000000000000000000000 default && != .fieldderef .fieldderef INST SOP2 SSRC0 255 I 32 != .fieldderef .fieldderef INST SOP2 SSRC1 255 I 32 Scalar ALU operations with one destination and two sources. Allowed patterns: SOP2 (32 bits) SOP2 + LITERAL (64 bits) ENCODING Encoding. 2 30 OP Opcode. 7 23 SDST Destination operand for instruction. 7 16 SSRC0 First source operand for instruction. 8 0 SSRC1 Second source operand for instruction. 8 8 ENC_SMEM 64 0000000000000000000000000000000011111100000001111110000000000000 0000000000000000000000000000000011110100000000000000000000000000 0000000000000000000000000000000011110100000000000010000000000000 0000000000000000000000000000000011110100000000000100000000000000 0000000000000000000000000000000011110100000000000110000000000000 0000000000000000000000000000000011110100000000001000000000000000 0000000000000000000000000000000011110100000000001010000000000000 0000000000000000000000000000000011110100000000010000000000000000 0000000000000000000000000000000011110100000000010010000000000000 0000000000000000000000000000000011110100000000010100000000000000 0000000000000000000000000000000011110100000000010110000000000000 0000000000000000000000000000000011110100000000100000000000000000 0000000000000000000000000000000011110100000000100010000000000000 0000000000000000000000000000000011110100000000100100000000000000 0000000000000000000000000000000011110100000000100110000000000000 0000000000000000000000000000000011110100000000101000000000000000 0000000000000000000000000000000011110100000000101010000000000000 0000000000000000000000000000000011110100000000110000000000000000 0000000000000000000000000000000011110100000000110010000000000000 0000000000000000000000000000000011110100000000110100000000000000 0000000000000000000000000000000011110100000000110110000000000000 0000000000000000000000000000000011110100000001000010000000000000 0000000000000000000000000000000011110100000001000100000000000000 0000000000000000000000000000000011110100000001000110000000000000 0000000000000000000000000000000011110100000001001000000000000000 0000000000000000000000000000000011110100000001001010000000000000 0000000000000000000000000000000011110100000001001100000000000000 0000000000000000000000000000000011110100000001001110000000000000 0000000000000000000000000000000011110100000001010000000000000000 default 1 U 1 Scalar memory operations. Examples: // Load some resource constants into SGPRs s_load_b128 s[20:23], s[0:1], 0x0 // pos resource buffer s_load_b128 s[24:27], s[0:1], 0x4 // color0 resource buffer s_load_b128 s[28:31], s[0:1], 0x10 // tex0 resource buffer Different ways of writing an offset: s_load_b32 s20, s[0:1], null // No offset applied s_load_b32 s20, s[0:1], s4 // Register-based offset only s_load_b32 s20, s[0:1], s4 offset:0x8 // Apply register and constant offsets To simplify porting older shader programs, instructions accepting a constant offset marked with the OPF_SMEM_LEGACY_OFFSET flag may either specify the offset as an operand (to match old architectures that did not allow for an SGPR offset) or they may use the offset: modifier. If a constant operand is used then the offset: modifier may not be used, and the instruction is translated to the current binary format as shown below: s_load_b32 s20, s[0:1], 0x8 // Constant offset only, in operand s_load_b32 s20, s[0:1], null offset:0x8 // Identical to above, using modifier, matches hardware The second example reflects the current binary encoding of the instruction. ENCODING Encoding. 6 26 IOFFSET Specifies a 24-bit SIGNED BYTE offset aligned to one DWORD. Negative offsets only work with instructions that use a raw 64-bit pointer in SBASE; when using a buffer descriptor, any negative offset results in MEMVIOL. 24 32 NV Non-volatile bit. 1 20 OP Opcode. 6 13 SBASE Bits [6:1] of an SGPR pair carrying a base address or buffer descriptor (SBASE has an implied LSB of zero). When providing a buffer descriptor, only the BASE, STRIDE and NUM_RECORDS fields are used for scalar instructions. The low-order bits are in the first SGPR. 6 0 1 0 SCOPE Scope of memory operation. 2 21 SDATA Destination operand for instruction. 7 6 SOFFSET If not set to SQ_NULL: Specifies an SGPR address that supplies an UNSIGNED BYTE offset aligned to one DWORD for the memory operation (see enumeration). If set to SQ_NULL: not used in the address calculation. 7 57 TH Temporal hints and atomic return request. 2 23 ENC_VOP1 32 11111110000000001111111000000000 01111110000000000000000000000000 01111110000000000000001000000000 01111110000000000000010000000000 01111110000000000000011000000000 01111110000000000000100000000000 01111110000000000000101000000000 01111110000000000000110000000000 01111110000000000000111000000000 01111110000000000001000000000000 01111110000000000001010000000000 01111110000000000001011000000000 01111110000000000001100000000000 01111110000000000001101000000000 01111110000000000001110000000000 01111110000000000001111000000000 01111110000000000010000000000000 01111110000000000010001000000000 01111110000000000010010000000000 01111110000000000010011000000000 01111110000000000010100000000000 01111110000000000010101000000000 01111110000000000010110000000000 01111110000000000010111000000000 01111110000000000011000000000000 01111110000000000011001000000000 01111110000000000011010000000000 01111110000000000011011000000000 01111110000000000011100000000000 01111110000000000100000000000000 01111110000000000100001000000000 01111110000000000100010000000000 01111110000000000100011000000000 01111110000000000100100000000000 01111110000000000100101000000000 01111110000000000100111000000000 01111110000000000101010000000000 01111110000000000101011000000000 01111110000000000101110000000000 01111110000000000101111000000000 01111110000000000110001000000000 01111110000000000110011000000000 01111110000000000110100000000000 01111110000000000110101000000000 01111110000000000110110000000000 01111110000000000110111000000000 01111110000000000111000000000000 01111110000000000111001000000000 01111110000000000111010000000000 01111110000000000111011000000000 01111110000000000111100000000000 01111110000000000111101000000000 01111110000000000111110000000000 01111110000000000111111000000000 01111110000000001000000000000000 01111110000000001000010000000000 01111110000000001000011000000000 01111110000000001000100000000000 01111110000000001001000000000000 01111110000000001010000000000000 01111110000000001010001000000000 01111110000000001010010000000000 01111110000000001010011000000000 01111110000000001010100000000000 01111110000000001010101000000000 01111110000000001010110000000000 01111110000000001010111000000000 01111110000000001011000000000000 01111110000000001011001000000000 01111110000000001011010000000000 01111110000000001011011000000000 01111110000000001011100000000000 01111110000000001011101000000000 01111110000000001011110000000000 01111110000000001011111000000000 01111110000000001100000000000000 01111110000000001100001000000000 01111110000000001100010000000000 01111110000000001100011000000000 01111110000000001100100000000000 01111110000000001100101000000000 01111110000000001100110000000000 01111110000000001100111000000000 01111110000000001101000000000000 01111110000000001101001000000000 01111110000000001101010000000000 01111110000000001101011000000000 01111110000000001101100000000000 01111110000000001101101000000000 01111110000000001101110000000000 01111110000000001101111000000000 default && && != .fieldderef .fieldderef INST VOP1 SRC0 250 I 32 && != .fieldderef .fieldderef INST VOP1 SRC0 233 I 32 != .fieldderef .fieldderef INST VOP1 SRC0 234 I 32 != .fieldderef .fieldderef INST VOP1 SRC0 255 I 32 Vector ALU operations with one destination and one source. Instructions in this encoding may be promoted to VOP3 unless otherwise noted. Instructions in this encoding may use DPP modes unless otherwise noted. Allowed patterns: VOP1 (32 bits) VOP1 + LITERAL (64 bits) VOP1 + DPP (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) ENCODING Encoding. 7 25 OP Opcode. 7 9 SRC0 First source operand for instruction. 9 0 VDST Destination operand for instruction. 8 17 ENC_VOPC 32 11111111111111100000000000000000 01111100000000100000000000000000 01111100000001000000000000000000 01111100000001100000000000000000 01111100000010000000000000000000 01111100000010100000000000000000 01111100000011000000000000000000 01111100000011100000000000000000 01111100000100000000000000000000 01111100000100100000000000000000 01111100000101000000000000000000 01111100000101100000000000000000 01111100000110000000000000000000 01111100000110100000000000000000 01111100000111000000000000000000 01111100001000100000000000000000 01111100001001000000000000000000 01111100001001100000000000000000 01111100001010000000000000000000 01111100001010100000000000000000 01111100001011000000000000000000 01111100001011100000000000000000 01111100001100000000000000000000 01111100001100100000000000000000 01111100001101000000000000000000 01111100001101100000000000000000 01111100001110000000000000000000 01111100001110100000000000000000 01111100001111000000000000000000 01111100010000100000000000000000 01111100010001000000000000000000 01111100010001100000000000000000 01111100010010000000000000000000 01111100010010100000000000000000 01111100010011000000000000000000 01111100010011100000000000000000 01111100010100000000000000000000 01111100010100100000000000000000 01111100010101000000000000000000 01111100010101100000000000000000 01111100010110000000000000000000 01111100010110100000000000000000 01111100010111000000000000000000 01111100011000100000000000000000 01111100011001000000000000000000 01111100011001100000000000000000 01111100011010000000000000000000 01111100011010100000000000000000 01111100011011000000000000000000 01111100011100100000000000000000 01111100011101000000000000000000 01111100011101100000000000000000 01111100011110000000000000000000 01111100011110100000000000000000 01111100011111000000000000000000 01111100100000100000000000000000 01111100100001000000000000000000 01111100100001100000000000000000 01111100100010000000000000000000 01111100100010100000000000000000 01111100100011000000000000000000 01111100100100100000000000000000 01111100100101000000000000000000 01111100100101100000000000000000 01111100100110000000000000000000 01111100100110100000000000000000 01111100100111000000000000000000 01111100101000100000000000000000 01111100101001000000000000000000 01111100101001100000000000000000 01111100101010000000000000000000 01111100101010100000000000000000 01111100101011000000000000000000 01111100101100100000000000000000 01111100101101000000000000000000 01111100101101100000000000000000 01111100101110000000000000000000 01111100101110100000000000000000 01111100101111000000000000000000 01111100111110100000000000000000 01111100111111000000000000000000 01111100111111100000000000000000 01111101000000100000000000000000 01111101000001000000000000000000 01111101000001100000000000000000 01111101000010000000000000000000 01111101000010100000000000000000 01111101000011000000000000000000 01111101000011100000000000000000 01111101000100000000000000000000 01111101000100100000000000000000 01111101000101000000000000000000 01111101000101100000000000000000 01111101000110000000000000000000 01111101000110100000000000000000 01111101000111000000000000000000 01111101001000100000000000000000 01111101001001000000000000000000 01111101001001100000000000000000 01111101001010000000000000000000 01111101001010100000000000000000 01111101001011000000000000000000 01111101001011100000000000000000 01111101001100000000000000000000 01111101001100100000000000000000 01111101001101000000000000000000 01111101001101100000000000000000 01111101001110000000000000000000 01111101001110100000000000000000 01111101001111000000000000000000 01111101010000100000000000000000 01111101010001000000000000000000 01111101010001100000000000000000 01111101010010000000000000000000 01111101010010100000000000000000 01111101010011000000000000000000 01111101010011100000000000000000 01111101010100000000000000000000 01111101010100100000000000000000 01111101010101000000000000000000 01111101010101100000000000000000 01111101010110000000000000000000 01111101010110100000000000000000 01111101010111000000000000000000 01111101011000100000000000000000 01111101011001000000000000000000 01111101011001100000000000000000 01111101011010000000000000000000 01111101011010100000000000000000 01111101011011000000000000000000 01111101011100100000000000000000 01111101011101000000000000000000 01111101011101100000000000000000 01111101011110000000000000000000 01111101011110100000000000000000 01111101011111000000000000000000 01111101100000100000000000000000 01111101100001000000000000000000 01111101100001100000000000000000 01111101100010000000000000000000 01111101100010100000000000000000 01111101100011000000000000000000 01111101100100100000000000000000 01111101100101000000000000000000 01111101100101100000000000000000 01111101100110000000000000000000 01111101100110100000000000000000 01111101100111000000000000000000 01111101101000100000000000000000 01111101101001000000000000000000 01111101101001100000000000000000 01111101101010000000000000000000 01111101101010100000000000000000 01111101101011000000000000000000 01111101101100100000000000000000 01111101101101000000000000000000 01111101101101100000000000000000 01111101101110000000000000000000 01111101101110100000000000000000 01111101101111000000000000000000 01111101111110100000000000000000 01111101111111000000000000000000 01111101111111100000000000000000 default && && != .fieldderef .fieldderef INST VOPC SRC0 250 I 32 && != .fieldderef .fieldderef INST VOPC SRC0 233 I 32 != .fieldderef .fieldderef INST VOPC SRC0 234 I 32 != .fieldderef .fieldderef INST VOPC SRC0 255 I 32 Vector ALU comparison operations with two sources. Instructions in this encoding may be promoted to VOP3 unless otherwise noted. Note that vector compare instructions respect the EXEC mask for computing the result in each thread; if EXEC[laneId] is zero, then the comparison for that thread is not performed and the result of VCC[laneId] is zero. Instructions in this encoding may use DPP modes unless otherwise noted. Allowed patterns: VOPC (32 bits) VOPC + LITERAL (64 bits) VOPC + DPP (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) ENCODING Encoding. 7 25 OP Opcode. 8 17 SRC0 First source operand for instruction. 9 0 VSRC1 Second source operand for instruction. 8 9 ENC_VOP2 32 11111110000000000000000000000000 00000010000000000000000000000000 00000100000000000000000000000000 00000110000000000000000000000000 00001000000000000000000000000000 00001010000000000000000000000000 00001100000000000000000000000000 00001110000000000000000000000000 00010000000000000000000000000000 00010010000000000000000000000000 00010100000000000000000000000000 00010110000000000000000000000000 00011000000000000000000000000000 00011010000000000000000000000000 00011100000000000000000000000000 00100010000000000000000000000000 00100100000000000000000000000000 00100110000000000000000000000000 00101000000000000000000000000000 00101010000000000000000000000000 00101100000000000000000000000000 00110000000000000000000000000000 00110010000000000000000000000000 00110100000000000000000000000000 00110110000000000000000000000000 00111000000000000000000000000000 00111010000000000000000000000000 00111100000000000000000000000000 00111110000000000000000000000000 01000000000000000000000000000000 01000010000000000000000000000000 01000100000000000000000000000000 01001010000000000000000000000000 01001100000000000000000000000000 01001110000000000000000000000000 01010110000000000000000000000000 01011110000000000000000000000000 01100000000000000000000000000000 01100010000000000000000000000000 01100100000000000000000000000000 01100110000000000000000000000000 01101000000000000000000000000000 01101010000000000000000000000000 01101100000000000000000000000000 01110110000000000000000000000000 01111000000000000000000000000000 default && && != .fieldderef .fieldderef INST VOP2 SRC0 250 I 32 && != .fieldderef .fieldderef INST VOP2 SRC0 233 I 32 != .fieldderef .fieldderef INST VOP2 SRC0 234 I 32 != .fieldderef .fieldderef INST VOP2 SRC0 255 I 32 Vector ALU operations with one destination and two sources. Instructions in this encoding may be promoted to VOP3 unless otherwise noted. DPP modes are allowed unless otherwise noted. Allowed patterns: VOP2 (32 bits) VOP2 + LITERAL (64 bits) VOP2 + DPP (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) ENCODING Encoding. 1 31 OP Opcode. 6 25 SRC0 First source operand for instruction. 9 0 VDST Destination operand for instruction. 8 17 VSRC1 Second source operand for instruction. 8 9 ENC_VOP3 64 0000000000000000000000000000000011111111111111110000000000000000 0000000000000000000000000000000011010101100000000000000000000000 0000000000000000000000000000000011010101100000010000000000000000 0000000000000000000000000000000011010101100000100000000000000000 0000000000000000000000000000000011010101100000110000000000000000 0000000000000000000000000000000011010101100001000000000000000000 0000000000000000000000000000000011010101100001010000000000000000 0000000000000000000000000000000011010101100001100000000000000000 0000000000000000000000000000000011010101100001110000000000000000 0000000000000000000000000000000011010101100010000000000000000000 0000000000000000000000000000000011010101100010100000000000000000 0000000000000000000000000000000011010101100010110000000000000000 0000000000000000000000000000000011010101100011000000000000000000 0000000000000000000000000000000011010101100011010000000000000000 0000000000000000000000000000000011010101100011100000000000000000 0000000000000000000000000000000011010101100011110000000000000000 0000000000000000000000000000000011010101100100000000000000000000 0000000000000000000000000000000011010101100100010000000000000000 0000000000000000000000000000000011010101100100100000000000000000 0000000000000000000000000000000011010101100100110000000000000000 0000000000000000000000000000000011010101100101000000000000000000 0000000000000000000000000000000011010101100101010000000000000000 0000000000000000000000000000000011010101100101100000000000000000 0000000000000000000000000000000011010101100101110000000000000000 0000000000000000000000000000000011010101100110000000000000000000 0000000000000000000000000000000011010101100110010000000000000000 0000000000000000000000000000000011010101100110100000000000000000 0000000000000000000000000000000011010101100110110000000000000000 0000000000000000000000000000000011010101100111000000000000000000 0000000000000000000000000000000011010101101000000000000000000000 0000000000000000000000000000000011010101101000010000000000000000 0000000000000000000000000000000011010101101000100000000000000000 0000000000000000000000000000000011010101101000110000000000000000 0000000000000000000000000000000011010101101001000000000000000000 0000000000000000000000000000000011010101101001010000000000000000 0000000000000000000000000000000011010101101001110000000000000000 0000000000000000000000000000000011010101101010100000000000000000 0000000000000000000000000000000011010101101010110000000000000000 0000000000000000000000000000000011010101101011100000000000000000 0000000000000000000000000000000011010101101011110000000000000000 0000000000000000000000000000000011010101101100010000000000000000 0000000000000000000000000000000011010101101100110000000000000000 0000000000000000000000000000000011010101101101000000000000000000 0000000000000000000000000000000011010101101101010000000000000000 0000000000000000000000000000000011010101101101100000000000000000 0000000000000000000000000000000011010101101101110000000000000000 0000000000000000000000000000000011010101101110000000000000000000 0000000000000000000000000000000011010101101110010000000000000000 0000000000000000000000000000000011010101101110100000000000000000 0000000000000000000000000000000011010101101110110000000000000000 0000000000000000000000000000000011010101101111000000000000000000 0000000000000000000000000000000011010101101111010000000000000000 0000000000000000000000000000000011010101101111100000000000000000 0000000000000000000000000000000011010101101111110000000000000000 0000000000000000000000000000000011010101110000000000000000000000 0000000000000000000000000000000011010101110000100000000000000000 0000000000000000000000000000000011010101110000110000000000000000 0000000000000000000000000000000011010101110001000000000000000000 0000000000000000000000000000000011010101110010000000000000000000 0000000000000000000000000000000011010101110100000000000000000000 0000000000000000000000000000000011010101110100010000000000000000 0000000000000000000000000000000011010101110100100000000000000000 0000000000000000000000000000000011010101110100110000000000000000 0000000000000000000000000000000011010101110101000000000000000000 0000000000000000000000000000000011010101110101010000000000000000 0000000000000000000000000000000011010101110101100000000000000000 0000000000000000000000000000000011010101110101110000000000000000 0000000000000000000000000000000011010101110110000000000000000000 0000000000000000000000000000000011010101110110010000000000000000 0000000000000000000000000000000011010101110110100000000000000000 0000000000000000000000000000000011010101110110110000000000000000 0000000000000000000000000000000011010101110111000000000000000000 0000000000000000000000000000000011010101110111010000000000000000 0000000000000000000000000000000011010101110111100000000000000000 0000000000000000000000000000000011010101110111110000000000000000 0000000000000000000000000000000011010101111000000000000000000000 0000000000000000000000000000000011010101111000010000000000000000 0000000000000000000000000000000011010101111000100000000000000000 0000000000000000000000000000000011010101111000110000000000000000 0000000000000000000000000000000011010101111001000000000000000000 0000000000000000000000000000000011010101111010010000000000000000 0000000000000000000000000000000011010101111010100000000000000000 0000000000000000000000000000000011010101111010110000000000000000 0000000000000000000000000000000011010101111011000000000000000000 0000000000000000000000000000000011010101111011010000000000000000 0000000000000000000000000000000011010101111011100000000000000000 0000000000000000000000000000000011010101111011110000000000000000 0000000000000000000000000000000011010101000000010000000000000000 0000000000000000000000000000000011010101000000100000000000000000 0000000000000000000000000000000011010101000000110000000000000000 0000000000000000000000000000000011010101000001000000000000000000 0000000000000000000000000000000011010101000001010000000000000000 0000000000000000000000000000000011010101000001100000000000000000 0000000000000000000000000000000011010101000001110000000000000000 0000000000000000000000000000000011010101000010000000000000000000 0000000000000000000000000000000011010101000010010000000000000000 0000000000000000000000000000000011010101000010100000000000000000 0000000000000000000000000000000011010101000010110000000000000000 0000000000000000000000000000000011010101000011000000000000000000 0000000000000000000000000000000011010101000011010000000000000000 0000000000000000000000000000000011010101000011100000000000000000 0000000000000000000000000000000011010101000100010000000000000000 0000000000000000000000000000000011010101000100100000000000000000 0000000000000000000000000000000011010101000100110000000000000000 0000000000000000000000000000000011010101000101000000000000000000 0000000000000000000000000000000011010101000101010000000000000000 0000000000000000000000000000000011010101000101100000000000000000 0000000000000000000000000000000011010101000110000000000000000000 0000000000000000000000000000000011010101000110010000000000000000 0000000000000000000000000000000011010101000110100000000000000000 0000000000000000000000000000000011010101000110110000000000000000 0000000000000000000000000000000011010101000111000000000000000000 0000000000000000000000000000000011010101000111010000000000000000 0000000000000000000000000000000011010101000111100000000000000000 0000000000000000000000000000000011010101000111110000000000000000 0000000000000000000000000000000011010101001001010000000000000000 0000000000000000000000000000000011010101001001100000000000000000 0000000000000000000000000000000011010101001001110000000000000000 0000000000000000000000000000000011010101001010110000000000000000 0000000000000000000000000000000011010101001011110000000000000000 0000000000000000000000000000000011010101001100000000000000000000 0000000000000000000000000000000011010101001100010000000000000000 0000000000000000000000000000000011010101001100100000000000000000 0000000000000000000000000000000011010101001100110000000000000000 0000000000000000000000000000000011010101001101000000000000000000 0000000000000000000000000000000011010101001101010000000000000000 0000000000000000000000000000000011010101001101100000000000000000 0000000000000000000000000000000011010101001110110000000000000000 0000000000000000000000000000000011010110000010010000000000000000 0000000000000000000000000000000011010110000010100000000000000000 0000000000000000000000000000000011010110000010110000000000000000 0000000000000000000000000000000011010110000011000000000000000000 0000000000000000000000000000000011010110000011010000000000000000 0000000000000000000000000000000011010110000011100000000000000000 0000000000000000000000000000000011010110000011110000000000000000 0000000000000000000000000000000011010110000100000000000000000000 0000000000000000000000000000000011010110000100010000000000000000 0000000000000000000000000000000011010110000100100000000000000000 0000000000000000000000000000000011010110000100110000000000000000 0000000000000000000000000000000011010110000101000000000000000000 0000000000000000000000000000000011010110000101010000000000000000 0000000000000000000000000000000011010110000101100000000000000000 0000000000000000000000000000000011010110000101110000000000000000 0000000000000000000000000000000011010110000110000000000000000000 0000000000000000000000000000000011010110000110100000000000000000 0000000000000000000000000000000011010110000110110000000000000000 0000000000000000000000000000000011010110000111010000000000000000 0000000000000000000000000000000011010110000111100000000000000000 0000000000000000000000000000000011010110001000000000000000000000 0000000000000000000000000000000011010110001000010000000000000000 0000000000000000000000000000000011010110001000100000000000000000 0000000000000000000000000000000011010110001000110000000000000000 0000000000000000000000000000000011010110001001000000000000000000 0000000000000000000000000000000011010110001001010000000000000000 0000000000000000000000000000000011010110001001100000000000000000 0000000000000000000000000000000011010110001001110000000000000000 0000000000000000000000000000000011010110001010000000000000000000 0000000000000000000000000000000011010110001010010000000000000000 0000000000000000000000000000000011010110001010100000000000000000 0000000000000000000000000000000011010110001010110000000000000000 0000000000000000000000000000000011010110001011000000000000000000 0000000000000000000000000000000011010110001011010000000000000000 0000000000000000000000000000000011010110001011100000000000000000 0000000000000000000000000000000011010110001011110000000000000000 0000000000000000000000000000000011010110001100000000000000000000 0000000000000000000000000000000011010110001100010000000000000000 0000000000000000000000000000000011010110001100100000000000000000 0000000000000000000000000000000011010110001101110000000000000000 0000000000000000000000000000000011010110001110000000000000000000 0000000000000000000000000000000011010110001110010000000000000000 0000000000000000000000000000000011010110001110100000000000000000 0000000000000000000000000000000011010110001110110000000000000000 0000000000000000000000000000000011010110001111010000000000000000 0000000000000000000000000000000011010110010000000000000000000000 0000000000000000000000000000000011010110010000010000000000000000 0000000000000000000000000000000011010110010001000000000000000000 0000000000000000000000000000000011010110010001010000000000000000 0000000000000000000000000000000011010110010001100000000000000000 0000000000000000000000000000000011010110010001110000000000000000 0000000000000000000000000000000011010110010010000000000000000000 0000000000000000000000000000000011010110010010100000000000000000 0000000000000000000000000000000011010110010010110000000000000000 0000000000000000000000000000000011010110010011010000000000000000 0000000000000000000000000000000011010110010011100000000000000000 0000000000000000000000000000000011010110010100000000000000000000 0000000000000000000000000000000011010110010100010000000000000000 0000000000000000000000000000000011010110010100110000000000000000 0000000000000000000000000000000011010110010101000000000000000000 0000000000000000000000000000000011010110010101010000000000000000 0000000000000000000000000000000011010110010101100000000000000000 0000000000000000000000000000000011010110010101110000000000000000 0000000000000000000000000000000011010110010110000000000000000000 0000000000000000000000000000000011010110010110010000000000000000 0000000000000000000000000000000011010110010110100000000000000000 0000000000000000000000000000000011010110010110110000000000000000 0000000000000000000000000000000011010110010111000000000000000000 0000000000000000000000000000000011010110010111010000000000000000 0000000000000000000000000000000011010110011000100000000000000000 0000000000000000000000000000000011010110011000110000000000000000 0000000000000000000000000000000011010110011001000000000000000000 0000000000000000000000000000000011010110011001010000000000000000 0000000000000000000000000000000011010110011001100000000000000000 0000000000000000000000000000000011010110011001110000000000000000 0000000000000000000000000000000011010110011010000000000000000000 0000000000000000000000000000000011010110011010010000000000000000 0000000000000000000000000000000011010110011010100000000000000000 0000000000000000000000000000000011010110011010110000000000000000 0000000000000000000000000000000011010110011011000000000000000000 0000000000000000000000000000000011010110011011010000000000000000 0000000000000000000000000000000011010110011011100000000000000000 0000000000000000000000000000000011010110011011110000000000000000 0000000000000000000000000000000011010110100000000000000000000000 0000000000000000000000000000000011010110100000010000000000000000 0000000000000000000000000000000011010110100000100000000000000000 0000000000000000000000000000000011010110100000110000000000000000 0000000000000000000000000000000011010110100001000000000000000000 0000000000000000000000000000000011010110100001010000000000000000 0000000000000000000000000000000011010110100001100000000000000000 0000000000000000000000000000000011010110100001110000000000000000 0000000000000000000000000000000011010110100010000000000000000000 0000000000000000000000000000000011010110100010010000000000000000 0000000000000000000000000000000011010111000000110000000000000000 0000000000000000000000000000000011010111000001000000000000000000 0000000000000000000000000000000011010111000001010000000000000000 0000000000000000000000000000000011010111000001100000000000000000 0000000000000000000000000000000011010111000001110000000000000000 0000000000000000000000000000000011010111000010010000000000000000 0000000000000000000000000000000011010111000010100000000000000000 0000000000000000000000000000000011010111000010110000000000000000 0000000000000000000000000000000011010111000011000000000000000000 0000000000000000000000000000000011010111000011010000000000000000 0000000000000000000000000000000011010111000011100000000000000000 0000000000000000000000000000000011010111000011110000000000000000 0000000000000000000000000000000011010111000100000000000000000000 0000000000000000000000000000000011010111000100010000000000000000 0000000000000000000000000000000011010111000100100000000000000000 0000000000000000000000000000000011010111000100110000000000000000 0000000000000000000000000000000011010111000111000000000000000000 0000000000000000000000000000000011010111000111010000000000000000 0000000000000000000000000000000011010111000111100000000000000000 0000000000000000000000000000000011010111000111110000000000000000 0000000000000000000000000000000011010111001000000000000000000000 0000000000000000000000000000000011010111001000010000000000000000 0000000000000000000000000000000011010111001000100000000000000000 0000000000000000000000000000000011010111001000110000000000000000 0000000000000000000000000000000011010111001001000000000000000000 0000000000000000000000000000000011010111001001010000000000000000 0000000000000000000000000000000011010111001001100000000000000000 0000000000000000000000000000000011010111001010110000000000000000 0000000000000000000000000000000011010111001011000000000000000000 0000000000000000000000000000000011010111001011010000000000000000 0000000000000000000000000000000011010111001011100000000000000000 0000000000000000000000000000000011010111001011110000000000000000 0000000000000000000000000000000011010111001110000000000000000000 0000000000000000000000000000000011010111001110010000000000000000 0000000000000000000000000000000011010111001110100000000000000000 0000000000000000000000000000000011010111001111010000000000000000 0000000000000000000000000000000011010111001111100000000000000000 0000000000000000000000000000000011010111010000010000000000000000 0000000000000000000000000000000011010111010000100000000000000000 0000000000000000000000000000000011010111011000000000000000000000 0000000000000000000000000000000011010111011000010000000000000000 0000000000000000000000000000000011010111011000100000000000000000 0000000000000000000000000000000011010111011000110000000000000000 0000000000000000000000000000000011010111011001000000000000000000 0000000000000000000000000000000011010111011001010000000000000000 0000000000000000000000000000000011010111011001100000000000000000 0000000000000000000000000000000011010111011001110000000000000000 0000000000000000000000000000000011010111011010000000000000000000 0000000000000000000000000000000011010111011010010000000000000000 0000000000000000000000000000000011010111011010100000000000000000 0000000000000000000000000000000011010111011010110000000000000000 0000000000000000000000000000000011010111011011000000000000000000 0000000000000000000000000000000011010100000000010000000000000000 0000000000000000000000000000000011010100000000100000000000000000 0000000000000000000000000000000011010100000000110000000000000000 0000000000000000000000000000000011010100000001000000000000000000 0000000000000000000000000000000011010100000001010000000000000000 0000000000000000000000000000000011010100000001100000000000000000 0000000000000000000000000000000011010100000001110000000000000000 0000000000000000000000000000000011010100000010000000000000000000 0000000000000000000000000000000011010100000010010000000000000000 0000000000000000000000000000000011010100000010100000000000000000 0000000000000000000000000000000011010100000010110000000000000000 0000000000000000000000000000000011010100000011000000000000000000 0000000000000000000000000000000011010100000011010000000000000000 0000000000000000000000000000000011010100000011100000000000000000 0000000000000000000000000000000011010100000100010000000000000000 0000000000000000000000000000000011010100000100100000000000000000 0000000000000000000000000000000011010100000100110000000000000000 0000000000000000000000000000000011010100000101000000000000000000 0000000000000000000000000000000011010100000101010000000000000000 0000000000000000000000000000000011010100000101100000000000000000 0000000000000000000000000000000011010100000101110000000000000000 0000000000000000000000000000000011010100000110000000000000000000 0000000000000000000000000000000011010100000110010000000000000000 0000000000000000000000000000000011010100000110100000000000000000 0000000000000000000000000000000011010100000110110000000000000000 0000000000000000000000000000000011010100000111000000000000000000 0000000000000000000000000000000011010100000111010000000000000000 0000000000000000000000000000000011010100000111100000000000000000 0000000000000000000000000000000011010100001000010000000000000000 0000000000000000000000000000000011010100001000100000000000000000 0000000000000000000000000000000011010100001000110000000000000000 0000000000000000000000000000000011010100001001000000000000000000 0000000000000000000000000000000011010100001001010000000000000000 0000000000000000000000000000000011010100001001100000000000000000 0000000000000000000000000000000011010100001001110000000000000000 0000000000000000000000000000000011010100001010000000000000000000 0000000000000000000000000000000011010100001010010000000000000000 0000000000000000000000000000000011010100001010100000000000000000 0000000000000000000000000000000011010100001010110000000000000000 0000000000000000000000000000000011010100001011000000000000000000 0000000000000000000000000000000011010100001011010000000000000000 0000000000000000000000000000000011010100001011100000000000000000 0000000000000000000000000000000011010100001100010000000000000000 0000000000000000000000000000000011010100001100100000000000000000 0000000000000000000000000000000011010100001100110000000000000000 0000000000000000000000000000000011010100001101000000000000000000 0000000000000000000000000000000011010100001101010000000000000000 0000000000000000000000000000000011010100001101100000000000000000 0000000000000000000000000000000011010100001110010000000000000000 0000000000000000000000000000000011010100001110100000000000000000 0000000000000000000000000000000011010100001110110000000000000000 0000000000000000000000000000000011010100001111000000000000000000 0000000000000000000000000000000011010100001111010000000000000000 0000000000000000000000000000000011010100001111100000000000000000 0000000000000000000000000000000011010100010000010000000000000000 0000000000000000000000000000000011010100010000100000000000000000 0000000000000000000000000000000011010100010000110000000000000000 0000000000000000000000000000000011010100010001000000000000000000 0000000000000000000000000000000011010100010001010000000000000000 0000000000000000000000000000000011010100010001100000000000000000 0000000000000000000000000000000011010100010010010000000000000000 0000000000000000000000000000000011010100010010100000000000000000 0000000000000000000000000000000011010100010010110000000000000000 0000000000000000000000000000000011010100010011000000000000000000 0000000000000000000000000000000011010100010011010000000000000000 0000000000000000000000000000000011010100010011100000000000000000 0000000000000000000000000000000011010100010100010000000000000000 0000000000000000000000000000000011010100010100100000000000000000 0000000000000000000000000000000011010100010100110000000000000000 0000000000000000000000000000000011010100010101000000000000000000 0000000000000000000000000000000011010100010101010000000000000000 0000000000000000000000000000000011010100010101100000000000000000 0000000000000000000000000000000011010100010110010000000000000000 0000000000000000000000000000000011010100010110100000000000000000 0000000000000000000000000000000011010100010110110000000000000000 0000000000000000000000000000000011010100010111000000000000000000 0000000000000000000000000000000011010100010111010000000000000000 0000000000000000000000000000000011010100010111100000000000000000 0000000000000000000000000000000011010100011111010000000000000000 0000000000000000000000000000000011010100011111100000000000000000 0000000000000000000000000000000011010100011111110000000000000000 0000000000000000000000000000000011010100100000010000000000000000 0000000000000000000000000000000011010100100000100000000000000000 0000000000000000000000000000000011010100100000110000000000000000 0000000000000000000000000000000011010100100001000000000000000000 0000000000000000000000000000000011010100100001010000000000000000 0000000000000000000000000000000011010100100001100000000000000000 0000000000000000000000000000000011010100100001110000000000000000 0000000000000000000000000000000011010100100010000000000000000000 0000000000000000000000000000000011010100100010010000000000000000 0000000000000000000000000000000011010100100010100000000000000000 0000000000000000000000000000000011010100100010110000000000000000 0000000000000000000000000000000011010100100011000000000000000000 0000000000000000000000000000000011010100100011010000000000000000 0000000000000000000000000000000011010100100011100000000000000000 0000000000000000000000000000000011010100100100010000000000000000 0000000000000000000000000000000011010100100100100000000000000000 0000000000000000000000000000000011010100100100110000000000000000 0000000000000000000000000000000011010100100101000000000000000000 0000000000000000000000000000000011010100100101010000000000000000 0000000000000000000000000000000011010100100101100000000000000000 0000000000000000000000000000000011010100100101110000000000000000 0000000000000000000000000000000011010100100110000000000000000000 0000000000000000000000000000000011010100100110010000000000000000 0000000000000000000000000000000011010100100110100000000000000000 0000000000000000000000000000000011010100100110110000000000000000 0000000000000000000000000000000011010100100111000000000000000000 0000000000000000000000000000000011010100100111010000000000000000 0000000000000000000000000000000011010100100111100000000000000000 0000000000000000000000000000000011010100101000010000000000000000 0000000000000000000000000000000011010100101000100000000000000000 0000000000000000000000000000000011010100101000110000000000000000 0000000000000000000000000000000011010100101001000000000000000000 0000000000000000000000000000000011010100101001010000000000000000 0000000000000000000000000000000011010100101001100000000000000000 0000000000000000000000000000000011010100101001110000000000000000 0000000000000000000000000000000011010100101010000000000000000000 0000000000000000000000000000000011010100101010010000000000000000 0000000000000000000000000000000011010100101010100000000000000000 0000000000000000000000000000000011010100101010110000000000000000 0000000000000000000000000000000011010100101011000000000000000000 0000000000000000000000000000000011010100101011010000000000000000 0000000000000000000000000000000011010100101011100000000000000000 0000000000000000000000000000000011010100101100010000000000000000 0000000000000000000000000000000011010100101100100000000000000000 0000000000000000000000000000000011010100101100110000000000000000 0000000000000000000000000000000011010100101101000000000000000000 0000000000000000000000000000000011010100101101010000000000000000 0000000000000000000000000000000011010100101101100000000000000000 0000000000000000000000000000000011010100101110010000000000000000 0000000000000000000000000000000011010100101110100000000000000000 0000000000000000000000000000000011010100101110110000000000000000 0000000000000000000000000000000011010100101111000000000000000000 0000000000000000000000000000000011010100101111010000000000000000 0000000000000000000000000000000011010100101111100000000000000000 0000000000000000000000000000000011010100110000010000000000000000 0000000000000000000000000000000011010100110000100000000000000000 0000000000000000000000000000000011010100110000110000000000000000 0000000000000000000000000000000011010100110001000000000000000000 0000000000000000000000000000000011010100110001010000000000000000 0000000000000000000000000000000011010100110001100000000000000000 0000000000000000000000000000000011010100110010010000000000000000 0000000000000000000000000000000011010100110010100000000000000000 0000000000000000000000000000000011010100110010110000000000000000 0000000000000000000000000000000011010100110011000000000000000000 0000000000000000000000000000000011010100110011010000000000000000 0000000000000000000000000000000011010100110011100000000000000000 0000000000000000000000000000000011010100110100010000000000000000 0000000000000000000000000000000011010100110100100000000000000000 0000000000000000000000000000000011010100110100110000000000000000 0000000000000000000000000000000011010100110101000000000000000000 0000000000000000000000000000000011010100110101010000000000000000 0000000000000000000000000000000011010100110101100000000000000000 0000000000000000000000000000000011010100110110010000000000000000 0000000000000000000000000000000011010100110110100000000000000000 0000000000000000000000000000000011010100110110110000000000000000 0000000000000000000000000000000011010100110111000000000000000000 0000000000000000000000000000000011010100110111010000000000000000 0000000000000000000000000000000011010100110111100000000000000000 0000000000000000000000000000000011010100111111010000000000000000 0000000000000000000000000000000011010100111111100000000000000000 0000000000000000000000000000000011010100111111110000000000000000 default && && && && != .fieldderef .fieldderef INST VOP3 SRC0 250 I 32 && != .fieldderef .fieldderef INST VOP3 SRC0 233 I 32 != .fieldderef .fieldderef INST VOP3 SRC0 234 I 32 != .fieldderef .fieldderef INST VOP3 SRC0 255 I 32 != .fieldderef .fieldderef INST VOP3 SRC1 255 I 32 != .fieldderef .fieldderef INST VOP3 SRC2 255 I 32 Vector ALU operations requiring three sources or modifiers; also some obscure opcodes that cannot be encoded in smaller instructions due to limited OP bits. Almost all VOP1, VOP2 and VOPC opcodes can be promoted to VOP3 to use input and output modifiers. Allowed patterns: VOP3 (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) ABS If ABS[N] set, take the floating-point absolute value of the N'th input operand. This is applied before negation. See OPF_NOABS for instructions that do not allow this input modifier. 3 8 CLAMP If set, clamp or signal on output. VCMP instructions enable signaling-compare. For floating point outputs, clamp to [0.0, 1.0]. For signed integer outputs, clamp to [MIN_INT, MAX_INT]. For unsigned integer outputs, clamp to [0, MAX_UINT]. Applied after output modifier. See OPF_NOCLAMP for a list of instructions that do not support clamping or signaling. 1 15 ENCODING Encoding. 6 26 NEG If NEG[N] set, take the floating-point negation of the N'th input operand. This is applied after absolute value. See OPF_NONEG for a list of instructions that do not support negation. 3 61 OMOD Output modifier for instruction. Only useful for floating point outputs; integer outputs should set this to OMOD_OFF. Applied before clamping. See OPF_NOOMOD for a list of instructions that do not support output modifier. Integer and packed float 16 results ignore the omod setting. Output modifiers are not compatible with output denorms. If omod is not OMOD_OFF then output denormals are flushed regardless of the denorm control setting. Output modifiers are not IEEE compliant (-0*x=+0). 2 59 OP Opcode. 10 16 OPSEL For 16-bit inputs, OPSEL[N] determines if lower [15:0] or upper [31:16] bits are selected for the N'th source operand. Similarly OPSEL[3] controls whether the lower [15:0] or upper [31:16] bits are written for a 16-bit destination (preserving the unwritten bits). 0 selects the lower order bits and 1 selects the higher order bits. 4 11 SRC0 First source operand for instruction. 9 32 SRC1 Second source operand for instruction. 9 41 SRC2 Third source operand for instruction. 9 50 VDST Destination operand for instruction. 8 0 ENC_VOP3P 64 0000000000000000000000000000000011111111011111110000000000000000 0000000000000000000000000000000011001100000000000000000000000000 0000000000000000000000000000000011001100000000010000000000000000 0000000000000000000000000000000011001100000000100000000000000000 0000000000000000000000000000000011001100000000110000000000000000 0000000000000000000000000000000011001100000001000000000000000000 0000000000000000000000000000000011001100000001010000000000000000 0000000000000000000000000000000011001100000001100000000000000000 0000000000000000000000000000000011001100000001110000000000000000 0000000000000000000000000000000011001100000010000000000000000000 0000000000000000000000000000000011001100000010010000000000000000 0000000000000000000000000000000011001100000010100000000000000000 0000000000000000000000000000000011001100000010110000000000000000 0000000000000000000000000000000011001100000011000000000000000000 0000000000000000000000000000000011001100000011010000000000000000 0000000000000000000000000000000011001100000011100000000000000000 0000000000000000000000000000000011001100000011110000000000000000 0000000000000000000000000000000011001100000100000000000000000000 0000000000000000000000000000000011001100000100110000000000000000 0000000000000000000000000000000011001100000101100000000000000000 0000000000000000000000000000000011001100000101110000000000000000 0000000000000000000000000000000011001100000110000000000000000000 0000000000000000000000000000000011001100000110010000000000000000 0000000000000000000000000000000011001100000110100000000000000000 0000000000000000000000000000000011001100000110110000000000000000 0000000000000000000000000000000011001100000111000000000000000000 0000000000000000000000000000000011001100000111010000000000000000 0000000000000000000000000000000011001100000111100000000000000000 0000000000000000000000000000000011001100001000000000000000000000 0000000000000000000000000000000011001100001000010000000000000000 0000000000000000000000000000000011001100001000100000000000000000 0000000000000000000000000000000011001100001001000000000000000000 0000000000000000000000000000000011001100001001010000000000000000 0000000000000000000000000000000011001100001001100000000000000000 0000000000000000000000000000000011001100001001110000000000000000 0000000000000000000000000000000011001100010000000000000000000000 0000000000000000000000000000000011001100010000010000000000000000 0000000000000000000000000000000011001100010000100000000000000000 0000000000000000000000000000000011001100010000110000000000000000 0000000000000000000000000000000011001100010001000000000000000000 0000000000000000000000000000000011001100010001010000000000000000 0000000000000000000000000000000011001100010001100000000000000000 0000000000000000000000000000000011001100010001110000000000000000 0000000000000000000000000000000011001100010010000000000000000000 0000000000000000000000000000000011001100010010010000000000000000 0000000000000000000000000000000011001100010010100000000000000000 0000000000000000000000000000000011001100010100000000000000000000 0000000000000000000000000000000011001100010100010000000000000000 0000000000000000000000000000000011001100010100100000000000000000 0000000000000000000000000000000011001100010100110000000000000000 0000000000000000000000000000000011001100010101000000000000000000 0000000000000000000000000000000011001100010101010000000000000000 0000000000000000000000000000000011001100010101100000000000000000 0000000000000000000000000000000011001100010101110000000000000000 0000000000000000000000000000000011001100010110000000000000000000 0000000000000000000000000000000011001100010110010000000000000000 0000000000000000000000000000000011001100010110100000000000000000 default && && && && != .fieldderef .fieldderef INST VOP3P SRC0 250 I 32 && != .fieldderef .fieldderef INST VOP3P SRC0 233 I 32 != .fieldderef .fieldderef INST VOP3P SRC0 234 I 32 != .fieldderef .fieldderef INST VOP3P SRC0 255 I 32 != .fieldderef .fieldderef INST VOP3P SRC1 255 I 32 != .fieldderef .fieldderef INST VOP3P SRC2 255 I 32 Packed math instructions: vector ALU operations requiring two or three sources and performing two 16 bit ops in one instruction. Allowed patterns: VOP3P (64 bits) VOP3P + LITERAL (96 bits) VOP3P + DPP (96 bits) CLAMP If set, clamp or signal on output. VCMP instructions enable signaling-compare. For floating point outputs, clamp to [0.0, 1.0]. For signed integer outputs, clamp to [MIN_INT, MAX_INT]. For unsigned integer outputs, clamp to [0, MAX_UINT]. Applied after output modifier. See OPF_NOCLAMP for a list of instructions that do not support clamping or signaling. 1 15 ENCODING Encoding. 8 24 NEG If NEG[N] set, take the floating-point negation of the N'th input LOW OPERATION. See OPF_NONEG for a list of instructions that do not support negation. 3 61 NEG_HI If NEG_HI[N] set, take the floating-point negation of the N'th input HIGH OPERATION. See OPF_NONEG for a list of instructions that do not support negation. 3 8 OP Opcode. 7 16 OPSEL OPSEL[N] determines if lower [15:0] or upper [31:16] bits are selected for the N'th source for the LOW OPERATION. 0 selects the lower order bits and 1 selects the higher order bits. 3 11 OPSEL_HI OPSEL_HI[N] determines if lower [15:0] or upper [31:16] bits are selected for the N'th source for the HIGH OPERATION. 0 selects the lower order bits and 1 selects the higher order bits. 2 59 OPSEL_HI_2 Bit 2 of OPSEL_HI. 1 14 SRC0 First source operand for instruction. 9 32 SRC1 Second source operand for instruction. 9 41 SRC2 Third source operand for instruction. 9 50 VDST Destination operand for instruction. 8 0 ENC_VINTERP 64 0000000000000000000000000000000011111111000111110000000000000000 0000000000000000000000000000000011001101000000000000000000000000 0000000000000000000000000000000011001101000000010000000000000000 0000000000000000000000000000000011001101000000100000000000000000 0000000000000000000000000000000011001101000000110000000000000000 0000000000000000000000000000000011001101000001000000000000000000 0000000000000000000000000000000011001101000001010000000000000000 default 1 U 1 Vector ALU interpolation. DPP or literal constants cannot be used with interpolation instructions. Parameter forwarding restrictions The VINTERP operands that are used to pass in parameter data do not support forwarding from other parts of the VALU pipeline. This has subtle effects on the shader program if parameter data does not originate from a VDSDIR instruction. Parameter data is passed in S0 and S2 of OPF_INTERP_P10-style opcodes and in S0 of OPF_INTERP_P2-style opcodes. If this data in any thread is the result of a VALU write then the only safe way to avoid a hazard is to insert a V_PIPEFLUSH instruction between the VALU write and the VINTERP instruction. Waiting for idle or for the VA_VDST counter is not sufficient to avoid this hazard. Parameter data loaded by DS (including VDSDIR) or VMEM instructions is not affected and does not require a V_PIPEFLUSH instruction. This is an intentional design restriction due to area and power requirements. CLAMP If set, clamp on output. Floating point outputs clamp to [0.0, 1.0]. 1 15 ENCODING Encoding. 8 24 NEG If NEG[N] set, take the floating-point negation of the N'th input operand. 3 61 OP Opcode. 5 16 OPSEL For 16-bit inputs, OPSEL[N] determines if lower [15:0] or upper [31:16] bits are selected for the N'th source operand. Similarly OPSEL[3] controls whether the lower [15:0] or upper [31:16] bits are written for a 16-bit destination (preserving the unwritten bits). 0 selects the lower order bits and 1 selects the higher order bits. 4 11 SRC0 First source operand for instruction. 9 32 SRC1 Second source operand for instruction. 9 41 SRC2 Third source operand for instruction. 9 50 VDST Destination operand for instruction. 8 0 WAIT_EXP Wait on the EXP counter before issuing this instruction. EXP must be less than or equal to this value before the instruction is issued. If set to 7, no wait is performed. 3 8 ENC_VDSDIR 32 11111111001100000000000000000000 11001110000000000000000000000000 11001110000100000000000000000000 default 1 U 1 DS direct operation. DPP or literal constants cannot be used with DS direct instructions. ATTR Attribute to interpolate. 6 10 ATTR_CHAN Attribute component to interpolate. 2 8 ENCODING Encoding. 8 24 OP Opcode. 2 20 VDST Destination operand for instruction. 8 0 WAIT_VA_VDST Specify a wait on the VA_VDST counter before issuing this instruction. VA_VDST must be less than or equal to this value before the instruction is issued. If set to 15, no wait is performed. 4 16 WAIT_VM_VSRC Specify a wait on the VM_VSRC counter before issuing this instruction. VM_VSRC must be less than or equal to this value before the instruction is issued. If set to 1, no wait is performed. 1 23 ENC_VDS 64 0000000000000000000000000000000011111111111111000000000000000000 0000000000000000000000000000000011011000000000000000000000000000 0000000000000000000000000000000011011000000001000000000000000000 0000000000000000000000000000000011011000000010000000000000000000 0000000000000000000000000000000011011000000011000000000000000000 0000000000000000000000000000000011011000000100000000000000000000 0000000000000000000000000000000011011000000101000000000000000000 0000000000000000000000000000000011011000000110000000000000000000 0000000000000000000000000000000011011000000111000000000000000000 0000000000000000000000000000000011011000001000000000000000000000 0000000000000000000000000000000011011000001001000000000000000000 0000000000000000000000000000000011011000001010000000000000000000 0000000000000000000000000000000011011000001011000000000000000000 0000000000000000000000000000000011011000001100000000000000000000 0000000000000000000000000000000011011000001101000000000000000000 0000000000000000000000000000000011011000001110000000000000000000 0000000000000000000000000000000011011000001111000000000000000000 0000000000000000000000000000000011011000010000000000000000000000 0000000000000000000000000000000011011000010010000000000000000000 0000000000000000000000000000000011011000010011000000000000000000 0000000000000000000000000000000011011000010100000000000000000000 0000000000000000000000000000000011011000010101000000000000000000 0000000000000000000000000000000011011000011110000000000000000000 0000000000000000000000000000000011011000011111000000000000000000 0000000000000000000000000000000011011000100000000000000000000000 0000000000000000000000000000000011011000100001000000000000000000 0000000000000000000000000000000011011000100010000000000000000000 0000000000000000000000000000000011011000100011000000000000000000 0000000000000000000000000000000011011000100100000000000000000000 0000000000000000000000000000000011011000100101000000000000000000 0000000000000000000000000000000011011000100110000000000000000000 0000000000000000000000000000000011011000100111000000000000000000 0000000000000000000000000000000011011000101000000000000000000000 0000000000000000000000000000000011011000101001000000000000000000 0000000000000000000000000000000011011000101010000000000000000000 0000000000000000000000000000000011011000101011000000000000000000 0000000000000000000000000000000011011000101100000000000000000000 0000000000000000000000000000000011011000101101000000000000000000 0000000000000000000000000000000011011000101110000000000000000000 0000000000000000000000000000000011011000101111000000000000000000 0000000000000000000000000000000011011000110000000000000000000000 0000000000000000000000000000000011011000110010000000000000000000 0000000000000000000000000000000011011000110011000000000000000000 0000000000000000000000000000000011011000110101000000000000000000 0000000000000000000000000000000011011000110110000000000000000000 0000000000000000000000000000000011011000110111000000000000000000 0000000000000000000000000000000011011000111000000000000000000000 0000000000000000000000000000000011011000111001000000000000000000 0000000000000000000000000000000011011000111010000000000000000000 0000000000000000000000000000000011011000111011000000000000000000 0000000000000000000000000000000011011000111100000000000000000000 0000000000000000000000000000000011011000111101000000000000000000 0000000000000000000000000000000011011000111110000000000000000000 0000000000000000000000000000000011011001000000000000000000000000 0000000000000000000000000000000011011001000001000000000000000000 0000000000000000000000000000000011011001000010000000000000000000 0000000000000000000000000000000011011001000011000000000000000000 0000000000000000000000000000000011011001000100000000000000000000 0000000000000000000000000000000011011001000101000000000000000000 0000000000000000000000000000000011011001000110000000000000000000 0000000000000000000000000000000011011001000111000000000000000000 0000000000000000000000000000000011011001001000000000000000000000 0000000000000000000000000000000011011001001001000000000000000000 0000000000000000000000000000000011011001001010000000000000000000 0000000000000000000000000000000011011001001011000000000000000000 0000000000000000000000000000000011011001001100000000000000000000 0000000000000000000000000000000011011001001101000000000000000000 0000000000000000000000000000000011011001001110000000000000000000 0000000000000000000000000000000011011001001111000000000000000000 0000000000000000000000000000000011011001010000000000000000000000 0000000000000000000000000000000011011001010010000000000000000000 0000000000000000000000000000000011011001010011000000000000000000 0000000000000000000000000000000011011001100000000000000000000000 0000000000000000000000000000000011011001100001000000000000000000 0000000000000000000000000000000011011001100010000000000000000000 0000000000000000000000000000000011011001100011000000000000000000 0000000000000000000000000000000011011001100100000000000000000000 0000000000000000000000000000000011011001100101000000000000000000 0000000000000000000000000000000011011001100110000000000000000000 0000000000000000000000000000000011011001100111000000000000000000 0000000000000000000000000000000011011001101000000000000000000000 0000000000000000000000000000000011011001101001000000000000000000 0000000000000000000000000000000011011001101010000000000000000000 0000000000000000000000000000000011011001101011000000000000000000 0000000000000000000000000000000011011001101100000000000000000000 0000000000000000000000000000000011011001101101000000000000000000 0000000000000000000000000000000011011001101110000000000000000000 0000000000000000000000000000000011011001101111000000000000000000 0000000000000000000000000000000011011001110000000000000000000000 0000000000000000000000000000000011011001110010000000000000000000 0000000000000000000000000000000011011001110011000000000000000000 0000000000000000000000000000000011011001110110000000000000000000 0000000000000000000000000000000011011001110111000000000000000000 0000000000000000000000000000000011011001111000000000000000000000 0000000000000000000000000000000011011001111001000000000000000000 0000000000000000000000000000000011011001111110000000000000000000 0000000000000000000000000000000011011010011000000000000000000000 0000000000000000000000000000000011011010011001000000000000000000 0000000000000000000000000000000011011010011010000000000000000000 0000000000000000000000000000000011011010011011000000000000000000 0000000000000000000000000000000011011010100000000000000000000000 0000000000000000000000000000000011011010100001000000000000000000 0000000000000000000000000000000011011010100010000000000000000000 0000000000000000000000000000000011011010100011000000000000000000 0000000000000000000000000000000011011010100100000000000000000000 0000000000000000000000000000000011011010100101000000000000000000 0000000000000000000000000000000011011010100110000000000000000000 0000000000000000000000000000000011011010100111000000000000000000 0000000000000000000000000000000011011010101000000000000000000000 0000000000000000000000000000000011011010101001000000000000000000 0000000000000000000000000000000011011010101010000000000000000000 0000000000000000000000000000000011011010101011000000000000000000 0000000000000000000000000000000011011010110000000000000000000000 0000000000000000000000000000000011011010110001000000000000000000 0000000000000000000000000000000011011010110010000000000000000000 0000000000000000000000000000000011011010110011000000000000000000 0000000000000000000000000000000011011011001101000000000000000000 0000000000000000000000000000000011011011011110000000000000000000 0000000000000000000000000000000011011011011111000000000000000000 0000000000000000000000000000000011011011100000000000000000000000 0000000000000000000000000000000011011011100001000000000000000000 0000000000000000000000000000000011011011100010000000000000000000 0000000000000000000000000000000011011011111110000000000000000000 0000000000000000000000000000000011011011111111000000000000000000 default 1 U 1 Local data share operations. Examples: // Write 1 dword from v22 into LDS at byte address v12 ds_write_b32 v12, v22 // Read 1 dword from LDS byte address (v3 + 120), write value to v2 ds_read_b32 v2, v3 offset:120 ADDR Source VGPR that provides the DS address. 8 32 DATA0 Source data 0. 8 40 DATA1 Source data 1. 8 48 ENCODING Encoding. 6 26 OFFSET0 First DS offset. See individual instructions for how this field is used. 8 0 OFFSET1 Second DS offset. See individual instructions for how this field is used. 8 8 OP Opcode. 8 18 VDST Destination VGPR. 8 56 ENC_VBUFFER 96 000000000000000000000000000000000000000000000000000000000000000011111100001111111100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000000000100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000000001000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000000001100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000000010100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000000011000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000000011100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000000100100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000000101000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000000101100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000000110100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000000111000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000000111100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000001000100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000001001000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000001001100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000001010100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000001011000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000001011100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000001100100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000001101000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000001101100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000001110100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000001111000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000001111100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000010000100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000010001000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000010001100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000010010100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000010011000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000010011100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000011001100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000011010100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000011011000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000011011100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000011100100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000011101000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000011101100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000011110100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000011111000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000011111100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000100000100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000100001000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000100001100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000100010100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000100011000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000100011100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000100100100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000100101000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000100101100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000100110100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000101000100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000101001000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000101011000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000101100100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100000101101000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100001000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100001000000100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100001000001000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100001000001100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100001000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100001000010100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100001000011000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100001000011100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100001000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100001000100100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100001000101000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100001000101100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100001000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100001000110100000000000000 000000000000000000000000000000000000000000000000000000000000000011000100001000111000000000000000 000000000000000000000000000000000000000000000000000000000000000011000100001000111100000000000000 default 1 U 1 Vector memory buffer operations. ENCODING Encoding. 6 26 FORMAT Format for typed buffer. 7 55 IDXEN If set, send VADDR as an index. If unset, treat the index as zero. 1 63 IOFFSET Signed byte offset. 24 72 NV Non-volatile bit. 1 7 OFFEN If set, send VADDR (or VADDR+1 if IDXEN=1) as an offset. 1 62 OP Opcode. 8 14 RSRC Specify SGPR that contains resource descriptor. Must be aligned to a multiple of 4. 9 41 SCOPE Scope of memory operation. 2 50 SOFFSET Scalar register supplying unsigned byte offset. 7 0 TFE Texture Fail Enable (for partially resident textures). 1 22 TH Temporal hints and atomic return request. 3 52 VADDR Address source - may carry an offset or an index. 8 64 VDATA Vector GPR to write result to. 8 32 ENC_VIMAGE 96 000000000000000000000000000000000000000000000000000000000000000011111100001111111100000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000000000100000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000000001000000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000000001100000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000000010100000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000000011000000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000000011100000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000000100100000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000000101000000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000000101100000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000000110100000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000000111000000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000000111100000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000001000100000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000001001000000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000001001100000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000001010100000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000001011000000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000001011100000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000001100100000000000000 000000000000000000000000000000000000000000000000000000000000000011010000000001101000000000000000 000000000000000000000000000000000000000000000000000000000000000011010000001000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010000001000000100000000000000 000000000000000000000000000000000000000000000000000000000000000011010000001000001100000000000000 000000000000000000000000000000000000000000000000000000000000000011010000001000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010000001000010100000000000000 000000000000000000000000000000000000000000000000000000000000000011010000001000011000000000000000 000000000000000000000000000000000000000000000000000000000000000011010000001000011100000000000000 default 1 U 1 Memory image (texture) operations without a sampler. A16 Texture coordinates size: 1 = 16 bit, 0 = 32 bit. 1 6 D16 If set, convert 32 bit floating point texture return data to 16 bit. 1 5 DIM Dimensionality of resource constant, used to determine the number of addresses supplied for this instruction. This should be set the same as bits [3:1] of the resource TYPE field. 3 0 DMASK Enable mask for image read/write data components. Normally bit 0 is red, 1 is green, 2 is blue and 3 is alpha. At least 1 bit must be on. Data is assumed to be packed into consecutive VGPRs. For reads, defines which components are returned. For writes, defines which components are written with data from VGPRs (missing components get 0). Enabled components come from consecutive VGPRs. For example when DMASK = 'b1001 the red component is in VDATA0 and the alpha component is in VDATA1. For D16 reads, DMASK indicates which components to return. For D16 writes, the DMASK indicates which components to write but has restrictions: Data is read out of consecutive VGPRs: LSBs of VDATA, then MSBs of VDATA, then LSBs of VDATA+1 and last if needed MSBs of VDATA+1. This is regardless of which DMASK bits are set, only how many bits are set. The position of the DMASK bits controls which components are written in memory. If DMASK==0, the TA overrides with DMASK = 1 and puts zeros in VGPR followed by LWE status if exists. TFE status is not generated since the fetch is dropped. For IMAGE_GATHER4* instructions, DMASK indicates which component (RGBA), and the number of VGPRs to use is determined automatically by hardware (4 VGPRs when D16=0, and 2 VGPRs when D16=1). 4 22 ENCODING Encoding. 6 26 NV Non-volatile bit. 1 7 OP Opcode. 8 14 R128 Texture resource size: 1 = 128 bit (4 SGPRs), 0 = 256 bit (8 SGPRs). 1 4 RSRC Specify SGPR that contains resource descriptor. Must be aligned to a multiple of 4. 9 41 SCOPE Scope of memory operation. 2 50 TFE Texture Fail Enable (for partially resident textures). 1 55 TH Temporal hints and atomic return request. 3 52 VADDR0 First VGPR component to provide address information. 8 64 VADDR1 Second address source. 8 72 VADDR2 Third address source. 8 80 VADDR3 Fourth address source. 8 88 VADDR4 Fifth address source. 8 56 VDATA Vector GPR to write result/read input from. 8 32 ENC_VSAMPLE 96 000000000000000000000000000000000000000000000000000000000000000011111100001111111100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000001101100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000001110100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000001111000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000001111100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000010000100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000010001000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000010001100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000010010100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000010011000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000010011100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000010100100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000010101000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000010101100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000010110100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000010111000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000010111100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000011000100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000011001000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000011001100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000011010100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000011011000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000011011100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000011100100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000011101000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000011101100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000100000100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000100001000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000100001100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000100010100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000100011000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000100011100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000100100100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000100101000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000100101100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000101010100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000101011000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000101111100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000110000100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000110001000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000110001100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011100100000110010100000000000000 000000000000000000000000000000000000000000000000000000000000000011100100001001000000000000000000 default 1 U 1 Memory image (texture) operations with a sampler. Example from textri: s_load_b256 s[20:27], s[2:3], 0x0 // load image resource buffer s_load_b128 s[28:31], s[4:5], 0x0 // load image sampler buffer s_wait_kmcnt 0 // v4..v7 -> incoming texture coordinates // v16..v19 -> the resulting sample from the texture image_sample v16, v4, s20, s28 dmask:0xf ... s_wait_vmcnt 0 // now safe to access v16..v19 NSA example: image_sample v[0:3], [v9,v4], s[20:27], s[28:31] // address component is NSA A16 Texture coordinates size: 1 = 16 bit, 0 = 32 bit. 1 6 D16 If set, convert 32 bit floating point texture return data to 16 bit. 1 5 DIM Dimensionality of resource constant, used to determine the number of addresses supplied for this instruction. This should be set the same as bits [3:1] of the resource TYPE field. 3 0 DMASK Enable mask for image read/write data components. Normally bit 0 is red, 1 is green, 2 is blue and 3 is alpha. At least 1 bit must be on. Data is assumed to be packed into consecutive VGPRs. For reads, defines which components are returned. For writes, defines which components are written with data from VGPRs (missing components get 0). Enabled components come from consecutive VGPRs. For example when DMASK = 'b1001 the red component is in VDATA0 and the alpha component is in VDATA1. For D16 reads, DMASK indicates which components to return. For D16 writes, the DMASK indicates which components to write but has restrictions: Data is read out of consecutive VGPRs: LSBs of VDATA, then MSBs of VDATA, then LSBs of VDATA+1 and last if needed MSBs of VDATA+1. This is regardless of which DMASK bits are set, only how many bits are set. The position of the DMASK bits controls which components are written in memory. If DMASK==0, the TA overrides with DMASK = 1 and puts zeros in VGPR followed by LWE status if exists. TFE status is not generated since the fetch is dropped. For IMAGE_GATHER4* instructions, DMASK indicates which component (RGBA), and the number of VGPRs to use is determined automatically by hardware (4 VGPRs when D16=0, and 2 VGPRs when D16=1). 4 22 ENCODING Encoding. 6 26 LWE LOD Warning Enable (for partially resident textures). 1 40 NV Non-volatile bit. 1 7 OP Opcode. 8 14 R128 Texture resource size: 1 = 128 bit (4 SGPRs), 0 = 256 bit (8 SGPRs). 1 4 RSRC Specify SGPR that contains resource descriptor. Must be aligned to a multiple of 4. 9 41 SAMP Scalar GPR that specifies the sampler constant. Must be aligned to a multiple of 4. 9 55 SCOPE Scope of memory operation. 2 50 TFE Texture Fail Enable (for partially resident textures). 1 3 TH Temporal hints and atomic return request. 3 52 UNORM Force address to be un-normalized regardless of texture sampler setting. Must be set to 1 for image store and atomic operations. 0: for image ops with samplers, (S,T,R) coordinates in [0.0, 1.0] span the entire texture map. 1: for image ops with samplers, (S,T,R) coordinates in [0.0, N] span the texture map where N is (width,height,depth). Array index, cube slice, LOD, bias etc. are not affected. Image ops without a sampler are not affected. UINT inputs are unnormalized. This bit is logically OR'd with the SAMP.FORCE_UNNORMALIZED when sent to the texture unit. 1 13 VADDR0 First VGPR component to provide address information. 8 64 VADDR1 Second address source. 8 72 VADDR2 Third address source. 8 80 VADDR3 Fourth address source. 8 88 VDATA Vector GPR to write result/read input from. 8 32 ENC_VEXPORT 64 0000000000000000000000000000000011111100000000000000000000000000 0000000000000000000000000000000011111000000000000000000000000000 default 1 U 1 Graphics export operations. Only one export may be performed per target. The EXEC register carries the valid mask for export targets. Example: // Export RGBA color and valid mask to render target 0, using values in v[0:3] exp mrt0, v0, v1, v2, v3 ... // Need to ensure the export has completed before modifying v0: s_wait_expcnt 0 v_mov_b32 v0, ... ... // Export a single colour to render target 1, and export the valid mask // Mark this as the final render target export with the 'done' flag: exp mrt1, v0, off, off, off done DONE If set, this is the last export of a given type. If this is set for a colour export (PS only), then the valid mask must be present in the EXEC register. 1 11 EN Indicate which VSRC selects should be exported to SX. 16-bit components: export half-dword enable. Valid values are 0, 1, 3. Bit 0 enables VSRC0: R,G from one VGPR (R in low bits, G high), bit 1 enables VSRC1: B,A from one VGPR (B in low bits, A high). 32-bit components: bit N enables VSRC[N]. EN must not be zero. 4 0 ENCODING Encoding. 6 26 ROW_EN If set, use M0 to provide the row number for mesh shader's POS and PRIM exports. 1 13 TGT Export target. 6 4 VSRC0 VGPR of the first data to export. 8 32 VSRC1 VGPR of the second data to export. 8 40 VSRC2 VGPR of the third data to export. 8 48 VSRC3 VGPR of the fourth data to export. 8 56 ENC_VFLAT 96 000000000000000000000000000000000000000000000000000000000000000011111111001111111100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000001000100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000001001000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000001001100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000001010100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000001011000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000001011100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000001100100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000001101000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000001101100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000001110100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000001111000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000001111100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000010000100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000010001000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000010001100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000010010100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000011001100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000011010100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000011011000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000011011100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000011100100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000011101000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000011101100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000011110100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000011111000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000011111100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000100000100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000100001000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000100001100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000100010100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000100011000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000100011100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000100100100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000100101000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000100101100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000100110100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000101000100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000101001000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000101011000000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000101100100000000000000 000000000000000000000000000000000000000000000000000000000000000011101100000101101000000000000000 default 1 U 1 Flat (DUA) addressing instructions. For basic flat instructions note that a VGPR address must be specified. SGPR addresses are not allowed. An aperture test is performed on the VGPR address operand to determine which of the modes to use on a per-thread basis: Global, LDS or scratch. * GLOBAL Use global GV addressing mode. * SCRATCH Addr = scratch_base + swizzle(VGPR ADDR + inst_offset - privateApertureBase, tid) * LDS LDS_ADDR = VGPR ADDR + inst_offset - sharedApertureBase FLAT instructions cannot return memory-read-data to LDS. ENCODING Encoding. 8 24 IOFFSET 24-bit signed offset. 24 72 NV Non-volatile bit. 1 7 OP Opcode. 8 14 SADDR Addr/offset SGPR. 7 0 SCOPE Scope of memory operation. 2 50 SVE Used to indicate which addressing mode is in use for SCRATCH opcodes. In conjunction with SADDR determines which of 4 addressing modes are in use for SCRATCH. 1 49 TH Temporal hints and atomic return request. 3 52 VADDR Source flat address VGPR. 8 64 VDST Destination VGPR. 8 32 VSRC Source data. 8 55 ENC_VSCRATCH 96 000000000000000000000000000000000000000000000000000000000000000011111111001111111100000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000001000100000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000001001000000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000001001100000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000001010100000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000001011000000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000001011100000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000001100100000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000001101000000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000001101100000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000001110100000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000001111000000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000001111100000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000010000100000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000010001000000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000010001100000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000010010100000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000101001100000000000000 000000000000000000000000000000000000000000000000000000000000000011101101000101010000000000000000 default 1 U 1 Scratch addressing instructions. Scratch instructions support 4 addressing modes: * SV Scratch Vector addressing. Specify "off" or "null" in sp3 for the SADDR operand. This generates "null" in the SADDR field and sets SVE=1. Addr = scratch_base + swizzle(ADDR + inst_offset, tid) * SS Scratch Scalar addressing. Specify "off" or "null" in sp3 for the ADDR operand. This sets SVE=0. Addr = scratch_base + swizzle(SADDR + inst_offset, tid) * ST Scratch with Thread ID only. Specify "off" or "null" in sp3 for the ADDR and SADDR operands. This generates "null" in the SADDR field and sets SVE=0. Addr = scratch_base + swizzle(0 + inst_offfset, tid) * SVS Scratch with Vector and Scalar addressing. Specify valid registers in sp3 for the ADDR and SADDR operands. This sets SVE=1. Addr = scratch_base + swizzle(ADDR + SADDR + inst_offset, tid) ENCODING Encoding. 8 24 IOFFSET 24-bit signed offset. 24 72 NV Non-volatile bit. 1 7 OP Opcode. 8 14 SADDR Addr/offset SGPR. 7 0 SCOPE Scope of memory operation. 2 50 SVE Used to indicate which addressing mode is in use for SCRATCH opcodes. In conjunction with SADDR determines which of 4 addressing modes are in use for SCRATCH. 1 49 TH Temporal hints and atomic return request. 3 52 VADDR Source flat address VGPR. 8 64 VDST Destination VGPR. 8 32 VSRC Source data. 8 55 ENC_VGLOBAL 96 000000000000000000000000000000000000000000000000000000000000000011111111001111111100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000001000100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000001001000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000001001100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000001010100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000001011000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000001011100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000001100100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000001101000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000001101100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000001110100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000001111000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000001111100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000010000100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000010001000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000010001100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000010010100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000010100100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000010101100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000011001100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000011010100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000011011000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000011011100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000011100100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000011101000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000011101100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000011110100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000011111000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000011111100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000100000100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000100001000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000100001100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000100010100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000100011000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000100011100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000100100100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000100101000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000100101100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000100110100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000100111100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000101000100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000101001000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000101001100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000101011000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000101011100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000101100100000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000101101000000000000000 000000000000000000000000000000000000000000000000000000000000000011101110000111001100000000000000 default 1 U 1 Global addressing instructions. For global instructions note that a VGPR address must be specified. In sp3 an SGPR address may be omitted by specifying "off" or "null" in the SADDR operand slot ("off" is a legacy syntax and is translated to "null"). The following addressing modes are supported: * GV Global Vector addressing. Specify "off" or "null" in sp3 for the SADDR operand. Addr = VGPR ADDR (64-bit) + inst_offset * GS Scratch with Vector and Scalar addressing. Specify a valid address in the SADDR operand to enable this mode. Addr = SADDR (64-bit) + VGPR ADDR (32-bit offset) + inst_offset * GT Global with Thread ID. This calculation is only used by instructions with the OPF_GLOBAL_ADDTID flag. Instructions with this flag do not use the other global modes. Addr = SADDR (64-bit) + inst_offset + tid * 4 ENCODING Encoding. 8 24 IOFFSET 24-bit signed offset. 24 72 NV Non-volatile bit. 1 7 OP Opcode. 8 14 SADDR Addr/offset SGPR. 7 0 SCOPE Scope of memory operation. 2 50 SVE Used to indicate which addressing mode is in use for SCRATCH opcodes. In conjunction with SADDR determines which of 4 addressing modes are in use for SCRATCH. 1 49 TH Temporal hints and atomic return request. 3 52 VADDR Source flat address VGPR. 8 64 VDST Destination VGPR. 8 32 VSRC Source data. 8 55 SOP1_INST_LITERAL 64 0000000000000000000000000000000011111111100000001111111100000000 0000000000000000000000000000000010111110100000000000000000000000 0000000000000000000000000000000010111110100000000000000100000000 0000000000000000000000000000000010111110100000000000001000000000 0000000000000000000000000000000010111110100000000000001100000000 0000000000000000000000000000000010111110100000000000010000000000 0000000000000000000000000000000010111110100000000000010100000000 0000000000000000000000000000000010111110100000000000100000000000 0000000000000000000000000000000010111110100000000000100100000000 0000000000000000000000000000000010111110100000000000101000000000 0000000000000000000000000000000010111110100000000000101100000000 0000000000000000000000000000000010111110100000000000110000000000 0000000000000000000000000000000010111110100000000000110100000000 0000000000000000000000000000000010111110100000000000111000000000 0000000000000000000000000000000010111110100000000000111100000000 0000000000000000000000000000000010111110100000000001000000000000 0000000000000000000000000000000010111110100000000001000100000000 0000000000000000000000000000000010111110100000000001001000000000 0000000000000000000000000000000010111110100000000001001100000000 0000000000000000000000000000000010111110100000000001010000000000 0000000000000000000000000000000010111110100000000001010100000000 0000000000000000000000000000000010111110100000000001011000000000 0000000000000000000000000000000010111110100000000001011100000000 0000000000000000000000000000000010111110100000000001100000000000 0000000000000000000000000000000010111110100000000001100100000000 0000000000000000000000000000000010111110100000000001101000000000 0000000000000000000000000000000010111110100000000001101100000000 0000000000000000000000000000000010111110100000000001110000000000 0000000000000000000000000000000010111110100000000001110100000000 0000000000000000000000000000000010111110100000000001111000000000 0000000000000000000000000000000010111110100000000001111100000000 0000000000000000000000000000000010111110100000000010000000000000 0000000000000000000000000000000010111110100000000010000100000000 0000000000000000000000000000000010111110100000000010001000000000 0000000000000000000000000000000010111110100000000010001100000000 0000000000000000000000000000000010111110100000000010010000000000 0000000000000000000000000000000010111110100000000010010100000000 0000000000000000000000000000000010111110100000000010011000000000 0000000000000000000000000000000010111110100000000010011100000000 0000000000000000000000000000000010111110100000000010100000000000 0000000000000000000000000000000010111110100000000010100100000000 0000000000000000000000000000000010111110100000000010101000000000 0000000000000000000000000000000010111110100000000010101100000000 0000000000000000000000000000000010111110100000000010110000000000 0000000000000000000000000000000010111110100000000010110100000000 0000000000000000000000000000000010111110100000000010111000000000 0000000000000000000000000000000010111110100000000010111100000000 0000000000000000000000000000000010111110100000000011000000000000 0000000000000000000000000000000010111110100000000011000100000000 0000000000000000000000000000000010111110100000000011001000000000 0000000000000000000000000000000010111110100000000011001100000000 0000000000000000000000000000000010111110100000000011010000000000 0000000000000000000000000000000010111110100000000011010100000000 0000000000000000000000000000000010111110100000000011011000000000 0000000000000000000000000000000010111110100000000011011100000000 0000000000000000000000000000000010111110100000000100001000000000 0000000000000000000000000000000010111110100000000100001100000000 0000000000000000000000000000000010111110100000000100101000000000 0000000000000000000000000000000010111110100000000100111000000000 0000000000000000000000000000000010111110100000000100111100000000 0000000000000000000000000000000010111110100000000101001100000000 0000000000000000000000000000000010111110100000000101100000000000 0000000000000000000000000000000010111110100000000110000000000000 0000000000000000000000000000000010111110100000000110000100000000 0000000000000000000000000000000010111110100000000110001000000000 0000000000000000000000000000000010111110100000000110001100000000 0000000000000000000000000000000010111110100000000110010000000000 0000000000000000000000000000000010111110100000000110010100000000 0000000000000000000000000000000010111110100000000110011000000000 0000000000000000000000000000000010111110100000000110011100000000 0000000000000000000000000000000010111110100000000110100000000000 0000000000000000000000000000000010111110100000000110100100000000 0000000000000000000000000000000010111110100000000110101000000000 0000000000000000000000000000000010111110100000000110101100000000 0000000000000000000000000000000010111110100000000110110000000000 0000000000000000000000000000000010111110100000000110110100000000 0000000000000000000000000000000010111110100000000110111000000000 has_lit_0 == .fieldderef .fieldderef INST SOP1 SSRC0 255 I 32 Scalar ALU operations with one destination and one source. Allowed patterns: SOP1 (32 bits) SOP1 + LITERAL (64 bits) ENCODING Encoding. 9 23 OP Opcode. 8 8 SDST Destination operand for instruction. 7 16 SIMM32 32-bit literal constant. 32 32 SSRC0 Source operand for instruction. 8 0 SOP2_INST_LITERAL 64 0000000000000000000000000000000011111111100000000000000000000000 0000000000000000000000000000000010000000000000000000000000000000 0000000000000000000000000000000010000000100000000000000000000000 0000000000000000000000000000000010000001000000000000000000000000 0000000000000000000000000000000010000001100000000000000000000000 0000000000000000000000000000000010000010000000000000000000000000 0000000000000000000000000000000010000010100000000000000000000000 0000000000000000000000000000000010000011000000000000000000000000 0000000000000000000000000000000010000100000000000000000000000000 0000000000000000000000000000000010000100100000000000000000000000 0000000000000000000000000000000010000101000000000000000000000000 0000000000000000000000000000000010000101100000000000000000000000 0000000000000000000000000000000010000110000000000000000000000000 0000000000000000000000000000000010000110100000000000000000000000 0000000000000000000000000000000010000111000000000000000000000000 0000000000000000000000000000000010000111100000000000000000000000 0000000000000000000000000000000010001000000000000000000000000000 0000000000000000000000000000000010001000100000000000000000000000 0000000000000000000000000000000010001001000000000000000000000000 0000000000000000000000000000000010001001100000000000000000000000 0000000000000000000000000000000010001010000000000000000000000000 0000000000000000000000000000000010001010100000000000000000000000 0000000000000000000000000000000010001011000000000000000000000000 0000000000000000000000000000000010001011100000000000000000000000 0000000000000000000000000000000010001100000000000000000000000000 0000000000000000000000000000000010001100100000000000000000000000 0000000000000000000000000000000010001101000000000000000000000000 0000000000000000000000000000000010001101100000000000000000000000 0000000000000000000000000000000010001110000000000000000000000000 0000000000000000000000000000000010001110100000000000000000000000 0000000000000000000000000000000010001111000000000000000000000000 0000000000000000000000000000000010001111100000000000000000000000 0000000000000000000000000000000010010000000000000000000000000000 0000000000000000000000000000000010010000100000000000000000000000 0000000000000000000000000000000010010001000000000000000000000000 0000000000000000000000000000000010010001100000000000000000000000 0000000000000000000000000000000010010010000000000000000000000000 0000000000000000000000000000000010010010100000000000000000000000 0000000000000000000000000000000010010011000000000000000000000000 0000000000000000000000000000000010010011100000000000000000000000 0000000000000000000000000000000010010100000000000000000000000000 0000000000000000000000000000000010010100100000000000000000000000 0000000000000000000000000000000010010101000000000000000000000000 0000000000000000000000000000000010010101100000000000000000000000 0000000000000000000000000000000010010110000000000000000000000000 0000000000000000000000000000000010010110100000000000000000000000 0000000000000000000000000000000010010111000000000000000000000000 0000000000000000000000000000000010011000000000000000000000000000 0000000000000000000000000000000010011000100000000000000000000000 0000000000000000000000000000000010011001000000000000000000000000 0000000000000000000000000000000010011001100000000000000000000000 0000000000000000000000000000000010011010000000000000000000000000 0000000000000000000000000000000010011010100000000000000000000000 0000000000000000000000000000000010100000000000000000000000000000 0000000000000000000000000000000010100000100000000000000000000000 0000000000000000000000000000000010100001000000000000000000000000 0000000000000000000000000000000010100001100000000000000000000000 0000000000000000000000000000000010100010000000000000000000000000 0000000000000000000000000000000010100010100000000000000000000000 0000000000000000000000000000000010100011000000000000000000000000 0000000000000000000000000000000010100011100000000000000000000000 0000000000000000000000000000000010100100000000000000000000000000 0000000000000000000000000000000010100100100000000000000000000000 0000000000000000000000000000000010100101000000000000000000000000 0000000000000000000000000000000010100101100000000000000000000000 0000000000000000000000000000000010100110000000000000000000000000 0000000000000000000000000000000010100110100000000000000000000000 0000000000000000000000000000000010100111000000000000000000000000 0000000000000000000000000000000010100111100000000000000000000000 0000000000000000000000000000000010101000000000000000000000000000 0000000000000000000000000000000010101000100000000000000000000000 0000000000000000000000000000000010101001000000000000000000000000 0000000000000000000000000000000010101001100000000000000000000000 0000000000000000000000000000000010101010000000000000000000000000 0000000000000000000000000000000010101010100000000000000000000000 has_lit_0 && == .fieldderef .fieldderef INST SOP2 SSRC0 255 I 32 != .fieldderef .fieldderef INST SOP2 SSRC1 255 I 32 has_lit_1 && != .fieldderef .fieldderef INST SOP2 SSRC0 255 I 32 == .fieldderef .fieldderef INST SOP2 SSRC1 255 I 32 has_lit_0_has_lit_1 && == .fieldderef .fieldderef INST SOP2 SSRC0 255 I 32 == .fieldderef .fieldderef INST SOP2 SSRC1 255 I 32 default && != .fieldderef .fieldderef INST SOP2 SSRC0 255 I 32 != .fieldderef .fieldderef INST SOP2 SSRC1 255 I 32 Scalar ALU operations with one destination and two sources. Allowed patterns: SOP2 (32 bits) SOP2 + LITERAL (64 bits) ENCODING Encoding. 2 30 OP Opcode. 7 23 SDST Destination operand for instruction. 7 16 SIMM32 32-bit literal constant. 32 32 SSRC0 First source operand for instruction. 8 0 SSRC1 Second source operand for instruction. 8 8 SOPC_INST_LITERAL 64 0000000000000000000000000000000011111111111111110000000000000000 0000000000000000000000000000000010111111000000000000000000000000 0000000000000000000000000000000010111111000000010000000000000000 0000000000000000000000000000000010111111000000100000000000000000 0000000000000000000000000000000010111111000000110000000000000000 0000000000000000000000000000000010111111000001000000000000000000 0000000000000000000000000000000010111111000001010000000000000000 0000000000000000000000000000000010111111000001100000000000000000 0000000000000000000000000000000010111111000001110000000000000000 0000000000000000000000000000000010111111000010000000000000000000 0000000000000000000000000000000010111111000010010000000000000000 0000000000000000000000000000000010111111000010100000000000000000 0000000000000000000000000000000010111111000010110000000000000000 0000000000000000000000000000000010111111000011000000000000000000 0000000000000000000000000000000010111111000011010000000000000000 0000000000000000000000000000000010111111000011100000000000000000 0000000000000000000000000000000010111111000011110000000000000000 0000000000000000000000000000000010111111000100000000000000000000 0000000000000000000000000000000010111111000100010000000000000000 0000000000000000000000000000000010111111010000010000000000000000 0000000000000000000000000000000010111111010100010000000000000000 0000000000000000000000000000000010111111010000100000000000000000 0000000000000000000000000000000010111111010100100000000000000000 0000000000000000000000000000000010111111010000110000000000000000 0000000000000000000000000000000010111111010100110000000000000000 0000000000000000000000000000000010111111010001000000000000000000 0000000000000000000000000000000010111111010101000000000000000000 0000000000000000000000000000000010111111010001010000000000000000 0000000000000000000000000000000010111111010101010000000000000000 0000000000000000000000000000000010111111010001100000000000000000 0000000000000000000000000000000010111111010101100000000000000000 0000000000000000000000000000000010111111010001110000000000000000 0000000000000000000000000000000010111111010101110000000000000000 0000000000000000000000000000000010111111010010000000000000000000 0000000000000000000000000000000010111111010110000000000000000000 0000000000000000000000000000000010111111010010010000000000000000 0000000000000000000000000000000010111111010110010000000000000000 0000000000000000000000000000000010111111010010100000000000000000 0000000000000000000000000000000010111111010110100000000000000000 0000000000000000000000000000000010111111010010110000000000000000 0000000000000000000000000000000010111111010110110000000000000000 0000000000000000000000000000000010111111010011000000000000000000 0000000000000000000000000000000010111111010111000000000000000000 0000000000000000000000000000000010111111010011010000000000000000 0000000000000000000000000000000010111111010111010000000000000000 0000000000000000000000000000000010111111010011100000000000000000 0000000000000000000000000000000010111111010111100000000000000000 has_lit_0 && == .fieldderef .fieldderef INST SOPC SSRC0 255 I 32 != .fieldderef .fieldderef INST SOPC SSRC1 255 I 32 has_lit_1 && != .fieldderef .fieldderef INST SOPC SSRC0 255 I 32 == .fieldderef .fieldderef INST SOPC SSRC1 255 I 32 has_lit_0_has_lit_1 && == .fieldderef .fieldderef INST SOPC SSRC0 255 I 32 == .fieldderef .fieldderef INST SOPC SSRC1 255 I 32 Scalar ALU comparison operations with two sources. Allowed patterns: SOPC (32 bits) SOPC + LITERAL (64 bits) ENCODING Encoding. 9 23 OP Opcode. 7 16 SIMM32 32-bit literal constant. 32 32 SSRC0 First source operand for instruction. 8 0 SSRC1 Second source operand for instruction. 8 8 SOPK_INST_LITERAL 64 0000000000000000000000000000000011111111100000000000000000000000 0000000000000000000000000000000010111001100000000000000000000000 default 1 U 1 Scalar ALU operations with one destination and one 16-bit constant source. Allowed patterns: SOPK (32 bits) SOPK + LITERAL (64 bits) ENCODING Encoding. 4 28 OP Opcode. 5 23 SDST Destination operand for instruction. 7 16 SIMM16 16-bit integer input for opcode. Signedness is determined by opcode. 16 0 SIMM32 32-bit literal constant. 32 32 VOP1_VOP_DPP16 64 0000000000000000000000000000000011111110000000001111111000000000 0000000000000000000000000000000001111110000000000000001000000000 0000000000000000000000000000000001111110000000000000101000000000 0000000000000000000000000000000001111110000000000000110000000000 0000000000000000000000000000000001111110000000000000111000000000 0000000000000000000000000000000001111110000000000001000000000000 0000000000000000000000000000000001111110000000000001010000000000 0000000000000000000000000000000001111110000000000001011000000000 0000000000000000000000000000000001111110000000000001100000000000 0000000000000000000000000000000001111110000000000001101000000000 0000000000000000000000000000000001111110000000000001110000000000 0000000000000000000000000000000001111110000000000010001000000000 0000000000000000000000000000000001111110000000000010010000000000 0000000000000000000000000000000001111110000000000010011000000000 0000000000000000000000000000000001111110000000000010100000000000 0000000000000000000000000000000001111110000000000011100000000000 0000000000000000000000000000000001111110000000000100000000000000 0000000000000000000000000000000001111110000000000100001000000000 0000000000000000000000000000000001111110000000000100010000000000 0000000000000000000000000000000001111110000000000100011000000000 0000000000000000000000000000000001111110000000000100100000000000 0000000000000000000000000000000001111110000000000100101000000000 0000000000000000000000000000000001111110000000000100111000000000 0000000000000000000000000000000001111110000000000101010000000000 0000000000000000000000000000000001111110000000000101011000000000 0000000000000000000000000000000001111110000000000101110000000000 0000000000000000000000000000000001111110000000000110011000000000 0000000000000000000000000000000001111110000000000110101000000000 0000000000000000000000000000000001111110000000000110110000000000 0000000000000000000000000000000001111110000000000110111000000000 0000000000000000000000000000000001111110000000000111000000000000 0000000000000000000000000000000001111110000000000111001000000000 0000000000000000000000000000000001111110000000000111010000000000 0000000000000000000000000000000001111110000000000111011000000000 0000000000000000000000000000000001111110000000000111111000000000 0000000000000000000000000000000001111110000000001000000000000000 0000000000000000000000000000000001111110000000001000010000000000 0000000000000000000000000000000001111110000000001000011000000000 0000000000000000000000000000000001111110000000001000100000000000 0000000000000000000000000000000001111110000000001001000000000000 0000000000000000000000000000000001111110000000001010000000000000 0000000000000000000000000000000001111110000000001010001000000000 0000000000000000000000000000000001111110000000001010010000000000 0000000000000000000000000000000001111110000000001010011000000000 0000000000000000000000000000000001111110000000001010100000000000 0000000000000000000000000000000001111110000000001010101000000000 0000000000000000000000000000000001111110000000001010110000000000 0000000000000000000000000000000001111110000000001010111000000000 0000000000000000000000000000000001111110000000001011000000000000 0000000000000000000000000000000001111110000000001011001000000000 0000000000000000000000000000000001111110000000001011010000000000 0000000000000000000000000000000001111110000000001011011000000000 0000000000000000000000000000000001111110000000001011100000000000 0000000000000000000000000000000001111110000000001011101000000000 0000000000000000000000000000000001111110000000001011110000000000 0000000000000000000000000000000001111110000000001011111000000000 0000000000000000000000000000000001111110000000001100000000000000 0000000000000000000000000000000001111110000000001100001000000000 0000000000000000000000000000000001111110000000001100010000000000 0000000000000000000000000000000001111110000000001100011000000000 0000000000000000000000000000000001111110000000001100100000000000 0000000000000000000000000000000001111110000000001101001000000000 0000000000000000000000000000000001111110000000001101010000000000 0000000000000000000000000000000001111110000000001101011000000000 0000000000000000000000000000000001111110000000001101100000000000 0000000000000000000000000000000001111110000000001101101000000000 default == .fieldderef .fieldderef INST VOP SRC0 250 I 32 Vector ALU operations with one destination and one source. Instructions in this encoding may be promoted to VOP3 unless otherwise noted. Instructions in this encoding may use DPP modes unless otherwise noted. Allowed patterns: VOP1 (32 bits) VOP1 + LITERAL (64 bits) VOP1 + DPP (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) BANK_MASK Bank enable mask. If bit N is set, then lanes[4*N:4*N+3, 4*N+16:4*N+19, 4*N+32:4*N+35, 4*N+48:4*N+51] are enabled. 4 56 BOUND_CTRL Specifies behaviour when shared data is invalid. 0: set write-enable on the affected lane to zero. 1: set source data for the affected lane to zero. 1 51 DPP_CTRL Data-parallel primitive control. 9 40 ENCODING Encoding. 7 25 FI If true, reads GPRs when fetching data from an inactive lane (Fetch Inactive mode). If false, reads ZERO when fetching data from an inactive lane. 1 50 OP Opcode. 7 9 ROW_MASK Row enable mask. If bit N is set, then lanes[16*N:16*N+15] are enabled. 4 60 SRC0 First source operand for instruction. 9 0 SRC0_ABS If true, take the floating-point absolute value of src0. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 53 SRC0_NEG If true, take the floating-point negation of src0. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 52 SRC1_ABS If true, take the floating-point absolute value of src1. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 55 SRC1_NEG If true, take the floating-point negation of src1. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 54 VDST Destination operand for instruction. 8 17 VSRC0 VGPR address for first source operand. 8 32 VOP1_VOP_DPP8 64 0000000000000000000000000000000011111110000000001111111000000000 0000000000000000000000000000000001111110000000000000001000000000 0000000000000000000000000000000001111110000000000000101000000000 0000000000000000000000000000000001111110000000000000110000000000 0000000000000000000000000000000001111110000000000000111000000000 0000000000000000000000000000000001111110000000000001000000000000 0000000000000000000000000000000001111110000000000001010000000000 0000000000000000000000000000000001111110000000000001011000000000 0000000000000000000000000000000001111110000000000001100000000000 0000000000000000000000000000000001111110000000000001101000000000 0000000000000000000000000000000001111110000000000001110000000000 0000000000000000000000000000000001111110000000000010001000000000 0000000000000000000000000000000001111110000000000010010000000000 0000000000000000000000000000000001111110000000000010011000000000 0000000000000000000000000000000001111110000000000010100000000000 0000000000000000000000000000000001111110000000000011100000000000 0000000000000000000000000000000001111110000000000100000000000000 0000000000000000000000000000000001111110000000000100001000000000 0000000000000000000000000000000001111110000000000100010000000000 0000000000000000000000000000000001111110000000000100011000000000 0000000000000000000000000000000001111110000000000100100000000000 0000000000000000000000000000000001111110000000000100101000000000 0000000000000000000000000000000001111110000000000100111000000000 0000000000000000000000000000000001111110000000000101010000000000 0000000000000000000000000000000001111110000000000101011000000000 0000000000000000000000000000000001111110000000000101110000000000 0000000000000000000000000000000001111110000000000110011000000000 0000000000000000000000000000000001111110000000000110101000000000 0000000000000000000000000000000001111110000000000110110000000000 0000000000000000000000000000000001111110000000000110111000000000 0000000000000000000000000000000001111110000000000111000000000000 0000000000000000000000000000000001111110000000000111001000000000 0000000000000000000000000000000001111110000000000111010000000000 0000000000000000000000000000000001111110000000000111011000000000 0000000000000000000000000000000001111110000000000111111000000000 0000000000000000000000000000000001111110000000001000000000000000 0000000000000000000000000000000001111110000000001000010000000000 0000000000000000000000000000000001111110000000001000011000000000 0000000000000000000000000000000001111110000000001000100000000000 0000000000000000000000000000000001111110000000001001000000000000 0000000000000000000000000000000001111110000000001010000000000000 0000000000000000000000000000000001111110000000001010001000000000 0000000000000000000000000000000001111110000000001010010000000000 0000000000000000000000000000000001111110000000001010011000000000 0000000000000000000000000000000001111110000000001010100000000000 0000000000000000000000000000000001111110000000001010101000000000 0000000000000000000000000000000001111110000000001010110000000000 0000000000000000000000000000000001111110000000001010111000000000 0000000000000000000000000000000001111110000000001011000000000000 0000000000000000000000000000000001111110000000001011001000000000 0000000000000000000000000000000001111110000000001011010000000000 0000000000000000000000000000000001111110000000001011011000000000 0000000000000000000000000000000001111110000000001011100000000000 0000000000000000000000000000000001111110000000001011101000000000 0000000000000000000000000000000001111110000000001011110000000000 0000000000000000000000000000000001111110000000001011111000000000 0000000000000000000000000000000001111110000000001100000000000000 0000000000000000000000000000000001111110000000001100001000000000 0000000000000000000000000000000001111110000000001100010000000000 0000000000000000000000000000000001111110000000001100011000000000 0000000000000000000000000000000001111110000000001100100000000000 0000000000000000000000000000000001111110000000001101001000000000 0000000000000000000000000000000001111110000000001101010000000000 0000000000000000000000000000000001111110000000001101011000000000 0000000000000000000000000000000001111110000000001101100000000000 0000000000000000000000000000000001111110000000001101101000000000 default || == .fieldderef .fieldderef INST VOP SRC0 233 I 32 == .fieldderef .fieldderef INST VOP SRC0 234 I 32 Vector ALU operations with one destination and one source. Instructions in this encoding may be promoted to VOP3 unless otherwise noted. Instructions in this encoding may use DPP modes unless otherwise noted. Allowed patterns: VOP1 (32 bits) VOP1 + LITERAL (64 bits) VOP1 + DPP (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) ENCODING Encoding. 7 25 LANE_SEL_0 Specify which lane to read from for first lane in the 8-lane group. 3 40 LANE_SEL_1 Specify which lane to read from for second lane in the 8-lane group. 3 43 LANE_SEL_2 Specify which lane to read from for third lane in the 8-lane group. 3 46 LANE_SEL_3 Specify which lane to read from for fourth lane in the 8-lane group. 3 49 LANE_SEL_4 Specify which lane to read from for fifth lane in the 8-lane group. 3 52 LANE_SEL_5 Specify which lane to read from for sixth lane in the 8-lane group. 3 55 LANE_SEL_6 Specify which lane to read from for seventh lane in the 8-lane group. 3 58 LANE_SEL_7 Specify which lane to read from for eighth lane in the 8-lane group. 3 61 OP Opcode. 7 9 SRC0 First source operand for instruction. 9 0 VDST Destination operand for instruction. 8 17 VSRC0 VGPR address for first source operand. 8 32 VOP1_INST_LITERAL 64 0000000000000000000000000000000011111110000000001111111000000000 0000000000000000000000000000000001111110000000000000001000000000 0000000000000000000000000000000001111110000000000000011000000000 0000000000000000000000000000000001111110000000000000100000000000 0000000000000000000000000000000001111110000000000000101000000000 0000000000000000000000000000000001111110000000000000110000000000 0000000000000000000000000000000001111110000000000000111000000000 0000000000000000000000000000000001111110000000000001000000000000 0000000000000000000000000000000001111110000000000001010000000000 0000000000000000000000000000000001111110000000000001011000000000 0000000000000000000000000000000001111110000000000001100000000000 0000000000000000000000000000000001111110000000000001101000000000 0000000000000000000000000000000001111110000000000001110000000000 0000000000000000000000000000000001111110000000000001111000000000 0000000000000000000000000000000001111110000000000010000000000000 0000000000000000000000000000000001111110000000000010001000000000 0000000000000000000000000000000001111110000000000010010000000000 0000000000000000000000000000000001111110000000000010011000000000 0000000000000000000000000000000001111110000000000010100000000000 0000000000000000000000000000000001111110000000000010101000000000 0000000000000000000000000000000001111110000000000010110000000000 0000000000000000000000000000000001111110000000000010111000000000 0000000000000000000000000000000001111110000000000011000000000000 0000000000000000000000000000000001111110000000000011001000000000 0000000000000000000000000000000001111110000000000011010000000000 0000000000000000000000000000000001111110000000000011100000000000 0000000000000000000000000000000001111110000000000100000000000000 0000000000000000000000000000000001111110000000000100001000000000 0000000000000000000000000000000001111110000000000100010000000000 0000000000000000000000000000000001111110000000000100011000000000 0000000000000000000000000000000001111110000000000100100000000000 0000000000000000000000000000000001111110000000000100101000000000 0000000000000000000000000000000001111110000000000100111000000000 0000000000000000000000000000000001111110000000000101010000000000 0000000000000000000000000000000001111110000000000101011000000000 0000000000000000000000000000000001111110000000000101110000000000 0000000000000000000000000000000001111110000000000101111000000000 0000000000000000000000000000000001111110000000000110001000000000 0000000000000000000000000000000001111110000000000110011000000000 0000000000000000000000000000000001111110000000000110100000000000 0000000000000000000000000000000001111110000000000110101000000000 0000000000000000000000000000000001111110000000000110110000000000 0000000000000000000000000000000001111110000000000110111000000000 0000000000000000000000000000000001111110000000000111000000000000 0000000000000000000000000000000001111110000000000111001000000000 0000000000000000000000000000000001111110000000000111010000000000 0000000000000000000000000000000001111110000000000111011000000000 0000000000000000000000000000000001111110000000000111100000000000 0000000000000000000000000000000001111110000000000111101000000000 0000000000000000000000000000000001111110000000000111110000000000 0000000000000000000000000000000001111110000000000111111000000000 0000000000000000000000000000000001111110000000001000000000000000 0000000000000000000000000000000001111110000000001000010000000000 0000000000000000000000000000000001111110000000001010000000000000 0000000000000000000000000000000001111110000000001010001000000000 0000000000000000000000000000000001111110000000001010010000000000 0000000000000000000000000000000001111110000000001010011000000000 0000000000000000000000000000000001111110000000001010100000000000 0000000000000000000000000000000001111110000000001010101000000000 0000000000000000000000000000000001111110000000001010110000000000 0000000000000000000000000000000001111110000000001010111000000000 0000000000000000000000000000000001111110000000001011000000000000 0000000000000000000000000000000001111110000000001011001000000000 0000000000000000000000000000000001111110000000001011010000000000 0000000000000000000000000000000001111110000000001011011000000000 0000000000000000000000000000000001111110000000001011100000000000 0000000000000000000000000000000001111110000000001011101000000000 0000000000000000000000000000000001111110000000001011110000000000 0000000000000000000000000000000001111110000000001011111000000000 0000000000000000000000000000000001111110000000001100000000000000 0000000000000000000000000000000001111110000000001100001000000000 0000000000000000000000000000000001111110000000001100010000000000 0000000000000000000000000000000001111110000000001100011000000000 0000000000000000000000000000000001111110000000001100100000000000 0000000000000000000000000000000001111110000000001100111000000000 0000000000000000000000000000000001111110000000001101001000000000 0000000000000000000000000000000001111110000000001101010000000000 0000000000000000000000000000000001111110000000001101011000000000 0000000000000000000000000000000001111110000000001101100000000000 0000000000000000000000000000000001111110000000001101101000000000 0000000000000000000000000000000001111110000000001101110000000000 0000000000000000000000000000000001111110000000001101111000000000 has_lit == .fieldderef .fieldderef INST VOP1 SRC0 255 I 32 Vector ALU operations with one destination and one source. Instructions in this encoding may be promoted to VOP3 unless otherwise noted. Instructions in this encoding may use DPP modes unless otherwise noted. Allowed patterns: VOP1 (32 bits) VOP1 + LITERAL (64 bits) VOP1 + DPP (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) ENCODING Encoding. 7 25 OP Opcode. 7 9 SIMM32 32-bit literal constant. 32 32 SRC0 First source operand for instruction. 9 0 VDST Destination operand for instruction. 8 17 VOP3_INST_LITERAL 96 000000000000000000000000000000000000000000000000000000000000000011111111111111110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101111110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110111110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000111110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110100000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110100000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110100000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110100000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110100001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110100001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110100001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110100001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110100010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110100010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000111110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111010000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111010000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100011111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100011111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100011111110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100111111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100111111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100111111110000000000000000 has_lit_0 && && == .fieldderef .fieldderef INST VOP3 SRC0 255 I 32 != .fieldderef .fieldderef INST VOP3 SRC1 255 I 32 != .fieldderef .fieldderef INST VOP3 SRC2 255 I 32 has_lit_1 && && && && != .fieldderef .fieldderef INST VOP3 SRC0 250 I 32 && != .fieldderef .fieldderef INST VOP3 SRC0 233 I 32 != .fieldderef .fieldderef INST VOP3 SRC0 234 I 32 != .fieldderef .fieldderef INST VOP3 SRC0 255 I 32 == .fieldderef .fieldderef INST VOP3 SRC1 255 I 32 != .fieldderef .fieldderef INST VOP3 SRC2 255 I 32 has_lit_0_has_lit_1 && && == .fieldderef .fieldderef INST VOP3 SRC0 255 I 32 == .fieldderef .fieldderef INST VOP3 SRC1 255 I 32 != .fieldderef .fieldderef INST VOP3 SRC2 255 I 32 has_lit_2 && && && && != .fieldderef .fieldderef INST VOP3 SRC0 250 I 32 && != .fieldderef .fieldderef INST VOP3 SRC0 233 I 32 != .fieldderef .fieldderef INST VOP3 SRC0 234 I 32 != .fieldderef .fieldderef INST VOP3 SRC0 255 I 32 != .fieldderef .fieldderef INST VOP3 SRC1 255 I 32 == .fieldderef .fieldderef INST VOP3 SRC2 255 I 32 has_lit_0_has_lit_2 && && == .fieldderef .fieldderef INST VOP3 SRC0 255 I 32 != .fieldderef .fieldderef INST VOP3 SRC1 255 I 32 == .fieldderef .fieldderef INST VOP3 SRC2 255 I 32 has_lit_1_has_lit_2 && && && && != .fieldderef .fieldderef INST VOP3 SRC0 250 I 32 && != .fieldderef .fieldderef INST VOP3 SRC0 233 I 32 != .fieldderef .fieldderef INST VOP3 SRC0 234 I 32 != .fieldderef .fieldderef INST VOP3 SRC0 255 I 32 == .fieldderef .fieldderef INST VOP3 SRC1 255 I 32 == .fieldderef .fieldderef INST VOP3 SRC2 255 I 32 has_lit_0_has_lit_1_has_lit_2 && && == .fieldderef .fieldderef INST VOP3 SRC0 255 I 32 == .fieldderef .fieldderef INST VOP3 SRC1 255 I 32 == .fieldderef .fieldderef INST VOP3 SRC2 255 I 32 Vector ALU operations requiring three sources or modifiers; also some obscure opcodes that cannot be encoded in smaller instructions due to limited OP bits. Almost all VOP1, VOP2 and VOPC opcodes can be promoted to VOP3 to use input and output modifiers. Allowed patterns: VOP3 (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) ABS If ABS[N] set, take the floating-point absolute value of the N'th input operand. This is applied before negation. See OPF_NOABS for instructions that do not allow this input modifier. 3 8 CLAMP If set, clamp or signal on output. VCMP instructions enable signaling-compare. For floating point outputs, clamp to [0.0, 1.0]. For signed integer outputs, clamp to [MIN_INT, MAX_INT]. For unsigned integer outputs, clamp to [0, MAX_UINT]. Applied after output modifier. See OPF_NOCLAMP for a list of instructions that do not support clamping or signaling. 1 15 ENCODING Encoding. 6 26 NEG If NEG[N] set, take the floating-point negation of the N'th input operand. This is applied after absolute value. See OPF_NONEG for a list of instructions that do not support negation. 3 61 OMOD Output modifier for instruction. Only useful for floating point outputs; integer outputs should set this to OMOD_OFF. Applied before clamping. See OPF_NOOMOD for a list of instructions that do not support output modifier. Integer and packed float 16 results ignore the omod setting. Output modifiers are not compatible with output denorms. If omod is not OMOD_OFF then output denormals are flushed regardless of the denorm control setting. Output modifiers are not IEEE compliant (-0*x=+0). 2 59 OP Opcode. 10 16 OPSEL For 16-bit inputs, OPSEL[N] determines if lower [15:0] or upper [31:16] bits are selected for the N'th source operand. Similarly OPSEL[3] controls whether the lower [15:0] or upper [31:16] bits are written for a 16-bit destination (preserving the unwritten bits). 0 selects the lower order bits and 1 selects the higher order bits. 4 11 SIMM32 32-bit literal constant. 32 64 SRC0 First source operand for instruction. 9 32 SRC1 Second source operand for instruction. 9 41 SRC2 Third source operand for instruction. 9 50 VDST Destination operand for instruction. 8 0 VOP3_VOP_DPP16 96 000000000000000000000000000000000000000000000000000000000000000011111111111111110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101111110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110111110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000111110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100011111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100011111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100111111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100111111100000000000000000 default == .fieldderef .fieldderef INST VOP SRC0 250 I 32 Vector ALU operations requiring three sources or modifiers; also some obscure opcodes that cannot be encoded in smaller instructions due to limited OP bits. Almost all VOP1, VOP2 and VOPC opcodes can be promoted to VOP3 to use input and output modifiers. Allowed patterns: VOP3 (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) ABS If ABS[N] set, take the floating-point absolute value of the N'th input operand. This is applied before negation. See OPF_NOABS for instructions that do not allow this input modifier. 3 8 BANK_MASK Bank enable mask. If bit N is set, then lanes[4*N:4*N+3, 4*N+16:4*N+19, 4*N+32:4*N+35, 4*N+48:4*N+51] are enabled. 4 88 BOUND_CTRL Specifies behaviour when shared data is invalid. 0: set write-enable on the affected lane to zero. 1: set source data for the affected lane to zero. 1 83 CLAMP If set, clamp or signal on output. VCMP instructions enable signaling-compare. For floating point outputs, clamp to [0.0, 1.0]. For signed integer outputs, clamp to [MIN_INT, MAX_INT]. For unsigned integer outputs, clamp to [0, MAX_UINT]. Applied after output modifier. See OPF_NOCLAMP for a list of instructions that do not support clamping or signaling. 1 15 DPP_CTRL Data-parallel primitive control. 9 72 ENCODING Encoding. 6 26 FI If true, reads GPRs when fetching data from an inactive lane (Fetch Inactive mode). If false, reads ZERO when fetching data from an inactive lane. 1 82 NEG If NEG[N] set, take the floating-point negation of the N'th input operand. This is applied after absolute value. See OPF_NONEG for a list of instructions that do not support negation. 3 61 OMOD Output modifier for instruction. Only useful for floating point outputs; integer outputs should set this to OMOD_OFF. Applied before clamping. See OPF_NOOMOD for a list of instructions that do not support output modifier. Integer and packed float 16 results ignore the omod setting. Output modifiers are not compatible with output denorms. If omod is not OMOD_OFF then output denormals are flushed regardless of the denorm control setting. Output modifiers are not IEEE compliant (-0*x=+0). 2 59 OP Opcode. 10 16 OPSEL For 16-bit inputs, OPSEL[N] determines if lower [15:0] or upper [31:16] bits are selected for the N'th source operand. Similarly OPSEL[3] controls whether the lower [15:0] or upper [31:16] bits are written for a 16-bit destination (preserving the unwritten bits). 0 selects the lower order bits and 1 selects the higher order bits. 4 11 ROW_MASK Row enable mask. If bit N is set, then lanes[16*N:16*N+15] are enabled. 4 92 SRC0 First source operand for instruction. 9 32 SRC0_ABS If true, take the floating-point absolute value of src0. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 85 SRC0_NEG If true, take the floating-point negation of src0. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 84 SRC1 Second source operand for instruction. 9 41 SRC1_ABS If true, take the floating-point absolute value of src1. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 87 SRC1_NEG If true, take the floating-point negation of src1. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 86 SRC2 Third source operand for instruction. 9 50 VDST Destination operand for instruction. 8 0 VSRC0 VGPR address for first source operand. 8 64 VOP3_VOP_DPP8 96 000000000000000000000000000000000000000000000000000000000000000011111111111111110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101100111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101101111110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101110111110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101111011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101000111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110000111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110001110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110010111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010110011011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000111110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111001110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111011011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100000111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100001111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100010011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100011111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100011111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100100111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101100010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101101010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100101111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100110011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100111111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010100111111100000000000000000 default || == .fieldderef .fieldderef INST VOP SRC0 233 I 32 == .fieldderef .fieldderef INST VOP SRC0 234 I 32 Vector ALU operations requiring three sources or modifiers; also some obscure opcodes that cannot be encoded in smaller instructions due to limited OP bits. Almost all VOP1, VOP2 and VOPC opcodes can be promoted to VOP3 to use input and output modifiers. Allowed patterns: VOP3 (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) ABS If ABS[N] set, take the floating-point absolute value of the N'th input operand. This is applied before negation. See OPF_NOABS for instructions that do not allow this input modifier. 3 8 CLAMP If set, clamp or signal on output. VCMP instructions enable signaling-compare. For floating point outputs, clamp to [0.0, 1.0]. For signed integer outputs, clamp to [MIN_INT, MAX_INT]. For unsigned integer outputs, clamp to [0, MAX_UINT]. Applied after output modifier. See OPF_NOCLAMP for a list of instructions that do not support clamping or signaling. 1 15 ENCODING Encoding. 6 26 LANE_SEL_0 Specify which lane to read from for first lane in the 8-lane group. 3 72 LANE_SEL_1 Specify which lane to read from for second lane in the 8-lane group. 3 75 LANE_SEL_2 Specify which lane to read from for third lane in the 8-lane group. 3 78 LANE_SEL_3 Specify which lane to read from for fourth lane in the 8-lane group. 3 81 LANE_SEL_4 Specify which lane to read from for fifth lane in the 8-lane group. 3 84 LANE_SEL_5 Specify which lane to read from for sixth lane in the 8-lane group. 3 87 LANE_SEL_6 Specify which lane to read from for seventh lane in the 8-lane group. 3 90 LANE_SEL_7 Specify which lane to read from for eighth lane in the 8-lane group. 3 93 NEG If NEG[N] set, take the floating-point negation of the N'th input operand. This is applied after absolute value. See OPF_NONEG for a list of instructions that do not support negation. 3 61 OMOD Output modifier for instruction. Only useful for floating point outputs; integer outputs should set this to OMOD_OFF. Applied before clamping. See OPF_NOOMOD for a list of instructions that do not support output modifier. Integer and packed float 16 results ignore the omod setting. Output modifiers are not compatible with output denorms. If omod is not OMOD_OFF then output denormals are flushed regardless of the denorm control setting. Output modifiers are not IEEE compliant (-0*x=+0). 2 59 OP Opcode. 10 16 OPSEL For 16-bit inputs, OPSEL[N] determines if lower [15:0] or upper [31:16] bits are selected for the N'th source operand. Similarly OPSEL[3] controls whether the lower [15:0] or upper [31:16] bits are written for a 16-bit destination (preserving the unwritten bits). 0 selects the lower order bits and 1 selects the higher order bits. 4 11 SRC0 First source operand for instruction. 9 32 SRC1 Second source operand for instruction. 9 41 SRC2 Third source operand for instruction. 9 50 VDST Destination operand for instruction. 8 0 VSRC0 VGPR address for first source operand. 8 64 VOP2_VOP_DPP16 64 0000000000000000000000000000000011111110000000000000000000000000 0000000000000000000000000000000000000010000000000000000000000000 0000000000000000000000000000000000000110000000000000000000000000 0000000000000000000000000000000000001000000000000000000000000000 0000000000000000000000000000000000001010000000000000000000000000 0000000000000000000000000000000000001110000000000000000000000000 0000000000000000000000000000000000010000000000000000000000000000 0000000000000000000000000000000000010010000000000000000000000000 0000000000000000000000000000000000010100000000000000000000000000 0000000000000000000000000000000000010110000000000000000000000000 0000000000000000000000000000000000011000000000000000000000000000 0000000000000000000000000000000000100010000000000000000000000000 0000000000000000000000000000000000100100000000000000000000000000 0000000000000000000000000000000000100110000000000000000000000000 0000000000000000000000000000000000101000000000000000000000000000 0000000000000000000000000000000000101010000000000000000000000000 0000000000000000000000000000000000101100000000000000000000000000 0000000000000000000000000000000000110000000000000000000000000000 0000000000000000000000000000000000110010000000000000000000000000 0000000000000000000000000000000000110100000000000000000000000000 0000000000000000000000000000000000110110000000000000000000000000 0000000000000000000000000000000000111000000000000000000000000000 0000000000000000000000000000000000111010000000000000000000000000 0000000000000000000000000000000000111100000000000000000000000000 0000000000000000000000000000000001000000000000000000000000000000 0000000000000000000000000000000001000010000000000000000000000000 0000000000000000000000000000000001000100000000000000000000000000 0000000000000000000000000000000001001010000000000000000000000000 0000000000000000000000000000000001001100000000000000000000000000 0000000000000000000000000000000001001110000000000000000000000000 0000000000000000000000000000000001010110000000000000000000000000 0000000000000000000000000000000001011110000000000000000000000000 0000000000000000000000000000000001100000000000000000000000000000 0000000000000000000000000000000001100010000000000000000000000000 0000000000000000000000000000000001100100000000000000000000000000 0000000000000000000000000000000001100110000000000000000000000000 0000000000000000000000000000000001101000000000000000000000000000 0000000000000000000000000000000001101010000000000000000000000000 0000000000000000000000000000000001101100000000000000000000000000 0000000000000000000000000000000001110110000000000000000000000000 default == .fieldderef .fieldderef INST VOP SRC0 250 I 32 Vector ALU operations with one destination and two sources. Instructions in this encoding may be promoted to VOP3 unless otherwise noted. DPP modes are allowed unless otherwise noted. Allowed patterns: VOP2 (32 bits) VOP2 + LITERAL (64 bits) VOP2 + DPP (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) BANK_MASK Bank enable mask. If bit N is set, then lanes[4*N:4*N+3, 4*N+16:4*N+19, 4*N+32:4*N+35, 4*N+48:4*N+51] are enabled. 4 56 BOUND_CTRL Specifies behaviour when shared data is invalid. 0: set write-enable on the affected lane to zero. 1: set source data for the affected lane to zero. 1 51 DPP_CTRL Data-parallel primitive control. 9 40 ENCODING Encoding. 1 31 FI If true, reads GPRs when fetching data from an inactive lane (Fetch Inactive mode). If false, reads ZERO when fetching data from an inactive lane. 1 50 OP Opcode. 6 25 ROW_MASK Row enable mask. If bit N is set, then lanes[16*N:16*N+15] are enabled. 4 60 SRC0 First source operand for instruction. 9 0 SRC0_ABS If true, take the floating-point absolute value of src0. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 53 SRC0_NEG If true, take the floating-point negation of src0. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 52 SRC1_ABS If true, take the floating-point absolute value of src1. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 55 SRC1_NEG If true, take the floating-point negation of src1. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 54 VDST Destination operand for instruction. 8 17 VSRC0 VGPR address for first source operand. 8 32 VSRC1 Second source operand for instruction. 8 9 VOP2_VOP_DPP8 64 0000000000000000000000000000000011111110000000000000000000000000 0000000000000000000000000000000000000010000000000000000000000000 0000000000000000000000000000000000000110000000000000000000000000 0000000000000000000000000000000000001000000000000000000000000000 0000000000000000000000000000000000001010000000000000000000000000 0000000000000000000000000000000000001110000000000000000000000000 0000000000000000000000000000000000010000000000000000000000000000 0000000000000000000000000000000000010010000000000000000000000000 0000000000000000000000000000000000010100000000000000000000000000 0000000000000000000000000000000000010110000000000000000000000000 0000000000000000000000000000000000011000000000000000000000000000 0000000000000000000000000000000000100010000000000000000000000000 0000000000000000000000000000000000100100000000000000000000000000 0000000000000000000000000000000000100110000000000000000000000000 0000000000000000000000000000000000101000000000000000000000000000 0000000000000000000000000000000000101010000000000000000000000000 0000000000000000000000000000000000101100000000000000000000000000 0000000000000000000000000000000000110000000000000000000000000000 0000000000000000000000000000000000110010000000000000000000000000 0000000000000000000000000000000000110100000000000000000000000000 0000000000000000000000000000000000110110000000000000000000000000 0000000000000000000000000000000000111000000000000000000000000000 0000000000000000000000000000000000111010000000000000000000000000 0000000000000000000000000000000000111100000000000000000000000000 0000000000000000000000000000000001000000000000000000000000000000 0000000000000000000000000000000001000010000000000000000000000000 0000000000000000000000000000000001000100000000000000000000000000 0000000000000000000000000000000001001010000000000000000000000000 0000000000000000000000000000000001001100000000000000000000000000 0000000000000000000000000000000001001110000000000000000000000000 0000000000000000000000000000000001010110000000000000000000000000 0000000000000000000000000000000001011110000000000000000000000000 0000000000000000000000000000000001100000000000000000000000000000 0000000000000000000000000000000001100010000000000000000000000000 0000000000000000000000000000000001100100000000000000000000000000 0000000000000000000000000000000001100110000000000000000000000000 0000000000000000000000000000000001101000000000000000000000000000 0000000000000000000000000000000001101010000000000000000000000000 0000000000000000000000000000000001101100000000000000000000000000 0000000000000000000000000000000001110110000000000000000000000000 default || == .fieldderef .fieldderef INST VOP SRC0 233 I 32 == .fieldderef .fieldderef INST VOP SRC0 234 I 32 Vector ALU operations with one destination and two sources. Instructions in this encoding may be promoted to VOP3 unless otherwise noted. DPP modes are allowed unless otherwise noted. Allowed patterns: VOP2 (32 bits) VOP2 + LITERAL (64 bits) VOP2 + DPP (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) ENCODING Encoding. 1 31 LANE_SEL_0 Specify which lane to read from for first lane in the 8-lane group. 3 40 LANE_SEL_1 Specify which lane to read from for second lane in the 8-lane group. 3 43 LANE_SEL_2 Specify which lane to read from for third lane in the 8-lane group. 3 46 LANE_SEL_3 Specify which lane to read from for fourth lane in the 8-lane group. 3 49 LANE_SEL_4 Specify which lane to read from for fifth lane in the 8-lane group. 3 52 LANE_SEL_5 Specify which lane to read from for sixth lane in the 8-lane group. 3 55 LANE_SEL_6 Specify which lane to read from for seventh lane in the 8-lane group. 3 58 LANE_SEL_7 Specify which lane to read from for eighth lane in the 8-lane group. 3 61 OP Opcode. 6 25 SRC0 First source operand for instruction. 9 0 VDST Destination operand for instruction. 8 17 VSRC0 VGPR address for first source operand. 8 32 VSRC1 Second source operand for instruction. 8 9 VOP2_INST_LITERAL 64 0000000000000000000000000000000011111110000000000000000000000000 0000000000000000000000000000000000000010000000000000000000000000 0000000000000000000000000000000000000100000000000000000000000000 0000000000000000000000000000000000000110000000000000000000000000 0000000000000000000000000000000000001000000000000000000000000000 0000000000000000000000000000000000001010000000000000000000000000 0000000000000000000000000000000000001100000000000000000000000000 0000000000000000000000000000000000001110000000000000000000000000 0000000000000000000000000000000000010000000000000000000000000000 0000000000000000000000000000000000010010000000000000000000000000 0000000000000000000000000000000000010100000000000000000000000000 0000000000000000000000000000000000010110000000000000000000000000 0000000000000000000000000000000000011000000000000000000000000000 0000000000000000000000000000000000011010000000000000000000000000 0000000000000000000000000000000000011100000000000000000000000000 0000000000000000000000000000000000100010000000000000000000000000 0000000000000000000000000000000000100100000000000000000000000000 0000000000000000000000000000000000100110000000000000000000000000 0000000000000000000000000000000000101000000000000000000000000000 0000000000000000000000000000000000101010000000000000000000000000 0000000000000000000000000000000000101100000000000000000000000000 0000000000000000000000000000000000110000000000000000000000000000 0000000000000000000000000000000000110010000000000000000000000000 0000000000000000000000000000000000110100000000000000000000000000 0000000000000000000000000000000000110110000000000000000000000000 0000000000000000000000000000000000111000000000000000000000000000 0000000000000000000000000000000000111010000000000000000000000000 0000000000000000000000000000000000111100000000000000000000000000 0000000000000000000000000000000000111110000000000000000000000000 0000000000000000000000000000000001000000000000000000000000000000 0000000000000000000000000000000001000010000000000000000000000000 0000000000000000000000000000000001000100000000000000000000000000 0000000000000000000000000000000001001010000000000000000000000000 0000000000000000000000000000000001001100000000000000000000000000 0000000000000000000000000000000001001110000000000000000000000000 0000000000000000000000000000000001010110000000000000000000000000 0000000000000000000000000000000001011000000000000000000000000000 0000000000000000000000000000000001011010000000000000000000000000 0000000000000000000000000000000001011110000000000000000000000000 0000000000000000000000000000000001100000000000000000000000000000 0000000000000000000000000000000001100010000000000000000000000000 0000000000000000000000000000000001100100000000000000000000000000 0000000000000000000000000000000001100110000000000000000000000000 0000000000000000000000000000000001101000000000000000000000000000 0000000000000000000000000000000001101010000000000000000000000000 0000000000000000000000000000000001101100000000000000000000000000 0000000000000000000000000000000001101110000000000000000000000000 0000000000000000000000000000000001110000000000000000000000000000 0000000000000000000000000000000001110110000000000000000000000000 0000000000000000000000000000000001111000000000000000000000000000 has_lit == .fieldderef .fieldderef INST VOP2 SRC0 255 I 32 default && && != .fieldderef .fieldderef INST VOP2 SRC0 250 I 32 && != .fieldderef .fieldderef INST VOP2 SRC0 233 I 32 != .fieldderef .fieldderef INST VOP2 SRC0 234 I 32 != .fieldderef .fieldderef INST VOP2 SRC0 255 I 32 Vector ALU operations with one destination and two sources. Instructions in this encoding may be promoted to VOP3 unless otherwise noted. DPP modes are allowed unless otherwise noted. Allowed patterns: VOP2 (32 bits) VOP2 + LITERAL (64 bits) VOP2 + DPP (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) ENCODING Encoding. 1 31 OP Opcode. 6 25 SIMM32 32-bit literal constant. 32 32 SRC0 First source operand for instruction. 9 0 VDST Destination operand for instruction. 8 17 VSRC1 Second source operand for instruction. 8 9 VOP3_SDST_ENC 64 0000000000000000000000000000000011111111111111110000000000000000 0000000000000000000000000000000011010101001000000000000000000000 0000000000000000000000000000000011010101001000010000000000000000 0000000000000000000000000000000011010101001000100000000000000000 0000000000000000000000000000000011010110111111000000000000000000 0000000000000000000000000000000011010110111111010000000000000000 0000000000000000000000000000000011010110111111100000000000000000 0000000000000000000000000000000011010110111111110000000000000000 0000000000000000000000000000000011010111000000000000000000000000 0000000000000000000000000000000011010111000000010000000000000000 0000000000000000000000000000000011010111000000100000000000000000 default && && && && != .fieldderef .fieldderef INST VOP3 SRC0 250 I 32 && != .fieldderef .fieldderef INST VOP3 SRC0 233 I 32 != .fieldderef .fieldderef INST VOP3 SRC0 234 I 32 != .fieldderef .fieldderef INST VOP3 SRC0 255 I 32 != .fieldderef .fieldderef INST VOP3 SRC1 255 I 32 != .fieldderef .fieldderef INST VOP3 SRC2 255 I 32 Vector ALU operations requiring three sources or modifiers; also some obscure opcodes that cannot be encoded in smaller instructions due to limited OP bits. Almost all VOP1, VOP2 and VOPC opcodes can be promoted to VOP3 to use input and output modifiers. Allowed patterns: VOP3 (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) CLAMP If set, clamp or signal on output. VCMP instructions enable signaling-compare. For floating point outputs, clamp to [0.0, 1.0]. For signed integer outputs, clamp to [MIN_INT, MAX_INT]. For unsigned integer outputs, clamp to [0, MAX_UINT]. Applied after output modifier. See OPF_NOCLAMP for a list of instructions that do not support clamping or signaling. 1 15 ENCODING Encoding. 6 26 NEG If NEG[N] set, take the floating-point negation of the N'th input operand. This is applied after absolute value. See OPF_NONEG for a list of instructions that do not support negation. 3 61 OMOD Output modifier for instruction. Only useful for floating point outputs; integer outputs should set this to OMOD_OFF. Applied before clamping. See OPF_NOOMOD for a list of instructions that do not support output modifier. Integer and packed float 16 results ignore the omod setting. Output modifiers are not compatible with output denorms. If omod is not OMOD_OFF then output denormals are flushed regardless of the denorm control setting. Output modifiers are not IEEE compliant (-0*x=+0). 2 59 OP Opcode. 10 16 SDST Destination operand for compare result. 7 8 SRC0 First source operand for instruction. 9 32 SRC1 Second source operand for instruction. 9 41 SRC2 Third source operand for instruction. 9 50 VDST Destination operand for instruction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has_lit_0 && && == .fieldderef .fieldderef INST VOP3 SRC0 255 I 32 != .fieldderef .fieldderef INST VOP3 SRC1 255 I 32 != .fieldderef .fieldderef INST VOP3 SRC2 255 I 32 has_lit_1 && && && && != .fieldderef .fieldderef INST VOP3 SRC0 250 I 32 && != .fieldderef .fieldderef INST VOP3 SRC0 233 I 32 != .fieldderef .fieldderef INST VOP3 SRC0 234 I 32 != .fieldderef .fieldderef INST VOP3 SRC0 255 I 32 == .fieldderef .fieldderef INST VOP3 SRC1 255 I 32 != .fieldderef .fieldderef INST VOP3 SRC2 255 I 32 has_lit_0_has_lit_1 && && == .fieldderef .fieldderef INST VOP3 SRC0 255 I 32 == .fieldderef .fieldderef INST VOP3 SRC1 255 I 32 != .fieldderef .fieldderef INST VOP3 SRC2 255 I 32 has_lit_2 && && && && != .fieldderef .fieldderef INST VOP3 SRC0 250 I 32 && != .fieldderef .fieldderef INST VOP3 SRC0 233 I 32 != .fieldderef .fieldderef INST VOP3 SRC0 234 I 32 != .fieldderef .fieldderef INST VOP3 SRC0 255 I 32 != .fieldderef .fieldderef INST VOP3 SRC1 255 I 32 == .fieldderef .fieldderef INST VOP3 SRC2 255 I 32 has_lit_0_has_lit_2 && && == .fieldderef .fieldderef INST VOP3 SRC0 255 I 32 != .fieldderef .fieldderef INST VOP3 SRC1 255 I 32 == .fieldderef .fieldderef INST VOP3 SRC2 255 I 32 has_lit_1_has_lit_2 && && && && != .fieldderef .fieldderef INST VOP3 SRC0 250 I 32 && != .fieldderef .fieldderef INST VOP3 SRC0 233 I 32 != .fieldderef .fieldderef INST VOP3 SRC0 234 I 32 != .fieldderef .fieldderef INST VOP3 SRC0 255 I 32 == .fieldderef .fieldderef INST VOP3 SRC1 255 I 32 == .fieldderef .fieldderef INST VOP3 SRC2 255 I 32 has_lit_0_has_lit_1_has_lit_2 && && == .fieldderef .fieldderef INST VOP3 SRC0 255 I 32 == .fieldderef .fieldderef INST VOP3 SRC1 255 I 32 == .fieldderef .fieldderef INST VOP3 SRC2 255 I 32 Vector ALU operations requiring three sources or modifiers; also some obscure opcodes that cannot be encoded in smaller instructions due to limited OP bits. Almost all VOP1, VOP2 and VOPC opcodes can be promoted to VOP3 to use input and output modifiers. Allowed patterns: VOP3 (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) CLAMP If set, clamp or signal on output. VCMP instructions enable signaling-compare. For floating point outputs, clamp to [0.0, 1.0]. For signed integer outputs, clamp to [MIN_INT, MAX_INT]. For unsigned integer outputs, clamp to [0, MAX_UINT]. Applied after output modifier. See OPF_NOCLAMP for a list of instructions that do not support clamping or signaling. 1 15 ENCODING Encoding. 6 26 NEG If NEG[N] set, take the floating-point negation of the N'th input operand. This is applied after absolute value. See OPF_NONEG for a list of instructions that do not support negation. 3 61 OMOD Output modifier for instruction. Only useful for floating point outputs; integer outputs should set this to OMOD_OFF. Applied before clamping. See OPF_NOOMOD for a list of instructions that do not support output modifier. Integer and packed float 16 results ignore the omod setting. Output modifiers are not compatible with output denorms. If omod is not OMOD_OFF then output denormals are flushed regardless of the denorm control setting. Output modifiers are not IEEE compliant (-0*x=+0). 2 59 OP Opcode. 10 16 SDST Destination operand for compare result. 7 8 SIMM32 32-bit literal constant. 32 64 SRC0 First source operand for instruction. 9 32 SRC1 Second source operand for instruction. 9 41 SRC2 Third source operand for instruction. 9 50 VDST Destination operand for instruction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default == .fieldderef .fieldderef INST VOP SRC0 250 I 32 Vector ALU operations requiring three sources or modifiers; also some obscure opcodes that cannot be encoded in smaller instructions due to limited OP bits. Almost all VOP1, VOP2 and VOPC opcodes can be promoted to VOP3 to use input and output modifiers. Allowed patterns: VOP3 (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) BANK_MASK Bank enable mask. If bit N is set, then lanes[4*N:4*N+3, 4*N+16:4*N+19, 4*N+32:4*N+35, 4*N+48:4*N+51] are enabled. 4 88 BOUND_CTRL Specifies behaviour when shared data is invalid. 0: set write-enable on the affected lane to zero. 1: set source data for the affected lane to zero. 1 83 CLAMP If set, clamp or signal on output. VCMP instructions enable signaling-compare. For floating point outputs, clamp to [0.0, 1.0]. For signed integer outputs, clamp to [MIN_INT, MAX_INT]. For unsigned integer outputs, clamp to [0, MAX_UINT]. Applied after output modifier. See OPF_NOCLAMP for a list of instructions that do not support clamping or signaling. 1 15 DPP_CTRL Data-parallel primitive control. 9 72 ENCODING Encoding. 6 26 FI If true, reads GPRs when fetching data from an inactive lane (Fetch Inactive mode). If false, reads ZERO when fetching data from an inactive lane. 1 82 NEG If NEG[N] set, take the floating-point negation of the N'th input operand. This is applied after absolute value. See OPF_NONEG for a list of instructions that do not support negation. 3 61 OMOD Output modifier for instruction. Only useful for floating point outputs; integer outputs should set this to OMOD_OFF. Applied before clamping. See OPF_NOOMOD for a list of instructions that do not support output modifier. Integer and packed float 16 results ignore the omod setting. Output modifiers are not compatible with output denorms. If omod is not OMOD_OFF then output denormals are flushed regardless of the denorm control setting. Output modifiers are not IEEE compliant (-0*x=+0). 2 59 OP Opcode. 10 16 ROW_MASK Row enable mask. If bit N is set, then lanes[16*N:16*N+15] are enabled. 4 92 SDST Destination operand for compare result. 7 8 SRC0 First source operand for instruction. 9 32 SRC0_ABS If true, take the floating-point absolute value of src0. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 85 SRC0_NEG If true, take the floating-point negation of src0. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 84 SRC1 Second source operand for instruction. 9 41 SRC1_ABS If true, take the floating-point absolute value of src1. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 87 SRC1_NEG If true, take the floating-point negation of src1. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 86 SRC2 Third source operand for instruction. 9 50 VDST Destination operand for instruction. 8 0 VSRC0 VGPR address for first source operand. 8 64 VOP3_SDST_ENC_VOP_DPP8 96 000000000000000000000000000000000000000000000000000000000000000011111111111111110000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010101001000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011010111000000100000000000000000 default || == .fieldderef .fieldderef INST VOP SRC0 233 I 32 == .fieldderef .fieldderef INST VOP SRC0 234 I 32 Vector ALU operations requiring three sources or modifiers; also some obscure opcodes that cannot be encoded in smaller instructions due to limited OP bits. Almost all VOP1, VOP2 and VOPC opcodes can be promoted to VOP3 to use input and output modifiers. Allowed patterns: VOP3 (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) CLAMP If set, clamp or signal on output. VCMP instructions enable signaling-compare. For floating point outputs, clamp to [0.0, 1.0]. For signed integer outputs, clamp to [MIN_INT, MAX_INT]. For unsigned integer outputs, clamp to [0, MAX_UINT]. Applied after output modifier. See OPF_NOCLAMP for a list of instructions that do not support clamping or signaling. 1 15 ENCODING Encoding. 6 26 LANE_SEL_0 Specify which lane to read from for first lane in the 8-lane group. 3 72 LANE_SEL_1 Specify which lane to read from for second lane in the 8-lane group. 3 75 LANE_SEL_2 Specify which lane to read from for third lane in the 8-lane group. 3 78 LANE_SEL_3 Specify which lane to read from for fourth lane in the 8-lane group. 3 81 LANE_SEL_4 Specify which lane to read from for fifth lane in the 8-lane group. 3 84 LANE_SEL_5 Specify which lane to read from for sixth lane in the 8-lane group. 3 87 LANE_SEL_6 Specify which lane to read from for seventh lane in the 8-lane group. 3 90 LANE_SEL_7 Specify which lane to read from for eighth lane in the 8-lane group. 3 93 NEG If NEG[N] set, take the floating-point negation of the N'th input operand. This is applied after absolute value. See OPF_NONEG for a list of instructions that do not support negation. 3 61 OMOD Output modifier for instruction. Only useful for floating point outputs; integer outputs should set this to OMOD_OFF. Applied before clamping. See OPF_NOOMOD for a list of instructions that do not support output modifier. Integer and packed float 16 results ignore the omod setting. Output modifiers are not compatible with output denorms. If omod is not OMOD_OFF then output denormals are flushed regardless of the denorm control setting. Output modifiers are not IEEE compliant (-0*x=+0). 2 59 OP Opcode. 10 16 SDST Destination operand for compare result. 7 8 SRC0 First source operand for instruction. 9 32 SRC1 Second source operand for instruction. 9 41 SRC2 Third source operand for instruction. 9 50 VDST Destination operand for instruction. 8 0 VSRC0 VGPR address for first source operand. 8 64 VOP3P_INST_LITERAL 96 000000000000000000000000000000000000000000000000000000000000000011111111011111110000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000000110000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000001110000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000010010000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000010110000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000011010000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000011110000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000101110000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000110010000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000110110000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000111000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000111010000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000111100000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100001000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100001000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100001000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100001001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100001001010000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100001001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100001001110000000000000000 has_lit_0 && && == .fieldderef .fieldderef INST VOP3P SRC0 255 I 32 != .fieldderef .fieldderef INST VOP3P SRC1 255 I 32 != .fieldderef .fieldderef INST VOP3P SRC2 255 I 32 has_lit_1 && && && && != .fieldderef .fieldderef INST VOP3P SRC0 250 I 32 && != .fieldderef .fieldderef INST VOP3P SRC0 233 I 32 != .fieldderef .fieldderef INST VOP3P SRC0 234 I 32 != .fieldderef .fieldderef INST VOP3P SRC0 255 I 32 == .fieldderef .fieldderef INST VOP3P SRC1 255 I 32 != .fieldderef .fieldderef INST VOP3P SRC2 255 I 32 has_lit_0_has_lit_1 && && == .fieldderef .fieldderef INST VOP3P SRC0 255 I 32 == .fieldderef .fieldderef INST VOP3P SRC1 255 I 32 != .fieldderef .fieldderef INST VOP3P SRC2 255 I 32 has_lit_2 && && && && != .fieldderef .fieldderef INST VOP3P SRC0 250 I 32 && != .fieldderef .fieldderef INST VOP3P SRC0 233 I 32 != .fieldderef .fieldderef INST VOP3P SRC0 234 I 32 != .fieldderef .fieldderef INST VOP3P SRC0 255 I 32 != .fieldderef .fieldderef INST VOP3P SRC1 255 I 32 == .fieldderef .fieldderef INST VOP3P SRC2 255 I 32 has_lit_0_has_lit_2 && && == .fieldderef .fieldderef INST VOP3P SRC0 255 I 32 != .fieldderef .fieldderef INST VOP3P SRC1 255 I 32 == .fieldderef .fieldderef INST VOP3P SRC2 255 I 32 has_lit_1_has_lit_2 && && && && != .fieldderef .fieldderef INST VOP3P SRC0 250 I 32 && != .fieldderef .fieldderef INST VOP3P SRC0 233 I 32 != .fieldderef .fieldderef INST VOP3P SRC0 234 I 32 != .fieldderef .fieldderef INST VOP3P SRC0 255 I 32 == .fieldderef .fieldderef INST VOP3P SRC1 255 I 32 == .fieldderef .fieldderef INST VOP3P SRC2 255 I 32 has_lit_0_has_lit_1_has_lit_2 && && == .fieldderef .fieldderef INST VOP3P SRC0 255 I 32 == .fieldderef .fieldderef INST VOP3P SRC1 255 I 32 == .fieldderef .fieldderef INST VOP3P SRC2 255 I 32 Packed math instructions: vector ALU operations requiring two or three sources and performing two 16 bit ops in one instruction. Allowed patterns: VOP3P (64 bits) VOP3P + LITERAL (96 bits) VOP3P + DPP (96 bits) CLAMP If set, clamp or signal on output. VCMP instructions enable signaling-compare. For floating point outputs, clamp to [0.0, 1.0]. For signed integer outputs, clamp to [MIN_INT, MAX_INT]. For unsigned integer outputs, clamp to [0, MAX_UINT]. Applied after output modifier. See OPF_NOCLAMP for a list of instructions that do not support clamping or signaling. 1 15 ENCODING Encoding. 8 24 NEG If NEG[N] set, take the floating-point negation of the N'th input LOW OPERATION. See OPF_NONEG for a list of instructions that do not support negation. 3 61 NEG_HI If NEG_HI[N] set, take the floating-point negation of the N'th input HIGH OPERATION. See OPF_NONEG for a list of instructions that do not support negation. 3 8 OP Opcode. 7 16 OPSEL OPSEL[N] determines if lower [15:0] or upper [31:16] bits are selected for the N'th source for the LOW OPERATION. 0 selects the lower order bits and 1 selects the higher order bits. 3 11 OPSEL_HI OPSEL_HI[N] determines if lower [15:0] or upper [31:16] bits are selected for the N'th source for the HIGH OPERATION. 0 selects the lower order bits and 1 selects the higher order bits. 2 59 OPSEL_HI_2 Bit 2 of OPSEL_HI. 1 14 SIMM32 32-bit literal constant. 32 64 SRC0 First source operand for instruction. 9 32 SRC1 Second source operand for instruction. 9 41 SRC2 Third source operand for instruction. 9 50 VDST Destination operand for instruction. 8 0 VOP3P_VOP_DPP16 96 000000000000000000000000000000000000000000000000000000000000000011111111011111110000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100001000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100001000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100001000100000000000000000 default == .fieldderef .fieldderef INST VOP SRC0 250 I 32 Packed math instructions: vector ALU operations requiring two or three sources and performing two 16 bit ops in one instruction. Allowed patterns: VOP3P (64 bits) VOP3P + LITERAL (96 bits) VOP3P + DPP (96 bits) BANK_MASK Bank enable mask. If bit N is set, then lanes[4*N:4*N+3, 4*N+16:4*N+19, 4*N+32:4*N+35, 4*N+48:4*N+51] are enabled. 4 88 BOUND_CTRL Specifies behaviour when shared data is invalid. 0: set write-enable on the affected lane to zero. 1: set source data for the affected lane to zero. 1 83 CLAMP If set, clamp or signal on output. VCMP instructions enable signaling-compare. For floating point outputs, clamp to [0.0, 1.0]. For signed integer outputs, clamp to [MIN_INT, MAX_INT]. For unsigned integer outputs, clamp to [0, MAX_UINT]. Applied after output modifier. See OPF_NOCLAMP for a list of instructions that do not support clamping or signaling. 1 15 DPP_CTRL Data-parallel primitive control. 9 72 ENCODING Encoding. 8 24 FI If true, reads GPRs when fetching data from an inactive lane (Fetch Inactive mode). If false, reads ZERO when fetching data from an inactive lane. 1 82 NEG If NEG[N] set, take the floating-point negation of the N'th input LOW OPERATION. See OPF_NONEG for a list of instructions that do not support negation. 3 61 NEG_HI If NEG_HI[N] set, take the floating-point negation of the N'th input HIGH OPERATION. See OPF_NONEG for a list of instructions that do not support negation. 3 8 OP Opcode. 7 16 OPSEL OPSEL[N] determines if lower [15:0] or upper [31:16] bits are selected for the N'th source for the LOW OPERATION. 0 selects the lower order bits and 1 selects the higher order bits. 3 11 OPSEL_HI OPSEL_HI[N] determines if lower [15:0] or upper [31:16] bits are selected for the N'th source for the HIGH OPERATION. 0 selects the lower order bits and 1 selects the higher order bits. 2 59 OPSEL_HI_2 Bit 2 of OPSEL_HI. 1 14 ROW_MASK Row enable mask. If bit N is set, then lanes[16*N:16*N+15] are enabled. 4 92 SRC0 First source operand for instruction. 9 32 SRC0_ABS If true, take the floating-point absolute value of src0. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 85 SRC0_NEG If true, take the floating-point negation of src0. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 84 SRC1 Second source operand for instruction. 9 41 SRC1_ABS If true, take the floating-point absolute value of src1. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 87 SRC1_NEG If true, take the floating-point negation of src1. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 86 SRC2 Third source operand for instruction. 9 50 VDST Destination operand for instruction. 8 0 VSRC0 VGPR address for first source operand. 8 64 VOP3P_VOP_DPP8 96 000000000000000000000000000000000000000000000000000000000000000011111111011111110000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000100110000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100001000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100001000010000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100001000100000000000000000 default || == .fieldderef .fieldderef INST VOP SRC0 233 I 32 == .fieldderef .fieldderef INST VOP SRC0 234 I 32 Packed math instructions: vector ALU operations requiring two or three sources and performing two 16 bit ops in one instruction. Allowed patterns: VOP3P (64 bits) VOP3P + LITERAL (96 bits) VOP3P + DPP (96 bits) CLAMP If set, clamp or signal on output. VCMP instructions enable signaling-compare. For floating point outputs, clamp to [0.0, 1.0]. For signed integer outputs, clamp to [MIN_INT, MAX_INT]. For unsigned integer outputs, clamp to [0, MAX_UINT]. Applied after output modifier. See OPF_NOCLAMP for a list of instructions that do not support clamping or signaling. 1 15 ENCODING Encoding. 8 24 LANE_SEL_0 Specify which lane to read from for first lane in the 8-lane group. 3 72 LANE_SEL_1 Specify which lane to read from for second lane in the 8-lane group. 3 75 LANE_SEL_2 Specify which lane to read from for third lane in the 8-lane group. 3 78 LANE_SEL_3 Specify which lane to read from for fourth lane in the 8-lane group. 3 81 LANE_SEL_4 Specify which lane to read from for fifth lane in the 8-lane group. 3 84 LANE_SEL_5 Specify which lane to read from for sixth lane in the 8-lane group. 3 87 LANE_SEL_6 Specify which lane to read from for seventh lane in the 8-lane group. 3 90 LANE_SEL_7 Specify which lane to read from for eighth lane in the 8-lane group. 3 93 NEG If NEG[N] set, take the floating-point negation of the N'th input LOW OPERATION. See OPF_NONEG for a list of instructions that do not support negation. 3 61 NEG_HI If NEG_HI[N] set, take the floating-point negation of the N'th input HIGH OPERATION. See OPF_NONEG for a list of instructions that do not support negation. 3 8 OP Opcode. 7 16 OPSEL OPSEL[N] determines if lower [15:0] or upper [31:16] bits are selected for the N'th source for the LOW OPERATION. 0 selects the lower order bits and 1 selects the higher order bits. 3 11 OPSEL_HI OPSEL_HI[N] determines if lower [15:0] or upper [31:16] bits are selected for the N'th source for the HIGH OPERATION. 0 selects the lower order bits and 1 selects the higher order bits. 2 59 OPSEL_HI_2 Bit 2 of OPSEL_HI. 1 14 SRC0 First source operand for instruction. 9 32 SRC1 Second source operand for instruction. 9 41 SRC2 Third source operand for instruction. 9 50 VDST Destination operand for instruction. 8 0 VSRC0 VGPR address for first source operand. 8 64 VOPC_VOP_DPP16 64 0000000000000000000000000000000011111111111111100000000000000000 0000000000000000000000000000000001111100000000100000000000000000 0000000000000000000000000000000001111100000001000000000000000000 0000000000000000000000000000000001111100000001100000000000000000 0000000000000000000000000000000001111100000010000000000000000000 0000000000000000000000000000000001111100000010100000000000000000 0000000000000000000000000000000001111100000011000000000000000000 0000000000000000000000000000000001111100000011100000000000000000 0000000000000000000000000000000001111100000100000000000000000000 0000000000000000000000000000000001111100000100100000000000000000 0000000000000000000000000000000001111100000101000000000000000000 0000000000000000000000000000000001111100000101100000000000000000 0000000000000000000000000000000001111100000110000000000000000000 0000000000000000000000000000000001111100000110100000000000000000 0000000000000000000000000000000001111100000111000000000000000000 0000000000000000000000000000000001111100001000100000000000000000 0000000000000000000000000000000001111100001001000000000000000000 0000000000000000000000000000000001111100001001100000000000000000 0000000000000000000000000000000001111100001010000000000000000000 0000000000000000000000000000000001111100001010100000000000000000 0000000000000000000000000000000001111100001011000000000000000000 0000000000000000000000000000000001111100001011100000000000000000 0000000000000000000000000000000001111100001100000000000000000000 0000000000000000000000000000000001111100001100100000000000000000 0000000000000000000000000000000001111100001101000000000000000000 0000000000000000000000000000000001111100001101100000000000000000 0000000000000000000000000000000001111100001110000000000000000000 0000000000000000000000000000000001111100001110100000000000000000 0000000000000000000000000000000001111100001111000000000000000000 0000000000000000000000000000000001111100011000100000000000000000 0000000000000000000000000000000001111100011001000000000000000000 0000000000000000000000000000000001111100011001100000000000000000 0000000000000000000000000000000001111100011010000000000000000000 0000000000000000000000000000000001111100011010100000000000000000 0000000000000000000000000000000001111100011011000000000000000000 0000000000000000000000000000000001111100011100100000000000000000 0000000000000000000000000000000001111100011101000000000000000000 0000000000000000000000000000000001111100011101100000000000000000 0000000000000000000000000000000001111100011110000000000000000000 0000000000000000000000000000000001111100011110100000000000000000 0000000000000000000000000000000001111100011111000000000000000000 0000000000000000000000000000000001111100100000100000000000000000 0000000000000000000000000000000001111100100001000000000000000000 0000000000000000000000000000000001111100100001100000000000000000 0000000000000000000000000000000001111100100010000000000000000000 0000000000000000000000000000000001111100100010100000000000000000 0000000000000000000000000000000001111100100011000000000000000000 0000000000000000000000000000000001111100100100100000000000000000 0000000000000000000000000000000001111100100101000000000000000000 0000000000000000000000000000000001111100100101100000000000000000 0000000000000000000000000000000001111100100110000000000000000000 0000000000000000000000000000000001111100100110100000000000000000 0000000000000000000000000000000001111100100111000000000000000000 0000000000000000000000000000000001111100111110100000000000000000 0000000000000000000000000000000001111100111111000000000000000000 0000000000000000000000000000000001111101000000100000000000000000 0000000000000000000000000000000001111101000001000000000000000000 0000000000000000000000000000000001111101000001100000000000000000 0000000000000000000000000000000001111101000010000000000000000000 0000000000000000000000000000000001111101000010100000000000000000 0000000000000000000000000000000001111101000011000000000000000000 0000000000000000000000000000000001111101000011100000000000000000 0000000000000000000000000000000001111101000100000000000000000000 0000000000000000000000000000000001111101000100100000000000000000 0000000000000000000000000000000001111101000101000000000000000000 0000000000000000000000000000000001111101000101100000000000000000 0000000000000000000000000000000001111101000110000000000000000000 0000000000000000000000000000000001111101000110100000000000000000 0000000000000000000000000000000001111101000111000000000000000000 0000000000000000000000000000000001111101001000100000000000000000 0000000000000000000000000000000001111101001001000000000000000000 0000000000000000000000000000000001111101001001100000000000000000 0000000000000000000000000000000001111101001010000000000000000000 0000000000000000000000000000000001111101001010100000000000000000 0000000000000000000000000000000001111101001011000000000000000000 0000000000000000000000000000000001111101001011100000000000000000 0000000000000000000000000000000001111101001100000000000000000000 0000000000000000000000000000000001111101001100100000000000000000 0000000000000000000000000000000001111101001101000000000000000000 0000000000000000000000000000000001111101001101100000000000000000 0000000000000000000000000000000001111101001110000000000000000000 0000000000000000000000000000000001111101001110100000000000000000 0000000000000000000000000000000001111101001111000000000000000000 0000000000000000000000000000000001111101011000100000000000000000 0000000000000000000000000000000001111101011001000000000000000000 0000000000000000000000000000000001111101011001100000000000000000 0000000000000000000000000000000001111101011010000000000000000000 0000000000000000000000000000000001111101011010100000000000000000 0000000000000000000000000000000001111101011011000000000000000000 0000000000000000000000000000000001111101011100100000000000000000 0000000000000000000000000000000001111101011101000000000000000000 0000000000000000000000000000000001111101011101100000000000000000 0000000000000000000000000000000001111101011110000000000000000000 0000000000000000000000000000000001111101011110100000000000000000 0000000000000000000000000000000001111101011111000000000000000000 0000000000000000000000000000000001111101100000100000000000000000 0000000000000000000000000000000001111101100001000000000000000000 0000000000000000000000000000000001111101100001100000000000000000 0000000000000000000000000000000001111101100010000000000000000000 0000000000000000000000000000000001111101100010100000000000000000 0000000000000000000000000000000001111101100011000000000000000000 0000000000000000000000000000000001111101100100100000000000000000 0000000000000000000000000000000001111101100101000000000000000000 0000000000000000000000000000000001111101100101100000000000000000 0000000000000000000000000000000001111101100110000000000000000000 0000000000000000000000000000000001111101100110100000000000000000 0000000000000000000000000000000001111101100111000000000000000000 0000000000000000000000000000000001111101111110100000000000000000 0000000000000000000000000000000001111101111111000000000000000000 default == .fieldderef .fieldderef INST VOP SRC0 250 I 32 Vector ALU comparison operations with two sources. Instructions in this encoding may be promoted to VOP3 unless otherwise noted. Note that vector compare instructions respect the EXEC mask for computing the result in each thread; if EXEC[laneId] is zero, then the comparison for that thread is not performed and the result of VCC[laneId] is zero. Instructions in this encoding may use DPP modes unless otherwise noted. Allowed patterns: VOPC (32 bits) VOPC + LITERAL (64 bits) VOPC + DPP (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) BANK_MASK Bank enable mask. If bit N is set, then lanes[4*N:4*N+3, 4*N+16:4*N+19, 4*N+32:4*N+35, 4*N+48:4*N+51] are enabled. 4 56 BOUND_CTRL Specifies behaviour when shared data is invalid. 0: set write-enable on the affected lane to zero. 1: set source data for the affected lane to zero. 1 51 DPP_CTRL Data-parallel primitive control. 9 40 ENCODING Encoding. 7 25 FI If true, reads GPRs when fetching data from an inactive lane (Fetch Inactive mode). If false, reads ZERO when fetching data from an inactive lane. 1 50 OP Opcode. 8 17 ROW_MASK Row enable mask. If bit N is set, then lanes[16*N:16*N+15] are enabled. 4 60 SRC0 First source operand for instruction. 9 0 SRC0_ABS If true, take the floating-point absolute value of src0. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 53 SRC0_NEG If true, take the floating-point negation of src0. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 52 SRC1_ABS If true, take the floating-point absolute value of src1. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 55 SRC1_NEG If true, take the floating-point negation of src1. Note: this field is ignored when DPP is used in VOP3/VOP3P encodings; use the fields provided by those encodings instead. 1 54 VSRC0 VGPR address for first source operand. 8 32 VSRC1 Second source operand for instruction. 8 9 VOPC_VOP_DPP8 64 0000000000000000000000000000000011111111111111100000000000000000 0000000000000000000000000000000001111100000000100000000000000000 0000000000000000000000000000000001111100000001000000000000000000 0000000000000000000000000000000001111100000001100000000000000000 0000000000000000000000000000000001111100000010000000000000000000 0000000000000000000000000000000001111100000010100000000000000000 0000000000000000000000000000000001111100000011000000000000000000 0000000000000000000000000000000001111100000011100000000000000000 0000000000000000000000000000000001111100000100000000000000000000 0000000000000000000000000000000001111100000100100000000000000000 0000000000000000000000000000000001111100000101000000000000000000 0000000000000000000000000000000001111100000101100000000000000000 0000000000000000000000000000000001111100000110000000000000000000 0000000000000000000000000000000001111100000110100000000000000000 0000000000000000000000000000000001111100000111000000000000000000 0000000000000000000000000000000001111100001000100000000000000000 0000000000000000000000000000000001111100001001000000000000000000 0000000000000000000000000000000001111100001001100000000000000000 0000000000000000000000000000000001111100001010000000000000000000 0000000000000000000000000000000001111100001010100000000000000000 0000000000000000000000000000000001111100001011000000000000000000 0000000000000000000000000000000001111100001011100000000000000000 0000000000000000000000000000000001111100001100000000000000000000 0000000000000000000000000000000001111100001100100000000000000000 0000000000000000000000000000000001111100001101000000000000000000 0000000000000000000000000000000001111100001101100000000000000000 0000000000000000000000000000000001111100001110000000000000000000 0000000000000000000000000000000001111100001110100000000000000000 0000000000000000000000000000000001111100001111000000000000000000 0000000000000000000000000000000001111100011000100000000000000000 0000000000000000000000000000000001111100011001000000000000000000 0000000000000000000000000000000001111100011001100000000000000000 0000000000000000000000000000000001111100011010000000000000000000 0000000000000000000000000000000001111100011010100000000000000000 0000000000000000000000000000000001111100011011000000000000000000 0000000000000000000000000000000001111100011100100000000000000000 0000000000000000000000000000000001111100011101000000000000000000 0000000000000000000000000000000001111100011101100000000000000000 0000000000000000000000000000000001111100011110000000000000000000 0000000000000000000000000000000001111100011110100000000000000000 0000000000000000000000000000000001111100011111000000000000000000 0000000000000000000000000000000001111100100000100000000000000000 0000000000000000000000000000000001111100100001000000000000000000 0000000000000000000000000000000001111100100001100000000000000000 0000000000000000000000000000000001111100100010000000000000000000 0000000000000000000000000000000001111100100010100000000000000000 0000000000000000000000000000000001111100100011000000000000000000 0000000000000000000000000000000001111100100100100000000000000000 0000000000000000000000000000000001111100100101000000000000000000 0000000000000000000000000000000001111100100101100000000000000000 0000000000000000000000000000000001111100100110000000000000000000 0000000000000000000000000000000001111100100110100000000000000000 0000000000000000000000000000000001111100100111000000000000000000 0000000000000000000000000000000001111100111110100000000000000000 0000000000000000000000000000000001111100111111000000000000000000 0000000000000000000000000000000001111101000000100000000000000000 0000000000000000000000000000000001111101000001000000000000000000 0000000000000000000000000000000001111101000001100000000000000000 0000000000000000000000000000000001111101000010000000000000000000 0000000000000000000000000000000001111101000010100000000000000000 0000000000000000000000000000000001111101000011000000000000000000 0000000000000000000000000000000001111101000011100000000000000000 0000000000000000000000000000000001111101000100000000000000000000 0000000000000000000000000000000001111101000100100000000000000000 0000000000000000000000000000000001111101000101000000000000000000 0000000000000000000000000000000001111101000101100000000000000000 0000000000000000000000000000000001111101000110000000000000000000 0000000000000000000000000000000001111101000110100000000000000000 0000000000000000000000000000000001111101000111000000000000000000 0000000000000000000000000000000001111101001000100000000000000000 0000000000000000000000000000000001111101001001000000000000000000 0000000000000000000000000000000001111101001001100000000000000000 0000000000000000000000000000000001111101001010000000000000000000 0000000000000000000000000000000001111101001010100000000000000000 0000000000000000000000000000000001111101001011000000000000000000 0000000000000000000000000000000001111101001011100000000000000000 0000000000000000000000000000000001111101001100000000000000000000 0000000000000000000000000000000001111101001100100000000000000000 0000000000000000000000000000000001111101001101000000000000000000 0000000000000000000000000000000001111101001101100000000000000000 0000000000000000000000000000000001111101001110000000000000000000 0000000000000000000000000000000001111101001110100000000000000000 0000000000000000000000000000000001111101001111000000000000000000 0000000000000000000000000000000001111101011000100000000000000000 0000000000000000000000000000000001111101011001000000000000000000 0000000000000000000000000000000001111101011001100000000000000000 0000000000000000000000000000000001111101011010000000000000000000 0000000000000000000000000000000001111101011010100000000000000000 0000000000000000000000000000000001111101011011000000000000000000 0000000000000000000000000000000001111101011100100000000000000000 0000000000000000000000000000000001111101011101000000000000000000 0000000000000000000000000000000001111101011101100000000000000000 0000000000000000000000000000000001111101011110000000000000000000 0000000000000000000000000000000001111101011110100000000000000000 0000000000000000000000000000000001111101011111000000000000000000 0000000000000000000000000000000001111101100000100000000000000000 0000000000000000000000000000000001111101100001000000000000000000 0000000000000000000000000000000001111101100001100000000000000000 0000000000000000000000000000000001111101100010000000000000000000 0000000000000000000000000000000001111101100010100000000000000000 0000000000000000000000000000000001111101100011000000000000000000 0000000000000000000000000000000001111101100100100000000000000000 0000000000000000000000000000000001111101100101000000000000000000 0000000000000000000000000000000001111101100101100000000000000000 0000000000000000000000000000000001111101100110000000000000000000 0000000000000000000000000000000001111101100110100000000000000000 0000000000000000000000000000000001111101100111000000000000000000 0000000000000000000000000000000001111101111110100000000000000000 0000000000000000000000000000000001111101111111000000000000000000 default || == .fieldderef .fieldderef INST VOP SRC0 233 I 32 == .fieldderef .fieldderef INST VOP SRC0 234 I 32 Vector ALU comparison operations with two sources. Instructions in this encoding may be promoted to VOP3 unless otherwise noted. Note that vector compare instructions respect the EXEC mask for computing the result in each thread; if EXEC[laneId] is zero, then the comparison for that thread is not performed and the result of VCC[laneId] is zero. Instructions in this encoding may use DPP modes unless otherwise noted. Allowed patterns: VOPC (32 bits) VOPC + LITERAL (64 bits) VOPC + DPP (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) ENCODING Encoding. 7 25 LANE_SEL_0 Specify which lane to read from for first lane in the 8-lane group. 3 40 LANE_SEL_1 Specify which lane to read from for second lane in the 8-lane group. 3 43 LANE_SEL_2 Specify which lane to read from for third lane in the 8-lane group. 3 46 LANE_SEL_3 Specify which lane to read from for fourth lane in the 8-lane group. 3 49 LANE_SEL_4 Specify which lane to read from for fifth lane in the 8-lane group. 3 52 LANE_SEL_5 Specify which lane to read from for sixth lane in the 8-lane group. 3 55 LANE_SEL_6 Specify which lane to read from for seventh lane in the 8-lane group. 3 58 LANE_SEL_7 Specify which lane to read from for eighth lane in the 8-lane group. 3 61 OP Opcode. 8 17 SRC0 First source operand for instruction. 9 0 VSRC0 VGPR address for first source operand. 8 32 VSRC1 Second source operand for instruction. 8 9 VOPC_INST_LITERAL 64 0000000000000000000000000000000011111111111111100000000000000000 0000000000000000000000000000000001111100000000100000000000000000 0000000000000000000000000000000001111100000001000000000000000000 0000000000000000000000000000000001111100000001100000000000000000 0000000000000000000000000000000001111100000010000000000000000000 0000000000000000000000000000000001111100000010100000000000000000 0000000000000000000000000000000001111100000011000000000000000000 0000000000000000000000000000000001111100000011100000000000000000 0000000000000000000000000000000001111100000100000000000000000000 0000000000000000000000000000000001111100000100100000000000000000 0000000000000000000000000000000001111100000101000000000000000000 0000000000000000000000000000000001111100000101100000000000000000 0000000000000000000000000000000001111100000110000000000000000000 0000000000000000000000000000000001111100000110100000000000000000 0000000000000000000000000000000001111100000111000000000000000000 0000000000000000000000000000000001111100001000100000000000000000 0000000000000000000000000000000001111100001001000000000000000000 0000000000000000000000000000000001111100001001100000000000000000 0000000000000000000000000000000001111100001010000000000000000000 0000000000000000000000000000000001111100001010100000000000000000 0000000000000000000000000000000001111100001011000000000000000000 0000000000000000000000000000000001111100001011100000000000000000 0000000000000000000000000000000001111100001100000000000000000000 0000000000000000000000000000000001111100001100100000000000000000 0000000000000000000000000000000001111100001101000000000000000000 0000000000000000000000000000000001111100001101100000000000000000 0000000000000000000000000000000001111100001110000000000000000000 0000000000000000000000000000000001111100001110100000000000000000 0000000000000000000000000000000001111100001111000000000000000000 0000000000000000000000000000000001111100010000100000000000000000 0000000000000000000000000000000001111100010001000000000000000000 0000000000000000000000000000000001111100010001100000000000000000 0000000000000000000000000000000001111100010010000000000000000000 0000000000000000000000000000000001111100010010100000000000000000 0000000000000000000000000000000001111100010011000000000000000000 0000000000000000000000000000000001111100010011100000000000000000 0000000000000000000000000000000001111100010100000000000000000000 0000000000000000000000000000000001111100010100100000000000000000 0000000000000000000000000000000001111100010101000000000000000000 0000000000000000000000000000000001111100010101100000000000000000 0000000000000000000000000000000001111100010110000000000000000000 0000000000000000000000000000000001111100010110100000000000000000 0000000000000000000000000000000001111100010111000000000000000000 0000000000000000000000000000000001111100011000100000000000000000 0000000000000000000000000000000001111100011001000000000000000000 0000000000000000000000000000000001111100011001100000000000000000 0000000000000000000000000000000001111100011010000000000000000000 0000000000000000000000000000000001111100011010100000000000000000 0000000000000000000000000000000001111100011011000000000000000000 0000000000000000000000000000000001111100011100100000000000000000 0000000000000000000000000000000001111100011101000000000000000000 0000000000000000000000000000000001111100011101100000000000000000 0000000000000000000000000000000001111100011110000000000000000000 0000000000000000000000000000000001111100011110100000000000000000 0000000000000000000000000000000001111100011111000000000000000000 0000000000000000000000000000000001111100100000100000000000000000 0000000000000000000000000000000001111100100001000000000000000000 0000000000000000000000000000000001111100100001100000000000000000 0000000000000000000000000000000001111100100010000000000000000000 0000000000000000000000000000000001111100100010100000000000000000 0000000000000000000000000000000001111100100011000000000000000000 0000000000000000000000000000000001111100100100100000000000000000 0000000000000000000000000000000001111100100101000000000000000000 0000000000000000000000000000000001111100100101100000000000000000 0000000000000000000000000000000001111100100110000000000000000000 0000000000000000000000000000000001111100100110100000000000000000 0000000000000000000000000000000001111100100111000000000000000000 0000000000000000000000000000000001111100101000100000000000000000 0000000000000000000000000000000001111100101001000000000000000000 0000000000000000000000000000000001111100101001100000000000000000 0000000000000000000000000000000001111100101010000000000000000000 0000000000000000000000000000000001111100101010100000000000000000 0000000000000000000000000000000001111100101011000000000000000000 0000000000000000000000000000000001111100101100100000000000000000 0000000000000000000000000000000001111100101101000000000000000000 0000000000000000000000000000000001111100101101100000000000000000 0000000000000000000000000000000001111100101110000000000000000000 0000000000000000000000000000000001111100101110100000000000000000 0000000000000000000000000000000001111100101111000000000000000000 0000000000000000000000000000000001111100111110100000000000000000 0000000000000000000000000000000001111100111111000000000000000000 0000000000000000000000000000000001111100111111100000000000000000 0000000000000000000000000000000001111101000000100000000000000000 0000000000000000000000000000000001111101000001000000000000000000 0000000000000000000000000000000001111101000001100000000000000000 0000000000000000000000000000000001111101000010000000000000000000 0000000000000000000000000000000001111101000010100000000000000000 0000000000000000000000000000000001111101000011000000000000000000 0000000000000000000000000000000001111101000011100000000000000000 0000000000000000000000000000000001111101000100000000000000000000 0000000000000000000000000000000001111101000100100000000000000000 0000000000000000000000000000000001111101000101000000000000000000 0000000000000000000000000000000001111101000101100000000000000000 0000000000000000000000000000000001111101000110000000000000000000 0000000000000000000000000000000001111101000110100000000000000000 0000000000000000000000000000000001111101000111000000000000000000 0000000000000000000000000000000001111101001000100000000000000000 0000000000000000000000000000000001111101001001000000000000000000 0000000000000000000000000000000001111101001001100000000000000000 0000000000000000000000000000000001111101001010000000000000000000 0000000000000000000000000000000001111101001010100000000000000000 0000000000000000000000000000000001111101001011000000000000000000 0000000000000000000000000000000001111101001011100000000000000000 0000000000000000000000000000000001111101001100000000000000000000 0000000000000000000000000000000001111101001100100000000000000000 0000000000000000000000000000000001111101001101000000000000000000 0000000000000000000000000000000001111101001101100000000000000000 0000000000000000000000000000000001111101001110000000000000000000 0000000000000000000000000000000001111101001110100000000000000000 0000000000000000000000000000000001111101001111000000000000000000 0000000000000000000000000000000001111101010000100000000000000000 0000000000000000000000000000000001111101010001000000000000000000 0000000000000000000000000000000001111101010001100000000000000000 0000000000000000000000000000000001111101010010000000000000000000 0000000000000000000000000000000001111101010010100000000000000000 0000000000000000000000000000000001111101010011000000000000000000 0000000000000000000000000000000001111101010011100000000000000000 0000000000000000000000000000000001111101010100000000000000000000 0000000000000000000000000000000001111101010100100000000000000000 0000000000000000000000000000000001111101010101000000000000000000 0000000000000000000000000000000001111101010101100000000000000000 0000000000000000000000000000000001111101010110000000000000000000 0000000000000000000000000000000001111101010110100000000000000000 0000000000000000000000000000000001111101010111000000000000000000 0000000000000000000000000000000001111101011000100000000000000000 0000000000000000000000000000000001111101011001000000000000000000 0000000000000000000000000000000001111101011001100000000000000000 0000000000000000000000000000000001111101011010000000000000000000 0000000000000000000000000000000001111101011010100000000000000000 0000000000000000000000000000000001111101011011000000000000000000 0000000000000000000000000000000001111101011100100000000000000000 0000000000000000000000000000000001111101011101000000000000000000 0000000000000000000000000000000001111101011101100000000000000000 0000000000000000000000000000000001111101011110000000000000000000 0000000000000000000000000000000001111101011110100000000000000000 0000000000000000000000000000000001111101011111000000000000000000 0000000000000000000000000000000001111101100000100000000000000000 0000000000000000000000000000000001111101100001000000000000000000 0000000000000000000000000000000001111101100001100000000000000000 0000000000000000000000000000000001111101100010000000000000000000 0000000000000000000000000000000001111101100010100000000000000000 0000000000000000000000000000000001111101100011000000000000000000 0000000000000000000000000000000001111101100100100000000000000000 0000000000000000000000000000000001111101100101000000000000000000 0000000000000000000000000000000001111101100101100000000000000000 0000000000000000000000000000000001111101100110000000000000000000 0000000000000000000000000000000001111101100110100000000000000000 0000000000000000000000000000000001111101100111000000000000000000 0000000000000000000000000000000001111101101000100000000000000000 0000000000000000000000000000000001111101101001000000000000000000 0000000000000000000000000000000001111101101001100000000000000000 0000000000000000000000000000000001111101101010000000000000000000 0000000000000000000000000000000001111101101010100000000000000000 0000000000000000000000000000000001111101101011000000000000000000 0000000000000000000000000000000001111101101100100000000000000000 0000000000000000000000000000000001111101101101000000000000000000 0000000000000000000000000000000001111101101101100000000000000000 0000000000000000000000000000000001111101101110000000000000000000 0000000000000000000000000000000001111101101110100000000000000000 0000000000000000000000000000000001111101101111000000000000000000 0000000000000000000000000000000001111101111110100000000000000000 0000000000000000000000000000000001111101111111000000000000000000 0000000000000000000000000000000001111101111111100000000000000000 has_lit == .fieldderef .fieldderef INST VOPC SRC0 255 I 32 Vector ALU comparison operations with two sources. Instructions in this encoding may be promoted to VOP3 unless otherwise noted. Note that vector compare instructions respect the EXEC mask for computing the result in each thread; if EXEC[laneId] is zero, then the comparison for that thread is not performed and the result of VCC[laneId] is zero. Instructions in this encoding may use DPP modes unless otherwise noted. Allowed patterns: VOPC (32 bits) VOPC + LITERAL (64 bits) VOPC + DPP (64 bits) VOP3 + LITERAL (96 bits) VOP3 + DPP (96 bits) ENCODING Encoding. 7 25 OP Opcode. 8 17 SIMM32 32-bit literal constant. 32 32 SRC0 First source operand for instruction. 9 0 VSRC1 Second source operand for instruction. 8 9 VOPDXY 64 0000000000000000000000000000000011111111110000000000000000000000 0000000000000000000000000000000011001000000000000000000000000000 0000000000000000000000000000000011001000110000000000000000000000 0000000000000000000000000000000011001000000001100000000000000000 0000000000000000000000000000000011001001000000000000000000000000 0000000000000000000000000000000011001000000010000000000000000000 0000000000000000000000000000000011001001010000000000000000000000 0000000000000000000000000000000011001000000010100000000000000000 0000000000000000000000000000000011001001100000000000000000000000 0000000000000000000000000000000011001000000011000000000000000000 0000000000000000000000000000000011001001110000000000000000000000 0000000000000000000000000000000011001000000011100000000000000000 0000000000000000000000000000000011001010000000000000000000000000 0000000000000000000000000000000011001000000100000000000000000000 0000000000000000000000000000000011001010010000000000000000000000 0000000000000000000000000000000011001000000100100000000000000000 0000000000000000000000000000000011001010100000000000000000000000 0000000000000000000000000000000011001000000101000000000000000000 0000000000000000000000000000000011001010110000000000000000000000 0000000000000000000000000000000011001000000101100000000000000000 0000000000000000000000000000000011001011000000000000000000000000 0000000000000000000000000000000011001000000110000000000000000000 0000000000000000000000000000000011001011010000000000000000000000 0000000000000000000000000000000011001000000110100000000000000000 0000000000000000000000000000000011001100000000000000000000000000 0000000000000000000000000000000011001000001000000000000000000000 0000000000000000000000000000000011001100010000000000000000000000 0000000000000000000000000000000011001000001000100000000000000000 0000000000000000000000000000000011001100100000000000000000000000 0000000000000000000000000000000011001000001001000000000000000000 default != .fieldderef .fieldderef INST VOPD SRCX0 255 I 32 "X" opcode of a VOPD-encoded instruction. All ENC_VOPDX opcodes may be promoted to the ENC_VOPDY encoding for use as "Y" opcodes. The VOPD instruction encoding allows a single shader instruction to encode two separate VALU operations which are executed in parallel, referred to as the "X" and "Y" opcodes. The "X" opcode is issued to the VALU core MACC and the "Y" opcode is issued to the VALU side MACC. The two operations must be independent of each other. This instruction has certain restrictions which must be met --- hardware does not function correctly if they are not. In sp3 the X and Y opcodes are specified as a pair of instructions concatenated with the :: (list concatenation) operator. The first instruction uses the ENC_VOPDX encoding and the second instruction uses the ENC_VOPDY encoding. A simple example is shown below: v_dual_mul_f32 v0, v1, v2 :: v_dual_fmac_f32 v3, v4, v5 // X opcode is MUL, Y opcode is FMAC. VOPD instructions are only legal for wave32 shaders. VOPD is not allowed in a wave64 shader program. There are restrictions on what types of operands may be specified for a VOPD instruction. No VALU modifiers are allowed. Some restrictions are listed below; see the Shader Programming Guide for a complete list of restrictions. Each of the two instructions may use up to 2 VGPRs. Allowed combinations of scalar inputs for the VOPD are: 2 SGPRs, 1 SGPR and one literal constant, or both ops share the same literal constant. For CNDMASK the implicit VCC read counts against the scalar input limit. If CNDMASK is used then the instruction pair may only use one additional scalar input (SGPR or literal). Instructions must not exceed the VGPR source-cache port limits. There are 4 VGPR banks (indexed by SRC[1:0]) and each bank has a cache. Each cache has 3 read ports: one dedicated to SRC0, one dedicated to SRC1 and one for SRC2. A cache can read all 3 ports at once, but it can't read two SRC0's at once (or SRC1/2). SRC2 VGPR reads are further restricted: if both operations read SRC2 then one must be an even VGPR and the other must be an odd VGPR. For the VGPR source-cache port limit, FMAMK_F32 VSRC1 counts as a SRC2 read. For the VGPR source-cache port limit, FMAC_F32, DOT2ACC_F32_F16 and DOT2ACC_F32_BF16 VDST counts as a SRC2 read. Special case: If OPY is MOV then it uses the SRC2 read port instead of SRC0. This allows two MOV instructions in one VOPD to avoid source bank conflicts. Examples of illegal combinations: v_dual_fmamk_f32 v0, v1, 1.0, v2 :: v_dual_fmamk_f32 v1, 1.0, v4 // v2 and v4 are SRC2 and both even v_dual_fmamk_f32 v0, v1, 1.0, v3 :: v_dual_fmac_f32 v1, v4, v5 // v3 and v1 are SRC2 and both odd Instructions which use 3 operands may run at full rate if mixed with 1-operand instructions. The limitation is that each source-cache (per VGPR-bank) has only one cache probe, and wave64 must probe 2 banks per VGPR-operand. One destination VGPR must be even and the other must be odd due to restrictions in the microcode encoding. The instructions must be independent of each other. The instructions must not use DPP. Allowed patterns: VOPD (64 bits) VOPD + LITERAL (96 bits) ENCODING Encoding. 6 26 OPX Opcode X. 4 22 OPY Opcode Y. 5 17 SRCX0 First source for opcode X. 9 0 SRCY0 First source for opcode Y. 9 32 VDSTX Destination for opcode X. 8 56 VDSTY Destination for opcode Y. Bits [7:1] are stored in this field and bit 0 is ~VDSTX[0]. 7 49 1 0 VSRCX1 Second source for opcode X. 8 9 VSRCY1 Second source for opcode Y. 8 41 VOPDXY_INST_LITERAL 96 000000000000000000000000000000000000000000000000000000000000000011111111110000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000000000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000010000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000000000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000100000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000000001000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000110000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000000001100000000000000000 000000000000000000000000000000000000000000000000000000000000000011001001000000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000000010000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001001010000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000000010100000000000000000 000000000000000000000000000000000000000000000000000000000000000011001001100000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000000011000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001001110000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000000011100000000000000000 000000000000000000000000000000000000000000000000000000000000000011001010000000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000000100000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001010010000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000000100100000000000000000 000000000000000000000000000000000000000000000000000000000000000011001010100000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000000101000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001010110000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000000101100000000000000000 000000000000000000000000000000000000000000000000000000000000000011001011000000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000000110000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001011010000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000000110100000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100000000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000001000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100010000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000001000100000000000000000 000000000000000000000000000000000000000000000000000000000000000011001100100000000000000000000000 000000000000000000000000000000000000000000000000000000000000000011001000001001000000000000000000 has_lit == .fieldderef .fieldderef INST VOPD SRCX0 255 I 32 default != .fieldderef .fieldderef INST VOPD SRCX0 255 I 32 "X" opcode of a VOPD-encoded instruction. All ENC_VOPDX opcodes may be promoted to the ENC_VOPDY encoding for use as "Y" opcodes. The VOPD instruction encoding allows a single shader instruction to encode two separate VALU operations which are executed in parallel, referred to as the "X" and "Y" opcodes. The "X" opcode is issued to the VALU core MACC and the "Y" opcode is issued to the VALU side MACC. The two operations must be independent of each other. This instruction has certain restrictions which must be met --- hardware does not function correctly if they are not. In sp3 the X and Y opcodes are specified as a pair of instructions concatenated with the :: (list concatenation) operator. The first instruction uses the ENC_VOPDX encoding and the second instruction uses the ENC_VOPDY encoding. A simple example is shown below: v_dual_mul_f32 v0, v1, v2 :: v_dual_fmac_f32 v3, v4, v5 // X opcode is MUL, Y opcode is FMAC. VOPD instructions are only legal for wave32 shaders. VOPD is not allowed in a wave64 shader program. There are restrictions on what types of operands may be specified for a VOPD instruction. No VALU modifiers are allowed. Some restrictions are listed below; see the Shader Programming Guide for a complete list of restrictions. Each of the two instructions may use up to 2 VGPRs. Allowed combinations of scalar inputs for the VOPD are: 2 SGPRs, 1 SGPR and one literal constant, or both ops share the same literal constant. For CNDMASK the implicit VCC read counts against the scalar input limit. If CNDMASK is used then the instruction pair may only use one additional scalar input (SGPR or literal). Instructions must not exceed the VGPR source-cache port limits. There are 4 VGPR banks (indexed by SRC[1:0]) and each bank has a cache. Each cache has 3 read ports: one dedicated to SRC0, one dedicated to SRC1 and one for SRC2. A cache can read all 3 ports at once, but it can't read two SRC0's at once (or SRC1/2). SRC2 VGPR reads are further restricted: if both operations read SRC2 then one must be an even VGPR and the other must be an odd VGPR. For the VGPR source-cache port limit, FMAMK_F32 VSRC1 counts as a SRC2 read. For the VGPR source-cache port limit, FMAC_F32, DOT2ACC_F32_F16 and DOT2ACC_F32_BF16 VDST counts as a SRC2 read. Special case: If OPY is MOV then it uses the SRC2 read port instead of SRC0. This allows two MOV instructions in one VOPD to avoid source bank conflicts. Examples of illegal combinations: v_dual_fmamk_f32 v0, v1, 1.0, v2 :: v_dual_fmamk_f32 v1, 1.0, v4 // v2 and v4 are SRC2 and both even v_dual_fmamk_f32 v0, v1, 1.0, v3 :: v_dual_fmac_f32 v1, v4, v5 // v3 and v1 are SRC2 and both odd Instructions which use 3 operands may run at full rate if mixed with 1-operand instructions. The limitation is that each source-cache (per VGPR-bank) has only one cache probe, and wave64 must probe 2 banks per VGPR-operand. One destination VGPR must be even and the other must be odd due to restrictions in the microcode encoding. The instructions must be independent of each other. The instructions must not use DPP. Allowed patterns: VOPD (64 bits) VOPD + LITERAL (96 bits) ENCODING Encoding. 6 26 OPX Opcode X. 4 22 OPY Opcode Y. 5 17 SIMM32 32-bit literal constant. 32 64 SRCX0 First source for opcode X. 9 0 SRCY0 First source for opcode Y. 9 32 VDSTX Destination for opcode X. 8 56 VDSTY Destination for opcode Y. Bits [7:1] are stored in this field and bit 0 is ~VDSTX[0]. 7 49 1 0 VSRCX1 Second source for opcode X. 8 9 VSRCY1 Second source for opcode Y. 8 41 FALSE FALSE FALSE FALSE FALSE S_LOAD_B32 S_LOAD_DWORD Load 32 bits of data from the scalar memory into a scalar register. ENC_SMEM default 0 SDATA FMT_ANY OPR_SREG 32 SBASE FMT_BUF OPR_SREG 64 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_B32 OPR_GPUMEM 32 SMEM FALSE FALSE FALSE FALSE FALSE S_LOAD_B64 S_LOAD_DWORDX2 Load 64 bits of data from the scalar memory into a scalar register. ENC_SMEM default 1 SDATA FMT_ANY OPR_SREG 64 SBASE FMT_BUF OPR_SREG 64 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_B64 OPR_GPUMEM 64 SMEM FALSE FALSE FALSE FALSE FALSE S_LOAD_B128 S_LOAD_DWORDX4 Load 128 bits of data from the scalar memory into a scalar register. ENC_SMEM default 2 SDATA FMT_ANY OPR_SREG 128 SBASE FMT_BUF OPR_SREG 64 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_B128 OPR_GPUMEM 128 SMEM FALSE FALSE FALSE FALSE FALSE S_LOAD_B256 S_LOAD_DWORDX8 Load 256 bits of data from the scalar memory into a scalar register. ENC_SMEM default 3 SDATA FMT_ANY OPR_SREG 256 SBASE FMT_BUF OPR_SREG 64 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_B256 OPR_GPUMEM 256 SMEM FALSE FALSE FALSE FALSE FALSE S_LOAD_B512 S_LOAD_DWORDX16 Load 512 bits of data from the scalar memory into a scalar register. ENC_SMEM default 4 SDATA FMT_ANY OPR_SREG 512 SBASE FMT_BUF OPR_SREG 64 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_B512 OPR_GPUMEM 512 SMEM FALSE FALSE FALSE FALSE FALSE S_LOAD_B96 Load 96 bits of data from the scalar memory into a scalar register. ENC_SMEM default 5 SDATA FMT_ANY OPR_SREG 96 SBASE FMT_BUF OPR_SREG 64 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_B96 OPR_GPUMEM 96 SMEM FALSE FALSE FALSE FALSE FALSE S_LOAD_I8 Load 8 bits of signed data from the scalar memory, sign extend to 32 bits and store the result into a scalar register. ENC_SMEM default 8 SDATA FMT_NUM_I32 OPR_SREG 32 SBASE FMT_BUF OPR_SREG 64 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_I8 OPR_GPUMEM 8 SMEM FALSE FALSE FALSE FALSE FALSE S_LOAD_U8 Load 8 bits of unsigned data from the scalar memory, zero extend to 32 bits and store the result into a scalar register. ENC_SMEM default 9 SDATA FMT_NUM_U32 OPR_SREG 32 SBASE FMT_BUF OPR_SREG 64 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_U8 OPR_GPUMEM 8 SMEM FALSE FALSE FALSE FALSE FALSE S_LOAD_I16 Load 16 bits of signed data from the scalar memory, sign extend to 32 bits and store the result into a scalar register. ENC_SMEM default 10 SDATA FMT_NUM_I32 OPR_SREG 32 SBASE FMT_BUF OPR_SREG 64 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_I16 OPR_GPUMEM 16 SMEM FALSE FALSE FALSE FALSE FALSE S_LOAD_U16 Load 16 bits of unsigned data from the scalar memory, zero extend to 32 bits and store the result into a scalar register. ENC_SMEM default 11 SDATA FMT_NUM_U32 OPR_SREG 32 SBASE FMT_BUF OPR_SREG 64 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_U16 OPR_GPUMEM 16 SMEM FALSE FALSE FALSE FALSE FALSE S_BUFFER_LOAD_B32 S_BUFFER_LOAD_DWORD Load 32 bits of data from a scalar buffer surface into a scalar register. ENC_SMEM default 16 SDATA FMT_ANY OPR_SREG 32 SBASE FMT_RSRC_SCALAR OPR_SREG 128 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_B32 OPR_GPUMEM 32 SMEM FALSE FALSE FALSE FALSE FALSE S_BUFFER_LOAD_B64 S_BUFFER_LOAD_DWORDX2 Load 64 bits of data from a scalar buffer surface into a scalar register. ENC_SMEM default 17 SDATA FMT_ANY OPR_SREG 64 SBASE FMT_RSRC_SCALAR OPR_SREG 128 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_B64 OPR_GPUMEM 64 SMEM FALSE FALSE FALSE FALSE FALSE S_BUFFER_LOAD_B128 S_BUFFER_LOAD_DWORDX4 Load 128 bits of data from a scalar buffer surface into a scalar register. ENC_SMEM default 18 SDATA FMT_ANY OPR_SREG 128 SBASE FMT_RSRC_SCALAR OPR_SREG 128 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_B128 OPR_GPUMEM 128 SMEM FALSE FALSE FALSE FALSE FALSE S_BUFFER_LOAD_B256 S_BUFFER_LOAD_DWORDX8 Load 256 bits of data from a scalar buffer surface into a scalar register. ENC_SMEM default 19 SDATA FMT_ANY OPR_SREG 256 SBASE FMT_RSRC_SCALAR OPR_SREG 128 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_B256 OPR_GPUMEM 256 SMEM FALSE FALSE FALSE FALSE FALSE S_BUFFER_LOAD_B512 S_BUFFER_LOAD_DWORDX16 Load 512 bits of data from a scalar buffer surface into a scalar register. ENC_SMEM default 20 SDATA FMT_ANY OPR_SREG 512 SBASE FMT_RSRC_SCALAR OPR_SREG 128 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_B512 OPR_GPUMEM 512 SMEM FALSE FALSE FALSE FALSE FALSE S_BUFFER_LOAD_B96 Load 96 bits of data from a scalar buffer surface into a scalar register. ENC_SMEM default 21 SDATA FMT_ANY OPR_SREG 96 SBASE FMT_RSRC_SCALAR OPR_SREG 128 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_B96 OPR_GPUMEM 96 SMEM FALSE FALSE FALSE FALSE FALSE S_BUFFER_LOAD_I8 Load 8 bits of signed data from a scalar buffer surface, sign extend to 32 bits and store the result into a scalar register. ENC_SMEM default 24 SDATA FMT_NUM_I32 OPR_SREG 32 SBASE FMT_RSRC_SCALAR OPR_SREG 128 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_I8 OPR_GPUMEM 8 SMEM FALSE FALSE FALSE FALSE FALSE S_BUFFER_LOAD_U8 Load 8 bits of unsigned data from a scalar buffer surface, zero extend to 32 bits and store the result into a scalar register. ENC_SMEM default 25 SDATA FMT_NUM_U32 OPR_SREG 32 SBASE FMT_RSRC_SCALAR OPR_SREG 128 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_U8 OPR_GPUMEM 8 SMEM FALSE FALSE FALSE FALSE FALSE S_BUFFER_LOAD_I16 Load 16 bits of signed data from a scalar buffer surface, sign extend to 32 bits and store the result into a scalar register. ENC_SMEM default 26 SDATA FMT_NUM_I32 OPR_SREG 32 SBASE FMT_RSRC_SCALAR OPR_SREG 128 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_I16 OPR_GPUMEM 16 SMEM FALSE FALSE FALSE FALSE FALSE S_BUFFER_LOAD_U16 Load 16 bits of unsigned data from a scalar buffer surface, zero extend to 32 bits and store the result into a scalar register. ENC_SMEM default 27 SDATA FMT_NUM_U32 OPR_SREG 32 SBASE FMT_RSRC_SCALAR OPR_SREG 128 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 FMT_NUM_U16 OPR_GPUMEM 16 SMEM FALSE FALSE FALSE FALSE FALSE S_DCACHE_INV Invalidate the scalar data L0 cache. ENC_SMEM default 33 SMEM FALSE FALSE FALSE FALSE FALSE S_ATC_PROBE Probe or prefetch an address into the scalar data cache. ENC_SMEM default 34 SDATA FMT_NUM_B8 OPR_SIMM8 8 SBASE FMT_BUF OPR_SREG 64 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 SMEM FALSE FALSE FALSE FALSE FALSE S_ATC_PROBE_BUFFER Probe or prefetch an address into the scalar data cache. ENC_SMEM default 35 SDATA FMT_NUM_B8 OPR_SIMM8 8 SBASE FMT_RSRC_SCALAR OPR_SREG 128 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET 32 SMEM FALSE FALSE FALSE FALSE FALSE S_PREFETCH_INST Prefetch instructions into the shader instruction cache, relative to a base address provided. ENC_SMEM default 36 SBASE FMT_NUM_B64 OPR_SREG 64 IOFFSET FMT_NUM_B32 OPR_SIMM24 32 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET_NOK 32 SDATA FMT_NUM_B32 OPR_SIMM5 32 SMEM FALSE FALSE FALSE FALSE FALSE S_PREFETCH_INST_PC_REL Prefetch instructions into the shader instruction cache, relative to the current PC address. ENC_SMEM default 37 IOFFSET FMT_NUM_B32 OPR_SIMM24 32 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET_NOK 32 SDATA FMT_NUM_B32 OPR_SIMM5 32 SMEM FALSE FALSE FALSE FALSE FALSE S_PREFETCH_DATA Prefetch data into the scalar data cache, relative to a base address provided. ENC_SMEM default 38 SBASE FMT_NUM_B64 OPR_SREG 64 IOFFSET FMT_NUM_B32 OPR_SIMM24 32 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET_NOK 32 SDATA FMT_NUM_B32 OPR_SIMM5 32 SMEM FALSE FALSE FALSE FALSE FALSE S_BUFFER_PREFETCH_DATA Prefetch data into the scalar data cache, relative to a base address provided in a resource descriptor constant. ENC_SMEM default 39 SBASE FMT_RSRC_SCALAR OPR_SREG 128 IOFFSET FMT_NUM_B32 OPR_SIMM24 32 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET_NOK 32 SDATA FMT_NUM_B32 OPR_SIMM5 32 SMEM FALSE FALSE FALSE FALSE FALSE S_PREFETCH_DATA_PC_REL Prefetch data into the scalar data cache, relative to the current PC address. ENC_SMEM default 40 IOFFSET FMT_NUM_B32 OPR_SIMM24 32 SOFFSET FMT_NUM_B32 OPR_SMEM_OFFSET_NOK 32 SDATA FMT_NUM_B32 OPR_SIMM5 32 SMEM FALSE FALSE FALSE FALSE FALSE S_MOV_B32 Move scalar input into a scalar register. ENC_SOP1 default 0 SDST FMT_NUM_B32 OPR_SDST 32 SSRC0 FMT_NUM_B32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 0 SDST FMT_NUM_B32 OPR_SDST 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SALU FALSE FALSE FALSE FALSE FALSE S_MOV_B64 Move scalar input into a scalar register. ENC_SOP1 default 1 SDST FMT_NUM_B64 OPR_SDST 64 SSRC0 FMT_NUM_B64 OPR_SSRC 64 SOP1_INST_LITERAL has_lit_0 1 SDST FMT_NUM_B64 OPR_SDST 64 SIMM32 FMT_NUM_B64 OPR_SSRC 64 SALU FALSE FALSE FALSE FALSE FALSE S_CMOV_B32 Move scalar input into a scalar register iff SCC is nonzero. ENC_SOP1 default 2 SDST FMT_NUM_B32 OPR_SDST 32 SSRC0 FMT_NUM_B32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP1_INST_LITERAL has_lit_0 2 SDST FMT_NUM_B32 OPR_SDST 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_CMOV_B64 Move scalar input into a scalar register iff SCC is nonzero. ENC_SOP1 default 3 SDST FMT_NUM_B64 OPR_SDST 64 SSRC0 FMT_NUM_B64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP1_INST_LITERAL has_lit_0 3 SDST FMT_NUM_B64 OPR_SDST 64 SIMM32 FMT_NUM_B64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_BREV_B32 Reverse the order of bits in a scalar input and store the result into a scalar register. ENC_SOP1 default 4 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 4 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SALU FALSE FALSE FALSE FALSE FALSE S_BREV_B64 Reverse the order of bits in a scalar input and store the result into a scalar register. ENC_SOP1 default 5 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SOP1_INST_LITERAL has_lit_0 5 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SALU FALSE FALSE FALSE FALSE FALSE S_CTZ_I32_B32 S_FF1_I32_B32 Count the number of trailing "0" bits before the first "1" in a scalar input and store the result into a scalar register. Store -1 if there are no "1" bits in the input. ENC_SOP1 default 8 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 8 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SALU FALSE FALSE FALSE FALSE FALSE S_CTZ_I32_B64 S_FF1_I32_B64 Count the number of trailing "0" bits before the first "1" in a scalar input and store the result into a scalar register. Store -1 if there are no "1" bits in the input. ENC_SOP1 default 9 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SOP1_INST_LITERAL has_lit_0 9 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SALU FALSE FALSE FALSE FALSE FALSE S_CLZ_I32_U32 S_FLBIT_I32_B32 Count the number of leading "0" bits before the first "1" in a scalar input and store the result into a scalar register. Store -1 if there are no "1" bits. ENC_SOP1 default 10 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 10 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SALU FALSE FALSE FALSE FALSE FALSE S_CLZ_I32_U64 S_FLBIT_I32_B64 Count the number of leading "0" bits before the first "1" in a scalar input and store the result into a scalar register. Store -1 if there are no "1" bits. ENC_SOP1 default 11 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SOP1_INST_LITERAL has_lit_0 11 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SALU FALSE FALSE FALSE FALSE FALSE S_CLS_I32 S_FLBIT_I32 Count the number of leading bits that are the same as the sign bit of a scalar input and store the result into a scalar register. Store -1 if all input bits are the same. ENC_SOP1 default 12 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 12 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SALU FALSE FALSE FALSE FALSE FALSE S_CLS_I32_I64 S_FLBIT_I32_I64 Count the number of leading bits that are the same as the sign bit of a scalar input and store the result into a scalar register. Store -1 if all input bits are the same. ENC_SOP1 default 13 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I64 OPR_SSRC 64 SOP1_INST_LITERAL has_lit_0 13 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I64 OPR_SSRC 64 SALU FALSE FALSE FALSE FALSE FALSE S_SEXT_I32_I8 Sign extend a signed 8 bit scalar input to 32 bits and store the result into a scalar register. ENC_SOP1 default 14 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I16 OPR_SSRC 16 SOP1_INST_LITERAL has_lit_0 14 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I16 OPR_SSRC 16 SALU FALSE FALSE FALSE FALSE FALSE S_SEXT_I32_I16 Sign extend a signed 16 bit scalar input to 32 bits and store the result into a scalar register. ENC_SOP1 default 15 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I16 OPR_SSRC 16 SOP1_INST_LITERAL has_lit_0 15 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I16 OPR_SSRC 16 SALU FALSE FALSE FALSE FALSE FALSE S_BITSET0_B32 Given a bit offset in a scalar input, set the indicated bit in the destination scalar register to 0. ENC_SOP1 default 16 SDST FMT_NUM_B32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 16 SDST FMT_NUM_B32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SALU FALSE FALSE FALSE FALSE FALSE S_BITSET0_B64 Given a bit offset in a scalar input, set the indicated bit in the destination scalar register to 0. ENC_SOP1 default 17 SDST FMT_NUM_B64 OPR_SDST 64 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 17 SDST FMT_NUM_B64 OPR_SDST 64 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SALU FALSE FALSE FALSE FALSE FALSE S_BITSET1_B32 Given a bit offset in a scalar input, set the indicated bit in the destination scalar register to 1. ENC_SOP1 default 18 SDST FMT_NUM_B32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 18 SDST FMT_NUM_B32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SALU FALSE FALSE FALSE FALSE FALSE S_BITSET1_B64 Given a bit offset in a scalar input, set the indicated bit in the destination scalar register to 1. ENC_SOP1 default 19 SDST FMT_NUM_B64 OPR_SDST 64 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 19 SDST FMT_NUM_B64 OPR_SDST 64 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SALU FALSE FALSE FALSE FALSE FALSE S_BITREPLICATE_B64_B32 Substitute each bit of a 32 bit scalar input with two instances of itself and store the result into a 64 bit scalar register. ENC_SOP1 default 20 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 20 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SALU FALSE FALSE FALSE FALSE FALSE S_ABS_I32 Compute the absolute value of a scalar input, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP1 default 21 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP1_INST_LITERAL has_lit_0 21 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_BCNT0_I32_B32 Count the number of "0" bits in a scalar input, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP1 default 22 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP1_INST_LITERAL has_lit_0 22 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_BCNT0_I32_B64 Count the number of "0" bits in a scalar input, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP1 default 23 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP1_INST_LITERAL has_lit_0 23 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_BCNT1_I32_B32 Count the number of "1" bits in a scalar input, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP1 default 24 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP1_INST_LITERAL has_lit_0 24 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_BCNT1_I32_B64 Count the number of "1" bits in a scalar input, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP1 default 25 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP1_INST_LITERAL has_lit_0 25 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_QUADMASK_B32 Reduce a pixel mask from the scalar input into a quad mask, store the result in a scalar register and set SCC iff the result is nonzero. ENC_SOP1 default 26 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP1_INST_LITERAL has_lit_0 26 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_QUADMASK_B64 Reduce a pixel mask from the scalar input into a quad mask, store the result in a scalar register and set SCC iff the result is nonzero. ENC_SOP1 default 27 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP1_INST_LITERAL has_lit_0 27 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_WQM_B32 Given an active pixel mask in a scalar input, calculate whole quad mode mask for that input, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP1 default 28 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP1_INST_LITERAL has_lit_0 28 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_WQM_B64 Given an active pixel mask in a scalar input, calculate whole quad mode mask for that input, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP1 default 29 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP1_INST_LITERAL has_lit_0 29 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_NOT_B32 Calculate bitwise negation on a scalar input, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP1 default 30 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP1_INST_LITERAL has_lit_0 30 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_NOT_B64 Calculate bitwise negation on a scalar input, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP1 default 31 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP1_INST_LITERAL has_lit_0 31 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_AND_SAVEEXEC_B32 Calculate bitwise AND on the scalar input and the EXEC mask, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 32 SDST FMT_NUM_U32 OPR_SREG 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 32 SDST FMT_NUM_U32 OPR_SREG 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_AND_SAVEEXEC_B64 Calculate bitwise AND on the scalar input and the EXEC mask, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 33 SDST FMT_NUM_U64 OPR_SREG 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 33 SDST FMT_NUM_U64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_OR_SAVEEXEC_B32 Calculate bitwise OR on the scalar input and the EXEC mask, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 34 SDST FMT_NUM_U32 OPR_SREG 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 34 SDST FMT_NUM_U32 OPR_SREG 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_OR_SAVEEXEC_B64 Calculate bitwise OR on the scalar input and the EXEC mask, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 35 SDST FMT_NUM_U64 OPR_SREG 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 35 SDST FMT_NUM_U64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_XOR_SAVEEXEC_B32 Calculate bitwise XOR on the scalar input and the EXEC mask, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 36 SDST FMT_NUM_U32 OPR_SREG 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 36 SDST FMT_NUM_U32 OPR_SREG 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_XOR_SAVEEXEC_B64 Calculate bitwise XOR on the scalar input and the EXEC mask, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 37 SDST FMT_NUM_U64 OPR_SREG 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 37 SDST FMT_NUM_U64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_NAND_SAVEEXEC_B32 Calculate bitwise NAND on the scalar input and the EXEC mask, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 38 SDST FMT_NUM_U32 OPR_SREG 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 38 SDST FMT_NUM_U32 OPR_SREG 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_NAND_SAVEEXEC_B64 Calculate bitwise NAND on the scalar input and the EXEC mask, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 39 SDST FMT_NUM_U64 OPR_SREG 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 39 SDST FMT_NUM_U64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_NOR_SAVEEXEC_B32 Calculate bitwise NOR on the scalar input and the EXEC mask, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 40 SDST FMT_NUM_U32 OPR_SREG 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 40 SDST FMT_NUM_U32 OPR_SREG 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_NOR_SAVEEXEC_B64 Calculate bitwise NOR on the scalar input and the EXEC mask, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 41 SDST FMT_NUM_U64 OPR_SREG 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 41 SDST FMT_NUM_U64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_XNOR_SAVEEXEC_B32 Calculate bitwise XNOR on the scalar input and the EXEC mask, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 42 SDST FMT_NUM_U32 OPR_SREG 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 42 SDST FMT_NUM_U32 OPR_SREG 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_XNOR_SAVEEXEC_B64 Calculate bitwise XNOR on the scalar input and the EXEC mask, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 43 SDST FMT_NUM_U64 OPR_SREG 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 43 SDST FMT_NUM_U64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_AND_NOT0_SAVEEXEC_B32 S_ANDN1_SAVEEXEC_B32 Calculate bitwise AND on the EXEC mask and the negation of the scalar input, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 44 SDST FMT_NUM_U32 OPR_SREG 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 44 SDST FMT_NUM_U32 OPR_SREG 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_AND_NOT0_SAVEEXEC_B64 S_ANDN1_SAVEEXEC_B64 Calculate bitwise AND on the EXEC mask and the negation of the scalar input, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 45 SDST FMT_NUM_U64 OPR_SREG 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 45 SDST FMT_NUM_U64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_OR_NOT0_SAVEEXEC_B32 S_ORN1_SAVEEXEC_B32 Calculate bitwise OR on the EXEC mask and the negation of the scalar input, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 46 SDST FMT_NUM_U32 OPR_SREG 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 46 SDST FMT_NUM_U32 OPR_SREG 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_OR_NOT0_SAVEEXEC_B64 S_ORN1_SAVEEXEC_B64 Calculate bitwise OR on the EXEC mask and the negation of the scalar input, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 47 SDST FMT_NUM_U64 OPR_SREG 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 47 SDST FMT_NUM_U64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_AND_NOT1_SAVEEXEC_B32 S_ANDN2_SAVEEXEC_B32 Calculate bitwise AND on the scalar input and the negation of the EXEC mask, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 48 SDST FMT_NUM_U32 OPR_SREG 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 48 SDST FMT_NUM_U32 OPR_SREG 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_AND_NOT1_SAVEEXEC_B64 S_ANDN2_SAVEEXEC_B64 Calculate bitwise AND on the scalar input and the negation of the EXEC mask, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 49 SDST FMT_NUM_U64 OPR_SREG 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 49 SDST FMT_NUM_U64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_OR_NOT1_SAVEEXEC_B32 S_ORN2_SAVEEXEC_B32 Calculate bitwise OR on the scalar input and the negation of the EXEC mask, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 50 SDST FMT_NUM_U32 OPR_SREG 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 50 SDST FMT_NUM_U32 OPR_SREG 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_OR_NOT1_SAVEEXEC_B64 S_ORN2_SAVEEXEC_B64 Calculate bitwise OR on the scalar input and the negation of the EXEC mask, store the calculated result into the EXEC mask, set SCC iff the calculated result is nonzero and store the original value of the EXEC mask into the scalar destination register. ENC_SOP1 default 51 SDST FMT_NUM_U64 OPR_SREG 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 51 SDST FMT_NUM_U64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_AND_NOT0_WREXEC_B32 S_ANDN1_WREXEC_B32 Calculate bitwise AND on the EXEC mask and the negation of the scalar input, store the calculated result into the EXEC mask and also into the scalar destination register, and set SCC iff the calculated result is nonzero. ENC_SOP1 default 52 SDST FMT_NUM_U32 OPR_SREG 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 52 SDST FMT_NUM_U32 OPR_SREG 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_AND_NOT0_WREXEC_B64 S_ANDN1_WREXEC_B64 Calculate bitwise AND on the EXEC mask and the negation of the scalar input, store the calculated result into the EXEC mask and also into the scalar destination register, and set SCC iff the calculated result is nonzero. ENC_SOP1 default 53 SDST FMT_NUM_U64 OPR_SREG 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 53 SDST FMT_NUM_U64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_AND_NOT1_WREXEC_B32 S_ANDN2_WREXEC_B32 Calculate bitwise AND on the scalar input and the negation of the EXEC mask, store the calculated result into the EXEC mask and also into the scalar destination register, and set SCC iff the calculated result is nonzero. ENC_SOP1 default 54 SDST FMT_NUM_U32 OPR_SREG 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 54 SDST FMT_NUM_U32 OPR_SREG 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_AND_NOT1_WREXEC_B64 S_ANDN2_WREXEC_B64 Calculate bitwise AND on the scalar input and the negation of the EXEC mask, store the calculated result into the EXEC mask and also into the scalar destination register, and set SCC iff the calculated result is nonzero. ENC_SOP1 default 55 SDST FMT_NUM_U64 OPR_SREG 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SOP1_INST_LITERAL has_lit_0 55 SDST FMT_NUM_U64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_M64 OPR_SDST_EXEC 64 SALU FALSE FALSE FALSE FALSE FALSE S_MOVRELS_B32 Move data from a relatively-indexed scalar register into another scalar register. ENC_SOP1 default 64 SDST FMT_NUM_B32 OPR_SDST 32 SSRC0 FMT_NUM_B32 OPR_SREG 32 FMT_NUM_B32 OPR_SDST_M0 32 SALU FALSE FALSE FALSE FALSE FALSE S_MOVRELS_B64 Move data from a relatively-indexed scalar register into another scalar register. ENC_SOP1 default 65 SDST FMT_NUM_B64 OPR_SDST 64 SSRC0 FMT_NUM_B64 OPR_SREG 64 FMT_NUM_B32 OPR_SDST_M0 32 SALU FALSE FALSE FALSE FALSE FALSE S_MOVRELD_B32 Move data from a scalar input into a relatively-indexed scalar register. ENC_SOP1 default 66 SDST FMT_NUM_B32 OPR_SREG 32 SSRC0 FMT_NUM_B32 OPR_SSRC 32 FMT_NUM_B32 OPR_SDST_M0 32 SOP1_INST_LITERAL has_lit_0 66 SDST FMT_NUM_B32 OPR_SREG 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 FMT_NUM_B32 OPR_SDST_M0 32 SALU FALSE FALSE FALSE FALSE FALSE S_MOVRELD_B64 Move data from a scalar input into a relatively-indexed scalar register. ENC_SOP1 default 67 SDST FMT_NUM_B64 OPR_SREG 64 SSRC0 FMT_NUM_B64 OPR_SSRC 64 FMT_NUM_B32 OPR_SDST_M0 32 SOP1_INST_LITERAL has_lit_0 67 SDST FMT_NUM_B64 OPR_SREG 64 SIMM32 FMT_NUM_B64 OPR_SSRC 64 FMT_NUM_B32 OPR_SDST_M0 32 SALU FALSE FALSE FALSE FALSE FALSE S_MOVRELSD_2_B32 Move data from a relatively-indexed scalar register into another relatively-indexed scalar register, using different offsets for each index. ENC_SOP1 default 68 SDST FMT_NUM_B32 OPR_SREG 32 SSRC0 FMT_NUM_B32 OPR_SREG 32 FMT_NUM_B32 OPR_SDST_M0 32 SALU FALSE FALSE FALSE FALSE FALSE S_GETPC_B64 Store the address of the next instruction to a scalar register. ENC_SOP1 default 71 SDST FMT_NUM_B64 OPR_SDST 64 FMT_NUM_B64 OPR_PC 64 SALU TRUE FALSE TRUE FALSE FALSE S_SETPC_B64 Jump to an address specified in a scalar register. ENC_SOP1 default 72 SSRC0 FMT_NUM_B64 OPR_SREG 64 FMT_NUM_B64 OPR_PC 64 SALU TRUE FALSE TRUE FALSE FALSE S_SWAPPC_B64 Store the address of the next instruction to a scalar register and then jump to an address specified in the scalar input. ENC_SOP1 default 73 SDST FMT_NUM_B64 OPR_SDST 64 SSRC0 FMT_NUM_B64 OPR_SREG 64 FMT_NUM_B64 OPR_PC 64 FMT_NUM_B64 OPR_PC 64 SALU TRUE FALSE TRUE FALSE FALSE S_RFE_B64 Return from the exception handler. Clear the wave's PRIV bit and then jump to an address specified by the scalar input. ENC_SOP1 default 74 SSRC0 FMT_NUM_B64 OPR_SREG 64 FMT_NUM_B64 OPR_PC 64 SOP1_INST_LITERAL has_lit_0 74 SIMM32 FMT_NUM_B64 OPR_SREG 64 FMT_NUM_B64 OPR_PC 64 TRAP FALSE FALSE FALSE FALSE FALSE S_SENDMSG_RTN_B32 Send a message to upstream control hardware. ENC_SOP1 default 76 SDST FMT_NUM_B32 OPR_SDST 32 SSRC0 FMT_NUM_B32 OPR_SENDMSG_RTN 32 MESSAGE FALSE FALSE FALSE FALSE FALSE S_SENDMSG_RTN_B64 Send a message to upstream control hardware. ENC_SOP1 default 77 SDST FMT_NUM_B64 OPR_SDST 64 SSRC0 FMT_NUM_B32 OPR_SENDMSG_RTN 32 MESSAGE FALSE FALSE FALSE FALSE FALSE S_BARRIER_SIGNAL Signal that a wave has arrived at a barrier. The argument specifies which barrier to signal. ENC_SOP1 default 78 SSRC0 FMT_NUM_B32 OPR_SSRC_BARRIER_ID 32 SOP1_INST_LITERAL has_lit_0 78 SIMM32 FMT_NUM_B32 OPR_SSRC_BARRIER_ID 32 MESSAGE FALSE FALSE FALSE FALSE FALSE S_BARRIER_SIGNAL_ISFIRST Signal that a wave has arrived at a barrier and set SCC to indicate if this is the first wave to signal the barrier. The argument specifies which barrier to signal. ENC_SOP1 default 79 SSRC0 FMT_NUM_B32 OPR_SSRC_BARRIER_ID 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP1_INST_LITERAL has_lit_0 79 SIMM32 FMT_NUM_B32 OPR_SSRC_BARRIER_ID 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 MESSAGE FALSE FALSE FALSE FALSE FALSE S_ALLOC_VGPR Attempt to set the wave's VGPR allocation to the specified number of VGPRs. The new VGPR count may be specified as a constant or in an SGPR. The request is rounded up to the next block size so a successful allocation may include more than the requested number of VGPRs. ENC_SOP1 default 83 SSRC0 FMT_NUM_B32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP1_INST_LITERAL has_lit_0 83 SIMM32 FMT_NUM_B32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_SLEEP_VAR Cause a wave to sleep for up to ~8000 clocks, or to sleep until an external event wakes the wave up. ENC_SOP1 default 88 SSRC0 FMT_NUM_B32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 88 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SALU FALSE FALSE FALSE FALSE FALSE S_CEIL_F32 Round the single-precision float input up to next integer and store the result in floating point format into a scalar register. ENC_SOP1 default 96 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 96 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_FLOOR_F32 Round the single-precision float input down to previous integer and store the result in floating point format into a scalar register. ENC_SOP1 default 97 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 97 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_TRUNC_F32 Compute the integer part of a single-precision float input using round toward zero semantics and store the result in floating point format into a scalar register. ENC_SOP1 default 98 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 98 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_RNDNE_F32 Round the single-precision float input to the nearest even integer and store the result in floating point format into a scalar register. ENC_SOP1 default 99 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 99 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CVT_F32_I32 Convert from a signed 32-bit integer input to a single-precision float value and store the result into a scalar register. ENC_SOP1 default 100 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 100 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CVT_F32_U32 Convert from an unsigned 32-bit integer input to a single-precision float value and store the result into a scalar register. ENC_SOP1 default 101 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 101 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CVT_I32_F32 Convert from a single-precision float input to a signed 32-bit integer value and store the result into a scalar register. ENC_SOP1 default 102 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 102 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CVT_U32_F32 Convert from a single-precision float input to an unsigned 32-bit integer value and store the result into a scalar register. ENC_SOP1 default 103 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 103 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CVT_F16_F32 Convert from a single-precision float input to a half-precision float value and store the result into a scalar register. ENC_SOP1 default 104 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SOP1_INST_LITERAL has_lit_0 104 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CVT_F32_F16 Convert from a half-precision float input to a single-precision float value and store the result into a scalar register. ENC_SOP1 default 105 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SOP1_INST_LITERAL has_lit_0 105 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CVT_HI_F32_F16 Convert from a half-precision float value in the high 16 bits of a scalar input to a single-precision float value and store the result into a scalar register. ENC_SOP1 default 106 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SOP1_INST_LITERAL has_lit_0 106 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CEIL_F16 Round the half-precision float input up to next integer and store the result in floating point format into a scalar register. ENC_SOP1 default 107 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SOP1_INST_LITERAL has_lit_0 107 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_FLOOR_F16 Round the half-precision float input down to previous integer and store the result in floating point format into a scalar register. ENC_SOP1 default 108 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SOP1_INST_LITERAL has_lit_0 108 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_TRUNC_F16 Compute the integer part of a half-precision float input using round toward zero semantics and store the result in floating point format into a scalar register. ENC_SOP1 default 109 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SOP1_INST_LITERAL has_lit_0 109 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_RNDNE_F16 Round the half-precision float input to the nearest even integer and store the result in floating point format into a scalar register. ENC_SOP1 default 110 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SOP1_INST_LITERAL has_lit_0 110 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_ADD_CO_U32 S_ADD_U32 Add two unsigned 32-bit integer inputs, store the result into a scalar register and store the carry-out bit into SCC. ENC_SOP2 default 0 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 0 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 0 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 0 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_SUB_CO_U32 S_SUB_U32 Subtract the second unsigned 32-bit integer input from the first input, store the result into a scalar register and store the carry-out bit into SCC. ENC_SOP2 default 1 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 1 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 1 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 1 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_ADD_CO_I32 S_ADD_I32 Add two signed 32-bit integer inputs, store the result into a scalar register and store the carry-out bit into SCC. ENC_SOP2 default 2 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 2 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 2 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 2 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_SUB_CO_I32 S_SUB_I32 Subtract the second signed 32-bit integer input from the first input, store the result into a scalar register and store the carry-out bit into SCC. ENC_SOP2 default 3 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 3 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 3 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 3 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_ADD_CO_CI_U32 S_ADDC_U32 Add two unsigned 32-bit integer inputs and a carry-in bit from SCC, store the result into a scalar register and store the carry-out bit into SCC. ENC_SOP2 default 4 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 4 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 4 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 4 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_SUB_CO_CI_U32 S_SUBB_U32 Subtract the second unsigned 32-bit integer input from the first input, subtract the carry-in bit, store the result into a scalar register and store the carry-out bit into SCC. ENC_SOP2 default 5 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 5 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 5 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 5 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_ABSDIFF_I32 Calculate the absolute value of difference between two scalar inputs, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP2 default 6 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 6 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 6 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 6 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_LSHL_B32 Given a shift count in the second scalar input, calculate the logical shift left of the first scalar input, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP2 default 8 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 8 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 8 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 8 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_LSHL_B64 Given a shift count in the second scalar input, calculate the logical shift left of the first scalar input, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP2 default 9 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 9 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 9 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 9 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_LSHR_B32 Given a shift count in the second scalar input, calculate the logical shift right of the first scalar input, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP2 default 10 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 10 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 10 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 10 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_LSHR_B64 Given a shift count in the second scalar input, calculate the logical shift right of the first scalar input, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP2 default 11 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 11 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 11 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 11 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_ASHR_I32 Given a shift count in the second scalar input, calculate the arithmetic shift right (preserving sign bit) of the first scalar input, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP2 default 12 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 12 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 12 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 12 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_ASHR_I64 Given a shift count in the second scalar input, calculate the arithmetic shift right (preserving sign bit) of the first scalar input, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP2 default 13 SDST FMT_NUM_I64 OPR_SDST 64 SSRC0 FMT_NUM_I64 OPR_SSRC 64 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 13 SDST FMT_NUM_I64 OPR_SDST 64 SIMM32 FMT_NUM_I64 OPR_SSRC 64 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 13 SDST FMT_NUM_I64 OPR_SDST 64 SSRC0 FMT_NUM_I64 OPR_SSRC 64 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 13 SDST FMT_NUM_I64 OPR_SDST 64 SIMM32 FMT_NUM_I64 OPR_SSRC 64 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_LSHL1_ADD_U32 Calculate the logical shift left of the first input by 1, then add the second input, store the result into a scalar register and set SCC iff the summation results in an unsigned overflow. ENC_SOP2 default 14 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 14 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 14 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 14 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_LSHL2_ADD_U32 Calculate the logical shift left of the first input by 2, then add the second input, store the result into a scalar register and set SCC iff the summation results in an unsigned overflow. ENC_SOP2 default 15 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 15 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 15 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 15 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_LSHL3_ADD_U32 Calculate the logical shift left of the first input by 3, then add the second input, store the result into a scalar register and set SCC iff the summation results in an unsigned overflow. ENC_SOP2 default 16 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 16 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 16 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 16 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_LSHL4_ADD_U32 Calculate the logical shift left of the first input by 4, then add the second input, store the result into a scalar register and set SCC iff the summation results in an unsigned overflow. ENC_SOP2 default 17 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 17 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 17 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 17 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_MIN_I32 Select the minimum of two signed 32-bit integer inputs, store the selected value into a scalar register and set SCC iff the first value is selected. ENC_SOP2 default 18 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 18 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 18 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 18 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_MIN_U32 Select the minimum of two unsigned 32-bit integer inputs, store the selected value into a scalar register and set SCC iff the first value is selected. ENC_SOP2 default 19 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 19 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 19 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 19 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_MAX_I32 Select the maximum of two signed 32-bit integer inputs, store the selected value into a scalar register and set SCC iff the first value is selected. ENC_SOP2 default 20 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 20 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 20 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 20 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_MAX_U32 Select the maximum of two unsigned 32-bit integer inputs, store the selected value into a scalar register and set SCC iff the first value is selected. ENC_SOP2 default 21 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 21 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 21 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 21 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_AND_B32 Calculate bitwise AND on two scalar inputs, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP2 default 22 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 22 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 22 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 22 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_AND_B64 Calculate bitwise AND on two scalar inputs, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP2 default 23 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 23 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 23 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 23 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_OR_B32 Calculate bitwise OR on two scalar inputs, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP2 default 24 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 24 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 24 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 24 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_OR_B64 Calculate bitwise OR on two scalar inputs, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP2 default 25 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 25 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 25 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 25 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_XOR_B32 Calculate bitwise XOR on two scalar inputs, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP2 default 26 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 26 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 26 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 26 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_XOR_B64 Calculate bitwise XOR on two scalar inputs, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP2 default 27 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 27 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 27 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 27 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_NAND_B32 Calculate bitwise NAND on two scalar inputs, store the result into a scalar register and set SCC if the result is nonzero. ENC_SOP2 default 28 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 28 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 28 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 28 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_NAND_B64 Calculate bitwise NAND on two scalar inputs, store the result into a scalar register and set SCC if the result is nonzero. ENC_SOP2 default 29 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 29 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 29 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 29 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_NOR_B32 Calculate bitwise NOR on two scalar inputs, store the result into a scalar register and set SCC if the result is nonzero. ENC_SOP2 default 30 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 30 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 30 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 30 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_NOR_B64 Calculate bitwise NOR on two scalar inputs, store the result into a scalar register and set SCC if the result is nonzero. ENC_SOP2 default 31 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 31 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 31 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 31 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_XNOR_B32 Calculate bitwise XNOR on two scalar inputs, store the result into a scalar register and set SCC if the result is nonzero. ENC_SOP2 default 32 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 32 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 32 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 32 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_XNOR_B64 Calculate bitwise XNOR on two scalar inputs, store the result into a scalar register and set SCC if the result is nonzero. ENC_SOP2 default 33 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 33 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 33 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 33 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_AND_NOT1_B32 S_ANDN2_B32 Calculate bitwise AND with the first input and the negation of the second input, store the result into a scalar register and set SCC if the result is nonzero. ENC_SOP2 default 34 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 34 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 34 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 34 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_AND_NOT1_B64 S_ANDN2_B64 Calculate bitwise AND with the first input and the negation of the second input, store the result into a scalar register and set SCC if the result is nonzero. ENC_SOP2 default 35 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 35 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 35 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 35 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_OR_NOT1_B32 S_ORN2_B32 Calculate bitwise OR with the first input and the negation of the second input, store the result into a scalar register and set SCC if the result is nonzero. ENC_SOP2 default 36 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 36 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 36 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 36 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_OR_NOT1_B64 S_ORN2_B64 Calculate bitwise OR with the first input and the negation of the second input, store the result into a scalar register and set SCC if the result is nonzero. ENC_SOP2 default 37 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 37 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 37 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 37 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_BFE_U32 Extract an unsigned bitfield from the first input using field offset and size encoded in the second input, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP2 default 38 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 38 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 38 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 38 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_BFE_I32 Extract a signed bitfield from the first input using field offset and size encoded in the second input, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP2 default 39 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 39 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 39 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 39 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_BFE_U64 Extract an unsigned bitfield from the first input using field offset and size encoded in the second input, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP2 default 40 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 40 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 40 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 40 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_BFE_I64 Extract a signed bitfield from the first input using field offset and size encoded in the second input, store the result into a scalar register and set SCC iff the result is nonzero. ENC_SOP2 default 41 SDST FMT_NUM_I64 OPR_SDST 64 SSRC0 FMT_NUM_I64 OPR_SSRC 64 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 41 SDST FMT_NUM_I64 OPR_SDST 64 SIMM32 FMT_NUM_I64 OPR_SSRC 64 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 41 SDST FMT_NUM_I64 OPR_SDST 64 SSRC0 FMT_NUM_I64 OPR_SSRC 64 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 41 SDST FMT_NUM_I64 OPR_SDST 64 SIMM32 FMT_NUM_I64 OPR_SSRC 64 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_BFM_B32 Calculate a bitfield mask given a field offset and size and store the result in a scalar register. ENC_SOP2 default 42 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0 42 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_1 42 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0_has_lit_1 42 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SALU FALSE FALSE FALSE FALSE FALSE S_BFM_B64 Calculate a bitfield mask given a field offset and size and store the result in a scalar register. ENC_SOP2 default 43 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0 43 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_1 43 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0_has_lit_1 43 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SALU FALSE FALSE FALSE FALSE FALSE S_MUL_I32 Multiply two signed 32-bit integer inputs and store the result into a scalar register. ENC_SOP2 default 44 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0 44 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_1 44 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0_has_lit_1 44 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SALU FALSE FALSE FALSE FALSE FALSE S_MUL_HI_U32 Multiply two unsigned integers and store the high 32 bits of the result into a scalar register. ENC_SOP2 default 45 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0 45 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_1 45 SDST FMT_NUM_U32 OPR_SDST 32 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0_has_lit_1 45 SDST FMT_NUM_U32 OPR_SDST 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SALU FALSE FALSE FALSE FALSE FALSE S_MUL_HI_I32 Multiply two signed integers and store the high 32 bits of the result into a scalar register. ENC_SOP2 default 46 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0 46 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_1 46 SDST FMT_NUM_I32 OPR_SDST 32 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0_has_lit_1 46 SDST FMT_NUM_I32 OPR_SDST 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SALU FALSE FALSE FALSE FALSE FALSE S_CSELECT_B32 Select the first input if SCC is true otherwise select the second input, then store the selected input into a scalar register. ENC_SOP2 default 48 SDST FMT_NUM_B32 OPR_SDST 32 SSRC0 FMT_NUM_B32 OPR_SSRC 32 SSRC1 FMT_NUM_B32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 48 SDST FMT_NUM_B32 OPR_SDST 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SSRC1 FMT_NUM_B32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 48 SDST FMT_NUM_B32 OPR_SDST 32 SSRC0 FMT_NUM_B32 OPR_SSRC 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 48 SDST FMT_NUM_B32 OPR_SDST 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_CSELECT_B64 Select the first input if SCC is true otherwise select the second input, then store the selected input into a scalar register. ENC_SOP2 default 49 SDST FMT_NUM_B64 OPR_SDST 64 SSRC0 FMT_NUM_B64 OPR_SSRC 64 SSRC1 FMT_NUM_B64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0 49 SDST FMT_NUM_B64 OPR_SDST 64 SIMM32 FMT_NUM_B64 OPR_SSRC 64 SSRC1 FMT_NUM_B64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_1 49 SDST FMT_NUM_B64 OPR_SDST 64 SSRC0 FMT_NUM_B64 OPR_SSRC 64 SIMM32 FMT_NUM_B64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOP2_INST_LITERAL has_lit_0_has_lit_1 49 SDST FMT_NUM_B64 OPR_SDST 64 SIMM32 FMT_NUM_B64 OPR_SSRC 64 SIMM32 FMT_NUM_B64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_PACK_LL_B32_B16 Pack two 16-bit scalar values into a scalar register. ENC_SOP2 default 50 SDST FMT_NUM_B32 OPR_SDST 32 SSRC0 FMT_NUM_B16 OPR_SSRC 16 SSRC1 FMT_NUM_B16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0 50 SDST FMT_NUM_B32 OPR_SDST 32 SIMM32 FMT_NUM_B16 OPR_SSRC 16 SSRC1 FMT_NUM_B16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_1 50 SDST FMT_NUM_B32 OPR_SDST 32 SSRC0 FMT_NUM_B16 OPR_SSRC 16 SIMM32 FMT_NUM_B16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0_has_lit_1 50 SDST FMT_NUM_B32 OPR_SDST 32 SIMM32 FMT_NUM_B16 OPR_SSRC 16 SIMM32 FMT_NUM_B16 OPR_SSRC 16 SALU FALSE FALSE FALSE FALSE FALSE S_PACK_LH_B32_B16 Pack two 16-bit scalar values into a scalar register. ENC_SOP2 default 51 SDST FMT_NUM_B32 OPR_SDST 32 SSRC0 FMT_NUM_B16 OPR_SSRC 16 SSRC1 FMT_NUM_B32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0 51 SDST FMT_NUM_B32 OPR_SDST 32 SIMM32 FMT_NUM_B16 OPR_SSRC 16 SSRC1 FMT_NUM_B32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_1 51 SDST FMT_NUM_B32 OPR_SDST 32 SSRC0 FMT_NUM_B16 OPR_SSRC 16 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0_has_lit_1 51 SDST FMT_NUM_B32 OPR_SDST 32 SIMM32 FMT_NUM_B16 OPR_SSRC 16 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SALU FALSE FALSE FALSE FALSE FALSE S_PACK_HH_B32_B16 Pack two 16-bit scalar values into a scalar register. ENC_SOP2 default 52 SDST FMT_NUM_B32 OPR_SDST 32 SSRC0 FMT_NUM_B32 OPR_SSRC 32 SSRC1 FMT_NUM_B32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0 52 SDST FMT_NUM_B32 OPR_SDST 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SSRC1 FMT_NUM_B32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_1 52 SDST FMT_NUM_B32 OPR_SDST 32 SSRC0 FMT_NUM_B32 OPR_SSRC 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0_has_lit_1 52 SDST FMT_NUM_B32 OPR_SDST 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SALU FALSE FALSE FALSE FALSE FALSE S_PACK_HL_B32_B16 Pack two 16-bit scalar values into a scalar register. ENC_SOP2 default 53 SDST FMT_NUM_B32 OPR_SDST 32 SSRC0 FMT_NUM_B32 OPR_SSRC 32 SSRC1 FMT_NUM_B16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0 53 SDST FMT_NUM_B32 OPR_SDST 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SSRC1 FMT_NUM_B16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_1 53 SDST FMT_NUM_B32 OPR_SDST 32 SSRC0 FMT_NUM_B32 OPR_SSRC 32 SIMM32 FMT_NUM_B16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0_has_lit_1 53 SDST FMT_NUM_B32 OPR_SDST 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SIMM32 FMT_NUM_B16 OPR_SSRC 16 SALU FALSE FALSE FALSE FALSE FALSE S_ADD_F32 Add two floating point inputs and store the result into a scalar register. ENC_SOP2 default 64 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0 64 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_1 64 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0_has_lit_1 64 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_SUB_F32 Subtract the second floating point input from the first input and store the result in a scalar register. ENC_SOP2 default 65 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0 65 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_1 65 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0_has_lit_1 65 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_MIN_NUM_F32 S_MIN_F32 Select the IEEE minimumNumber() of two single-precision float inputs and store the result into a scalar register. ENC_SOP2 default 66 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0 66 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_1 66 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0_has_lit_1 66 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_MAX_NUM_F32 S_MAX_F32 Select the IEEE maximumNumber() of two single-precision float inputs and store the result into a scalar register. ENC_SOP2 default 67 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0 67 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_1 67 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0_has_lit_1 67 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_MUL_F32 Multiply two floating point inputs and store the result into a scalar register. ENC_SOP2 default 68 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0 68 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_1 68 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0_has_lit_1 68 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_FMAAK_F32 Multiply two floating point inputs and add a literal constant using fused multiply add, and store the result into a scalar register. SOP2_INST_LITERAL default 69 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_F32 OPR_SIMM32 32 SOP2_INST_LITERAL has_lit_0 69 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_F32 OPR_SIMM32 32 SOP2_INST_LITERAL has_lit_1 69 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_F32 OPR_SIMM32 32 SOP2_INST_LITERAL has_lit_0_has_lit_1 69 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_F32 OPR_SIMM32 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_FMAMK_F32 Multiply a floating point input with a literal constant and add a second floating point input using fused multiply add, and store the result into a scalar register. SOP2_INST_LITERAL default 70 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_F32 OPR_SIMM32 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0 70 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_F32 OPR_SIMM32 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_1 70 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SIMM32 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0_has_lit_1 70 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SIMM32 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_FMAC_F32 Compute the fused multiply add of floating point inputs and accumulate with the destination operand, and store the result into the destination. ENC_SOP2 default 71 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0 71 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_1 71 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0_has_lit_1 71 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CVT_PK_RTZ_F16_F32 Convert two single-precision float inputs into a packed half-precision float result using round toward zero semantics (ignore the current rounding mode), and store the result into a scalar register. ENC_SOP2 default 72 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0 72 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_1 72 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0_has_lit_1 72 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_ADD_F16 Add two floating point inputs and store the result into a scalar register. ENC_SOP2 default 73 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0 73 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_1 73 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0_has_lit_1 73 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_SUB_F16 Subtract the second floating point input from the first input and store the result in a scalar register. ENC_SOP2 default 74 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0 74 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_1 74 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0_has_lit_1 74 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_MIN_NUM_F16 S_MIN_F16 Select the IEEE minimumNumber() of two half-precision float inputs and store the result into a scalar register. ENC_SOP2 default 75 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0 75 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_1 75 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0_has_lit_1 75 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_MAX_NUM_F16 S_MAX_F16 Select the IEEE maximumNumber() of two half-precision float inputs and store the result into a scalar register. ENC_SOP2 default 76 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0 76 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_1 76 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0_has_lit_1 76 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_MUL_F16 Multiply two floating point inputs and store the result into a scalar register. ENC_SOP2 default 77 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0 77 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_1 77 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0_has_lit_1 77 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_FMAC_F16 Compute the fused multiply add of floating point inputs and accumulate with the destination operand, and store the result into the destination. ENC_SOP2 default 78 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0 78 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_1 78 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0_has_lit_1 78 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_MINIMUM_F32 Select the IEEE minimum() of two single-precision float inputs and store the result into a scalar register. ENC_SOP2 default 79 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0 79 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_1 79 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0_has_lit_1 79 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_MAXIMUM_F32 Select the IEEE maximum() of two single-precision float inputs and store the result into a scalar register. ENC_SOP2 default 80 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0 80 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_1 80 SDST FMT_NUM_F32 OPR_SDST 32 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SOP2_INST_LITERAL has_lit_0_has_lit_1 80 SDST FMT_NUM_F32 OPR_SDST 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_MINIMUM_F16 Select the IEEE minimum() of two half-precision float inputs and store the result into a scalar register. ENC_SOP2 default 81 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0 81 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_1 81 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0_has_lit_1 81 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_MAXIMUM_F16 Select the IEEE maximum() of two half-precision float inputs and store the result into a scalar register. ENC_SOP2 default 82 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0 82 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_1 82 SDST FMT_NUM_F16 OPR_SDST 16 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SOP2_INST_LITERAL has_lit_0_has_lit_1 82 SDST FMT_NUM_F16 OPR_SDST 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_ADD_NC_U64 S_ADD_U64 Add two unsigned 64-bit integer inputs and store the result into a scalar register. ENC_SOP2 default 83 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 SOP2_INST_LITERAL has_lit_0 83 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 SOP2_INST_LITERAL has_lit_1 83 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SOP2_INST_LITERAL has_lit_0_has_lit_1 83 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SALU FALSE FALSE FALSE FALSE FALSE S_SUB_NC_U64 S_SUB_U64 Subtract the second unsigned 64-bit integer input from the first input and store the result into a scalar register. ENC_SOP2 default 84 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 SOP2_INST_LITERAL has_lit_0 84 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 SOP2_INST_LITERAL has_lit_1 84 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SOP2_INST_LITERAL has_lit_0_has_lit_1 84 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SALU FALSE FALSE FALSE FALSE FALSE S_MUL_U64 Multiply two unsigned 64-bit integer inputs and store the result into a scalar register. ENC_SOP2 default 85 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 SOP2_INST_LITERAL has_lit_0 85 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 SOP2_INST_LITERAL has_lit_1 85 SDST FMT_NUM_U64 OPR_SDST 64 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SOP2_INST_LITERAL has_lit_0_has_lit_1 85 SDST FMT_NUM_U64 OPR_SDST 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SALU FALSE FALSE FALSE FALSE FALSE S_CMP_EQ_I32 Set SCC to 1 iff the first scalar input is equal to the second scalar input. ENC_SOPC default 0 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 0 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 0 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 0 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_CMP_LG_I32 Set SCC to 1 iff the first scalar input is less than or greater than the second scalar input. ENC_SOPC default 1 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 1 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 1 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 1 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_CMP_GT_I32 Set SCC to 1 iff the first scalar input is greater than the second scalar input. ENC_SOPC default 2 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 2 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 2 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 2 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_CMP_GE_I32 Set SCC to 1 iff the first scalar input is greater than or equal to the second scalar input. ENC_SOPC default 3 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 3 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 3 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 3 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_CMP_LT_I32 Set SCC to 1 iff the first scalar input is less than the second scalar input. ENC_SOPC default 4 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 4 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 4 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 4 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_CMP_LE_I32 Set SCC to 1 iff the first scalar input is less than or equal to the second scalar input. ENC_SOPC default 5 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 5 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SSRC1 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 5 SSRC0 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 5 SIMM32 FMT_NUM_I32 OPR_SSRC 32 SIMM32 FMT_NUM_I32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_CMP_EQ_U32 Set SCC to 1 iff the first scalar input is equal to the second scalar input. ENC_SOPC default 6 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 6 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 6 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 6 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_CMP_LG_U32 Set SCC to 1 iff the first scalar input is less than or greater than the second scalar input. ENC_SOPC default 7 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 7 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 7 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 7 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_CMP_GT_U32 Set SCC to 1 iff the first scalar input is greater than the second scalar input. ENC_SOPC default 8 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 8 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 8 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 8 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_CMP_GE_U32 Set SCC to 1 iff the first scalar input is greater than or equal to the second scalar input. ENC_SOPC default 9 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 9 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 9 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 9 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_CMP_LT_U32 Set SCC to 1 iff the first scalar input is less than the second scalar input. ENC_SOPC default 10 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 10 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 10 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 10 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_CMP_LE_U32 Set SCC to 1 iff the first scalar input is less than or equal to the second scalar input. ENC_SOPC default 11 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 11 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 11 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 11 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_BITCMP0_B32 Extract a bit from the first scalar input based on an index in the second scalar input, and set SCC to 1 iff the extracted bit is equal to 0. ENC_SOPC default 12 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 12 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 12 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 12 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_BITCMP1_B32 Extract a bit from the first scalar input based on an index in the second scalar input, and set SCC to 1 iff the extracted bit is equal to 1. ENC_SOPC default 13 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 13 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 13 SSRC0 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 13 SIMM32 FMT_NUM_U32 OPR_SSRC 32 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_BITCMP0_B64 Extract a bit from the first scalar input based on an index in the second scalar input, and set SCC to 1 iff the extracted bit is equal to 0. ENC_SOPC default 14 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 14 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 14 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 14 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_BITCMP1_B64 Extract a bit from the first scalar input based on an index in the second scalar input, and set SCC to 1 iff the extracted bit is equal to 1. ENC_SOPC default 15 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 15 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 15 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 15 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_CMP_EQ_U64 Set SCC to 1 iff the first scalar input is equal to the second scalar input. ENC_SOPC default 16 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 16 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 16 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 16 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_CMP_LG_U64 Set SCC to 1 iff the first scalar input is less than or greater than the second scalar input. ENC_SOPC default 17 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 17 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SSRC1 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 17 SSRC0 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 17 SIMM32 FMT_NUM_U64 OPR_SSRC 64 SIMM32 FMT_NUM_U64 OPR_SSRC 64 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_CMP_LT_F32 Set SCC to 1 iff the first scalar input is less than the second scalar input. ENC_SOPC default 65 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 65 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 65 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 65 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_LT_F16 Set SCC to 1 iff the first scalar input is less than the second scalar input. ENC_SOPC default 81 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 81 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 81 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 81 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_EQ_F32 Set SCC to 1 iff the first scalar input is equal to the second scalar input. ENC_SOPC default 66 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 66 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 66 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 66 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_EQ_F16 Set SCC to 1 iff the first scalar input is equal to the second scalar input. ENC_SOPC default 82 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 82 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 82 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 82 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_LE_F32 Set SCC to 1 iff the first scalar input is less than or equal to the second scalar input. ENC_SOPC default 67 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 67 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 67 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 67 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_LE_F16 Set SCC to 1 iff the first scalar input is less than or equal to the second scalar input. ENC_SOPC default 83 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 83 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 83 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 83 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_GT_F32 Set SCC to 1 iff the first scalar input is greater than the second scalar input. ENC_SOPC default 68 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 68 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 68 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 68 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_GT_F16 Set SCC to 1 iff the first scalar input is greater than the second scalar input. ENC_SOPC default 84 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 84 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 84 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 84 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_LG_F32 Set SCC to 1 iff the first scalar input is less than or greater than the second scalar input. ENC_SOPC default 69 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 69 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 69 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 69 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_LG_F16 Set SCC to 1 iff the first scalar input is less than or greater than the second scalar input. ENC_SOPC default 85 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 85 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 85 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 85 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_GE_F32 Set SCC to 1 iff the first scalar input is greater than or equal to the second scalar input. ENC_SOPC default 70 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 70 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 70 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 70 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_GE_F16 Set SCC to 1 iff the first scalar input is greater than or equal to the second scalar input. ENC_SOPC default 86 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 86 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 86 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 86 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_O_F32 Set SCC to 1 iff the first scalar input is orderable to the second scalar input. ENC_SOPC default 71 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 71 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 71 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 71 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_O_F16 Set SCC to 1 iff the first scalar input is orderable to the second scalar input. ENC_SOPC default 87 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 87 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 87 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 87 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_U_F32 Set SCC to 1 iff the first scalar input is not orderable to the second scalar input. ENC_SOPC default 72 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 72 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 72 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 72 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_U_F16 Set SCC to 1 iff the first scalar input is not orderable to the second scalar input. ENC_SOPC default 88 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 88 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 88 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 88 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_NGE_F32 Set SCC to 1 iff the first scalar input is not greater than or equal to the second scalar input. ENC_SOPC default 73 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 73 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 73 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 73 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_NGE_F16 Set SCC to 1 iff the first scalar input is not greater than or equal to the second scalar input. ENC_SOPC default 89 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 89 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 89 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 89 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_NLG_F32 Set SCC to 1 iff the first scalar input is not less than or greater than the second scalar input. ENC_SOPC default 74 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 74 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 74 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 74 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_NLG_F16 Set SCC to 1 iff the first scalar input is not less than or greater than the second scalar input. ENC_SOPC default 90 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 90 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 90 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 90 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_NGT_F32 Set SCC to 1 iff the first scalar input is not greater than the second scalar input. ENC_SOPC default 75 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 75 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 75 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 75 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_NGT_F16 Set SCC to 1 iff the first scalar input is not greater than the second scalar input. ENC_SOPC default 91 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 91 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 91 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 91 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_NLE_F32 Set SCC to 1 iff the first scalar input is not less than or equal to the second scalar input. ENC_SOPC default 76 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 76 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 76 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 76 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_NLE_F16 Set SCC to 1 iff the first scalar input is not less than or equal to the second scalar input. ENC_SOPC default 92 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 92 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 92 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 92 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_NEQ_F32 Set SCC to 1 iff the first scalar input is not equal to the second scalar input. ENC_SOPC default 77 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 77 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 77 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 77 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_NEQ_F16 Set SCC to 1 iff the first scalar input is not equal to the second scalar input. ENC_SOPC default 93 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 93 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 93 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 93 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_NLT_F32 Set SCC to 1 iff the first scalar input is not less than the second scalar input. ENC_SOPC default 78 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 78 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SSRC1 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 78 SSRC0 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 78 SIMM32 FMT_NUM_F32 OPR_SSRC 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_CMP_NLT_F16 Set SCC to 1 iff the first scalar input is not less than the second scalar input. ENC_SOPC default 94 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0 94 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SSRC1 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_1 94 SSRC0 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SOPC_INST_LITERAL has_lit_0_has_lit_1 94 SIMM32 FMT_NUM_F16 OPR_SSRC 16 SIMM32 FMT_NUM_F16 OPR_SSRC 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FLOATING_POINT FALSE FALSE FALSE FALSE FALSE S_MOVK_I32 Sign extend a literal 16-bit constant and store the result into a scalar register. ENC_SOPK default 0 SDST FMT_NUM_I32 OPR_SDST 32 SIMM16 FMT_NUM_I16 OPR_SIMM16 16 SALU FALSE FALSE FALSE FALSE TRUE S_VERSION Do nothing. This opcode is used to specify the microcode version for tools that interpret shader microcode. ENC_SOPK default 1 SIMM16 FMT_NUM_B16 OPR_VERSION 16 WAVE_CONTROL FALSE FALSE FALSE FALSE FALSE S_CMOVK_I32 Move the sign extension of a literal 16-bit constant into a scalar register iff SCC is nonzero. ENC_SOPK default 2 SDST FMT_NUM_I32 OPR_SDST 32 SIMM16 FMT_NUM_I16 OPR_SIMM16 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_ADDK_CO_I32 S_ADDK_I32 Add a scalar input and the sign extension of a literal 16-bit constant, store the result into a scalar register and store the carry-out bit into SCC. ENC_SOPK default 15 SDST FMT_NUM_I32 OPR_SDST 32 SIMM16 FMT_NUM_I16 OPR_SIMM16 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 SALU FALSE FALSE FALSE FALSE FALSE S_MULK_I32 Multiply a scalar input with the sign extension of a literal 16-bit constant and store the result into a scalar register. ENC_SOPK default 16 SDST FMT_NUM_I32 OPR_SDST 32 SIMM16 FMT_NUM_I16 OPR_SIMM16 16 SALU FALSE FALSE FALSE FALSE FALSE S_GETREG_B32 Read some or all of a hardware register into the LSBs of destination. ENC_SOPK default 17 SDST FMT_NUM_B32 OPR_SDST 32 SIMM16 FMT_NUM_B16 OPR_HWREG 16 SALU FALSE FALSE FALSE FALSE FALSE S_SETREG_B32 Write some or all of the LSBs of source argument into a hardware register. ENC_SOPK default 18 SIMM16 FMT_NUM_B16 OPR_HWREG 16 SDST FMT_NUM_B32 OPR_SDST 32 SALU FALSE FALSE FALSE FALSE FALSE S_SETREG_IMM32_B32 Write some or all of the LSBs of a 32-bit literal constant into a hardware register; this instruction requires a 32-bit literal constant. SOPK_INST_LITERAL default 19 SIMM16 FMT_NUM_B16 OPR_HWREG 16 FMT_NUM_B32 OPR_SIMM32 32 SALU TRUE FALSE FALSE FALSE FALSE S_CALL_B64 Store the address of the next instruction to a scalar register and then jump to a constant offset relative to the current PC. ENC_SOPK default 20 SDST FMT_NUM_B64 OPR_SDST 64 SIMM16 FMT_NUM_I16 OPR_LABEL 16 FMT_NUM_B64 OPR_PC 64 FMT_NUM_B64 OPR_PC 64 SALU FALSE FALSE FALSE FALSE TRUE S_NOP Do nothing. Delay issue of next instruction by a small, fixed amount. ENC_SOPP default 0 SIMM16 FMT_NUM_B16 OPR_SIMM16 16 WAVE_CONTROL FALSE FALSE FALSE FALSE TRUE S_SETKILL Kill this wave if the least significant bit of the immediate constant is 1. ENC_SOPP default 1 SIMM16 FMT_NUM_B16 OPR_SIMM16 16 WAVE_CONTROL FALSE FALSE FALSE FALSE TRUE S_SETHALT Set or clear the HALT or FATAL_HALT status bits. ENC_SOPP default 2 SIMM16 FMT_NUM_B16 OPR_SIMM16 16 WAVE_CONTROL FALSE FALSE FALSE FALSE TRUE S_SLEEP Cause a wave to sleep for up to ~8000 clocks, or to sleep until an external event wakes the wave up. ENC_SOPP default 3 SIMM16 FMT_NUM_B16 OPR_SLEEP 16 WAVE_CONTROL FALSE FALSE FALSE FALSE TRUE S_CLAUSE Mark the beginning of a clause. ENC_SOPP default 5 SIMM16 FMT_NUM_B16 OPR_CLAUSE 16 WAVE_CONTROL FALSE FALSE FALSE FALSE TRUE S_DELAY_ALU Insert delay between dependent SALU/VALU instructions. ENC_SOPP default 7 SIMM16 FMT_NUM_B16 OPR_DELAY 16 WAVE_CONTROL FALSE FALSE FALSE FALSE TRUE S_WAIT_ALU S_WAITCNT_DEPCTR Wait for one or more ALU-centric counters to fall below specified values. ENC_SOPP default 8 SIMM16 FMT_NUM_B16 OPR_WAIT_ALU 16 WAVE_CONTROL FALSE FALSE FALSE FALSE TRUE S_WAIT_IDLE Wait for all activity in the wave to be complete (all dependency and memory counters at zero). ENC_SOPP default 10 WAVE_CONTROL FALSE FALSE FALSE FALSE TRUE S_WAIT_EVENT Wait for an event to occur or a condition to be satisfied before continuing. The SIMM16 argument specifies which event(s) to wait on. ENC_SOPP default 11 SIMM16 FMT_NUM_B16 OPR_WAIT_EVENT 16 WAVE_CONTROL TRUE FALSE TRUE FALSE FALSE S_TRAP Enter the trap handler. ENC_SOPP default 16 SIMM16 FMT_NUM_B16 OPR_SIMM16 16 TRAP FALSE FALSE FALSE FALSE FALSE S_ROUND_MODE Set floating point round mode using an immediate constant. ENC_SOPP default 17 SIMM16 FMT_NUM_B16 OPR_SIMM16 16 SALU FALSE FALSE FALSE FALSE FALSE S_DENORM_MODE Set floating point denormal mode using an immediate constant. ENC_SOPP default 18 SIMM16 FMT_NUM_B16 OPR_SIMM16 16 SALU FALSE FALSE FALSE FALSE TRUE S_BARRIER_WAIT Wait for a barrier to complete. The SIMM16 argument specifies which barrier to wait on. ENC_SOPP default 20 SIMM16 FMT_NUM_B16 OPR_SIMM16 16 WAVE_CONTROL TRUE FALSE TRUE FALSE FALSE S_CODE_END Generate an illegal instruction interrupt. This instruction is used to mark the end of a shader buffer for debug tools. ENC_SOPP default 31 SALU TRUE FALSE FALSE FALSE FALSE S_BRANCH Jump to a constant offset relative to the current PC. ENC_SOPP default 32 SIMM16 FMT_NUM_I16 OPR_LABEL 16 BRANCH STATIC TRUE TRUE FALSE FALSE FALSE S_CBRANCH_SCC0 If SCC is 0 then jump to a constant offset relative to the current PC. ENC_SOPP default 33 SIMM16 FMT_NUM_I16 OPR_LABEL 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 BRANCH STATIC TRUE TRUE FALSE FALSE FALSE S_CBRANCH_SCC1 If SCC is 1 then jump to a constant offset relative to the current PC. ENC_SOPP default 34 SIMM16 FMT_NUM_I16 OPR_LABEL 16 FMT_NUM_B1 OPR_SSRC_SPECIAL_SCC 1 BRANCH STATIC TRUE TRUE FALSE FALSE FALSE S_CBRANCH_VCCZ If VCCZ is 1 then jump to a constant offset relative to the current PC. ENC_SOPP default 35 SIMM16 FMT_NUM_I16 OPR_LABEL 16 FMT_NUM_M64 OPR_VCC 64 BRANCH STATIC TRUE TRUE FALSE FALSE FALSE S_CBRANCH_VCCNZ If VCCZ is 0 then jump to a constant offset relative to the current PC. ENC_SOPP default 36 SIMM16 FMT_NUM_I16 OPR_LABEL 16 FMT_NUM_M64 OPR_VCC 64 BRANCH STATIC TRUE TRUE FALSE FALSE FALSE S_CBRANCH_EXECZ If EXECZ is 1 then jump to a constant offset relative to the current PC. ENC_SOPP default 37 SIMM16 FMT_NUM_I16 OPR_LABEL 16 FMT_NUM_M64 OPR_SDST_EXEC 64 BRANCH STATIC TRUE TRUE FALSE FALSE FALSE S_CBRANCH_EXECNZ If EXECZ is 0 then jump to a constant offset relative to the current PC. ENC_SOPP default 38 SIMM16 FMT_NUM_I16 OPR_LABEL 16 FMT_NUM_M64 OPR_SDST_EXEC 64 BRANCH STATIC FALSE FALSE FALSE TRUE FALSE S_ENDPGM End of program; terminate wavefront. ENC_SOPP default 48 MESSAGE FALSE FALSE FALSE TRUE FALSE S_ENDPGM_SAVED End of program; signal that a wave has been saved by the context-switch trap handler and terminate wavefront. ENC_SOPP default 49 MESSAGE FALSE FALSE FALSE FALSE FALSE S_WAKEUP Allow a wave to 'ping' all the other waves in its threadgroup to force them to wake up early from an S_SLEEP instruction. ENC_SOPP default 52 BRANCH STATIC FALSE FALSE FALSE FALSE FALSE S_SETPRIO Change wave user priority. ENC_SOPP default 53 SIMM16 FMT_NUM_B16 OPR_SIMM16 16 BRANCH STATIC FALSE FALSE FALSE FALSE FALSE S_SENDMSG Send a message to upstream control hardware. ENC_SOPP default 54 SIMM16 FMT_NUM_B16 OPR_SENDMSG 16 FMT_NUM_B32 OPR_SDST_M0 32 MESSAGE FALSE FALSE FALSE FALSE FALSE S_SENDMSGHALT Send a message to upstream control hardware and then HALT the wavefront; see S_SENDMSG for details. ENC_SOPP default 55 SIMM16 FMT_NUM_B16 OPR_SENDMSG 16 FMT_NUM_B32 OPR_SDST_M0 32 MESSAGE FALSE FALSE FALSE FALSE FALSE S_INCPERFLEVEL Increment performance counter specified in SIMM16[3:0] by 1. ENC_SOPP default 56 SIMM16 FMT_NUM_B16 OPR_SIMM16 16 BRANCH STATIC FALSE FALSE FALSE FALSE FALSE S_DECPERFLEVEL Decrement performance counter specified in SIMM16[3:0] by 1. ENC_SOPP default 57 SIMM16 FMT_NUM_B16 OPR_SIMM16 16 BRANCH STATIC FALSE FALSE FALSE FALSE FALSE S_TTRACEDATA Send M0 as user data to the thread trace stream. ENC_SOPP default 58 FMT_NUM_B32 OPR_SDST_M0 32 BRANCH STATIC FALSE FALSE FALSE FALSE FALSE S_TTRACEDATA_IMM Send SIMM16[7:0] as user data to the thread trace stream. ENC_SOPP default 59 SIMM16 FMT_NUM_B16 OPR_SIMM16 16 FMT_NUM_B32 OPR_SDST_M0 32 BRANCH STATIC FALSE FALSE FALSE FALSE FALSE S_ICACHE_INV Invalidate entire first level instruction cache. ENC_SOPP default 60 BRANCH STATIC FALSE FALSE FALSE FALSE TRUE S_WAIT_LOADCNT Wait until LOADCNT is less than or equal to SIMM16[5:0]. ENC_SOPP default 64 SIMM16 FMT_NUM_B16 OPR_SIMM16 16 WAVE_CONTROL FALSE FALSE FALSE FALSE TRUE S_WAIT_STORECNT Wait until STORECNT is less than or equal to SIMM16[5:0]. ENC_SOPP default 65 SIMM16 FMT_NUM_B16 OPR_SIMM16 16 WAVE_CONTROL FALSE FALSE FALSE FALSE TRUE S_WAIT_SAMPLECNT Wait until SAMPLECNT is less than or equal to SIMM16[5:0]. ENC_SOPP default 66 SIMM16 FMT_NUM_B16 OPR_SIMM16 16 WAVE_CONTROL FALSE FALSE FALSE FALSE TRUE S_WAIT_BVHCNT Wait until BVHCNT is less than or equal to SIMM16[2:0]. ENC_SOPP default 67 SIMM16 FMT_NUM_B16 OPR_SIMM16 16 WAVE_CONTROL FALSE FALSE FALSE FALSE TRUE S_WAIT_EXPCNT Wait until EXPCNT is less than or equal to SIMM16[2:0]. ENC_SOPP default 68 SIMM16 FMT_NUM_B16 OPR_SIMM16 16 WAVE_CONTROL FALSE FALSE FALSE FALSE TRUE S_WAIT_DSCNT Wait until DSCNT is less than or equal to SIMM16[5:0]. ENC_SOPP default 70 SIMM16 FMT_NUM_B16 OPR_SIMM16 16 WAVE_CONTROL FALSE FALSE FALSE FALSE TRUE S_WAIT_KMCNT Wait until KMCNT is less than or equal to SIMM16[4:0]. ENC_SOPP default 71 SIMM16 FMT_NUM_B16 OPR_SIMM16 16 WAVE_CONTROL FALSE FALSE FALSE FALSE TRUE S_WAIT_LOADCNT_DSCNT Wait until LOADCNT is less than or equal to SIMM16[13:8] and DSCNT is less than or equal to SIMM16[5:0]. ENC_SOPP default 72 SIMM16 FMT_NUM_B16 OPR_WAIT_MEM_DS 16 WAVE_CONTROL FALSE FALSE FALSE FALSE TRUE S_WAIT_STORECNT_DSCNT Wait until STORECNT is less than or equal to SIMM16[13:8] and DSCNT is less than or equal to SIMM16[5:0]. ENC_SOPP default 73 SIMM16 FMT_NUM_B16 OPR_WAIT_MEM_DS 16 WAVE_CONTROL FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_FORMAT_X Load 1-component formatted data from a buffer surface, convert the data to 32 bit integral or floating point format, then store the result into a vector register. The resource descriptor specifies the data format of the surface. ENC_VBUFFER default 0 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM LOAD FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_FORMAT_XY Load 2-component formatted data from a buffer surface, convert the data to 32 bit integral or floating point format, then store the result into a vector register. The resource descriptor specifies the data format of the surface. ENC_VBUFFER default 1 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B64 OPR_GPUMEM 64 VMEM LOAD FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_FORMAT_XYZ Load 3-component formatted data from a buffer surface, convert the data to 32 bit integral or floating point format, then store the result into a vector register. The resource descriptor specifies the data format of the surface. ENC_VBUFFER default 2 VDATA FMT_ANY OPR_VGPR 96 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B96 OPR_GPUMEM 96 VMEM LOAD FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_FORMAT_XYZW Load 4-component formatted data from a buffer surface, convert the data to 32 bit integral or floating point format, then store the result into a vector register. The resource descriptor specifies the data format of the surface. ENC_VBUFFER default 3 VDATA FMT_ANY OPR_VGPR 128 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B128 OPR_GPUMEM 128 VMEM LOAD FALSE FALSE FALSE FALSE FALSE BUFFER_STORE_FORMAT_X Convert 32 bits of data from vector input registers into 1-component formatted data and store the data into a buffer surface. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 4 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM STORE FALSE FALSE FALSE FALSE FALSE BUFFER_STORE_FORMAT_XY Convert 64 bits of data from vector input registers into 2-component formatted data and store the data into a buffer surface. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 5 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B64 OPR_GPUMEM 64 VMEM STORE FALSE FALSE FALSE FALSE FALSE BUFFER_STORE_FORMAT_XYZ Convert 96 bits of data from vector input registers into 3-component formatted data and store the data into a buffer surface. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 6 VDATA FMT_ANY OPR_VGPR 96 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B96 OPR_GPUMEM 96 VMEM STORE FALSE FALSE FALSE FALSE FALSE BUFFER_STORE_FORMAT_XYZW Convert 128 bits of data from vector input registers into 4-component formatted data and store the data into a buffer surface. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 7 VDATA FMT_ANY OPR_VGPR 128 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B128 OPR_GPUMEM 128 VMEM STORE FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_D16_FORMAT_X BUFFER_LOAD_FORMAT_D16_X Load 1-component formatted data from a buffer surface, convert the data to packed 16 bit integral or floating point format, then store the result into the low 16 bits of a 32-bit vector register. The resource descriptor specifies the data format of the surface. ENC_VBUFFER default 8 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM LOAD FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_D16_FORMAT_XY BUFFER_LOAD_FORMAT_D16_XY Load 2-component formatted data from a buffer surface, convert the data to packed 16 bit integral or floating point format, then store the result into a vector register. The resource descriptor specifies the data format of the surface. ENC_VBUFFER default 9 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B64 OPR_GPUMEM 64 VMEM LOAD FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_D16_FORMAT_XYZ BUFFER_LOAD_FORMAT_D16_XYZ Load 3-component formatted data from a buffer surface, convert the data to packed 16 bit integral or floating point format, then store the result into a vector register. The resource descriptor specifies the data format of the surface. ENC_VBUFFER default 10 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B96 OPR_GPUMEM 96 VMEM LOAD FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_D16_FORMAT_XYZW BUFFER_LOAD_FORMAT_D16_XYZW Load 4-component formatted data from a buffer surface, convert the data to packed 16 bit integral or floating point format, then store the result into a vector register. The resource descriptor specifies the data format of the surface. ENC_VBUFFER default 11 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B128 OPR_GPUMEM 128 VMEM LOAD FALSE FALSE FALSE FALSE FALSE BUFFER_STORE_D16_FORMAT_X BUFFER_STORE_FORMAT_D16_X Convert 16 bits of data from the low 16 bits of a 32-bit vector input register into 1-component formatted data and store the data into a buffer surface. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 12 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM STORE FALSE FALSE FALSE FALSE FALSE BUFFER_STORE_D16_FORMAT_XY BUFFER_STORE_FORMAT_D16_XY Convert 32 bits of data from vector input registers into 2-component formatted data and store the data into a buffer surface. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 13 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B64 OPR_GPUMEM 64 VMEM STORE FALSE FALSE FALSE FALSE FALSE BUFFER_STORE_D16_FORMAT_XYZ BUFFER_STORE_FORMAT_D16_XYZ Convert 48 bits of data from vector input registers into 3-component formatted data and store the data into a buffer surface. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 14 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B96 OPR_GPUMEM 96 VMEM STORE FALSE FALSE FALSE FALSE FALSE BUFFER_STORE_D16_FORMAT_XYZW BUFFER_STORE_FORMAT_D16_XYZW Convert 64 bits of data from vector input registers into 4-component formatted data and store the data into a buffer surface. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 15 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B128 OPR_GPUMEM 128 VMEM STORE FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_U8 BUFFER_LOAD_UBYTE Load 8 bits of unsigned data from a buffer surface, zero extend to 32 bits and store the result into a vector register. ENC_VBUFFER default 16 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U8 OPR_GPUMEM 8 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_I8 BUFFER_LOAD_SBYTE Load 8 bits of signed data from a buffer surface, sign extend to 32 bits and store the result into a vector register. ENC_VBUFFER default 17 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_I8 OPR_GPUMEM 8 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_U16 BUFFER_LOAD_USHORT Load 16 bits of unsigned data from a buffer surface, zero extend to 32 bits and store the result into a vector register. ENC_VBUFFER default 18 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U16 OPR_GPUMEM 16 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_I16 BUFFER_LOAD_SSHORT Load 16 bits of signed data from a buffer surface, sign extend to 32 bits and store the result into a vector register. ENC_VBUFFER default 19 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_I16 OPR_GPUMEM 16 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_B32 BUFFER_LOAD_DWORD Load 32 bits of data from a buffer surface into a vector register. ENC_VBUFFER default 20 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_B64 BUFFER_LOAD_DWORDX2 Load 64 bits of data from a buffer surface into a vector register. ENC_VBUFFER default 21 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B64 OPR_GPUMEM 64 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_B96 BUFFER_LOAD_DWORDX3 Load 96 bits of data from a buffer surface into a vector register. ENC_VBUFFER default 22 VDATA FMT_ANY OPR_VGPR 96 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B96 OPR_GPUMEM 96 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_B128 BUFFER_LOAD_DWORDX4 Load 128 bits of data from a buffer surface into a vector register. ENC_VBUFFER default 23 VDATA FMT_ANY OPR_VGPR 128 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B128 OPR_GPUMEM 128 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_STORE_B8 BUFFER_STORE_BYTE Store 8 bits of data from a vector register into a buffer surface. ENC_VBUFFER default 24 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B8 OPR_GPUMEM 8 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_STORE_B16 BUFFER_STORE_SHORT Store 16 bits of data from a vector register into a buffer surface. ENC_VBUFFER default 25 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B16 OPR_GPUMEM 16 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_STORE_B32 BUFFER_STORE_DWORD Store 32 bits of data from vector input registers into a buffer surface. ENC_VBUFFER default 26 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_STORE_B64 BUFFER_STORE_DWORDX2 Store 64 bits of data from vector input registers into a buffer surface. ENC_VBUFFER default 27 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B64 OPR_GPUMEM 64 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_STORE_B96 BUFFER_STORE_DWORDX3 Store 96 bits of data from vector input registers into a buffer surface. ENC_VBUFFER default 28 VDATA FMT_ANY OPR_VGPR 96 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B96 OPR_GPUMEM 96 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_STORE_B128 BUFFER_STORE_DWORDX4 Store 128 bits of data from vector input registers into a buffer surface. ENC_VBUFFER default 29 VDATA FMT_ANY OPR_VGPR 128 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B128 OPR_GPUMEM 128 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_D16_U8 BUFFER_LOAD_UBYTE_D16 Load 8 bits of unsigned data from a buffer surface, zero extend to 16 bits and store the result into the low 16 bits of a 32-bit vector register. ENC_VBUFFER default 30 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U8 OPR_GPUMEM 8 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_D16_I8 BUFFER_LOAD_SBYTE_D16 Load 8 bits of signed data from a buffer surface, sign extend to 16 bits and store the result into the low 16 bits of a 32-bit vector register. ENC_VBUFFER default 31 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_I8 OPR_GPUMEM 8 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_D16_B16 BUFFER_LOAD_SHORT_D16 Load 16 bits of unsigned data from a buffer surface and store the result into the low 16 bits of a 32-bit vector register. ENC_VBUFFER default 32 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B16 OPR_GPUMEM 16 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_D16_HI_U8 BUFFER_LOAD_UBYTE_D16_HI Load 8 bits of unsigned data from a buffer surface, zero extend to 16 bits and store the result into the high 16 bits of a 32-bit vector register. ENC_VBUFFER default 33 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U8 OPR_GPUMEM 8 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_D16_HI_I8 BUFFER_LOAD_SBYTE_D16_HI Load 8 bits of signed data from a buffer surface, sign extend to 16 bits and store the result into the high 16 bits of a 32-bit vector register. ENC_VBUFFER default 34 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_I8 OPR_GPUMEM 8 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_D16_HI_B16 BUFFER_LOAD_SHORT_D16_HI Load 16 bits of unsigned data from a buffer surface and store the result into the high 16 bits of a 32-bit vector register. ENC_VBUFFER default 35 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B16 OPR_GPUMEM 16 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_STORE_D16_HI_B8 BUFFER_STORE_BYTE_D16_HI Store 8 bits of data from the high 16 bits of a 32-bit vector register into a buffer surface. ENC_VBUFFER default 36 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B8 OPR_GPUMEM 8 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_STORE_D16_HI_B16 BUFFER_STORE_SHORT_D16_HI Store 16 bits of data from the high 16 bits of a 32-bit vector register into a buffer surface. ENC_VBUFFER default 37 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B16 OPR_GPUMEM 16 VMEM BUFFER FALSE FALSE FALSE FALSE FALSE BUFFER_LOAD_D16_HI_FORMAT_X BUFFER_LOAD_FORMAT_D16_HI_X Load 1-component formatted data from a buffer surface, convert the data to packed 16 bit integral or floating point format, then store the result into the high 16 bits of a 32-bit vector register. The resource descriptor specifies the data format of the surface. ENC_VBUFFER default 38 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM LOAD FALSE FALSE FALSE FALSE FALSE BUFFER_STORE_D16_HI_FORMAT_X BUFFER_STORE_FORMAT_D16_HI_X Convert 16 bits of data from the high 16 bits of a 32-bit vector input register into 1-component formatted data and store the data into a buffer surface. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 39 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM STORE FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_SWAP_B32 BUFFER_ATOMIC_SWAP Swap an unsigned 32-bit integer value in the data register with a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 51 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B32 OPR_GPUMEM 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_CMPSWAP_B32 BUFFER_ATOMIC_CMPSWAP Compare two unsigned 32-bit integer values stored in the data comparison register and a location in a buffer surface. Modify the memory location with a value in the data source register iff the comparison is equal. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 52 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_ADD_U32 BUFFER_ATOMIC_ADD Add two unsigned 32-bit integer values stored in the data register and a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 53 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_SUB_U32 BUFFER_ATOMIC_SUB Subtract an unsigned 32-bit integer value stored in the data register from a value stored in a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 54 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_SUB_CLAMP_U32 BUFFER_ATOMIC_CSUB BUFFER_ATOMIC_CSUB_U32 Subtract an unsigned 32-bit integer location in a buffer surface from a value in the data register and clamp the result to zero. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 55 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_MIN_I32 BUFFER_ATOMIC_SMIN Select the minimum of two signed 32-bit integer inputs, given two values stored in the data register and a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 56 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_I32 OPR_GPUMEM 32 FMT_NUM_I32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_MIN_U32 BUFFER_ATOMIC_UMIN Select the minimum of two unsigned 32-bit integer inputs, given two values stored in the data register and a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 57 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_MAX_I32 BUFFER_ATOMIC_SMAX Select the maximum of two signed 32-bit integer inputs, given two values stored in the data register and a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 58 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_I32 OPR_GPUMEM 32 FMT_NUM_I32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_MAX_U32 BUFFER_ATOMIC_UMAX Select the maximum of two unsigned 32-bit integer inputs, given two values stored in the data register and a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 59 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_AND_B32 BUFFER_ATOMIC_AND Calculate bitwise AND given two unsigned 32-bit integer values stored in the data register and a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 60 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B32 OPR_GPUMEM 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_OR_B32 BUFFER_ATOMIC_OR Calculate bitwise OR given two unsigned 32-bit integer values stored in the data register and a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 61 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B32 OPR_GPUMEM 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_XOR_B32 BUFFER_ATOMIC_XOR Calculate bitwise XOR given two unsigned 32-bit integer values stored in the data register and a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 62 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B32 OPR_GPUMEM 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_INC_U32 BUFFER_ATOMIC_INC Increment an unsigned 32-bit integer value from a location in a buffer surface with wraparound to 0 if the value exceeds a value in the data register. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 63 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_DEC_U32 BUFFER_ATOMIC_DEC Decrement an unsigned 32-bit integer value from a location in a buffer surface with wraparound to a value in the data register if the decrement yields a negative value. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 64 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_SWAP_B64 BUFFER_ATOMIC_SWAP_X2 Swap an unsigned 64-bit integer value in the data register with a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 65 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B64 OPR_GPUMEM 64 FMT_NUM_B64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_CMPSWAP_B64 BUFFER_ATOMIC_CMPSWAP_X2 Compare two unsigned 64-bit integer values stored in the data comparison register and a location in a buffer surface. Modify the memory location with a value in the data source register iff the comparison is equal. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 66 VDATA FMT_ANY OPR_VGPR 128 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_ADD_U64 BUFFER_ATOMIC_ADD_X2 Add two unsigned 64-bit integer values stored in the data register and a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 67 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_SUB_U64 BUFFER_ATOMIC_SUB_X2 Subtract an unsigned 64-bit integer value stored in the data register from a value stored in a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 68 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_MIN_I64 BUFFER_ATOMIC_SMIN_X2 Select the minimum of two signed 64-bit integer inputs, given two values stored in the data register and a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 69 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_I64 OPR_GPUMEM 64 FMT_NUM_I64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_MIN_U64 BUFFER_ATOMIC_UMIN_X2 Select the minimum of two unsigned 64-bit integer inputs, given two values stored in the data register and a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 70 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_MAX_I64 BUFFER_ATOMIC_SMAX_X2 Select the maximum of two signed 64-bit integer inputs, given two values stored in the data register and a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 71 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_I64 OPR_GPUMEM 64 FMT_NUM_I64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_MAX_U64 BUFFER_ATOMIC_UMAX_X2 Select the maximum of two unsigned 64-bit integer inputs, given two values stored in the data register and a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 72 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_AND_B64 BUFFER_ATOMIC_AND_X2 Calculate bitwise AND given two unsigned 64-bit integer values stored in the data register and a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 73 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B64 OPR_GPUMEM 64 FMT_NUM_B64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_OR_B64 BUFFER_ATOMIC_OR_X2 Calculate bitwise OR given two unsigned 64-bit integer values stored in the data register and a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 74 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B64 OPR_GPUMEM 64 FMT_NUM_B64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_XOR_B64 BUFFER_ATOMIC_XOR_X2 Calculate bitwise XOR given two unsigned 64-bit integer values stored in the data register and a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 75 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B64 OPR_GPUMEM 64 FMT_NUM_B64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_INC_U64 BUFFER_ATOMIC_INC_X2 Increment an unsigned 64-bit integer value from a location in a buffer surface with wraparound to 0 if the value exceeds a value in the data register. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 76 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_DEC_U64 BUFFER_ATOMIC_DEC_X2 Decrement an unsigned 64-bit integer value from a location in a buffer surface with wraparound to a value in the data register if the decrement yields a negative value. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 77 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_COND_SUB_U32 Subtract an unsigned 32-bit integer value in the data register from a location in a buffer surface only if the memory value is greater than or equal to the data register value. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 80 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_MIN_NUM_F32 BUFFER_ATOMIC_FMIN BUFFER_ATOMIC_MIN_F32 Select the IEEE minimumNumber() of two single-precision float inputs, given two values stored in the data register and a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 81 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_F32 OPR_GPUMEM 32 FMT_NUM_F32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_MAX_NUM_F32 BUFFER_ATOMIC_FMAX BUFFER_ATOMIC_MAX_F32 Select the IEEE maximumNumber() of two single-precision float inputs, given two values stored in the data register and a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 82 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_F32 OPR_GPUMEM 32 FMT_NUM_F32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_ADD_F32 Add two single-precision float values stored in the data register and a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 86 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_F32 OPR_GPUMEM 32 FMT_NUM_F32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_PK_ADD_F16 Add a packed 2-component half-precision float value from the data register to a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 89 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_PK2_F16 OPR_GPUMEM 32 FMT_NUM_PK2_F16 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE BUFFER_ATOMIC_PK_ADD_BF16 Add a packed 2-component BF16 float value from the data register to a location in a buffer surface. Store the original value from buffer surface into a vector register iff the temporal hint enables atomic return. ENC_VBUFFER default 90 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_SCRATCH OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_PK2_BF16 OPR_GPUMEM 32 FMT_NUM_PK2_BF16 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE TBUFFER_LOAD_FORMAT_X Load 1-component formatted data from a buffer surface, convert the data to 32 bit integral or floating point format, then store the result into a vector register. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 128 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM LOAD FALSE FALSE FALSE FALSE FALSE TBUFFER_LOAD_FORMAT_XY Load 2-component formatted data from a buffer surface, convert the data to 32 bit integral or floating point format, then store the result into a vector register. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 129 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B64 OPR_GPUMEM 64 VMEM LOAD FALSE FALSE FALSE FALSE FALSE TBUFFER_LOAD_FORMAT_XYZ Load 3-component formatted data from a buffer surface, convert the data to 32 bit integral or floating point format, then store the result into a vector register. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 130 VDATA FMT_ANY OPR_VGPR 96 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B96 OPR_GPUMEM 96 VMEM LOAD FALSE FALSE FALSE FALSE FALSE TBUFFER_LOAD_FORMAT_XYZW Load 4-component formatted data from a buffer surface, convert the data to 32 bit integral or floating point format, then store the result into a vector register. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 131 VDATA FMT_ANY OPR_VGPR 128 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B128 OPR_GPUMEM 128 VMEM LOAD FALSE FALSE FALSE FALSE FALSE TBUFFER_STORE_FORMAT_X Convert 32 bits of data from vector input registers into 1-component formatted data and store the data into a buffer surface. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 132 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM STORE FALSE FALSE FALSE FALSE FALSE TBUFFER_STORE_FORMAT_XY Convert 64 bits of data from vector input registers into 2-component formatted data and store the data into a buffer surface. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 133 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B64 OPR_GPUMEM 64 VMEM STORE FALSE FALSE FALSE FALSE FALSE TBUFFER_STORE_FORMAT_XYZ Convert 96 bits of data from vector input registers into 3-component formatted data and store the data into a buffer surface. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 134 VDATA FMT_ANY OPR_VGPR 96 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B96 OPR_GPUMEM 96 VMEM STORE FALSE FALSE FALSE FALSE FALSE TBUFFER_STORE_FORMAT_XYZW Convert 128 bits of data from vector input registers into 4-component formatted data and store the data into a buffer surface. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 135 VDATA FMT_ANY OPR_VGPR 128 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B128 OPR_GPUMEM 128 VMEM STORE FALSE FALSE FALSE FALSE FALSE TBUFFER_LOAD_D16_FORMAT_X TBUFFER_LOAD_FORMAT_D16_X Load 1-component formatted data from a buffer surface, convert the data to packed 16 bit integral or floating point format, then store the result into a vector register. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 136 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM LOAD FALSE FALSE FALSE FALSE FALSE TBUFFER_LOAD_D16_FORMAT_XY TBUFFER_LOAD_FORMAT_D16_XY Load 2-component formatted data from a buffer surface, convert the data to packed 16 bit integral or floating point format, then store the result into a vector register. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 137 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B64 OPR_GPUMEM 64 VMEM LOAD FALSE FALSE FALSE FALSE FALSE TBUFFER_LOAD_D16_FORMAT_XYZ TBUFFER_LOAD_FORMAT_D16_XYZ Load 3-component formatted data from a buffer surface, convert the data to packed 16 bit integral or floating point format, then store the result into a vector register. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 138 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B96 OPR_GPUMEM 96 VMEM LOAD FALSE FALSE FALSE FALSE FALSE TBUFFER_LOAD_D16_FORMAT_XYZW TBUFFER_LOAD_FORMAT_D16_XYZW Load 4-component formatted data from a buffer surface, convert the data to packed 16 bit integral or floating point format, then store the result into a vector register. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 139 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B128 OPR_GPUMEM 128 VMEM LOAD FALSE FALSE FALSE FALSE FALSE TBUFFER_STORE_D16_FORMAT_X TBUFFER_STORE_FORMAT_D16_X Convert 16 bits of data from vector input registers into 1-component formatted data and store the data into a buffer surface. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 140 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM STORE FALSE FALSE FALSE FALSE FALSE TBUFFER_STORE_D16_FORMAT_XY TBUFFER_STORE_FORMAT_D16_XY Convert 32 bits of data from vector input registers into 2-component formatted data and store the data into a buffer surface. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 141 VDATA FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B64 OPR_GPUMEM 64 VMEM STORE FALSE FALSE FALSE FALSE FALSE TBUFFER_STORE_D16_FORMAT_XYZ TBUFFER_STORE_FORMAT_D16_XYZ Convert 48 bits of data from vector input registers into 3-component formatted data and store the data into a buffer surface. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 142 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B96 OPR_GPUMEM 96 VMEM STORE FALSE FALSE FALSE FALSE FALSE TBUFFER_STORE_D16_FORMAT_XYZW TBUFFER_STORE_FORMAT_D16_XYZW Convert 64 bits of data from vector input registers into 4-component formatted data and store the data into a buffer surface. The instruction specifies the data format of the surface, overriding the resource descriptor. ENC_VBUFFER default 143 VDATA FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 RSRC FMT_RSRC_TYPED OPR_SREG 128 SOFFSET FMT_ANY OPR_SREG_M0 32 FMT_NUM_B128 OPR_GPUMEM 128 VMEM STORE FALSE FALSE FALSE FALSE FALSE DS_ADD_U32 Add two unsigned 32-bit integer values stored in the data register and a location in a data share. ENC_VDS default 0 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_U32 OPR_DSMEM 32 FMT_NUM_U32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_SUB_U32 Subtract an unsigned 32-bit integer value stored in the data register from a value stored in a location in a data share. ENC_VDS default 1 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_U32 OPR_DSMEM 32 FMT_NUM_U32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_RSUB_U32 DS_SUBREV_U32 Subtract an unsigned 32-bit integer value stored in a location in a data share from a value stored in the data register. ENC_VDS default 2 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_U32 OPR_DSMEM 32 FMT_NUM_U32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_INC_U32 Increment an unsigned 32-bit integer value from a location in a data share with wraparound to 0 if the value exceeds a value in the data register. ENC_VDS default 3 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_U32 OPR_DSMEM 32 FMT_NUM_U32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_DEC_U32 Decrement an unsigned 32-bit integer value from a location in a data share with wraparound to a value in the data register if the decrement yields a negative value. ENC_VDS default 4 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_U32 OPR_DSMEM 32 FMT_NUM_U32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MIN_I32 Select the minimum of two signed 32-bit integer inputs, given two values stored in the data register and a location in a data share. ENC_VDS default 5 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_I32 OPR_DSMEM 32 FMT_NUM_I32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MAX_I32 Select the maximum of two signed 32-bit integer inputs, given two values stored in the data register and a location in a data share. ENC_VDS default 6 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_I32 OPR_DSMEM 32 FMT_NUM_I32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MIN_U32 Select the minimum of two unsigned 32-bit integer inputs, given two values stored in the data register and a location in a data share. ENC_VDS default 7 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_U32 OPR_DSMEM 32 FMT_NUM_U32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MAX_U32 Select the maximum of two unsigned 32-bit integer inputs, given two values stored in the data register and a location in a data share. ENC_VDS default 8 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_U32 OPR_DSMEM 32 FMT_NUM_U32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_AND_B32 Calculate bitwise AND given two unsigned 32-bit integer values stored in the data register and a location in a data share. ENC_VDS default 9 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_OR_B32 Calculate bitwise OR given two unsigned 32-bit integer values stored in the data register and a location in a data share. ENC_VDS default 10 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_XOR_B32 Calculate bitwise XOR given two unsigned 32-bit integer values stored in the data register and a location in a data share. ENC_VDS default 11 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MSKOR_B32 Calculate masked bitwise OR on an unsigned 32-bit integer location in a data share, given mask value and bits to OR in the data registers. ENC_VDS default 12 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 DATA1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_STORE_B32 DS_WRITE_B32 Store 32 bits of data from a vector input register into a data share. ENC_VDS default 13 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_STORE_2ADDR_B32 DS_WRITE2_B32 Store 32 bits of data from one vector input register and then 32 bits of data from a second vector input register into a data share. ENC_VDS default 14 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 DATA1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_STORE_2ADDR_STRIDE64_B32 DS_WRITE2ST64_B32 Store 32 bits of data from one vector input register and then 32 bits of data from a second vector input register into a data share. Treat each offset as an index and multiply by a stride of 64 elements (256 bytes) to generate an offset for each DS address. ENC_VDS default 15 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 DATA1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_CMPSTORE_B32 DS_CMPST_B32 Compare an unsigned 32-bit integer value in the data comparison register with a location in a data share, and modify the memory location with a value in the data source register if the comparison is equal. ENC_VDS default 16 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 DATA1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MIN_NUM_F32 DS_MIN_F32 Select the IEEE minimumNumber() of two single-precision float inputs, given two values stored in the data register and a location in a data share. ENC_VDS default 18 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_F32 OPR_DSMEM 32 FMT_NUM_F32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MAX_NUM_F32 DS_MAX_F32 Select the IEEE maximumNumber() of two single-precision float inputs, given two values stored in the data register and a location in a data share. ENC_VDS default 19 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_F32 OPR_DSMEM 32 FMT_NUM_F32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_NOP Do nothing. ENC_VDS default 20 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_ADD_F32 Add two single-precision float values stored in the data register and a location in a data share. ENC_VDS default 21 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_F32 OPR_DSMEM 32 FMT_NUM_F32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_STORE_B8 DS_WRITE_B8 Store 8 bits of data from a vector register into a data share. ENC_VDS default 30 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B8 OPR_VGPR 8 FMT_NUM_B8 OPR_DSMEM 8 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_STORE_B16 DS_WRITE_B16 Store 16 bits of data from a vector register into a data share. ENC_VDS default 31 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B16 OPR_VGPR 16 FMT_NUM_B16 OPR_DSMEM 16 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_ADD_RTN_U32 Add two unsigned 32-bit integer values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 32 VDST FMT_NUM_U32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_U32 OPR_DSMEM 32 FMT_NUM_U32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_SUB_RTN_U32 Subtract an unsigned 32-bit integer value stored in the data register from a value stored in a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 33 VDST FMT_NUM_U32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_U32 OPR_DSMEM 32 FMT_NUM_U32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_RSUB_RTN_U32 DS_SUBREV_RTN_U32 Subtract an unsigned 32-bit integer value stored in a location in a data share from a value stored in the data register. Store the original value from data share into a vector register. ENC_VDS default 34 VDST FMT_NUM_U32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_U32 OPR_DSMEM 32 FMT_NUM_U32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_INC_RTN_U32 Increment an unsigned 32-bit integer value from a location in a data share with wraparound to 0 if the value exceeds a value in the data register. Store the original value from data share into a vector register. ENC_VDS default 35 VDST FMT_NUM_U32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_U32 OPR_DSMEM 32 FMT_NUM_U32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_DEC_RTN_U32 Decrement an unsigned 32-bit integer value from a location in a data share with wraparound to a value in the data register if the decrement yields a negative value. Store the original value from data share into a vector register. ENC_VDS default 36 VDST FMT_NUM_U32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_U32 OPR_DSMEM 32 FMT_NUM_U32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MIN_RTN_I32 Select the minimum of two signed 32-bit integer inputs, given two values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 37 VDST FMT_NUM_I32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_I32 OPR_DSMEM 32 FMT_NUM_I32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MAX_RTN_I32 Select the maximum of two signed 32-bit integer inputs, given two values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 38 VDST FMT_NUM_I32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_I32 OPR_DSMEM 32 FMT_NUM_I32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MIN_RTN_U32 Select the minimum of two unsigned 32-bit integer inputs, given two values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 39 VDST FMT_NUM_U32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_U32 OPR_DSMEM 32 FMT_NUM_U32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MAX_RTN_U32 Select the maximum of two unsigned 32-bit integer inputs, given two values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 40 VDST FMT_NUM_U32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_U32 OPR_DSMEM 32 FMT_NUM_U32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_AND_RTN_B32 Calculate bitwise AND given two unsigned 32-bit integer values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 41 VDST FMT_NUM_B32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_OR_RTN_B32 Calculate bitwise OR given two unsigned 32-bit integer values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 42 VDST FMT_NUM_B32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_XOR_RTN_B32 Calculate bitwise XOR given two unsigned 32-bit integer values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 43 VDST FMT_NUM_B32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MSKOR_RTN_B32 Calculate masked bitwise OR on an unsigned 32-bit integer location in a data share, given mask value and bits to OR in the data registers. ENC_VDS default 44 VDST FMT_NUM_B32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 DATA1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_STOREXCHG_RTN_B32 DS_WRXCHG_RTN_B32 Swap an unsigned 32-bit integer value in the data register with a location in a data share. ENC_VDS default 45 VDST FMT_NUM_B32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_STOREXCHG_2ADDR_RTN_B32 DS_WRXCHG2_RTN_B32 Swap two unsigned 32-bit integer values in the data registers with two locations in a data share. ENC_VDS default 46 VDST FMT_NUM_PK2_B32 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 DATA1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_STOREXCHG_2ADDR_STRIDE64_RTN_B32 DS_WRXCHG2ST64_RTN_B32 Swap two unsigned 32-bit integer values in the data registers with two locations in a data share. Treat each offset as an index and multiply by a stride of 64 elements (256 bytes) to generate an offset for each DS address. ENC_VDS default 47 VDST FMT_NUM_PK2_B32 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 DATA1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_CMPSTORE_RTN_B32 DS_CMPST_RTN_B32 Compare an unsigned 32-bit integer value in the data comparison register with a location in a data share, and modify the memory location with a value in the data source register if the comparison is equal. ENC_VDS default 48 VDST FMT_NUM_B32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 DATA1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MIN_NUM_RTN_F32 DS_MIN_RTN_F32 Select the IEEE minimumNumber() of two single-precision float inputs, given two values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 50 VDST FMT_NUM_F32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_F32 OPR_DSMEM 32 FMT_NUM_F32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MAX_NUM_RTN_F32 DS_MAX_RTN_F32 Select the IEEE maximumNumber() of two single-precision float inputs, given two values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 51 VDST FMT_NUM_F32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_F32 OPR_DSMEM 32 FMT_NUM_F32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_SWIZZLE_B32 Dword swizzle, no data is written to LDS memory. ENC_VDS default 53 VDST FMT_NUM_B32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_LOAD_B32 DS_READ_B32 Load 32 bits of data from a data share into a vector register. ENC_VDS default 54 VDST FMT_NUM_B32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_LOAD_2ADDR_B32 DS_READ2_B32 Load 32 bits of data from one location in a data share and then 32 bits of data from a second location in a data share and store the results into a 64-bit vector register. ENC_VDS default 55 VDST FMT_NUM_PK2_B32 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_LOAD_2ADDR_STRIDE64_B32 DS_READ2ST64_B32 Load 32 bits of data from one location in a data share and then 32 bits of data from a second location in a data share and store the results into a 64-bit vector register. Treat each offset as an index and multiply by a stride of 64 elements (256 bytes) to generate an offset for each DS address. ENC_VDS default 56 VDST FMT_NUM_PK2_B32 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_LOAD_I8 DS_READ_I8 Load 8 bits of signed data from a data share, sign extend to 32 bits and store the result into a vector register. ENC_VDS default 57 VDST FMT_NUM_I32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B8 OPR_DSMEM 8 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_LOAD_U8 DS_READ_U8 Load 8 bits of unsigned data from a data share, zero extend to 32 bits and store the result into a vector register. ENC_VDS default 58 VDST FMT_NUM_U32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B8 OPR_DSMEM 8 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_LOAD_I16 DS_READ_I16 Load 16 bits of signed data from a data share, sign extend to 32 bits and store the result into a vector register. ENC_VDS default 59 VDST FMT_NUM_I32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B16 OPR_DSMEM 16 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_LOAD_U16 DS_READ_U16 Load 16 bits of unsigned data from a data share, zero extend to 32 bits and store the result into a vector register. ENC_VDS default 60 VDST FMT_NUM_U32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B16 OPR_DSMEM 16 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_CONSUME Subtract (count_bits(exec_mask)) from the value stored in DS memory at (M0.base + instr_offset). Return the pre-operation value to VGPRs. ENC_VDS default 61 VDST FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_APPEND Add (count_bits(exec_mask)) to the value stored in DS memory at (M0.base + instr_offset). Return the pre-operation value to VGPRs. ENC_VDS default 62 VDST FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 FMT_NUM_B32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_ADD_U64 Add two unsigned 64-bit integer values stored in the data register and a location in a data share. ENC_VDS default 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_U64 OPR_DSMEM 64 FMT_NUM_U64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_SUB_U64 Subtract an unsigned 64-bit integer value stored in the data register from a value stored in a location in a data share. ENC_VDS default 65 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_U64 OPR_DSMEM 64 FMT_NUM_U64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_RSUB_U64 DS_SUBREV_U64 Subtract an unsigned 64-bit integer value stored in a location in a data share from a value stored in the data register. ENC_VDS default 66 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_U64 OPR_DSMEM 64 FMT_NUM_U64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_INC_U64 Increment an unsigned 64-bit integer value from a location in a data share with wraparound to 0 if the value exceeds a value in the data register. ENC_VDS default 67 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_U64 OPR_DSMEM 64 FMT_NUM_U64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_DEC_U64 Decrement an unsigned 64-bit integer value from a location in a data share with wraparound to a value in the data register if the decrement yields a negative value. ENC_VDS default 68 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_U64 OPR_DSMEM 64 FMT_NUM_U64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MIN_I64 Select the minimum of two signed 64-bit integer inputs, given two values stored in the data register and a location in a data share. ENC_VDS default 69 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_I64 OPR_VGPR 64 FMT_NUM_I64 OPR_DSMEM 64 FMT_NUM_I64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MAX_I64 Select the maximum of two signed 64-bit integer inputs, given two values stored in the data register and a location in a data share. ENC_VDS default 70 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_I64 OPR_VGPR 64 FMT_NUM_I64 OPR_DSMEM 64 FMT_NUM_I64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MIN_U64 Select the minimum of two unsigned 64-bit integer inputs, given two values stored in the data register and a location in a data share. ENC_VDS default 71 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_U64 OPR_DSMEM 64 FMT_NUM_U64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MAX_U64 Select the maximum of two unsigned 64-bit integer inputs, given two values stored in the data register and a location in a data share. ENC_VDS default 72 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_U64 OPR_DSMEM 64 FMT_NUM_U64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_AND_B64 Calculate bitwise AND given two unsigned 64-bit integer values stored in the data register and a location in a data share. ENC_VDS default 73 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B64 OPR_VGPR 64 FMT_NUM_B64 OPR_DSMEM 64 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_OR_B64 Calculate bitwise OR given two unsigned 64-bit integer values stored in the data register and a location in a data share. ENC_VDS default 74 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B64 OPR_VGPR 64 FMT_NUM_B64 OPR_DSMEM 64 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_XOR_B64 Calculate bitwise XOR given two unsigned 64-bit integer values stored in the data register and a location in a data share. ENC_VDS default 75 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B64 OPR_VGPR 64 FMT_NUM_B64 OPR_DSMEM 64 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MSKOR_B64 Calculate masked bitwise OR on an unsigned 64-bit integer location in a data share, given mask value and bits to OR in the data registers. ENC_VDS default 76 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B64 OPR_VGPR 64 DATA1 FMT_NUM_B64 OPR_VGPR 64 FMT_NUM_B64 OPR_DSMEM 64 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_STORE_B64 DS_WRITE_B64 Store 64 bits of data from a vector input register into a data share. ENC_VDS default 77 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B64 OPR_VGPR 64 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_STORE_2ADDR_B64 DS_WRITE2_B64 Store 64 bits of data from one vector input register and then 64 bits of data from a second vector input register into a data share. ENC_VDS default 78 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B64 OPR_VGPR 64 DATA1 FMT_NUM_B64 OPR_VGPR 64 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_STORE_2ADDR_STRIDE64_B64 DS_WRITE2ST64_B64 Store 64 bits of data from one vector input register and then 64 bits of data from a second vector input register into a data share. Treat each offset as an index and multiply by a stride of 64 elements (256 bytes) to generate an offset for each DS address. ENC_VDS default 79 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B64 OPR_VGPR 64 DATA1 FMT_NUM_B64 OPR_VGPR 64 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_CMPSTORE_B64 DS_CMPST_B64 Compare an unsigned 64-bit integer value in the data comparison register with a location in a data share, and modify the memory location with a value in the data source register if the comparison is equal. ENC_VDS default 80 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B64 OPR_VGPR 64 DATA1 FMT_NUM_B64 OPR_VGPR 64 FMT_NUM_B64 OPR_DSMEM 64 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MIN_NUM_F64 DS_MIN_F64 Select the IEEE minimumNumber() of two double-precision float inputs, given two values stored in the data register and a location in a data share. ENC_VDS default 82 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_F64 OPR_DSMEM 64 FMT_NUM_F64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MAX_NUM_F64 DS_MAX_F64 Select the IEEE maximumNumber() of two double-precision float inputs, given two values stored in the data register and a location in a data share. ENC_VDS default 83 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_F64 OPR_DSMEM 64 FMT_NUM_F64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_ADD_RTN_U64 Add two unsigned 64-bit integer values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 96 VDST FMT_NUM_U64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_U64 OPR_DSMEM 64 FMT_NUM_U64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_SUB_RTN_U64 Subtract an unsigned 64-bit integer value stored in the data register from a value stored in a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 97 VDST FMT_NUM_U64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_U64 OPR_DSMEM 64 FMT_NUM_U64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_RSUB_RTN_U64 DS_SUBREV_RTN_U64 Subtract an unsigned 64-bit integer value stored in a location in a data share from a value stored in the data register. Store the original value from data share into a vector register. ENC_VDS default 98 VDST FMT_NUM_U64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_U64 OPR_DSMEM 64 FMT_NUM_U64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_INC_RTN_U64 Increment an unsigned 64-bit integer value from a location in a data share with wraparound to 0 if the value exceeds a value in the data register. Store the original value from data share into a vector register. ENC_VDS default 99 VDST FMT_NUM_U64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_U64 OPR_DSMEM 64 FMT_NUM_U64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_DEC_RTN_U64 Decrement an unsigned 64-bit integer value from a location in a data share with wraparound to a value in the data register if the decrement yields a negative value. Store the original value from data share into a vector register. ENC_VDS default 100 VDST FMT_NUM_U64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_U64 OPR_DSMEM 64 FMT_NUM_U64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MIN_RTN_I64 Select the minimum of two signed 64-bit integer inputs, given two values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 101 VDST FMT_NUM_I64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_I64 OPR_VGPR 64 FMT_NUM_I64 OPR_DSMEM 64 FMT_NUM_I64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MAX_RTN_I64 Select the maximum of two signed 64-bit integer inputs, given two values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 102 VDST FMT_NUM_I64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_I64 OPR_VGPR 64 FMT_NUM_I64 OPR_DSMEM 64 FMT_NUM_I64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MIN_RTN_U64 Select the minimum of two unsigned 64-bit integer inputs, given two values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 103 VDST FMT_NUM_U64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_U64 OPR_DSMEM 64 FMT_NUM_U64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MAX_RTN_U64 Select the maximum of two unsigned 64-bit integer inputs, given two values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 104 VDST FMT_NUM_U64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_U64 OPR_DSMEM 64 FMT_NUM_U64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_AND_RTN_B64 Calculate bitwise AND given two unsigned 64-bit integer values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 105 VDST FMT_NUM_B64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B64 OPR_VGPR 64 FMT_NUM_B64 OPR_DSMEM 64 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_OR_RTN_B64 Calculate bitwise OR given two unsigned 64-bit integer values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 106 VDST FMT_NUM_B64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B64 OPR_VGPR 64 FMT_NUM_B64 OPR_DSMEM 64 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_XOR_RTN_B64 Calculate bitwise XOR given two unsigned 64-bit integer values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 107 VDST FMT_NUM_B64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B64 OPR_VGPR 64 FMT_NUM_B64 OPR_DSMEM 64 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MSKOR_RTN_B64 Calculate masked bitwise OR on an unsigned 64-bit integer location in a data share, given mask value and bits to OR in the data registers. ENC_VDS default 108 VDST FMT_NUM_B64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B64 OPR_VGPR 64 DATA1 FMT_NUM_B64 OPR_VGPR 64 FMT_NUM_B64 OPR_DSMEM 64 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_STOREXCHG_RTN_B64 DS_WRXCHG_RTN_B64 Swap an unsigned 64-bit integer value in the data register with a location in a data share. ENC_VDS default 109 VDST FMT_NUM_B64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B64 OPR_VGPR 64 FMT_NUM_B64 OPR_DSMEM 64 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_STOREXCHG_2ADDR_RTN_B64 DS_WRXCHG2_RTN_B64 Swap two unsigned 64-bit integer values in the data registers with two locations in a data share. ENC_VDS default 110 VDST FMT_NUM_PK2_B64 OPR_VGPR 128 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B64 OPR_VGPR 64 DATA1 FMT_NUM_B64 OPR_VGPR 64 FMT_NUM_B64 OPR_DSMEM 64 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_STOREXCHG_2ADDR_STRIDE64_RTN_B64 DS_WRXCHG2ST64_RTN_B64 Swap two unsigned 64-bit integer values in the data registers with two locations in a data share. Treat each offset as an index and multiply by a stride of 64 elements (256 bytes) to generate an offset for each DS address. ENC_VDS default 111 VDST FMT_NUM_PK2_B64 OPR_VGPR 128 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B64 OPR_VGPR 64 DATA1 FMT_NUM_B64 OPR_VGPR 64 FMT_NUM_B64 OPR_DSMEM 64 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_CMPSTORE_RTN_B64 DS_CMPST_RTN_B64 Compare an unsigned 64-bit integer value in the data comparison register with a location in a data share, and modify the memory location with a value in the data source register if the comparison is equal. ENC_VDS default 112 VDST FMT_NUM_B64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B64 OPR_VGPR 64 DATA1 FMT_NUM_B64 OPR_VGPR 64 FMT_NUM_B64 OPR_DSMEM 64 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MIN_NUM_RTN_F64 DS_MIN_RTN_F64 Select the IEEE minimumNumber() of two double-precision float inputs, given two values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 114 VDST FMT_NUM_F64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_F64 OPR_DSMEM 64 FMT_NUM_F64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_MAX_NUM_RTN_F64 DS_MAX_RTN_F64 Select the IEEE maximumNumber() of two double-precision float inputs, given two values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 115 VDST FMT_NUM_F64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_F64 OPR_DSMEM 64 FMT_NUM_F64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_LOAD_B64 DS_READ_B64 Load 64 bits of data from a data share into a vector register. ENC_VDS default 118 VDST FMT_NUM_B64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_LOAD_2ADDR_B64 DS_READ2_B64 Load 64 bits of data from one location in a data share and then 64 bits of data from a second location in a data share and store the results into a 128-bit vector register. ENC_VDS default 119 VDST FMT_NUM_PK2_B64 OPR_VGPR 128 ADDR FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_LOAD_2ADDR_STRIDE64_B64 DS_READ2ST64_B64 Load 64 bits of data from one location in a data share and then 64 bits of data from a second location in a data share and store the results into a 128-bit vector register. Treat each offset as an index and multiply by a stride of 64 elements (256 bytes) to generate an offset for each DS address. ENC_VDS default 120 VDST FMT_NUM_PK2_B64 OPR_VGPR 128 ADDR FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_ADD_RTN_F32 Add two single-precision float values stored in the data register and a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 121 VDST FMT_NUM_F32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_F32 OPR_DSMEM 32 FMT_NUM_F32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_CONDXCHG32_RTN_B64 Perform 2 conditional write exchanges, where each conditional write exchange writes a 32 bit value from a data register to a location in data share iff the most significant bit of the data value is set. ENC_VDS default 126 VDST FMT_NUM_B64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B64 OPR_VGPR 64 FMT_NUM_B64 OPR_DSMEM 64 FMT_NUM_B64 OPR_DSMEM 64 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_COND_SUB_U32 Subtract an unsigned 32-bit integer value in the data register from a location in a data share only if the memory value is greater than or equal to the data register value. ENC_VDS default 152 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_U32 OPR_DSMEM 32 FMT_NUM_U32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_SUB_CLAMP_U32 Subtract an unsigned 32-bit integer location in a data share from a value in the data register and clamp the result to zero. ENC_VDS default 153 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_U32 OPR_DSMEM 32 FMT_NUM_U32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_PK_ADD_F16 Add a packed 2-component half-precision float value from the data register to a location in a data share. ENC_VDS default 154 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_PK2_F16 OPR_VGPR 32 FMT_NUM_PK2_F16 OPR_DSMEM 32 FMT_NUM_PK2_F16 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_PK_ADD_BF16 Add a packed 2-component BF16 float value from the data register to a location in a data share. ENC_VDS default 155 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_PK2_BF16 OPR_VGPR 32 FMT_NUM_PK2_BF16 OPR_DSMEM 32 FMT_NUM_PK2_BF16 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_STORE_B8_D16_HI DS_WRITE_B8_D16_HI Store 8 bits of data from the high bits of a vector register into a data share. ENC_VDS default 160 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_PK4_B8 OPR_VGPR 32 FMT_NUM_B8 OPR_DSMEM 8 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_STORE_B16_D16_HI DS_WRITE_B16_D16_HI Store 16 bits of data from the high bits of a vector register into a data share. ENC_VDS default 161 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_PK2_B16 OPR_VGPR 32 FMT_NUM_B16 OPR_DSMEM 16 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_LOAD_U8_D16 DS_READ_U8_D16 Load 8 bits of unsigned data from a data share, zero extend to 16 bits and store the result into the low 16 bits of a vector register. ENC_VDS default 162 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B8 OPR_DSMEM 8 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_LOAD_U8_D16_HI DS_READ_U8_D16_HI Load 8 bits of unsigned data from a data share, zero extend to 16 bits and store the result into the high 16 bits of a vector register. ENC_VDS default 163 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B8 OPR_DSMEM 8 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_LOAD_I8_D16 DS_READ_I8_D16 Load 8 bits of signed data from a data share, sign extend to 16 bits and store the result into the low 16 bits of a vector register. ENC_VDS default 164 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B8 OPR_DSMEM 8 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_LOAD_I8_D16_HI DS_READ_I8_D16_HI Load 8 bits of signed data from a data share, sign extend to 16 bits and store the result into the high 16 bits of a vector register. ENC_VDS default 165 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B8 OPR_DSMEM 8 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_LOAD_U16_D16 DS_READ_U16_D16 Load 16 bits of unsigned data from a data share and store the result into the low 16 bits of a vector register. ENC_VDS default 166 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B16 OPR_DSMEM 16 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_LOAD_U16_D16_HI DS_READ_U16_D16_HI Load 16 bits of unsigned data from a data share and store the result into the high 16 bits of a vector register. ENC_VDS default 167 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B16 OPR_DSMEM 16 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_COND_SUB_RTN_U32 Subtract an unsigned 32-bit integer value in the data register from a location in a data share only if the memory value is greater than or equal to the data register value. Store the original value from data share into a vector register. ENC_VDS default 168 VDST FMT_NUM_U32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_U32 OPR_DSMEM 32 FMT_NUM_U32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_SUB_CLAMP_RTN_U32 Subtract an unsigned 32-bit integer location in a data share from a value in the data register and clamp the result to zero. Store the original value from data share into a vector register. ENC_VDS default 169 VDST FMT_NUM_U32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_U32 OPR_DSMEM 32 FMT_NUM_U32 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_PK_ADD_RTN_F16 Add a packed 2-component half-precision float value from the data register to a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 170 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_PK2_F16 OPR_VGPR 32 FMT_NUM_PK2_F16 OPR_DSMEM 32 FMT_NUM_PK2_F16 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_PK_ADD_RTN_BF16 Add a packed 2-component BF16 float value from the data register to a location in a data share. Store the original value from data share into a vector register. ENC_VDS default 171 VDST FMT_NUM_PK2_BF16 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_PK2_BF16 OPR_VGPR 32 FMT_NUM_PK2_BF16 OPR_DSMEM 32 FMT_NUM_PK2_BF16 OPR_DSMEM 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_STORE_ADDTID_B32 DS_WRITE_ADDTID_B32 Store 32 bits of data from a vector input register into a data share. The memory base address is provided as an immediate value and the lane ID is used as an offset. ENC_VDS default 176 DATA0 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 FMT_NUM_B32 OPR_SDST_M0 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_LOAD_ADDTID_B32 DS_READ_ADDTID_B32 Load 32 bits of data from a data share into a vector register. The memory base address is provided as an immediate value and the lane ID is used as an offset. ENC_VDS default 177 VDST FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 FMT_NUM_B32 OPR_SDST_M0 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_PERMUTE_B32 Forward permute. This does not access LDS memory and may be called even if no LDS memory is allocated to the wave. It uses LDS to implement an arbitrary swizzle across threads in a wavefront. ENC_VDS default 178 VDST FMT_NUM_B32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_BPERMUTE_B32 Backward permute. This does not access LDS memory and may be called even if no LDS memory is allocated to the wave. It uses LDS hardware to implement an arbitrary swizzle across threads in a wavefront. ENC_VDS default 179 VDST FMT_NUM_B32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_BPERMUTE_FI_B32 Backward permute and fetch data for invalid lanes. This does not access LDS memory and may be called even if no LDS memory is allocated to the wave. It uses LDS hardware to implement an arbitrary swizzle across threads in a wavefront. ENC_VDS default 205 VDST FMT_NUM_B32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_STORE_B96 DS_WRITE_B96 Store 96 bits of data from a vector input register into a data share. ENC_VDS default 222 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B96 OPR_VGPR 96 FMT_NUM_B96 OPR_DSMEM 96 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_STORE_B128 DS_WRITE_B128 Store 128 bits of data from a vector input register into a data share. ENC_VDS default 223 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B128 OPR_VGPR 128 FMT_NUM_B128 OPR_DSMEM 128 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_BVH_STACK_PUSH4_POP1_RTN_B32 Ray tracing involves traversing a BVH which is a kind of tree where nodes have up to 4 children. Each shader thread processes one child at a time, and overflow nodes are stored temporarily in LDS using a stack. This instruction supports pushing/popping the stack to reduce the number of VALU instructions required per traversal and reduce VMEM bandwidth requirements. ENC_VDS default 224 VDST FMT_NUM_B32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 DATA1 FMT_NUM_B128 OPR_VGPR 128 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_BVH_STACK_PUSH8_POP1_RTN_B32 Ray tracing involves traversing a BVH which is a kind of tree where nodes have up to 4 children. Each shader thread processes one child at a time, and overflow nodes are stored temporarily in LDS using a stack. This instruction supports pushing/popping the stack to reduce the number of VALU instructions required per traversal and reduce VMEM bandwidth requirements. ENC_VDS default 225 VDST FMT_NUM_B32 OPR_VGPR 32 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 DATA1 FMT_NUM_B256 OPR_VGPR 256 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_BVH_STACK_PUSH8_POP2_RTN_B64 Ray tracing involves traversing a BVH which is a kind of tree where nodes have up to 4 children. Each shader thread processes one child at a time, and overflow nodes are stored temporarily in LDS using a stack. This instruction supports pushing/popping the stack to reduce the number of VALU instructions required per traversal and reduce VMEM bandwidth requirements. ENC_VDS default 226 VDST FMT_NUM_B64 OPR_VGPR 64 ADDR FMT_NUM_B32 OPR_VGPR 32 DATA0 FMT_NUM_B32 OPR_VGPR 32 DATA1 FMT_NUM_B256 OPR_VGPR 256 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_LOAD_B96 DS_READ_B96 Load 96 bits of data from a data share into a vector register. ENC_VDS default 254 VDST FMT_NUM_B96 OPR_VGPR 96 ADDR FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B96 OPR_DSMEM 96 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_LOAD_B128 DS_READ_B128 Load 128 bits of data from a data share into a vector register. ENC_VDS default 255 VDST FMT_NUM_B128 OPR_VGPR 128 ADDR FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B128 OPR_DSMEM 128 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_PARAM_LOAD LDS_PARAM_LOAD Transfer parameter data from LDS to VGPRs and expand data in LDS using the NewPrimMask (provided in M0) to place per-quad data into lanes 0-3 of each quad as follows: ENC_VDSDIR default 0 VDST FMT_NUM_B32 OPR_VGPR 32 ATTR FMT_NUM_B32 OPR_ATTR 32 FMT_NUM_B32 OPR_DSMEM 32 FMT_NUM_B32 OPR_SDST_M0 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE DS_DIRECT_LOAD LDS_DIRECT_LOAD Read a single 32-bit value from LDS to all lanes. A single DWORD is read from LDS memory at ADDR[M0[15:0]], where M0[15:0] is a byte address and is dword-aligned. M0[18:16] specify the data type for the read and may be 0=UBYTE, 1=USHORT, 2=DWORD, 4=SBYTE, 5=SSHORT. ENC_VDSDIR default 1 VDST FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_DSMEM 32 FMT_NUM_B32 OPR_SDST_M0 32 VMEM DATA_SHARE FALSE FALSE FALSE FALSE FALSE EXPORT EXP Export graphics data to the next stage of the render pipeline. The target and up to four channels of data are specified as operands. ENC_VEXPORT default 0 TGT FMT_ANY OPR_TGT 128 VSRC0 FMT_ANY OPR_VGPR 32 VSRC1 FMT_ANY OPR_VGPR 32 VSRC2 FMT_ANY OPR_VGPR 32 VSRC3 FMT_ANY OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 FMT_NUM_B32 OPR_SDST_M0 32 EXPORT FALSE FALSE FALSE FALSE FALSE FLAT_LOAD_U8 FLAT_LOAD_UBYTE Load 8 bits of unsigned data from the flat aperture, zero extend to 32 bits and store the result into a vector register. ENC_VFLAT default 16 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 FMT_NUM_U8 OPR_GPUMEM 8 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_LOAD_I8 FLAT_LOAD_SBYTE Load 8 bits of signed data from the flat aperture, sign extend to 32 bits and store the result into a vector register. ENC_VFLAT default 17 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 FMT_NUM_I8 OPR_GPUMEM 8 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_LOAD_U16 FLAT_LOAD_USHORT Load 16 bits of unsigned data from the flat aperture, zero extend to 32 bits and store the result into a vector register. ENC_VFLAT default 18 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 FMT_NUM_U16 OPR_GPUMEM 16 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_LOAD_I16 FLAT_LOAD_SSHORT Load 16 bits of signed data from the flat aperture, sign extend to 32 bits and store the result into a vector register. ENC_VFLAT default 19 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 FMT_NUM_I16 OPR_GPUMEM 16 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_LOAD_B32 FLAT_LOAD_DWORD Load 32 bits of data from the flat aperture into a vector register. ENC_VFLAT default 20 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 FMT_NUM_B32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_LOAD_B64 FLAT_LOAD_DWORDX2 Load 64 bits of data from the flat aperture into a vector register. ENC_VFLAT default 21 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 FMT_NUM_B64 OPR_GPUMEM 64 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_LOAD_B96 FLAT_LOAD_DWORDX3 Load 96 bits of data from the flat aperture into a vector register. ENC_VFLAT default 22 VDST FMT_ANY OPR_VGPR 96 VADDR FMT_ANY OPR_VGPR 64 FMT_NUM_B96 OPR_GPUMEM 96 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_LOAD_B128 FLAT_LOAD_DWORDX4 Load 128 bits of data from the flat aperture into a vector register. ENC_VFLAT default 23 VDST FMT_ANY OPR_VGPR 128 VADDR FMT_ANY OPR_VGPR 64 FMT_NUM_B128 OPR_GPUMEM 128 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_STORE_B8 FLAT_STORE_BYTE Store 8 bits of data from a vector register into the flat aperture. ENC_VFLAT default 24 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_B8 OPR_GPUMEM 8 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_STORE_B16 FLAT_STORE_SHORT Store 16 bits of data from a vector register into the flat aperture. ENC_VFLAT default 25 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_B16 OPR_GPUMEM 16 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_STORE_B32 FLAT_STORE_DWORD Store 32 bits of data from vector input registers into the flat aperture. ENC_VFLAT default 26 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_STORE_B64 FLAT_STORE_DWORDX2 Store 64 bits of data from vector input registers into the flat aperture. ENC_VFLAT default 27 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 FMT_NUM_B64 OPR_GPUMEM 64 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_STORE_B96 FLAT_STORE_DWORDX3 Store 96 bits of data from vector input registers into the flat aperture. ENC_VFLAT default 28 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 96 FMT_NUM_B96 OPR_GPUMEM 96 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_STORE_B128 FLAT_STORE_DWORDX4 Store 128 bits of data from vector input registers into the flat aperture. ENC_VFLAT default 29 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 128 FMT_NUM_B128 OPR_GPUMEM 128 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_LOAD_D16_U8 FLAT_LOAD_UBYTE_D16 Load 8 bits of unsigned data from the flat aperture, zero extend to 16 bits and store the result into the low 16 bits of a 32-bit vector register. ENC_VFLAT default 30 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 FMT_NUM_U8 OPR_GPUMEM 8 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_LOAD_D16_I8 FLAT_LOAD_SBYTE_D16 Load 8 bits of signed data from the flat aperture, sign extend to 16 bits and store the result into the low 16 bits of a 32-bit vector register. ENC_VFLAT default 31 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 FMT_NUM_I8 OPR_GPUMEM 8 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_LOAD_D16_B16 FLAT_LOAD_SHORT_D16 Load 16 bits of unsigned data from the flat aperture and store the result into the low 16 bits of a 32-bit vector register. ENC_VFLAT default 32 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 FMT_NUM_B16 OPR_GPUMEM 16 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_LOAD_D16_HI_U8 FLAT_LOAD_UBYTE_D16_HI Load 8 bits of unsigned data from the flat aperture, zero extend to 16 bits and store the result into the high 16 bits of a 32-bit vector register. ENC_VFLAT default 33 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 FMT_NUM_U8 OPR_GPUMEM 8 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_LOAD_D16_HI_I8 FLAT_LOAD_SBYTE_D16_HI Load 8 bits of signed data from the flat aperture, sign extend to 16 bits and store the result into the high 16 bits of a 32-bit vector register. ENC_VFLAT default 34 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 FMT_NUM_I8 OPR_GPUMEM 8 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_LOAD_D16_HI_B16 FLAT_LOAD_SHORT_D16_HI Load 16 bits of unsigned data from the flat aperture and store the result into the high 16 bits of a 32-bit vector register. ENC_VFLAT default 35 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 FMT_NUM_B16 OPR_GPUMEM 16 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_STORE_D16_HI_B8 FLAT_STORE_BYTE_D16_HI Store 8 bits of data from the high 16 bits of a 32-bit vector register into the flat aperture. ENC_VFLAT default 36 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_B8 OPR_GPUMEM 8 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_STORE_D16_HI_B16 FLAT_STORE_SHORT_D16_HI Store 16 bits of data from the high 16 bits of a 32-bit vector register into the flat aperture. ENC_VFLAT default 37 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_B16 OPR_GPUMEM 16 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_SWAP_B32 FLAT_ATOMIC_SWAP Swap an unsigned 32-bit integer value in the data register with a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 51 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_B32 OPR_GPUMEM 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_CMPSWAP_B32 FLAT_ATOMIC_CMPSWAP Compare two unsigned 32-bit integer values stored in the data comparison register and a location in the flat aperture. Modify the memory location with a value in the data source register iff the comparison is equal. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 52 VDST FMT_NUM_U32 OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_ADD_U32 FLAT_ATOMIC_ADD Add two unsigned 32-bit integer values stored in the data register and a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 53 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_SUB_U32 FLAT_ATOMIC_SUB Subtract an unsigned 32-bit integer value stored in the data register from a value stored in a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 54 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_SUB_CLAMP_U32 FLAT_ATOMIC_CSUB_U32 Subtract an unsigned 32-bit integer location in the flat aperture from a value in the data register and clamp the result to zero. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 55 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_MIN_I32 FLAT_ATOMIC_SMIN Select the minimum of two signed 32-bit integer inputs, given two values stored in the data register and a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 56 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_I32 OPR_GPUMEM 32 FMT_NUM_I32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_MIN_U32 FLAT_ATOMIC_UMIN Select the minimum of two unsigned 32-bit integer inputs, given two values stored in the data register and a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 57 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_MAX_I32 FLAT_ATOMIC_SMAX Select the maximum of two signed 32-bit integer inputs, given two values stored in the data register and a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 58 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_I32 OPR_GPUMEM 32 FMT_NUM_I32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_MAX_U32 FLAT_ATOMIC_UMAX Select the maximum of two unsigned 32-bit integer inputs, given two values stored in the data register and a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 59 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_AND_B32 FLAT_ATOMIC_AND Calculate bitwise AND given two unsigned 32-bit integer values stored in the data register and a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 60 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_B32 OPR_GPUMEM 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_OR_B32 FLAT_ATOMIC_OR Calculate bitwise OR given two unsigned 32-bit integer values stored in the data register and a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 61 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_B32 OPR_GPUMEM 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_XOR_B32 FLAT_ATOMIC_XOR Calculate bitwise XOR given two unsigned 32-bit integer values stored in the data register and a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 62 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_B32 OPR_GPUMEM 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_INC_U32 FLAT_ATOMIC_INC Increment an unsigned 32-bit integer value from a location in the flat aperture with wraparound to 0 if the value exceeds a value in the data register. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 63 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_DEC_U32 FLAT_ATOMIC_DEC Decrement an unsigned 32-bit integer value from a location in the flat aperture with wraparound to a value in the data register if the decrement yields a negative value. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 64 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_SWAP_B64 FLAT_ATOMIC_SWAP_X2 Swap an unsigned 64-bit integer value in the data register with a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 65 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 FMT_NUM_B64 OPR_GPUMEM 64 FMT_NUM_B64 OPR_GPUMEM 64 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_CMPSWAP_B64 FLAT_ATOMIC_CMPSWAP_X2 Compare two unsigned 64-bit integer values stored in the data comparison register and a location in the flat aperture. Modify the memory location with a value in the data source register iff the comparison is equal. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 66 VDST FMT_NUM_U64 OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 128 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_ADD_U64 FLAT_ATOMIC_ADD_X2 Add two unsigned 64-bit integer values stored in the data register and a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 67 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_SUB_U64 FLAT_ATOMIC_SUB_X2 Subtract an unsigned 64-bit integer value stored in the data register from a value stored in a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 68 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_MIN_I64 FLAT_ATOMIC_SMIN_X2 Select the minimum of two signed 64-bit integer inputs, given two values stored in the data register and a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 69 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 FMT_NUM_I64 OPR_GPUMEM 64 FMT_NUM_I64 OPR_GPUMEM 64 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_MIN_U64 FLAT_ATOMIC_UMIN_X2 Select the minimum of two unsigned 64-bit integer inputs, given two values stored in the data register and a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 70 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_MAX_I64 FLAT_ATOMIC_SMAX_X2 Select the maximum of two signed 64-bit integer inputs, given two values stored in the data register and a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 71 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 FMT_NUM_I64 OPR_GPUMEM 64 FMT_NUM_I64 OPR_GPUMEM 64 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_MAX_U64 FLAT_ATOMIC_UMAX_X2 Select the maximum of two unsigned 64-bit integer inputs, given two values stored in the data register and a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 72 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_AND_B64 FLAT_ATOMIC_AND_X2 Calculate bitwise AND given two unsigned 64-bit integer values stored in the data register and a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 73 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 FMT_NUM_B64 OPR_GPUMEM 64 FMT_NUM_B64 OPR_GPUMEM 64 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_OR_B64 FLAT_ATOMIC_OR_X2 Calculate bitwise OR given two unsigned 64-bit integer values stored in the data register and a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 74 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 FMT_NUM_B64 OPR_GPUMEM 64 FMT_NUM_B64 OPR_GPUMEM 64 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_XOR_B64 FLAT_ATOMIC_XOR_X2 Calculate bitwise XOR given two unsigned 64-bit integer values stored in the data register and a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 75 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 FMT_NUM_B64 OPR_GPUMEM 64 FMT_NUM_B64 OPR_GPUMEM 64 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_INC_U64 FLAT_ATOMIC_INC_X2 Increment an unsigned 64-bit integer value from a location in the flat aperture with wraparound to 0 if the value exceeds a value in the data register. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 76 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_DEC_U64 FLAT_ATOMIC_DEC_X2 Decrement an unsigned 64-bit integer value from a location in the flat aperture with wraparound to a value in the data register if the decrement yields a negative value. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 77 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_COND_SUB_U32 Subtract an unsigned 32-bit integer value in the data register from a location in the flat aperture only if the memory value is greater than or equal to the data register value. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 80 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_MIN_NUM_F32 FLAT_ATOMIC_FMIN FLAT_ATOMIC_MIN_F32 Select the IEEE minimumNumber() of two single-precision float inputs, given two values stored in the data register and a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 81 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_F32 OPR_GPUMEM 32 FMT_NUM_F32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_MAX_NUM_F32 FLAT_ATOMIC_FMAX FLAT_ATOMIC_MAX_F32 Select the IEEE maximumNumber() of two single-precision float inputs, given two values stored in the data register and a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 82 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_F32 OPR_GPUMEM 32 FMT_NUM_F32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_ADD_F32 Add two single-precision float values stored in the data register and a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 86 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_F32 OPR_GPUMEM 32 FMT_NUM_F32 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_PK_ADD_F16 Add a packed 2-component half-precision float value from the data register to a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 89 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_PK2_F16 OPR_GPUMEM 32 FMT_NUM_PK2_F16 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE FLAT_ATOMIC_PK_ADD_BF16 Add a packed 2-component BF16 float value from the data register to a location in the flat aperture. Store the original value from flat aperture into a vector register iff the temporal hint enables atomic return. ENC_VFLAT default 90 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 FMT_NUM_PK2_BF16 OPR_GPUMEM 32 FMT_NUM_PK2_BF16 OPR_GPUMEM 32 VMEM FLAT FALSE FALSE FALSE FALSE FALSE GLOBAL_LOAD_U8 GLOBAL_LOAD_UBYTE Load 8 bits of unsigned data from the global aperture, zero extend to 32 bits and store the result into a vector register. ENC_VGLOBAL default 16 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U8 OPR_GPUMEM 8 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_LOAD_I8 GLOBAL_LOAD_SBYTE Load 8 bits of signed data from the global aperture, sign extend to 32 bits and store the result into a vector register. ENC_VGLOBAL default 17 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_I8 OPR_GPUMEM 8 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_LOAD_U16 GLOBAL_LOAD_USHORT Load 16 bits of unsigned data from the global aperture, zero extend to 32 bits and store the result into a vector register. ENC_VGLOBAL default 18 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U16 OPR_GPUMEM 16 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_LOAD_I16 GLOBAL_LOAD_SSHORT Load 16 bits of signed data from the global aperture, sign extend to 32 bits and store the result into a vector register. ENC_VGLOBAL default 19 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_I16 OPR_GPUMEM 16 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_LOAD_B32 GLOBAL_LOAD_DWORD Load 32 bits of data from the global aperture into a vector register. ENC_VGLOBAL default 20 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B32 OPR_GPUMEM 32 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_LOAD_B64 GLOBAL_LOAD_DWORDX2 Load 64 bits of data from the global aperture into a vector register. ENC_VGLOBAL default 21 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B64 OPR_GPUMEM 64 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_LOAD_B96 GLOBAL_LOAD_DWORDX3 Load 96 bits of data from the global aperture into a vector register. ENC_VGLOBAL default 22 VDST FMT_ANY OPR_VGPR 96 VADDR FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B96 OPR_GPUMEM 96 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_LOAD_B128 GLOBAL_LOAD_DWORDX4 Load 128 bits of data from the global aperture into a vector register. ENC_VGLOBAL default 23 VDST FMT_ANY OPR_VGPR 128 VADDR FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B128 OPR_GPUMEM 128 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_STORE_B8 GLOBAL_STORE_BYTE Store 8 bits of data from a vector register into the global aperture. ENC_VGLOBAL default 24 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B8 OPR_GPUMEM 8 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_STORE_B16 GLOBAL_STORE_SHORT Store 16 bits of data from a vector register into the global aperture. ENC_VGLOBAL default 25 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B16 OPR_GPUMEM 16 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_STORE_B32 GLOBAL_STORE_DWORD Store 32 bits of data from vector input registers into the global aperture. ENC_VGLOBAL default 26 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B32 OPR_GPUMEM 32 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_STORE_B64 GLOBAL_STORE_DWORDX2 Store 64 bits of data from vector input registers into the global aperture. ENC_VGLOBAL default 27 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B64 OPR_GPUMEM 64 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_STORE_B96 GLOBAL_STORE_DWORDX3 Store 96 bits of data from vector input registers into the global aperture. ENC_VGLOBAL default 28 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 96 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B96 OPR_GPUMEM 96 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_STORE_B128 GLOBAL_STORE_DWORDX4 Store 128 bits of data from vector input registers into the global aperture. ENC_VGLOBAL default 29 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 128 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B128 OPR_GPUMEM 128 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_LOAD_D16_U8 GLOBAL_LOAD_UBYTE_D16 Load 8 bits of unsigned data from the global aperture, zero extend to 16 bits and store the result into the low 16 bits of a 32-bit vector register. ENC_VGLOBAL default 30 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U8 OPR_GPUMEM 8 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_LOAD_D16_I8 GLOBAL_LOAD_SBYTE_D16 Load 8 bits of signed data from the global aperture, sign extend to 16 bits and store the result into the low 16 bits of a 32-bit vector register. ENC_VGLOBAL default 31 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_I8 OPR_GPUMEM 8 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_LOAD_D16_B16 GLOBAL_LOAD_SHORT_D16 Load 16 bits of unsigned data from the global aperture and store the result into the low 16 bits of a 32-bit vector register. ENC_VGLOBAL default 32 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B16 OPR_GPUMEM 16 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_LOAD_D16_HI_U8 GLOBAL_LOAD_UBYTE_D16_HI Load 8 bits of unsigned data from the global aperture, zero extend to 16 bits and store the result into the high 16 bits of a 32-bit vector register. ENC_VGLOBAL default 33 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U8 OPR_GPUMEM 8 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_LOAD_D16_HI_I8 GLOBAL_LOAD_SBYTE_D16_HI Load 8 bits of signed data from the global aperture, sign extend to 16 bits and store the result into the high 16 bits of a 32-bit vector register. ENC_VGLOBAL default 34 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_I8 OPR_GPUMEM 8 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_LOAD_D16_HI_B16 GLOBAL_LOAD_SHORT_D16_HI Load 16 bits of unsigned data from the global aperture and store the result into the high 16 bits of a 32-bit vector register. ENC_VGLOBAL default 35 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B16 OPR_GPUMEM 16 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_STORE_D16_HI_B8 GLOBAL_STORE_BYTE_D16_HI Store 8 bits of data from the high 16 bits of a 32-bit vector register into the global aperture. ENC_VGLOBAL default 36 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B8 OPR_GPUMEM 8 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_STORE_D16_HI_B16 GLOBAL_STORE_SHORT_D16_HI Store 16 bits of data from the high 16 bits of a 32-bit vector register into the global aperture. ENC_VGLOBAL default 37 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B16 OPR_GPUMEM 16 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_LOAD_ADDTID_B32 GLOBAL_LOAD_DWORD_ADDTID Load 32 bits of data from the global aperture into a vector register. The memory base address is provided in a scalar register and the lane ID is used as an offset. ENC_VGLOBAL default 40 VDST FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B32 OPR_GPUMEM 32 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_STORE_ADDTID_B32 GLOBAL_STORE_DWORD_ADDTID Store 32 bits of data from a vector input register into the global aperture. The memory base address is provided as an immediate value and the lane ID is used as an offset. ENC_VGLOBAL default 41 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B32 OPR_GPUMEM 32 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_INV Invalidate cache lines based on the SCOPE field. Increments/decrements LOAD_CNT. ENC_VGLOBAL default 43 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_WB Write back dirty cache lines based on the SCOPE field. Increments/decrements STORE_CNT. ENC_VGLOBAL default 44 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_SWAP_B32 GLOBAL_ATOMIC_SWAP Swap an unsigned 32-bit integer value in the data register with a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 51 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B32 OPR_GPUMEM 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_CMPSWAP_B32 GLOBAL_ATOMIC_CMPSWAP Compare two unsigned 32-bit integer values stored in the data comparison register and a location in the global aperture. Modify the memory location with a value in the data source register iff the comparison is equal. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 52 VDST FMT_NUM_U32 OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_ADD_U32 GLOBAL_ATOMIC_ADD Add two unsigned 32-bit integer values stored in the data register and a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 53 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_SUB_U32 GLOBAL_ATOMIC_SUB Subtract an unsigned 32-bit integer value stored in the data register from a value stored in a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 54 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_SUB_CLAMP_U32 GLOBAL_ATOMIC_CSUB GLOBAL_ATOMIC_CSUB_U32 Subtract an unsigned 32-bit integer location in the global aperture from a value in the data register and clamp the result to zero. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 55 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_MIN_I32 GLOBAL_ATOMIC_SMIN Select the minimum of two signed 32-bit integer inputs, given two values stored in the data register and a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 56 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_I32 OPR_GPUMEM 32 FMT_NUM_I32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_MIN_U32 GLOBAL_ATOMIC_UMIN Select the minimum of two unsigned 32-bit integer inputs, given two values stored in the data register and a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 57 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_MAX_I32 GLOBAL_ATOMIC_SMAX Select the maximum of two signed 32-bit integer inputs, given two values stored in the data register and a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 58 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_I32 OPR_GPUMEM 32 FMT_NUM_I32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_MAX_U32 GLOBAL_ATOMIC_UMAX Select the maximum of two unsigned 32-bit integer inputs, given two values stored in the data register and a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 59 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_AND_B32 GLOBAL_ATOMIC_AND Calculate bitwise AND given two unsigned 32-bit integer values stored in the data register and a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 60 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B32 OPR_GPUMEM 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_OR_B32 GLOBAL_ATOMIC_OR Calculate bitwise OR given two unsigned 32-bit integer values stored in the data register and a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 61 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B32 OPR_GPUMEM 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_XOR_B32 GLOBAL_ATOMIC_XOR Calculate bitwise XOR given two unsigned 32-bit integer values stored in the data register and a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 62 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B32 OPR_GPUMEM 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_INC_U32 GLOBAL_ATOMIC_INC Increment an unsigned 32-bit integer value from a location in the global aperture with wraparound to 0 if the value exceeds a value in the data register. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 63 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_DEC_U32 GLOBAL_ATOMIC_DEC Decrement an unsigned 32-bit integer value from a location in the global aperture with wraparound to a value in the data register if the decrement yields a negative value. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 64 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_SWAP_B64 GLOBAL_ATOMIC_SWAP_X2 Swap an unsigned 64-bit integer value in the data register with a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 65 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B64 OPR_GPUMEM 64 FMT_NUM_B64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_CMPSWAP_B64 GLOBAL_ATOMIC_CMPSWAP_X2 Compare two unsigned 64-bit integer values stored in the data comparison register and a location in the global aperture. Modify the memory location with a value in the data source register iff the comparison is equal. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 66 VDST FMT_NUM_U64 OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 128 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_ADD_U64 GLOBAL_ATOMIC_ADD_X2 Add two unsigned 64-bit integer values stored in the data register and a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 67 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_SUB_U64 GLOBAL_ATOMIC_SUB_X2 Subtract an unsigned 64-bit integer value stored in the data register from a value stored in a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 68 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_MIN_I64 GLOBAL_ATOMIC_SMIN_X2 Select the minimum of two signed 64-bit integer inputs, given two values stored in the data register and a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 69 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_I64 OPR_GPUMEM 64 FMT_NUM_I64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_MIN_U64 GLOBAL_ATOMIC_UMIN_X2 Select the minimum of two unsigned 64-bit integer inputs, given two values stored in the data register and a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 70 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_MAX_I64 GLOBAL_ATOMIC_SMAX_X2 Select the maximum of two signed 64-bit integer inputs, given two values stored in the data register and a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 71 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_I64 OPR_GPUMEM 64 FMT_NUM_I64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_MAX_U64 GLOBAL_ATOMIC_UMAX_X2 Select the maximum of two unsigned 64-bit integer inputs, given two values stored in the data register and a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 72 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_AND_B64 GLOBAL_ATOMIC_AND_X2 Calculate bitwise AND given two unsigned 64-bit integer values stored in the data register and a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 73 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B64 OPR_GPUMEM 64 FMT_NUM_B64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_OR_B64 GLOBAL_ATOMIC_OR_X2 Calculate bitwise OR given two unsigned 64-bit integer values stored in the data register and a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 74 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B64 OPR_GPUMEM 64 FMT_NUM_B64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_XOR_B64 GLOBAL_ATOMIC_XOR_X2 Calculate bitwise XOR given two unsigned 64-bit integer values stored in the data register and a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 75 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B64 OPR_GPUMEM 64 FMT_NUM_B64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_INC_U64 GLOBAL_ATOMIC_INC_X2 Increment an unsigned 64-bit integer value from a location in the global aperture with wraparound to 0 if the value exceeds a value in the data register. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 76 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_DEC_U64 GLOBAL_ATOMIC_DEC_X2 Decrement an unsigned 64-bit integer value from a location in the global aperture with wraparound to a value in the data register if the decrement yields a negative value. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 77 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U64 OPR_GPUMEM 64 FMT_NUM_U64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_WBINV Write back and invalidate cache lines based on the SCOPE field. Increments/decrements STORE_CNT. ENC_VGLOBAL default 79 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_COND_SUB_U32 Subtract an unsigned 32-bit integer value in the data register from a location in the global aperture only if the memory value is greater than or equal to the data register value. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 80 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_MIN_NUM_F32 GLOBAL_ATOMIC_FMIN GLOBAL_ATOMIC_MIN_F32 Select the IEEE minimumNumber() of two single-precision float inputs, given two values stored in the data register and a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 81 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_F32 OPR_GPUMEM 32 FMT_NUM_F32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_MAX_NUM_F32 GLOBAL_ATOMIC_FMAX GLOBAL_ATOMIC_MAX_F32 Select the IEEE maximumNumber() of two single-precision float inputs, given two values stored in the data register and a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 82 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_F32 OPR_GPUMEM 32 FMT_NUM_F32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_LOAD_BLOCK Load a block of data from the global aperture. ENC_VGLOBAL default 83 VDST FMT_ANY OPR_VGPR 1024 VADDR FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B1024 OPR_GPUMEM 1024 FMT_NUM_B32 OPR_SDST_M0 32 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_STORE_BLOCK Store a block of data to the global aperture. ENC_VGLOBAL default 84 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 1024 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B1024 OPR_GPUMEM 1024 FMT_NUM_B32 OPR_SDST_M0 32 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_ADD_F32 Add two single-precision float values stored in the data register and a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 86 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_F32 OPR_GPUMEM 32 FMT_NUM_F32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_LOAD_TR_B128 Load a 16x16 matrix of 16-bit data from the global aperture, transpose data between row-major and column-major order, and store the result into a 128-bit vector register. ENC_VGLOBAL default 87 VDST FMT_ANY OPR_VGPR 128 VADDR FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B128 OPR_GPUMEM 128 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_LOAD_TR_B64 Load a 16x16 matrix of 8-bit data from the global aperture, transpose data between row-major and column-major order, and store the result into a 64-bit vector register. ENC_VGLOBAL default 88 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B64 OPR_GPUMEM 64 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_PK_ADD_F16 Add a packed 2-component half-precision float value from the data register to a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 89 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_PK2_F16 OPR_GPUMEM 32 FMT_NUM_PK2_F16 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_PK_ADD_BF16 Add a packed 2-component BF16 float value from the data register to a location in the global aperture. Store the original value from global aperture into a vector register iff the temporal hint enables atomic return. ENC_VGLOBAL default 90 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_PK2_BF16 OPR_GPUMEM 32 FMT_NUM_PK2_BF16 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE GLOBAL_ATOMIC_ORDERED_ADD_B64 Given an (ID, value) pair in memory, increment the value by a given amount if the ID matches an ID provided by the shader. ENC_VGLOBAL default 115 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 64 VSRC FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 64 FMT_NUM_B64 OPR_GPUMEM 64 FMT_NUM_B64 OPR_GPUMEM 64 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE IMAGE_LOAD Load a texel from the largest miplevel in an image surface and store the result into a vector register. Perform the format conversion specified by the resource descriptor. No sampling is performed. ENC_VIMAGE default 0 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE IMAGE_LOAD_MIP Load a texel from a user-specified miplevel in an image surface and store the result into a vector register. Perform the format conversion specified by the resource descriptor. No sampling is performed. ENC_VIMAGE default 1 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE IMAGE_LOAD_PCK Load a texel from the largest miplevel in an image surface and store the result into a vector register. 8- and 16-bit components are zero-extended. The format specified in the resource descriptor is ignored. No sampling is performed. ENC_VIMAGE default 2 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE IMAGE_LOAD_PCK_SGN Load a texel from the largest miplevel in an image surface and store the result into a vector register. 8- and 16-bit components are sign-extended. The format specified in the resource descriptor is ignored. No sampling is performed. ENC_VIMAGE default 3 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE IMAGE_LOAD_MIP_PCK Load a texel from a user-specified miplevel in an image surface and store the result into a vector register. 8- and 16-bit components are zero-extended. The format specified in the resource descriptor is ignored. No sampling is performed. ENC_VIMAGE default 4 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE IMAGE_LOAD_MIP_PCK_SGN Load a texel from a user-specified miplevel in an image surface and store the result into a vector register. 8- and 16-bit components are sign-extended. The format specified in the resource descriptor is ignored. No sampling is performed. ENC_VIMAGE default 5 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE IMAGE_STORE Store a texel from a vector register to the largest miplevel in an image surface. The texel data is converted using the format conversion specified by the resource descriptor prior to storage. ENC_VIMAGE default 6 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE IMAGE_STORE_MIP Store a texel from a vector register to a user-specified miplevel in an image surface. The texel data is converted using the format conversion specified by the resource descriptor prior to storage. ENC_VIMAGE default 7 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE IMAGE_STORE_PCK Store a texel from a vector register to the largest miplevel in an image surface. The texel data is already packed and the format specified in the resource descriptor is ignored. ENC_VIMAGE default 8 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE IMAGE_STORE_MIP_PCK Store a texel from a vector register to a user-specified miplevel in an image surface. The texel data is already packed and the format specified in the resource descriptor is ignored. ENC_VIMAGE default 9 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE IMAGE_ATOMIC_SWAP Swap an unsigned 32-bit integer value in the data register with a location in an image surface. Store the original value from image surface into a vector register iff the temporal hint enables atomic return. ENC_VIMAGE default 10 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 FMT_NUM_B32 OPR_GPUMEM 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE IMAGE_ATOMIC_CMPSWAP Compare two unsigned 32-bit integer values stored in the data comparison register and a location in an image surface. Modify the memory location with a value in the data source register iff the comparison is equal. Store the original value from image surface into a vector register iff the temporal hint enables atomic return. ENC_VIMAGE default 11 VDATA FMT_NUM_U32 OPR_VGPR 32 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE IMAGE_ATOMIC_ADD_UINT IMAGE_ATOMIC_ADD Add two unsigned 32-bit integer values stored in the data register and a location in an image surface. Store the original value from image surface into a vector register iff the temporal hint enables atomic return. ENC_VIMAGE default 12 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE IMAGE_ATOMIC_SUB_UINT IMAGE_ATOMIC_SUB Subtract an unsigned 32-bit integer value stored in the data register from a value stored in a location in an image surface. Store the original value from image surface into a vector register iff the temporal hint enables atomic return. ENC_VIMAGE default 13 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE IMAGE_ATOMIC_MIN_INT IMAGE_ATOMIC_SMIN Select the minimum of two signed 32-bit integer inputs, given two values stored in the data register and a location in an image surface. Store the original value from image surface into a vector register iff the temporal hint enables atomic return. ENC_VIMAGE default 14 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 FMT_NUM_I32 OPR_GPUMEM 32 FMT_NUM_I32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE IMAGE_ATOMIC_MIN_UINT IMAGE_ATOMIC_UMIN Select the minimum of two unsigned 32-bit integer inputs, given two values stored in the data register and a location in an image surface. Store the original value from image surface into a vector register iff the temporal hint enables atomic return. ENC_VIMAGE default 15 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE IMAGE_ATOMIC_MAX_INT IMAGE_ATOMIC_SMAX Select the maximum of two signed 32-bit integer inputs, given two values stored in the data register and a location in an image surface. Store the original value from image surface into a vector register iff the temporal hint enables atomic return. ENC_VIMAGE default 16 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 FMT_NUM_I32 OPR_GPUMEM 32 FMT_NUM_I32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE IMAGE_ATOMIC_MAX_UINT IMAGE_ATOMIC_UMAX Select the maximum of two unsigned 32-bit integer inputs, given two values stored in the data register and a location in an image surface. Store the original value from image surface into a vector register iff the temporal hint enables atomic return. ENC_VIMAGE default 17 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE IMAGE_ATOMIC_AND Calculate bitwise AND given two unsigned 32-bit integer values stored in the data register and a location in an image surface. Store the original value from image surface into a vector register iff the temporal hint enables atomic return. ENC_VIMAGE default 18 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 FMT_NUM_B32 OPR_GPUMEM 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE IMAGE_ATOMIC_OR Calculate bitwise OR given two unsigned 32-bit integer values stored in the data register and a location in an image surface. Store the original value from image surface into a vector register iff the temporal hint enables atomic return. ENC_VIMAGE default 19 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 FMT_NUM_B32 OPR_GPUMEM 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE IMAGE_ATOMIC_XOR Calculate bitwise XOR given two unsigned 32-bit integer values stored in the data register and a location in an image surface. Store the original value from image surface into a vector register iff the temporal hint enables atomic return. ENC_VIMAGE default 20 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 FMT_NUM_B32 OPR_GPUMEM 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE IMAGE_ATOMIC_INC_UINT IMAGE_ATOMIC_INC Increment an unsigned 32-bit integer value from a location in an image surface with wraparound to 0 if the value exceeds a value in the data register. Store the original value from image surface into a vector register iff the temporal hint enables atomic return. ENC_VIMAGE default 21 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE IMAGE_ATOMIC_DEC_UINT IMAGE_ATOMIC_DEC Decrement an unsigned 32-bit integer value from a location in an image surface with wraparound to a value in the data register if the decrement yields a negative value. Store the original value from image surface into a vector register iff the temporal hint enables atomic return. ENC_VIMAGE default 22 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 FMT_NUM_U32 OPR_GPUMEM 32 FMT_NUM_U32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE IMAGE_GET_RESINFO Gather resource information for a given miplevel provided in the address register. Returns 4 integer values into registers 3:0 as { num_mip_levels, depth, height, width }. No memory access is performed. ENC_VIMAGE default 23 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 32 RSRC FMT_IMG OPR_SREG 256 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE IMAGE_BVH_INTERSECT_RAY BVH_INTERSECT_RAY Test the intersection of rays with either box nodes or triangle nodes within a bounded volume hierarchy using 32 bit node pointers. Store the results of the test into a vector register. This instruction does not take a sampler constant. ENC_VIMAGE default 25 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 352 RSRC FMT_IMG_BVH OPR_SREG 128 VMEM BVH FALSE FALSE FALSE FALSE FALSE IMAGE_BVH64_INTERSECT_RAY BVH64_INTERSECT_RAY Test the intersection of rays with either box nodes or triangle nodes within a bounded volume hierarchy using 64 bit node pointers. Store the results of the test into a vector register. This instruction does not take a sampler constant. ENC_VIMAGE default 26 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 384 RSRC FMT_IMG_BVH OPR_SREG 128 VMEM BVH FALSE FALSE FALSE FALSE FALSE IMAGE_BVH_DUAL_INTERSECT_RAY BVH_DUAL_INTERSECT_RAY This instruction supports testing two QBVH nodes against the same ray per lane using both intersection engines. It is typically used to implement the BVH4x2 traversal algorithm. ENC_VIMAGE default 128 VDATA FMT_ANY OPR_VGPR 320 FMT_ANY OPR_VGPR 384 RSRC FMT_IMG_BVH OPR_SREG 128 VMEM BVH FALSE FALSE FALSE FALSE FALSE IMAGE_BVH8_INTERSECT_RAY BVH8_INTERSECT_RAY This instruction supports testing one BVH8 node against one ray per lane using both intersection engines. ENC_VIMAGE default 129 VDATA FMT_ANY OPR_VGPR 320 FMT_ANY OPR_VGPR 352 RSRC FMT_IMG_BVH OPR_SREG 128 VMEM BVH FALSE FALSE FALSE FALSE FALSE IMAGE_ATOMIC_ADD_FLT Add two single-precision float values stored in the data register and a location in an image surface. Store the original value from image surface into a vector register iff the temporal hint enables atomic return. ENC_VIMAGE default 131 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 FMT_NUM_F32 OPR_GPUMEM 32 FMT_NUM_F32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE IMAGE_ATOMIC_MIN_FLT IMAGE_ATOMIC_FMIN IMAGE_ATOMIC_MIN_NUM_FLT Select the IEEE minimumNumber() of two single-precision float inputs, given two values stored in the data register and a location in an image surface. Store the original value from image surface into a vector register iff the temporal hint enables atomic return. ENC_VIMAGE default 132 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 FMT_NUM_F32 OPR_GPUMEM 32 FMT_NUM_F32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE IMAGE_ATOMIC_MAX_FLT IMAGE_ATOMIC_FMAX IMAGE_ATOMIC_MAX_NUM_FLT Select the IEEE maximumNumber() of two single-precision float inputs, given two values stored in the data register and a location in an image surface. Store the original value from image surface into a vector register iff the temporal hint enables atomic return. ENC_VIMAGE default 133 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 FMT_NUM_F32 OPR_GPUMEM 32 FMT_NUM_F32 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE IMAGE_ATOMIC_PK_ADD_F16 Add a packed 2-component half-precision float value from the data register to a location in an image surface. Store the original value from image surface into a vector register iff the temporal hint enables atomic return. ENC_VIMAGE default 134 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 FMT_NUM_PK2_F16 OPR_GPUMEM 32 FMT_NUM_PK2_F16 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE IMAGE_ATOMIC_PK_ADD_BF16 Add a packed 2-component BF16 float value from the data register to a location in an image surface. Store the original value from image surface into a vector register iff the temporal hint enables atomic return. ENC_VIMAGE default 135 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 FMT_NUM_PK2_BF16 OPR_GPUMEM 32 FMT_NUM_PK2_BF16 OPR_GPUMEM 32 VMEM ATOMIC FALSE FALSE FALSE FALSE FALSE V_INTERP_P10_F32 Given the P10 parameter of an attribute, the I coordinate and the P0 parameter as single-precision float inputs, compute the first part of parameter interpolation and store the intermediate result into a vector register. Use V_INTERP_P2_F32 to complete the operation. ENC_VINTERP default 0 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC_VGPR 32 SRC2 FMT_NUM_F32 OPR_SRC_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_INTERP_P2_F32 V_INTERP_P2_NEW_F32 Given the P20 parameter of an attribute, the J coordinate and the result of a prior V_INTERP_P10_F32 instruction as single-precision float inputs, compute the second part of parameter interpolation and store the final result into a vector register. ENC_VINTERP default 1 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC_VGPR 32 SRC2 FMT_NUM_F32 OPR_SRC_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_INTERP_P10_F16_F32 Given a half-precision float P10 parameter of an attribute, a single-precision float I coordinate and a half-precision float P0 parameter as inputs, compute the first part of parameter interpolation and store the intermediate result in single-precision float format into a vector register. Use V_INTERP_P2_F16_F32 to complete the operation. ENC_VINTERP default 2 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F16 OPR_SRC_VGPR 16 SRC1 FMT_NUM_F32 OPR_SRC_VGPR 32 SRC2 FMT_NUM_F16 OPR_SRC_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_INTERP_P2_F16_F32 Given a half-precision float P20 parameter of an attribute, a single-precision float J coordinate and the result of a prior V_INTERP_P10_F16_F32 instruction as inputs, compute the second part of parameter interpolation and store the final result into a vector register. ENC_VINTERP default 3 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC_VGPR 16 SRC1 FMT_NUM_F32 OPR_SRC_VGPR 32 SRC2 FMT_NUM_F32 OPR_SRC_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_INTERP_P10_RTZ_F16_F32 Given a half-precision float P10 parameter of an attribute, a single-precision float I coordinate and a half-precision float P0 parameter as inputs, compute the first part of parameter interpolation using round toward zero semantics and store the intermediate result in single-precision float format into a vector register. Use V_INTERP_P2_RTZ_F16_F32 to complete the operation. ENC_VINTERP default 4 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F16 OPR_SRC_VGPR 16 SRC1 FMT_NUM_F32 OPR_SRC_VGPR 32 SRC2 FMT_NUM_F16 OPR_SRC_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_INTERP_P2_RTZ_F16_F32 Given a half-precision float P20 parameter of an attribute, a single-precision float J coordinate and the result of a prior V_INTERP_P10_RTZ_F16_F32 instruction as inputs, compute the second part of parameter interpolation using round toward zero semantics and store the final result into a vector register. ENC_VINTERP default 5 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC_VGPR 16 SRC1 FMT_NUM_F32 OPR_SRC_VGPR 32 SRC2 FMT_NUM_F32 OPR_SRC_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_NOP Do nothing. ENC_VOP1 default 0 ENC_VOP3 default 384 VALU FALSE FALSE FALSE FALSE FALSE V_MOV_B32 Move 32-bit data from a vector input into a vector register. VOP1_VOP_DPP16 default 1 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 VOP1_VOP_DPP8 default 1 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 ENC_VOP1 default 1 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 1 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 ENC_VOP3 default 385 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 385 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_VOP_DPP16 default 385 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 VOP3_VOP_DPP8 default 385 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_READFIRSTLANE_B32 Read the scalar value in the lowest active lane of the input vector register and store it into a scalar register. ENC_VOP1 default 2 VDST FMT_NUM_B32 OPR_SREG 32 SRC0 FMT_NUM_B32 OPR_VGPR 32 ENC_VOP3 default 386 VDST FMT_NUM_B32 OPR_SREG 32 SRC0 FMT_NUM_B32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_I32_F64 Convert from a double-precision float input to a signed 32-bit integer value and store the result into a vector register. ENC_VOP1 default 3 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP1_INST_LITERAL has_lit 3 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOP3 default 387 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 387 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_F64_I32 Convert from a signed 32-bit integer input to a double-precision float value and store the result into a vector register. ENC_VOP1 default 4 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_I32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 4 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 ENC_VOP3 default 388 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 388 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_F32_I32 Convert from a signed 32-bit integer input to a single-precision float value and store the result into a vector register. VOP1_VOP_DPP16 default 5 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VOP1_VOP_DPP8 default 5 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 ENC_VOP1 default 5 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 5 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 ENC_VOP3 default 389 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 389 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 389 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VOP3_VOP_DPP8 default 389 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_F32_U32 Convert from an unsigned 32-bit integer input to a single-precision float value and store the result into a vector register. VOP1_VOP_DPP16 default 6 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VOP1_VOP_DPP8 default 6 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP1 default 6 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 6 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 ENC_VOP3 default 390 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 390 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 390 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VOP3_VOP_DPP8 default 390 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_U32_F32 Convert from a single-precision float input to an unsigned 32-bit integer value and store the result into a vector register. VOP1_VOP_DPP16 default 7 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 7 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 7 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 7 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 391 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 391 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 391 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 391 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_I32_F32 Convert from a single-precision float input to a signed 32-bit integer value and store the result into a vector register. VOP1_VOP_DPP16 default 8 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 8 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 8 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 8 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 392 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 392 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 392 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 392 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_F16_F32 Convert from a single-precision float input to a half-precision float value and store the result into a vector register. VOP1_VOP_DPP16 default 10 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 10 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 10 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 10 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 394 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 394 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 394 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 394 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_F32_F16 Convert from a half-precision float input to a single-precision float value and store the result into a vector register. VOP1_VOP_DPP16 default 11 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP1_VOP_DPP8 default 11 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP1 default 11 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 11 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F16 OPR_SRC 16 ENC_VOP3 default 395 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 395 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 395 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 395 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_NEAREST_I32_F32 V_CVT_RPI_I32_F32 Convert from a single-precision float input to a signed 32-bit integer value using round to nearest integer semantics (ignore the default rounding mode) and store the result into a vector register. VOP1_VOP_DPP16 default 12 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 12 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 12 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 12 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 396 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 396 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 396 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 396 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_FLOOR_I32_F32 V_CVT_FLR_I32_F32 Convert from a single-precision float input to a signed 32-bit integer value using round-down semantics (ignore the default rounding mode) and store the result into a vector register. VOP1_VOP_DPP16 default 13 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 13 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 13 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 13 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 397 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 397 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 397 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 397 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_OFF_F32_I4 Convert from a signed 4-bit integer input to a single-precision float value using an offset table and store the result into a vector register. VOP1_VOP_DPP16 default 14 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VOP1_VOP_DPP8 default 14 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 ENC_VOP1 default 14 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 14 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 ENC_VOP3 default 398 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 398 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 398 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VOP3_VOP_DPP8 default 398 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_F32_F64 Convert from a double-precision float input to a single-precision float value and store the result into a vector register. ENC_VOP1 default 15 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP1_INST_LITERAL has_lit 15 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOP3 default 399 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 399 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_F64_F32 Convert from a single-precision float input to a double-precision float value and store the result into a vector register. ENC_VOP1 default 16 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 16 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 400 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 400 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_F32_UBYTE0 Convert an unsigned byte in byte 0 of the input to a single-precision float value and store the result into a vector register. VOP1_VOP_DPP16 default 17 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VOP1_VOP_DPP8 default 17 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP1 default 17 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 17 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 ENC_VOP3 default 401 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 401 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 401 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VOP3_VOP_DPP8 default 401 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_F32_UBYTE1 Convert an unsigned byte in byte 1 of the input to a single-precision float value and store the result into a vector register. VOP1_VOP_DPP16 default 18 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VOP1_VOP_DPP8 default 18 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP1 default 18 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 18 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 ENC_VOP3 default 402 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 402 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 402 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VOP3_VOP_DPP8 default 402 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_F32_UBYTE2 Convert an unsigned byte in byte 2 of the input to a single-precision float value and store the result into a vector register. VOP1_VOP_DPP16 default 19 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VOP1_VOP_DPP8 default 19 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP1 default 19 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 19 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 ENC_VOP3 default 403 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 403 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 403 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VOP3_VOP_DPP8 default 403 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_F32_UBYTE3 Convert an unsigned byte in byte 3 of the input to a single-precision float value and store the result into a vector register. VOP1_VOP_DPP16 default 20 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VOP1_VOP_DPP8 default 20 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP1 default 20 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 20 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 ENC_VOP3 default 404 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 404 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 404 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VOP3_VOP_DPP8 default 404 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_U32_F64 Convert from a double-precision float input to an unsigned 32-bit integer value and store the result into a vector register. ENC_VOP1 default 21 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP1_INST_LITERAL has_lit 21 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOP3 default 405 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 405 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_F64_U32 Convert from an unsigned 32-bit integer input to a double-precision float value and store the result into a vector register. ENC_VOP1 default 22 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_U32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 22 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 ENC_VOP3 default 406 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 406 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_TRUNC_F64 Compute the integer part of a double-precision float input using round toward zero semantics and store the result in floating point format into a vector register. ENC_VOP1 default 23 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP1_INST_LITERAL has_lit 23 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOP3 default 407 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 407 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CEIL_F64 Round the double-precision float input up to next integer and store the result in floating point format into a vector register. ENC_VOP1 default 24 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP1_INST_LITERAL has_lit 24 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOP3 default 408 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 408 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_RNDNE_F64 Round the double-precision float input to the nearest even integer and store the result in floating point format into a vector register. ENC_VOP1 default 25 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP1_INST_LITERAL has_lit 25 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOP3 default 409 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 409 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_FLOOR_F64 Round the double-precision float input down to previous integer and store the result in floating point format into a vector register. ENC_VOP1 default 26 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP1_INST_LITERAL has_lit 26 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOP3 default 410 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 410 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_PIPEFLUSH Flush the vector ALU pipeline through the destination cache. ENC_VOP1 default 27 ENC_VOP3 default 411 VALU FALSE FALSE FALSE FALSE FALSE V_MOV_B16 Move 16-bit data from a vector input into a vector register. VOP1_VOP_DPP16 default 28 VDST FMT_NUM_B16 OPR_VGPR 16 VSRC0 FMT_NUM_B16 OPR_VGPR 16 VOP1_VOP_DPP8 default 28 VDST FMT_NUM_B16 OPR_VGPR 16 VSRC0 FMT_NUM_B16 OPR_VGPR 16 ENC_VOP1 default 28 VDST FMT_NUM_B16 OPR_VGPR 16 SRC0 FMT_NUM_B16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 28 VDST FMT_NUM_B16 OPR_VGPR 16 SIMM32 FMT_NUM_B16 OPR_SRC 16 ENC_VOP3 default 412 VDST FMT_NUM_B16 OPR_VGPR 16 SRC0 FMT_NUM_B16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 412 VDST FMT_NUM_B16 OPR_VGPR 16 SIMM32 FMT_NUM_B16 OPR_SRC 16 VOP3_VOP_DPP16 default 412 VDST FMT_NUM_B16 OPR_VGPR 16 VSRC0 FMT_NUM_B16 OPR_VGPR 16 VOP3_VOP_DPP8 default 412 VDST FMT_NUM_B16 OPR_VGPR 16 VSRC0 FMT_NUM_B16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_FRACT_F32 Compute the fractional portion of a single-precision float input and store the result in floating point format into a vector register. VOP1_VOP_DPP16 default 32 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 32 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 32 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 32 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 416 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 416 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 416 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 416 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_TRUNC_F32 Compute the integer part of a single-precision float input using round toward zero semantics and store the result in floating point format into a vector register. VOP1_VOP_DPP16 default 33 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 33 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 33 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 33 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 417 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 417 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 417 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 417 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CEIL_F32 Round the single-precision float input up to next integer and store the result in floating point format into a vector register. VOP1_VOP_DPP16 default 34 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 34 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 34 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 34 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 418 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 418 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 418 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 418 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_RNDNE_F32 Round the single-precision float input to the nearest even integer and store the result in floating point format into a vector register. VOP1_VOP_DPP16 default 35 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 35 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 35 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 35 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 419 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 419 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 419 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 419 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_FLOOR_F32 Round the single-precision float input down to previous integer and store the result in floating point format into a vector register. VOP1_VOP_DPP16 default 36 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 36 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 36 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 36 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 420 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 420 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 420 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 420 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_EXP_F32 Calculate 2 raised to the power of the single-precision float input and store the result into a vector register. VOP1_VOP_DPP16 default 37 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 37 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 37 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 37 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 421 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 421 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 421 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 421 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_LOG_F32 Calculate the base 2 logarithm of the single-precision float input and store the result into a vector register. VOP1_VOP_DPP16 default 39 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 39 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 39 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 39 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 423 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 423 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 423 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 423 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_RCP_F32 Calculate the reciprocal of the single-precision float input using IEEE rules and store the result into a vector register. VOP1_VOP_DPP16 default 42 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 42 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 42 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 42 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 426 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 426 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 426 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 426 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_RCP_IFLAG_F32 Calculate the reciprocal of the vector float input in a manner suitable for integer division and store the result into a vector register. This opcode is intended for use as part of an integer division macro. VOP1_VOP_DPP16 default 43 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 43 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 43 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 43 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 427 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 427 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 427 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 427 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_RSQ_F32 Calculate the reciprocal of the square root of the single-precision float input using IEEE rules and store the result into a vector register. VOP1_VOP_DPP16 default 46 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 46 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 46 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 46 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 430 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 430 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 430 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 430 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_RCP_F64 Calculate the reciprocal of the double-precision float input using IEEE rules and store the result into a vector register. ENC_VOP1 default 47 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP1_INST_LITERAL has_lit 47 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOP3 default 431 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 431 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_RSQ_F64 Calculate the reciprocal of the square root of the double-precision float input using IEEE rules and store the result into a vector register. ENC_VOP1 default 49 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP1_INST_LITERAL has_lit 49 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOP3 default 433 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 433 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_SQRT_F32 Calculate the square root of the single-precision float input using IEEE rules and store the result into a vector register. VOP1_VOP_DPP16 default 51 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 51 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 51 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 51 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 435 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 435 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 435 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 435 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_SQRT_F64 Calculate the square root of the double-precision float input using IEEE rules and store the result into a vector register. ENC_VOP1 default 52 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP1_INST_LITERAL has_lit 52 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOP3 default 436 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 436 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_SIN_F32 Calculate the trigonometric sine of a single-precision float value using IEEE rules and store the result into a vector register. The operand is calculated by scaling the vector input by 2 PI. VOP1_VOP_DPP16 default 53 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 53 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 53 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 53 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 437 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 437 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 437 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 437 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_COS_F32 Calculate the trigonometric cosine of a single-precision float value using IEEE rules and store the result into a vector register. The operand is calculated by scaling the vector input by 2 PI. VOP1_VOP_DPP16 default 54 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 54 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 54 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 54 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 438 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 438 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 438 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 438 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_NOT_B32 Calculate bitwise negation on a vector input and store the result into a vector register. VOP1_VOP_DPP16 default 55 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VOP1_VOP_DPP8 default 55 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP1 default 55 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 55 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 ENC_VOP3 default 439 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 439 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 439 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VOP3_VOP_DPP8 default 439 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_BFREV_B32 Reverse the order of bits in a vector input and store the result into a vector register. VOP1_VOP_DPP16 default 56 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 VOP1_VOP_DPP8 default 56 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 ENC_VOP1 default 56 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 56 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 ENC_VOP3 default 440 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 440 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_VOP_DPP16 default 440 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 VOP3_VOP_DPP8 default 440 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CLZ_I32_U32 V_FFBH_U32 Count the number of leading "0" bits before the first "1" in a vector input and store the result into a vector register. Store -1 if there are no "1" bits. VOP1_VOP_DPP16 default 57 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VOP1_VOP_DPP8 default 57 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP1 default 57 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 57 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 ENC_VOP3 default 441 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 441 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 441 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VOP3_VOP_DPP8 default 441 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CTZ_I32_B32 V_FFBL_B32 Count the number of trailing "0" bits before the first "1" in a vector input and store the result into a vector register. Store -1 if there are no "1" bits in the input. VOP1_VOP_DPP16 default 58 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VOP1_VOP_DPP8 default 58 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP1 default 58 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 58 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 ENC_VOP3 default 442 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 442 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 442 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VOP3_VOP_DPP8 default 442 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CLS_I32 V_FFBH_I32 Count the number of leading bits that are the same as the sign bit of a vector input and store the result into a vector register. Store -1 if all input bits are the same. VOP1_VOP_DPP16 default 59 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VOP1_VOP_DPP8 default 59 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 ENC_VOP1 default 59 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 59 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 ENC_VOP3 default 443 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 443 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 443 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VOP3_VOP_DPP8 default 443 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_FREXP_EXP_I32_F64 Extract the exponent of a double-precision float input and store the result as a signed 32-bit integer into a vector register. ENC_VOP1 default 60 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP1_INST_LITERAL has_lit 60 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOP3 default 444 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 444 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_FREXP_MANT_F64 Extract the binary significand, or mantissa, of a double-precision float input and store the result as a double-precision float into a vector register. ENC_VOP1 default 61 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP1_INST_LITERAL has_lit 61 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOP3 default 445 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 445 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_FRACT_F64 Compute the fractional portion of a double-precision float input and store the result in floating point format into a vector register. ENC_VOP1 default 62 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP1_INST_LITERAL has_lit 62 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOP3 default 446 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 446 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_FREXP_EXP_I32_F32 Extract the exponent of a single-precision float input and store the result as a signed 32-bit integer into a vector register. VOP1_VOP_DPP16 default 63 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 63 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 63 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 63 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 447 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 447 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 447 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 447 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_FREXP_MANT_F32 Extract the binary significand, or mantissa, of a single-precision float input and store the result as a single-precision float into a vector register. VOP1_VOP_DPP16 default 64 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP1_VOP_DPP8 default 64 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP1 default 64 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP1_INST_LITERAL has_lit 64 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 ENC_VOP3 default 448 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 448 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 448 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 448 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_MOVRELD_B32 Move data from a vector input into a relatively-indexed vector register. VOP1_VOP_DPP16 default 66 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 VOP1_VOP_DPP8 default 66 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 ENC_VOP1 default 66 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 FMT_NUM_B32 OPR_SDST_M0 32 VOP1_INST_LITERAL has_lit 66 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 FMT_NUM_B32 OPR_SDST_M0 32 ENC_VOP3 default 450 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 FMT_NUM_B32 OPR_SDST_M0 32 VOP3_INST_LITERAL has_lit_0 450 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 FMT_NUM_B32 OPR_SDST_M0 32 VOP3_VOP_DPP16 default 450 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 VOP3_VOP_DPP8 default 450 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 VALU FALSE FALSE FALSE FALSE FALSE V_MOVRELS_B32 Move data from a relatively-indexed vector register into another vector register. VOP1_VOP_DPP16 default 67 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 VOP1_VOP_DPP8 default 67 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 ENC_VOP1 default 67 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 ENC_VOP3 default 451 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 VOP3_VOP_DPP16 default 451 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 VOP3_VOP_DPP8 default 451 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 VALU FALSE FALSE FALSE FALSE FALSE V_MOVRELSD_B32 Move data from a relatively-indexed vector register into another relatively-indexed vector register. VOP1_VOP_DPP16 default 68 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 VOP1_VOP_DPP8 default 68 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 ENC_VOP1 default 68 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 ENC_VOP3 default 452 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 VOP3_VOP_DPP16 default 452 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 VOP3_VOP_DPP8 default 452 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 VALU FALSE FALSE FALSE FALSE FALSE V_MOVRELSD_2_B32 Move data from a relatively-indexed vector register into another relatively-indexed vector register, using different offsets for each index. VOP1_VOP_DPP16 default 72 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 VOP1_VOP_DPP8 default 72 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 ENC_VOP1 default 72 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 ENC_VOP3 default 456 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 VOP3_VOP_DPP16 default 456 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 VOP3_VOP_DPP8 default 456 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_F16_U16 Convert from an unsigned 16-bit integer input to a half-precision float value and store the result into a vector register. VOP1_VOP_DPP16 default 80 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VOP1_VOP_DPP8 default 80 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 ENC_VOP1 default 80 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 80 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 ENC_VOP3 default 464 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 464 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 464 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VOP3_VOP_DPP8 default 464 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_F16_I16 Convert from a signed 16-bit integer input to a half-precision float value and store the result into a vector register. VOP1_VOP_DPP16 default 81 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VOP1_VOP_DPP8 default 81 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 ENC_VOP1 default 81 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 81 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 ENC_VOP3 default 465 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 465 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP16 default 465 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VOP3_VOP_DPP8 default 465 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_U16_F16 Convert from a half-precision float input to an unsigned 16-bit integer value and store the result into a vector register. VOP1_VOP_DPP16 default 82 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP1_VOP_DPP8 default 82 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP1 default 82 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 82 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 ENC_VOP3 default 466 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 466 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 466 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 466 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_I16_F16 Convert from a half-precision float input to a signed 16-bit integer value and store the result into a vector register. VOP1_VOP_DPP16 default 83 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP1_VOP_DPP8 default 83 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP1 default 83 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 83 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 ENC_VOP3 default 467 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 467 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 467 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 467 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_RCP_F16 Calculate the reciprocal of the half-precision float input using IEEE rules and store the result into a vector register. VOP1_VOP_DPP16 default 84 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP1_VOP_DPP8 default 84 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP1 default 84 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 84 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 ENC_VOP3 default 468 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 468 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 468 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 468 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_SQRT_F16 Calculate the square root of the half-precision float input using IEEE rules and store the result into a vector register. VOP1_VOP_DPP16 default 85 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP1_VOP_DPP8 default 85 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP1 default 85 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 85 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 ENC_VOP3 default 469 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 469 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 469 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 469 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_RSQ_F16 Calculate the reciprocal of the square root of the half-precision float input using IEEE rules and store the result into a vector register. VOP1_VOP_DPP16 default 86 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP1_VOP_DPP8 default 86 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP1 default 86 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 86 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 ENC_VOP3 default 470 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 470 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 470 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 470 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_LOG_F16 Calculate the base 2 logarithm of the half-precision float input and store the result into a vector register. VOP1_VOP_DPP16 default 87 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP1_VOP_DPP8 default 87 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP1 default 87 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 87 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 ENC_VOP3 default 471 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 471 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 471 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 471 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_EXP_F16 Calculate 2 raised to the power of the half-precision float input and store the result into a vector register. VOP1_VOP_DPP16 default 88 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP1_VOP_DPP8 default 88 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP1 default 88 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 88 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 ENC_VOP3 default 472 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 472 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 472 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 472 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_FREXP_MANT_F16 Extract the binary significand, or mantissa, of a half-precision float input and store the result as a half-precision float into a vector register. VOP1_VOP_DPP16 default 89 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP1_VOP_DPP8 default 89 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP1 default 89 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 89 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 ENC_VOP3 default 473 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 473 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 473 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 473 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_FREXP_EXP_I16_F16 Extract the exponent of a half-precision float input and store the result as a signed 16-bit integer into a vector register. VOP1_VOP_DPP16 default 90 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP1_VOP_DPP8 default 90 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP1 default 90 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 90 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 ENC_VOP3 default 474 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 474 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 474 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 474 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_FLOOR_F16 Round the half-precision float input down to previous integer and store the result in floating point format into a vector register. VOP1_VOP_DPP16 default 91 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP1_VOP_DPP8 default 91 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP1 default 91 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 91 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 ENC_VOP3 default 475 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 475 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 475 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 475 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CEIL_F16 Round the half-precision float input up to next integer and store the result in floating point format into a vector register. VOP1_VOP_DPP16 default 92 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP1_VOP_DPP8 default 92 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP1 default 92 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 92 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 ENC_VOP3 default 476 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 476 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 476 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 476 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_TRUNC_F16 Compute the integer part of a half-precision float input using round toward zero semantics and store the result in floating point format into a vector register. VOP1_VOP_DPP16 default 93 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP1_VOP_DPP8 default 93 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP1 default 93 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 93 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 ENC_VOP3 default 477 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 477 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 477 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 477 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_RNDNE_F16 Round the half-precision float input to the nearest even integer and store the result in floating point format into a vector register. VOP1_VOP_DPP16 default 94 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP1_VOP_DPP8 default 94 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP1 default 94 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 94 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 ENC_VOP3 default 478 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 478 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 478 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 478 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_FRACT_F16 Compute the fractional portion of a half-precision float input and store the result in floating point format into a vector register. VOP1_VOP_DPP16 default 95 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP1_VOP_DPP8 default 95 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP1 default 95 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 95 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 ENC_VOP3 default 479 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 479 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 479 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 479 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_SIN_F16 Calculate the trigonometric sine of a half-precision float value using IEEE rules and store the result into a vector register. The operand is calculated by scaling the vector input by 2 PI. VOP1_VOP_DPP16 default 96 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP1_VOP_DPP8 default 96 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP1 default 96 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 96 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 ENC_VOP3 default 480 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 480 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 480 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 480 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_COS_F16 Calculate the trigonometric cosine of a half-precision float value using IEEE rules and store the result into a vector register. The operand is calculated by scaling the vector input by 2 PI. VOP1_VOP_DPP16 default 97 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP1_VOP_DPP8 default 97 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP1 default 97 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 97 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 ENC_VOP3 default 481 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 481 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 481 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 481 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_SAT_PK_U8_I16 Given two 16-bit signed integer inputs, saturate each input over an 8-bit unsigned range, pack the resulting values into a 16-bit word and store the result into a vector register. VOP1_VOP_DPP16 default 98 VDST FMT_NUM_PK2_U8 OPR_VGPR 16 VSRC0 FMT_NUM_PK2_I16 OPR_VGPR 32 VOP1_VOP_DPP8 default 98 VDST FMT_NUM_PK2_U8 OPR_VGPR 16 VSRC0 FMT_NUM_PK2_I16 OPR_VGPR 32 ENC_VOP1 default 98 VDST FMT_NUM_PK2_U8 OPR_VGPR 16 SRC0 FMT_NUM_PK2_I16 OPR_SRC 32 VOP1_INST_LITERAL has_lit 98 VDST FMT_NUM_PK2_U8 OPR_VGPR 16 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 ENC_VOP3 default 482 VDST FMT_NUM_PK2_U8 OPR_VGPR 16 SRC0 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 482 VDST FMT_NUM_PK2_U8 OPR_VGPR 16 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3_VOP_DPP16 default 482 VDST FMT_NUM_PK2_U8 OPR_VGPR 16 VSRC0 FMT_NUM_PK2_I16 OPR_VGPR 32 VOP3_VOP_DPP8 default 482 VDST FMT_NUM_PK2_U8 OPR_VGPR 16 VSRC0 FMT_NUM_PK2_I16 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_NORM_I16_F16 Convert from a half-precision float input to a signed normalized short and store the result into a vector register. VOP1_VOP_DPP16 default 99 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP1_VOP_DPP8 default 99 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP1 default 99 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 99 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 ENC_VOP3 default 483 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 483 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 483 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 483 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_NORM_U16_F16 Convert from a half-precision float input to an unsigned normalized short and store the result into a vector register. VOP1_VOP_DPP16 default 100 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP1_VOP_DPP8 default 100 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP1 default 100 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 100 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 ENC_VOP3 default 484 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 484 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 484 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 484 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_SWAP_B32 Swap the values in two vector registers. ENC_VOP1 default 101 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_SWAP_B16 Swap the values in two vector registers. ENC_VOP1 default 102 VDST FMT_NUM_B16 OPR_VGPR 16 SRC0 FMT_NUM_B16 OPR_SRC_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_PERMLANE64_B32 Perform a specific permutation across lanes where the high half and low half of a wave64 are swapped. Performs no operation in wave32 mode. ENC_VOP1 default 103 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 VOP1_INST_LITERAL has_lit 103 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_SWAPREL_B32 Swap the values in two relatively-indexed vector registers. ENC_VOP1 default 104 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 FMT_NUM_B32 OPR_SDST_M0 32 VALU FALSE FALSE FALSE FALSE FALSE V_NOT_B16 Calculate bitwise negation on a vector input and store the result into a vector register. VOP1_VOP_DPP16 default 105 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VOP1_VOP_DPP8 default 105 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 ENC_VOP1 default 105 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 105 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 ENC_VOP3 default 489 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 489 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 489 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VOP3_VOP_DPP8 default 489 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_I32_I16 Convert from a signed 16-bit integer input to a signed 32-bit integer value using sign extension and store the result into a vector register. VOP1_VOP_DPP16 default 106 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VOP1_VOP_DPP8 default 106 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I16 OPR_VGPR 16 ENC_VOP1 default 106 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 106 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I16 OPR_SRC 16 ENC_VOP3 default 490 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 490 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP16 default 490 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VOP3_VOP_DPP8 default 490 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_U32_U16 Convert from an unsigned 16-bit integer input to an unsigned 32-bit integer value using zero extension and store the result into a vector register. VOP1_VOP_DPP16 default 107 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VOP1_VOP_DPP8 default 107 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U16 OPR_VGPR 16 ENC_VOP1 default 107 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U16 OPR_SRC 16 VOP1_INST_LITERAL has_lit 107 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U16 OPR_SRC 16 ENC_VOP3 default 491 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 491 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 491 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VOP3_VOP_DPP8 default 491 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_F32_FP8 Convert from an FP8 float input to a single-precision float value and store the result into a vector register. VOP1_VOP_DPP16 default 108 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_FP8 OPR_VGPR 8 VOP1_VOP_DPP8 default 108 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_FP8 OPR_VGPR 8 ENC_VOP1 default 108 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_FP8 OPR_SRC 8 VOP1_INST_LITERAL has_lit 108 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_FP8 OPR_SRC 8 ENC_VOP3 default 492 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_FP8 OPR_SRC 8 VOP3_INST_LITERAL has_lit_0 492 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_FP8 OPR_SRC 8 VOP3_VOP_DPP16 default 492 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_FP8 OPR_VGPR 8 VOP3_VOP_DPP8 default 492 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_FP8 OPR_VGPR 8 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_F32_BF8 Convert from a BF8 float input to a single-precision float value and store the result into a vector register. VOP1_VOP_DPP16 default 109 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_BF8 OPR_VGPR 8 VOP1_VOP_DPP8 default 109 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_BF8 OPR_VGPR 8 ENC_VOP1 default 109 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_BF8 OPR_SRC 8 VOP1_INST_LITERAL has_lit 109 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_BF8 OPR_SRC 8 ENC_VOP3 default 493 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_BF8 OPR_SRC 8 VOP3_INST_LITERAL has_lit_0 493 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_BF8 OPR_SRC 8 VOP3_VOP_DPP16 default 493 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_BF8 OPR_VGPR 8 VOP3_VOP_DPP8 default 493 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_BF8 OPR_VGPR 8 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_PK_F32_FP8 Convert from a packed 2-component FP8 float input to a packed single-precision float value and store the result into a vector register. ENC_VOP1 default 110 VDST FMT_NUM_PK2_F32 OPR_VGPR 64 SRC0 FMT_NUM_PK2_FP8 OPR_SRC 16 VOP1_INST_LITERAL has_lit 110 VDST FMT_NUM_PK2_F32 OPR_VGPR 64 SIMM32 FMT_NUM_PK2_FP8 OPR_SRC 16 ENC_VOP3 default 494 VDST FMT_NUM_PK2_F32 OPR_VGPR 64 SRC0 FMT_NUM_PK2_FP8 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 494 VDST FMT_NUM_PK2_F32 OPR_VGPR 64 SIMM32 FMT_NUM_PK2_FP8 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_PK_F32_BF8 Convert from a packed 2-component BF8 float input to a packed single-precision float value and store the result into a vector register. ENC_VOP1 default 111 VDST FMT_NUM_PK2_F32 OPR_VGPR 64 SRC0 FMT_NUM_PK2_BF8 OPR_SRC 16 VOP1_INST_LITERAL has_lit 111 VDST FMT_NUM_PK2_F32 OPR_VGPR 64 SIMM32 FMT_NUM_PK2_BF8 OPR_SRC 16 ENC_VOP3 default 495 VDST FMT_NUM_PK2_F32 OPR_VGPR 64 SRC0 FMT_NUM_PK2_BF8 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 495 VDST FMT_NUM_PK2_F32 OPR_VGPR 64 SIMM32 FMT_NUM_PK2_BF8 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_CNDMASK_B32 Copy data from one of two inputs based on the per-lane condition code and store the result into a vector register. VOP2_VOP_DPP16 default 1 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 VSRC1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VOP2_VOP_DPP8 default 1 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 VSRC1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 ENC_VOP2 default 1 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 VSRC1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VOP2_INST_LITERAL has_lit 1 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VSRC1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 ENC_VOP3 default 257 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_INST_LITERAL has_lit_0 257 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_INST_LITERAL has_lit_1 257 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 257 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_VOP_DPP16 default 257 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_VOP_DPP8 default 257 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VALU FALSE FALSE FALSE FALSE FALSE V_ADD_F64 Add two floating point inputs and store the result into a vector register. ENC_VOP2 default 2 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VOP2_INST_LITERAL has_lit 2 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 ENC_VOP3 default 258 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 258 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 258 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 258 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_ADD_F32 Add two floating point inputs and store the result into a vector register. VOP2_VOP_DPP16 default 3 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOP2_VOP_DPP8 default 3 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP2 default 3 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 3 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 259 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 259 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 259 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 259 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 259 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 259 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_SUB_F32 Subtract the second floating point input from the first input and store the result into a vector register. VOP2_VOP_DPP16 default 4 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOP2_VOP_DPP8 default 4 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP2 default 4 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 4 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 260 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 260 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 260 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 260 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 260 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 260 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_SUBREV_F32 Subtract the first floating point input from the second input and store the result into a vector register. VOP2_VOP_DPP16 default 5 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOP2_VOP_DPP8 default 5 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP2 default 5 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 5 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 261 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 261 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 261 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 261 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 261 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_VGPR 32 VOP3_VOP_DPP8 default 261 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_MUL_F64 Multiply two floating point inputs and store the result into a vector register. ENC_VOP2 default 6 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VOP2_INST_LITERAL has_lit 6 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 ENC_VOP3 default 262 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 262 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 262 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 262 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_MUL_DX9_ZERO_F32 V_MUL_LEGACY_F32 Multiply two floating point inputs and store the result into a vector register. Follows DX9 rules where 0.0 times anything produces 0.0 (this differs from other APIs when the other input is infinity or NaN). VOP2_VOP_DPP16 default 7 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOP2_VOP_DPP8 default 7 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP2 default 7 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 7 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 263 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 263 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 263 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 263 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 263 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 263 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MUL_F32 Multiply two floating point inputs and store the result into a vector register. VOP2_VOP_DPP16 default 8 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOP2_VOP_DPP8 default 8 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP2 default 8 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 8 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 264 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 264 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 264 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 264 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 264 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 264 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MUL_I32_I24 Multiply two signed 24-bit integer inputs and store the result as a signed 32-bit integer into a vector register. VOP2_VOP_DPP16 default 9 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOP2_VOP_DPP8 default 9 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 ENC_VOP2 default 9 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 9 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 ENC_VOP3 default 265 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 265 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 265 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 265 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 265 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 265 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MUL_HI_I32_I24 Multiply two signed 24-bit integer inputs and store the high 32 bits of the result as a signed 32-bit integer into a vector register. VOP2_VOP_DPP16 default 10 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOP2_VOP_DPP8 default 10 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 ENC_VOP2 default 10 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 10 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 ENC_VOP3 default 266 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 266 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 266 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 266 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 266 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 266 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MUL_U32_U24 Multiply two unsigned 24-bit integer inputs and store the result as an unsigned 32-bit integer into a vector register. VOP2_VOP_DPP16 default 11 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_VOP_DPP8 default 11 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP2 default 11 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 11 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP3 default 267 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 267 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 267 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 267 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 267 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 267 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MUL_HI_U32_U24 Multiply two unsigned 24-bit integer inputs and store the high 32 bits of the result as an unsigned 32-bit integer into a vector register. VOP2_VOP_DPP16 default 12 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_VOP_DPP8 default 12 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP2 default 12 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 12 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP3 default 268 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 268 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 268 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 268 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 268 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 268 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MIN_NUM_F64 V_MIN_F64 Select the IEEE minimumNumber() of two double-precision float inputs and store the result into a vector register. ENC_VOP2 default 13 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VOP2_INST_LITERAL has_lit 13 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 ENC_VOP3 default 269 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 269 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 269 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 269 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_MAX_NUM_F64 V_MAX_F64 Select the IEEE maximumNumber() of two double-precision float inputs and store the result into a vector register. ENC_VOP2 default 14 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VOP2_INST_LITERAL has_lit 14 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 ENC_VOP3 default 270 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 270 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 270 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 270 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_MIN_I32 Select the minimum of two signed 32-bit integer inputs and store the selected value into a vector register. VOP2_VOP_DPP16 default 17 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOP2_VOP_DPP8 default 17 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 ENC_VOP2 default 17 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 17 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 ENC_VOP3 default 273 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 273 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 273 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 273 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 273 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 273 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MAX_I32 Select the maximum of two signed 32-bit integer inputs and store the selected value into a vector register. VOP2_VOP_DPP16 default 18 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOP2_VOP_DPP8 default 18 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 ENC_VOP2 default 18 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 18 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 ENC_VOP3 default 274 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 274 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 274 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 274 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 274 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 274 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MIN_U32 Select the minimum of two unsigned 32-bit integer inputs and store the selected value into a vector register. VOP2_VOP_DPP16 default 19 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_VOP_DPP8 default 19 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP2 default 19 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 19 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP3 default 275 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 275 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 275 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 275 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 275 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 275 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MAX_U32 Select the maximum of two unsigned 32-bit integer inputs and store the selected value into a vector register. VOP2_VOP_DPP16 default 20 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_VOP_DPP8 default 20 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP2 default 20 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 20 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP3 default 276 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 276 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 276 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 276 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 276 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 276 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MIN_NUM_F32 V_MIN_F32 Select the IEEE minimumNumber() of two single-precision float inputs and store the result into a vector register. VOP2_VOP_DPP16 default 21 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOP2_VOP_DPP8 default 21 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP2 default 21 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 21 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 277 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 277 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 277 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 277 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 277 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 277 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MAX_NUM_F32 V_MAX_F32 Select the IEEE maximumNumber() of two single-precision float inputs and store the result into a vector register. VOP2_VOP_DPP16 default 22 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOP2_VOP_DPP8 default 22 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP2 default 22 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 22 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 278 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 278 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 278 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 278 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 278 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 278 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_LSHLREV_B32 Given a shift count in the first vector input, calculate the logical shift left of the second vector input and store the result into a vector register. VOP2_VOP_DPP16 default 24 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_VOP_DPP8 default 24 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP2 default 24 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 24 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP3 default 280 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 280 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 280 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 280 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 280 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 280 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_LSHRREV_B32 Given a shift count in the first vector input, calculate the logical shift right of the second vector input and store the result into a vector register. VOP2_VOP_DPP16 default 25 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_VOP_DPP8 default 25 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP2 default 25 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 25 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP3 default 281 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 281 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 281 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 281 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 281 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 281 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_ASHRREV_I32 Given a shift count in the first vector input, calculate the arithmetic shift right (preserving sign bit) of the second vector input and store the result into a vector register. VOP2_VOP_DPP16 default 26 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOP2_VOP_DPP8 default 26 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 ENC_VOP2 default 26 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 26 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 ENC_VOP3 default 282 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 282 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 282 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 282 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 282 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 282 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_AND_B32 Calculate bitwise AND on two vector inputs and store the result into a vector register. VOP2_VOP_DPP16 default 27 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_VOP_DPP8 default 27 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP2 default 27 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 27 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP3 default 283 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 283 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 283 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 283 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 283 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 283 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_OR_B32 Calculate bitwise OR on two vector inputs and store the result into a vector register. VOP2_VOP_DPP16 default 28 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_VOP_DPP8 default 28 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP2 default 28 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 28 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP3 default 284 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 284 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 284 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 284 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 284 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 284 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_XOR_B32 Calculate bitwise XOR on two vector inputs and store the result into a vector register. VOP2_VOP_DPP16 default 29 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_VOP_DPP8 default 29 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP2 default 29 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 29 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP3 default 285 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 285 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 285 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 285 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 285 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 285 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_XNOR_B32 Calculate bitwise XNOR on two vector inputs and store the result into a vector register. VOP2_VOP_DPP16 default 30 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_VOP_DPP8 default 30 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP2 default 30 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 30 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP3 default 286 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 286 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 286 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 286 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 286 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 286 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_LSHLREV_B64 Given a shift count in the first vector input, calculate the logical shift left of the second vector input and store the result into a vector register. ENC_VOP2 default 31 VDST FMT_NUM_U64 OPR_VGPR 64 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U64 OPR_VGPR 64 VOP2_INST_LITERAL has_lit 31 VDST FMT_NUM_U64 OPR_VGPR 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U64 OPR_VGPR 64 ENC_VOP3 default 287 VDST FMT_NUM_U64 OPR_VGPR 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 287 VDST FMT_NUM_U64 OPR_VGPR 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 287 VDST FMT_NUM_U64 OPR_VGPR 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 287 VDST FMT_NUM_U64 OPR_VGPR 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_ADD_CO_CI_U32 Add two unsigned 32-bit integer inputs and a bit from a carry-in mask, store the result into a vector register and store the carry-out mask into a scalar register. VOP2_VOP_DPP16 default 32 VDST FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VOP2_VOP_DPP8 default 32 VDST FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 ENC_VOP2 default 32 VDST FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VOP2_INST_LITERAL has_lit 32 VDST FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VOP3_SDST_ENC default 288 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_SDST_ENC_INST_LITERAL has_lit_0 288 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_SDST_ENC_INST_LITERAL has_lit_1 288 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_SDST_ENC_INST_LITERAL has_lit_0_has_lit_1 288 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_SDST_ENC_VOP_DPP16 default 288 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_SDST_ENC_VOP_DPP8 default 288 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VALU FALSE FALSE FALSE FALSE FALSE V_SUB_CO_CI_U32 Subtract the second unsigned 32-bit integer input from the first input, subtract a bit from the carry-in mask, store the result into a vector register and store the carry-out mask into a scalar register. VOP2_VOP_DPP16 default 33 VDST FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VOP2_VOP_DPP8 default 33 VDST FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 ENC_VOP2 default 33 VDST FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VOP2_INST_LITERAL has_lit 33 VDST FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VOP3_SDST_ENC default 289 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_SDST_ENC_INST_LITERAL has_lit_0 289 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_SDST_ENC_INST_LITERAL has_lit_1 289 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_SDST_ENC_INST_LITERAL has_lit_0_has_lit_1 289 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_SDST_ENC_VOP_DPP16 default 289 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_SDST_ENC_VOP_DPP8 default 289 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VALU FALSE FALSE FALSE FALSE FALSE V_SUBREV_CO_CI_U32 Subtract the first unsigned 32-bit integer input from the second input, subtract a bit from the carry-in mask, store the result into a vector register and store the carry-out mask into a scalar register. VOP2_VOP_DPP16 default 34 VDST FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VOP2_VOP_DPP8 default 34 VDST FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 ENC_VOP2 default 34 VDST FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VOP2_INST_LITERAL has_lit 34 VDST FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VOP3_SDST_ENC default 290 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_SDST_ENC_INST_LITERAL has_lit_0 290 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_SDST_ENC_INST_LITERAL has_lit_1 290 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_SDST_ENC_INST_LITERAL has_lit_0_has_lit_1 290 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_SDST_ENC_VOP_DPP16 default 290 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_VGPR 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_SDST_ENC_VOP_DPP8 default 290 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_VGPR 32 SRC2 FMT_NUM_M64 OPR_SREG 64 VALU FALSE FALSE FALSE FALSE FALSE V_ADD_NC_U32 Add two unsigned 32-bit integer inputs and store the result into a vector register. No carry-in or carry-out support. VOP2_VOP_DPP16 default 37 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_VOP_DPP8 default 37 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP2 default 37 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 37 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP3 default 293 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 293 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 293 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 293 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 293 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 293 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_SUB_NC_U32 Subtract the second unsigned 32-bit integer input from the first input and store the result into a vector register. No carry-in or carry-out support. VOP2_VOP_DPP16 default 38 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_VOP_DPP8 default 38 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP2 default 38 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 38 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP3 default 294 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 294 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 294 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 294 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 294 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 294 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_SUBREV_NC_U32 Subtract the first unsigned 32-bit integer input from the second input and store the result into a vector register. No carry-in or carry-out support. VOP2_VOP_DPP16 default 39 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_VOP_DPP8 default 39 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP2 default 39 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 39 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP3 default 295 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 295 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 295 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 295 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 295 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_VGPR 32 VOP3_VOP_DPP8 default 295 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_FMAC_F32 Multiply two floating point inputs and accumulate the result into the destination register using fused multiply add. VOP2_VOP_DPP16 default 43 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOP2_VOP_DPP8 default 43 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP2 default 43 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 43 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 299 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 299 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 299 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 299 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 299 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 299 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_FMAMK_F32 Multiply a single-precision float input with a literal constant and add a second single-precision float input using fused multiply add, and store the result into a vector register. VOP2_INST_LITERAL default 44 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_F32 OPR_SIMM32 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 44 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_F32 OPR_SIMM32 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_FMAAK_F32 Multiply two single-precision float inputs and add a literal constant using fused multiply add, and store the result into a vector register. VOP2_INST_LITERAL default 45 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_F32 OPR_SIMM32 32 VOP2_INST_LITERAL has_lit 45 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_F32 OPR_SIMM32 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_PK_RTZ_F16_F32 V_CVT_PKRTZ_F16_F32 Convert two single-precision float inputs to a packed half-precision float value using round toward zero semantics (ignore the current rounding mode), and store the result into a vector register. VOP2_VOP_DPP16 default 47 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOP2_VOP_DPP8 default 47 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP2 default 47 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 47 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 303 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 303 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 303 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 303 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 303 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 303 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MIN_NUM_F16 V_MIN_F16 Select the IEEE minimumNumber() of two half-precision float inputs and store the result into a vector register. VOP2_VOP_DPP16 default 48 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOP2_VOP_DPP8 default 48 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP2 default 48 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOP2_INST_LITERAL has_lit 48 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 304 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 304 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 304 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 304 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 304 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 304 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MAX_NUM_F16 V_MAX_F16 Select the IEEE maximumNumber() of two half-precision float inputs and store the result into a vector register. VOP2_VOP_DPP16 default 49 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOP2_VOP_DPP8 default 49 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP2 default 49 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOP2_INST_LITERAL has_lit 49 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 305 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 305 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 305 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 305 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 305 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 305 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_ADD_F16 Add two floating point inputs and store the result into a vector register. VOP2_VOP_DPP16 default 50 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOP2_VOP_DPP8 default 50 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP2 default 50 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOP2_INST_LITERAL has_lit 50 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 306 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 306 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 306 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 306 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 306 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 306 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_SUB_F16 Subtract the second floating point input from the first input and store the result into a vector register. VOP2_VOP_DPP16 default 51 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOP2_VOP_DPP8 default 51 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP2 default 51 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOP2_INST_LITERAL has_lit 51 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 307 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 307 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 307 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 307 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 307 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 307 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_SUBREV_F16 Subtract the first floating point input from the second input and store the result into a vector register. VOP2_VOP_DPP16 default 52 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOP2_VOP_DPP8 default 52 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP2 default 52 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOP2_INST_LITERAL has_lit 52 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 308 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 308 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 308 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 308 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 308 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_VGPR 16 VOP3_VOP_DPP8 default 308 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_MUL_F16 Multiply two floating point inputs and store the result into a vector register. VOP2_VOP_DPP16 default 53 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOP2_VOP_DPP8 default 53 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP2 default 53 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOP2_INST_LITERAL has_lit 53 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 309 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 309 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 309 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 309 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 309 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 309 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_FMAC_F16 Multiply two floating point inputs and accumulate the result into the destination register using fused multiply add. VOP2_VOP_DPP16 default 54 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOP2_VOP_DPP8 default 54 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP2 default 54 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOP2_INST_LITERAL has_lit 54 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 310 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 310 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 310 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 310 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 310 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 310 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_FMAMK_F16 Multiply a half-precision float input with a literal constant and add a second half-precision float input using fused multiply add, and store the result into a vector register. VOP2_INST_LITERAL default 55 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_F16 OPR_SIMM32 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOP2_INST_LITERAL has_lit 55 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_F16 OPR_SIMM32 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_FMAAK_F16 Multiply two half-precision float inputs and add a literal constant using fused multiply add, and store the result into a vector register. VOP2_INST_LITERAL default 56 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_F16 OPR_SIMM32 32 VOP2_INST_LITERAL has_lit 56 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_F16 OPR_SIMM32 32 VALU FALSE FALSE FALSE FALSE FALSE V_LDEXP_F16 Multiply the first input, a floating point value, by an integral power of 2 specified in the second input, a signed integer value, and store the floating point result into a vector register. VOP2_VOP_DPP16 default 59 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOP2_VOP_DPP8 default 59 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP2 default 59 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOP2_INST_LITERAL has_lit 59 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 315 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 315 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 315 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 315 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 315 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 315 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_PK_FMAC_F16 Multiply two packed half-precision float inputs component-wise and accumulate the result into the destination register using fused multiply add. ENC_VOP2 default 60 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 VSRC1 FMT_NUM_PK2_F16 OPR_VGPR 32 VOP2_INST_LITERAL has_lit 60 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 VSRC1 FMT_NUM_PK2_F16 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_FMA_DX9_ZERO_F32 V_FMA_LEGACY_F32 Multiply and add single-precision values. Follows DX9 rules where 0.0 times anything produces 0.0. ENC_VOP3 default 521 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 521 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 521 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 521 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 521 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 521 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 521 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 521 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 521 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 521 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MAD_I32_I24 Multiply two signed 24-bit integer inputs in the signed 32-bit integer domain, add a signed 32-bit integer value from a third input, and store the result as a signed 32-bit integer into a vector register. ENC_VOP3 default 522 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I24 OPR_SRC 24 SRC1 FMT_NUM_I24 OPR_SRC 24 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 522 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I24 OPR_SRC 24 SRC1 FMT_NUM_I24 OPR_SRC 24 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 522 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I24 OPR_SRC 24 SIMM32 FMT_NUM_I24 OPR_SRC 24 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 522 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I24 OPR_SRC 24 SIMM32 FMT_NUM_I24 OPR_SRC 24 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 522 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I24 OPR_SRC 24 SRC1 FMT_NUM_I24 OPR_SRC 24 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 522 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I24 OPR_SRC 24 SRC1 FMT_NUM_I24 OPR_SRC 24 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 522 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I24 OPR_SRC 24 SIMM32 FMT_NUM_I24 OPR_SRC 24 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 522 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I24 OPR_SRC 24 SIMM32 FMT_NUM_I24 OPR_SRC 24 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 522 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I24 OPR_VGPR 24 SRC1 FMT_NUM_I24 OPR_SRC 24 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 522 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I24 OPR_VGPR 24 SRC1 FMT_NUM_I24 OPR_SRC 24 SRC2 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MAD_U32_U24 Multiply two unsigned 24-bit integer inputs in the unsigned 32-bit integer domain, add a unsigned 32-bit integer value from a third input, and store the result as an unsigned 32-bit integer into a vector register. ENC_VOP3 default 523 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U24 OPR_SRC 24 SRC1 FMT_NUM_U24 OPR_SRC 24 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 523 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U24 OPR_SRC 24 SRC1 FMT_NUM_U24 OPR_SRC 24 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 523 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U24 OPR_SRC 24 SIMM32 FMT_NUM_U24 OPR_SRC 24 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 523 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U24 OPR_SRC 24 SIMM32 FMT_NUM_U24 OPR_SRC 24 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 523 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U24 OPR_SRC 24 SRC1 FMT_NUM_U24 OPR_SRC 24 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 523 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U24 OPR_SRC 24 SRC1 FMT_NUM_U24 OPR_SRC 24 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 523 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U24 OPR_SRC 24 SIMM32 FMT_NUM_U24 OPR_SRC 24 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 523 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U24 OPR_SRC 24 SIMM32 FMT_NUM_U24 OPR_SRC 24 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 523 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U24 OPR_VGPR 24 SRC1 FMT_NUM_U24 OPR_SRC 24 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 523 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U24 OPR_VGPR 24 SRC1 FMT_NUM_U24 OPR_SRC 24 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_CUBEID_F32 Compute the cubemap face ID of a 3D coordinate specified as three single-precision float inputs. Store the result in single-precision float format into a vector register. ENC_VOP3 default 524 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 524 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 524 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 524 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 524 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 524 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 524 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 524 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 524 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 524 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_CUBESC_F32 Compute the cubemap S coordinate of a 3D coordinate specified as three single-precision float inputs. Store the result in single-precision float format into a vector register. ENC_VOP3 default 525 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 525 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 525 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 525 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 525 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 525 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 525 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 525 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 525 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 525 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_CUBETC_F32 Compute the cubemap T coordinate of a 3D coordinate specified as three single-precision float inputs. Store the result in single-precision float format into a vector register. ENC_VOP3 default 526 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 526 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 526 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 526 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 526 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 526 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 526 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 526 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 526 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 526 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_CUBEMA_F32 Compute the cubemap major axis coordinate of a 3D coordinate specified as three single-precision float inputs. Store the result in single-precision float format into a vector register. ENC_VOP3 default 527 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 527 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 527 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 527 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 527 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 527 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 527 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 527 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 527 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 527 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_BFE_U32 Extract an unsigned bitfield from the first input using field offset from the second input and size from the third input, then store the result into a vector register. ENC_VOP3 default 528 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 528 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 528 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 528 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 528 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 528 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 528 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 528 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 528 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 528 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_BFE_I32 Extract a signed bitfield from the first input using field offset from the second input and size from the third input, then store the result into a vector register. ENC_VOP3 default 529 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 529 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 529 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 529 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 529 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 529 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 529 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 529 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 529 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 529 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_BFI_B32 Overwrite a bitfield in the third input with a bitfield from the second input using a mask from the first input, then store the result into a vector register. ENC_VOP3 default 530 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 530 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 530 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 530 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 530 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 530 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 530 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 530 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 530 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 530 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_FMA_F32 Multiply two single-precision float inputs and add a third input using fused multiply add, and store the result into a vector register. ENC_VOP3 default 531 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 531 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 531 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 531 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 531 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 531 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 531 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 531 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 531 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 531 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_FMA_F64 Multiply two double-precision float inputs and add a third input using fused multiply add, and store the result into a vector register. ENC_VOP3 default 532 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 SRC2 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 532 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 SRC2 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 532 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC2 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 532 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC2 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_2 532 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_2 532 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1_has_lit_2 532 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 532 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_LERP_U8 Average two 4-D vectors stored as packed bytes in the first two inputs with rounding control provided by the third input, then store the result into a vector register. Each byte in the third input acts as a rounding mode for the corresponding element; if the LSB is set then 0.5 rounds up, otherwise 0.5 truncates. ENC_VOP3 default 533 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 533 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 533 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 533 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 533 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 533 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 533 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 533 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_VOP_DPP16 default 533 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_VOP_DPP8 default 533 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_ALIGNBIT_B32 Align a 64-bit value encoded in the first two inputs to a bit position specified in the third input, then store the result into a 32-bit vector register. ENC_VOP3 default 534 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 534 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 534 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 534 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 534 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 534 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 534 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 534 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 534 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 534 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_ALIGNBYTE_B32 Align a 64-bit value encoded in the first two inputs to a byte position specified in the third input, then store the result into a 32-bit vector register. ENC_VOP3 default 535 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 535 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 535 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 535 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 535 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 535 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 535 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 535 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 535 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 535 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MULLIT_F32 Multiply two floating point inputs and store the result into a vector register. Specific rules apply to accommodate lighting calculations: 0.0 * x = 0.0 and alternate INF, NAN, overflow rules apply. ENC_VOP3 default 536 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 536 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 536 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 536 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 536 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 536 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 536 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 536 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 536 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 536 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MIN3_I32 Select the minimum of three signed 32-bit integer inputs and store the selected value into a vector register. ENC_VOP3 default 538 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 538 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 538 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 538 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 538 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 538 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 538 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 538 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 538 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 538 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MIN3_U32 Select the minimum of three unsigned 32-bit integer inputs and store the selected value into a vector register. ENC_VOP3 default 539 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 539 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 539 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 539 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 539 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 539 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 539 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 539 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 539 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 539 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MAX3_I32 Select the maximum of three signed 32-bit integer inputs and store the selected value into a vector register. ENC_VOP3 default 541 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 541 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 541 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 541 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 541 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 541 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 541 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 541 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 541 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 541 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MAX3_U32 Select the maximum of three unsigned 32-bit integer inputs and store the selected value into a vector register. ENC_VOP3 default 542 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 542 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 542 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 542 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 542 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 542 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 542 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 542 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 542 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 542 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MED3_I32 Select the median of three signed 32-bit integer values and store the selected value into a vector register. ENC_VOP3 default 544 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 544 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 544 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 544 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 544 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 544 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 544 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 544 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 544 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 544 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MED3_U32 Select the median of three unsigned 32-bit integer values and store the selected value into a vector register. ENC_VOP3 default 545 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 545 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 545 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 545 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 545 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 545 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 545 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 545 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 545 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 545 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_SAD_U8 Calculate the sum of absolute differences of elements in two packed 4-component unsigned 8-bit integer inputs, add an unsigned 32-bit integer value from the third input and store the result into a vector register. ENC_VOP3 default 546 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_U8 OPR_SRC 32 SRC1 FMT_NUM_PK4_U8 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 546 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SRC1 FMT_NUM_PK4_U8 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 546 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 546 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 546 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_U8 OPR_SRC 32 SRC1 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 546 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SRC1 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 546 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 546 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 546 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_PK4_U8 OPR_VGPR 32 SRC1 FMT_NUM_PK4_U8 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 546 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_PK4_U8 OPR_VGPR 32 SRC1 FMT_NUM_PK4_U8 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_SAD_HI_U8 Calculate the sum of absolute differences of elements in two packed 4-component unsigned 8-bit integer inputs, shift the sum left by 16 bits, add an unsigned 32-bit integer value from the third input and store the result into a vector register. ENC_VOP3 default 547 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_U8 OPR_SRC 32 SRC1 FMT_NUM_PK4_U8 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 547 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SRC1 FMT_NUM_PK4_U8 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 547 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 547 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 547 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_U8 OPR_SRC 32 SRC1 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 547 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SRC1 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 547 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 547 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 547 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_PK4_U8 OPR_VGPR 32 SRC1 FMT_NUM_PK4_U8 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 547 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_PK4_U8 OPR_VGPR 32 SRC1 FMT_NUM_PK4_U8 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_SAD_U16 Calculate the sum of absolute differences of elements in two packed 2-component unsigned 16-bit integer inputs, add an unsigned 32-bit integer value from the third input and store the result into a vector register. ENC_VOP3 default 548 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 548 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 548 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 548 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 548 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 548 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 548 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 548 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 548 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_PK2_U16 OPR_VGPR 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 548 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_PK2_U16 OPR_VGPR 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_SAD_U32 Calculate the absolute difference of two unsigned 32-bit integer inputs, add an unsigned 32-bit integer value from the third input and store the result into a vector register. ENC_VOP3 default 549 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 549 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 549 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 549 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 549 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 549 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 549 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 549 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 549 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 549 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_PK_U8_F32 Convert a single-precision float value from the first input to an unsigned 8-bit integer value and pack the result into one byte of the third input using the second input as a byte select. Store the result into a vector register. ENC_VOP3 default 550 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 550 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 550 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 550 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 550 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 550 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 550 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 550 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_VOP_DPP16 default 550 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_VOP_DPP8 default 550 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_DIV_FIXUP_F32 Given a single-precision float quotient in the first input, a denominator in the second input and a numerator in the third input, detect and apply corner cases related to division, including divide by zero, NaN inputs and overflow, and modify the quotient accordingly. Generate any invalid, denormal and divide-by-zero exceptions that are a result of the division. Store the modified quotient into a vector register. ENC_VOP3 default 551 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 551 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 551 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 551 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 551 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 551 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 551 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 551 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_DIV_FIXUP_F64 Given a double-precision float quotient in the first input, a denominator in the second input and a numerator in the third input, detect and apply corner cases related to division, including divide by zero, NaN inputs and overflow, and modify the quotient accordingly. Generate any invalid, denormal and divide-by-zero exceptions that are a result of the division. Store the modified quotient into a vector register. ENC_VOP3 default 552 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 SRC2 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 552 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 SRC2 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 552 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC2 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 552 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC2 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_2 552 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_2 552 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1_has_lit_2 552 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 552 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_MIN3_NUM_F32 V_MIN3_F32 Select the IEEE minimumNumber() of three single-precision float inputs and store the result into a vector register. ENC_VOP3 default 553 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 553 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 553 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 553 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 553 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 553 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 553 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 553 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 553 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 553 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MAX3_NUM_F32 V_MAX3_F32 Select the IEEE maximumNumber() of three single-precision float inputs and store the result into a vector register. ENC_VOP3 default 554 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 554 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 554 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 554 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 554 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 554 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 554 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 554 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 554 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 554 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MIN3_NUM_F16 V_MIN3_F16 Select the IEEE minimumNumber() of three half-precision float inputs and store the result into a vector register. ENC_VOP3 default 555 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 555 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 555 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 555 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 555 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 555 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 555 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 555 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 555 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 555 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MAX3_NUM_F16 V_MAX3_F16 Select the IEEE maximumNumber() of three half-precision float inputs and store the result into a vector register. ENC_VOP3 default 556 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 556 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 556 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 556 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 556 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 556 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 556 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 556 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 556 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 556 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MINIMUM3_F32 Select the IEEE minimum() of three single-precision float inputs and store the result into a vector register. ENC_VOP3 default 557 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 557 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 557 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 557 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 557 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 557 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 557 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 557 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 557 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 557 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MAXIMUM3_F32 Select the IEEE maximum() of three single-precision float inputs and store the result into a vector register. ENC_VOP3 default 558 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 558 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 558 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 558 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 558 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 558 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 558 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 558 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 558 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 558 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MINIMUM3_F16 Select the IEEE minimum() of three half-precision float inputs and store the result into a vector register. ENC_VOP3 default 559 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 559 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 559 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 559 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 559 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 559 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 559 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 559 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 559 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 559 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MAXIMUM3_F16 Select the IEEE maximum() of three half-precision float inputs and store the result into a vector register. ENC_VOP3 default 560 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 560 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 560 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 560 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 560 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 560 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 560 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 560 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 560 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 560 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MED3_NUM_F32 V_MED3_F32 Select the median of three single-precision float inputs and store the result into a vector register. ENC_VOP3 default 561 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 561 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 561 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 561 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 561 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 561 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 561 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 561 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 561 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 561 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MED3_NUM_F16 V_MED3_F16 Select the median of three half-precision float inputs and store the result into a vector register. ENC_VOP3 default 562 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 562 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 562 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 562 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 562 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 562 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 562 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 562 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 562 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 562 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_DIV_FMAS_F32 Multiply two single-precision float inputs and add a third input using fused multiply add, then scale the exponent of the result by a fixed factor if the vector condition code is set. Store the result into a vector register. ENC_VOP3 default 567 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_VCC 64 VOP3_INST_LITERAL has_lit_0 567 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_VCC 64 VOP3_INST_LITERAL has_lit_1 567 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_VCC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 567 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_VCC 64 VOP3_INST_LITERAL has_lit_2 567 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_VCC 64 VOP3_INST_LITERAL has_lit_0_has_lit_2 567 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_VCC 64 VOP3_INST_LITERAL has_lit_1_has_lit_2 567 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_VCC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 567 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_VCC 64 VALU FALSE FALSE FALSE FALSE FALSE V_DIV_FMAS_F64 Multiply two double-precision float inputs and add a third input using fused multiply add, then scale the exponent of the result by a fixed factor if the vector condition code is set. Store the result into a vector register. ENC_VOP3 default 568 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 SRC2 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_VCC 64 VOP3_INST_LITERAL has_lit_0 568 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 SRC2 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_VCC 64 VOP3_INST_LITERAL has_lit_1 568 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC2 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_VCC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 568 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC2 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_VCC 64 VOP3_INST_LITERAL has_lit_2 568 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_VCC 64 VOP3_INST_LITERAL has_lit_0_has_lit_2 568 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_VCC 64 VOP3_INST_LITERAL has_lit_1_has_lit_2 568 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_VCC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 568 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_VCC 64 VALU FALSE FALSE FALSE FALSE FALSE V_MSAD_U8 Calculate the sum of absolute differences of elements in two packed 4-component unsigned 8-bit integer inputs, except that elements where the second input (known as the reference input) is zero are not included in the sum. Add an unsigned 32-bit integer value from the third input and store the result into a vector register. ENC_VOP3 default 569 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 569 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 569 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 569 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 569 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 569 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 569 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 569 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_VOP_DPP16 default 569 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_VOP_DPP8 default 569 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_QSAD_PK_U16_U8 Perform the V_SAD_U8 operation four times using different slices of the first array, all entries of the second array and each entry of the third array. Truncate each result to 16 bits, pack the values into a 4-entry array and store the array into a vector register. The first input is an 8-entry array of unsigned 8-bit integers, the second input is a 4-entry array of unsigned 8-bit integers and the third input is a 4-entry array of unsigned 16-bit integers. ENC_VOP3 default 570 VDST FMT_NUM_B64 OPR_VGPR 64 SRC0 FMT_NUM_B64 OPR_SRC 64 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 570 VDST FMT_NUM_B64 OPR_VGPR 64 SIMM32 FMT_NUM_B64 OPR_SRC 64 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 570 VDST FMT_NUM_B64 OPR_VGPR 64 SRC0 FMT_NUM_B64 OPR_SRC 64 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 570 VDST FMT_NUM_B64 OPR_VGPR 64 SIMM32 FMT_NUM_B64 OPR_SRC 64 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_2 570 VDST FMT_NUM_B64 OPR_VGPR 64 SRC0 FMT_NUM_B64 OPR_SRC 64 SRC1 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_2 570 VDST FMT_NUM_B64 OPR_VGPR 64 SIMM32 FMT_NUM_B64 OPR_SRC 64 SRC1 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1_has_lit_2 570 VDST FMT_NUM_B64 OPR_VGPR 64 SRC0 FMT_NUM_B64 OPR_SRC 64 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 570 VDST FMT_NUM_B64 OPR_VGPR 64 SIMM32 FMT_NUM_B64 OPR_SRC 64 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_MQSAD_PK_U16_U8 Perform the V_MSAD_U8 operation four times using different slices of the first array, all entries of the second array and each entry of the third array. Truncate each result to 16 bits, pack the values into a 4-entry array and store the array into a vector register. The first input is an 8-entry array of unsigned 8-bit integers, the second input is a 4-entry array of unsigned 8-bit integers and the third input is a 4-entry array of unsigned 16-bit integers. ENC_VOP3 default 571 VDST FMT_NUM_B64 OPR_VGPR 64 SRC0 FMT_NUM_B64 OPR_SRC 64 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 571 VDST FMT_NUM_B64 OPR_VGPR 64 SIMM32 FMT_NUM_B64 OPR_SRC 64 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 571 VDST FMT_NUM_B64 OPR_VGPR 64 SRC0 FMT_NUM_B64 OPR_SRC 64 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 571 VDST FMT_NUM_B64 OPR_VGPR 64 SIMM32 FMT_NUM_B64 OPR_SRC 64 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_2 571 VDST FMT_NUM_B64 OPR_VGPR 64 SRC0 FMT_NUM_B64 OPR_SRC 64 SRC1 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_2 571 VDST FMT_NUM_B64 OPR_VGPR 64 SIMM32 FMT_NUM_B64 OPR_SRC 64 SRC1 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1_has_lit_2 571 VDST FMT_NUM_B64 OPR_VGPR 64 SRC0 FMT_NUM_B64 OPR_SRC 64 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 571 VDST FMT_NUM_B64 OPR_VGPR 64 SIMM32 FMT_NUM_B64 OPR_SRC 64 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_MQSAD_U32_U8 Perform the V_MSAD_U8 operation four times using different slices of the first array, all entries of the second array and each entry of the third array. Pack each 32-bit value into a 4-entry array and store the array into a vector register. The first input is an 8-entry array of unsigned 8-bit integers, the second input is a 4-entry array of unsigned 8-bit integers and the third input is a 4-entry array of unsigned 32-bit integers. ENC_VOP3 default 573 VDST FMT_NUM_B128 OPR_VGPR 128 SRC0 FMT_NUM_B64 OPR_SRC 64 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B128 OPR_SRC_VGPR 128 VOP3_INST_LITERAL has_lit_0 573 VDST FMT_NUM_B128 OPR_VGPR 128 SIMM32 FMT_NUM_B64 OPR_SRC 64 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B128 OPR_SRC_VGPR 128 VOP3_INST_LITERAL has_lit_1 573 VDST FMT_NUM_B128 OPR_VGPR 128 SRC0 FMT_NUM_B64 OPR_SRC 64 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B128 OPR_SRC_VGPR 128 VOP3_INST_LITERAL has_lit_0_has_lit_1 573 VDST FMT_NUM_B128 OPR_VGPR 128 SIMM32 FMT_NUM_B64 OPR_SRC 64 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B128 OPR_SRC_VGPR 128 VOP3_INST_LITERAL has_lit_2 573 VDST FMT_NUM_B128 OPR_VGPR 128 SRC0 FMT_NUM_B64 OPR_SRC 64 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B128 OPR_SRC_VGPR 128 VOP3_INST_LITERAL has_lit_0_has_lit_2 573 VDST FMT_NUM_B128 OPR_VGPR 128 SIMM32 FMT_NUM_B64 OPR_SRC 64 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B128 OPR_SRC_VGPR 128 VOP3_INST_LITERAL has_lit_1_has_lit_2 573 VDST FMT_NUM_B128 OPR_VGPR 128 SRC0 FMT_NUM_B64 OPR_SRC 64 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B128 OPR_SRC_VGPR 128 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 573 VDST FMT_NUM_B128 OPR_VGPR 128 SIMM32 FMT_NUM_B64 OPR_SRC 64 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B128 OPR_SRC_VGPR 128 VALU FALSE FALSE FALSE FALSE FALSE V_XOR3_B32 Calculate the bitwise XOR of three vector inputs and store the result into a vector register. ENC_VOP3 default 576 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 576 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 576 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 576 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 576 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 576 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 576 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 576 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 576 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 576 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MAD_U16 Multiply two unsigned 16-bit integer inputs, add an unsigned 16-bit integer value from a third input, and store the result into a vector register. ENC_VOP3 default 577 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 577 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 577 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 577 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 577 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 577 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 577 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 577 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 577 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 577 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_PERM_B32 Permute a 64-bit value constructed from two vector inputs (most significant bits come from the first input) using a per-lane selector from the third input. The lane selector allows each byte of the result to choose from any of the 8 input bytes, perform sign extension or pad with 0/1 bits. Store the result into a vector register. ENC_VOP3 default 580 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 580 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 580 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 580 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 580 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 580 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 580 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 580 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_VOP_DPP16 default 580 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3_VOP_DPP8 default 580 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_XAD_U32 Calculate bitwise XOR of the first two vector inputs, then add the third vector input to the intermediate result, then store the final result into a vector register. ENC_VOP3 default 581 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 581 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 581 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 581 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 581 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 581 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 581 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 581 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 581 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 581 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_LSHL_ADD_U32 Given a shift count in the second input, calculate the logical shift left of the first input, then add the third input to the intermediate result, then store the final result into a vector register. ENC_VOP3 default 582 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 582 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 582 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 582 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 582 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 582 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 582 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 582 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 582 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 582 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_ADD_LSHL_U32 Add the first two integer inputs, then given a shift count in the third input, calculate the logical shift left of the intermediate result, then store the final result into a vector register. ENC_VOP3 default 583 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 583 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 583 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 583 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 583 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 583 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 583 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 583 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 583 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 583 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_FMA_F16 Multiply two half-precision float inputs and add a third input using fused multiply add, and store the result into a vector register. ENC_VOP3 default 584 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 584 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 584 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 584 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 584 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 584 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 584 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 584 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 584 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 584 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MIN3_I16 Select the minimum of three signed 16-bit integer inputs and store the selected value into a vector register. ENC_VOP3 default 586 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 586 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 586 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 586 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 586 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 586 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 586 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 586 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP16 default 586 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP8 default 586 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MIN3_U16 Select the minimum of three unsigned 16-bit integer inputs and store the selected value into a vector register. ENC_VOP3 default 587 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 587 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 587 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 587 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 587 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 587 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 587 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 587 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 587 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 587 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MAX3_I16 Select the maximum of three signed 16-bit integer inputs and store the selected value into a vector register. ENC_VOP3 default 589 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 589 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 589 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 589 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 589 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 589 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 589 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 589 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP16 default 589 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP8 default 589 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MAX3_U16 Select the maximum of three unsigned 16-bit integer inputs and store the selected value into a vector register. ENC_VOP3 default 590 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 590 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 590 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 590 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 590 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 590 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 590 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 590 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 590 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 590 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MED3_I16 Select the median of three signed 16-bit integer values and store the selected value into a vector register. ENC_VOP3 default 592 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 592 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 592 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 592 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 592 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 592 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 592 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 592 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP16 default 592 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP8 default 592 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MED3_U16 Select the median of three unsigned 16-bit integer values and store the selected value into a vector register. ENC_VOP3 default 593 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 593 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 593 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 593 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 593 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 593 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 593 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 593 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 593 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 593 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MAD_I16 Multiply two signed 16-bit integer inputs, add a signed 16-bit integer value from a third input, and store the result into a vector register. ENC_VOP3 default 595 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 595 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 595 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 595 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 595 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 595 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 595 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 595 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP16 default 595 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP8 default 595 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_DIV_FIXUP_F16 Given a half-precision float quotient in the first input, a denominator in the second input and a numerator in the third input, detect and apply corner cases related to division, including divide by zero, NaN inputs and overflow, and modify the quotient accordingly. Generate any invalid, denormal and divide-by-zero exceptions that are a result of the division. Store the modified quotient into a vector register. ENC_VOP3 default 596 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 596 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 596 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 596 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 596 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 596 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 596 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 596 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 596 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 596 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_ADD3_U32 V_ADD3_NC_U32 Add three unsigned inputs and store the result into a vector register. No carry-in or carry-out support. ENC_VOP3 default 597 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 597 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 597 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 597 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 597 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 597 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 597 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 597 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 597 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 597 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_LSHL_OR_B32 Given a shift count in the second input, calculate the logical shift left of the first input, then calculate the bitwise OR of the intermediate result and the third input, then store the final result into a vector register. ENC_VOP3 default 598 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 598 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 598 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 598 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 598 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 598 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 598 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 598 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 598 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 598 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_AND_OR_B32 Calculate bitwise AND on the first two vector inputs, then compute the bitwise OR of the intermediate result and the third vector input, then store the final result into a vector register. ENC_VOP3 default 599 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 599 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 599 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 599 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 599 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 599 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 599 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 599 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 599 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 599 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_OR3_B32 Calculate the bitwise OR of three vector inputs and store the result into a vector register. ENC_VOP3 default 600 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 600 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 600 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 600 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 600 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 600 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 600 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 600 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 600 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 600 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MAD_U32_U16 Multiply two unsigned 16-bit integer inputs in the unsigned 32-bit integer domain, add an unsigned 32-bit integer value from a third input, and store the result as an unsigned 32-bit integer into a vector register. ENC_VOP3 default 601 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 601 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 601 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 601 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 601 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 601 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 601 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 601 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 601 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 601 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MAD_I32_I16 Multiply two signed 16-bit integer inputs in the signed 32-bit integer domain, add a signed 32-bit integer value from a third input, and store the result as a signed 32-bit integer into a vector register. ENC_VOP3 default 602 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 602 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 602 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 602 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 602 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 602 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 602 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 602 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 602 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 602 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 SRC2 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PERMLANE16_B32 Perform arbitrary gather-style operation within a row (16 contiguous lanes). ENC_VOP3 default 603 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 SRC1 FMT_NUM_B32 OPR_SSRC 32 SRC2 FMT_NUM_B32 OPR_SSRC 32 VOP3_INST_LITERAL has_lit_0 603 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC_VGPR 32 SRC1 FMT_NUM_B32 OPR_SSRC 32 SRC2 FMT_NUM_B32 OPR_SSRC 32 VOP3_INST_LITERAL has_lit_1 603 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SRC2 FMT_NUM_B32 OPR_SSRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 603 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SRC2 FMT_NUM_B32 OPR_SSRC 32 VOP3_INST_LITERAL has_lit_2 603 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 SRC1 FMT_NUM_B32 OPR_SSRC 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 603 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC_VGPR 32 SRC1 FMT_NUM_B32 OPR_SSRC 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 603 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 603 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PERMLANEX16_B32 Perform arbitrary gather-style operation across two rows (each row is 16 contiguous lanes). ENC_VOP3 default 604 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 SRC1 FMT_NUM_B32 OPR_SSRC 32 SRC2 FMT_NUM_B32 OPR_SSRC 32 VOP3_INST_LITERAL has_lit_0 604 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC_VGPR 32 SRC1 FMT_NUM_B32 OPR_SSRC 32 SRC2 FMT_NUM_B32 OPR_SSRC 32 VOP3_INST_LITERAL has_lit_1 604 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SRC2 FMT_NUM_B32 OPR_SSRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 604 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SRC2 FMT_NUM_B32 OPR_SSRC 32 VOP3_INST_LITERAL has_lit_2 604 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 SRC1 FMT_NUM_B32 OPR_SSRC 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 604 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC_VGPR 32 SRC1 FMT_NUM_B32 OPR_SSRC 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 604 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 604 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_CNDMASK_B16 Copy data from one of two inputs based on the per-lane condition code and store the result into a vector register. ENC_VOP3 default 605 VDST FMT_NUM_B16 OPR_VGPR 16 SRC0 FMT_NUM_B16 OPR_SRC 16 SRC1 FMT_NUM_B16 OPR_SRC 16 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_INST_LITERAL has_lit_0 605 VDST FMT_NUM_B16 OPR_VGPR 16 SIMM32 FMT_NUM_B16 OPR_SRC 16 SRC1 FMT_NUM_B16 OPR_SRC 16 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_INST_LITERAL has_lit_1 605 VDST FMT_NUM_B16 OPR_VGPR 16 SRC0 FMT_NUM_B16 OPR_SRC 16 SIMM32 FMT_NUM_B16 OPR_SRC 16 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 605 VDST FMT_NUM_B16 OPR_VGPR 16 SIMM32 FMT_NUM_B16 OPR_SRC 16 SIMM32 FMT_NUM_B16 OPR_SRC 16 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_INST_LITERAL has_lit_2 605 VDST FMT_NUM_B16 OPR_VGPR 16 SRC0 FMT_NUM_B16 OPR_SRC 16 SRC1 FMT_NUM_B16 OPR_SRC 16 SIMM32 FMT_NUM_M64 OPR_SREG 64 VOP3_INST_LITERAL has_lit_0_has_lit_2 605 VDST FMT_NUM_B16 OPR_VGPR 16 SIMM32 FMT_NUM_B16 OPR_SRC 16 SRC1 FMT_NUM_B16 OPR_SRC 16 SIMM32 FMT_NUM_M64 OPR_SREG 64 VOP3_INST_LITERAL has_lit_1_has_lit_2 605 VDST FMT_NUM_B16 OPR_VGPR 16 SRC0 FMT_NUM_B16 OPR_SRC 16 SIMM32 FMT_NUM_B16 OPR_SRC 16 SIMM32 FMT_NUM_M64 OPR_SREG 64 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 605 VDST FMT_NUM_B16 OPR_VGPR 16 SIMM32 FMT_NUM_B16 OPR_SRC 16 SIMM32 FMT_NUM_B16 OPR_SRC 16 SIMM32 FMT_NUM_M64 OPR_SREG 64 VOP3_VOP_DPP16 default 605 VDST FMT_NUM_B16 OPR_VGPR 16 VSRC0 FMT_NUM_B16 OPR_VGPR 16 SRC1 FMT_NUM_B16 OPR_SRC 16 SRC2 FMT_NUM_M64 OPR_SREG 64 VOP3_VOP_DPP8 default 605 VDST FMT_NUM_B16 OPR_VGPR 16 VSRC0 FMT_NUM_B16 OPR_VGPR 16 SRC1 FMT_NUM_B16 OPR_SRC 16 SRC2 FMT_NUM_M64 OPR_SREG 64 VALU FALSE FALSE FALSE FALSE FALSE V_MAXMIN_U32 Select the maximum of the first two unsigned 32-bit integer inputs and then select the minimum of that result and third unsigned 32-bit integer input. Store the final result into a vector register. ENC_VOP3 default 610 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 610 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 610 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 610 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 610 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 610 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 610 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 610 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 610 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 610 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MINMAX_U32 Select the minimum of the first two unsigned 32-bit integer inputs and then select the maximum of that result and third unsigned 32-bit integer input. Store the final result into a vector register. ENC_VOP3 default 611 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 611 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 611 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 611 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 611 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 611 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 611 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 611 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 611 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 611 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MAXMIN_I32 Select the maximum of the first two signed 32-bit integer inputs and then select the minimum of that result and third signed 32-bit integer input. Store the final result into a vector register. ENC_VOP3 default 612 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 612 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 612 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 612 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 612 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 612 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 612 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 612 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 612 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 612 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MINMAX_I32 Select the minimum of the first two signed 32-bit integer inputs and then select the maximum of that result and third signed 32-bit integer input. Store the final result into a vector register. ENC_VOP3 default 613 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 613 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 613 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 613 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 613 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 613 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 613 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 613 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 613 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 613 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_DOT2_F16_F16 Compute the dot product of two packed 2-D half-precision float inputs, add the third input and store the result into a vector register. ENC_VOP3 default 614 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 614 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 614 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 614 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 614 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 614 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 614 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 614 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 614 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_PK2_F16 OPR_VGPR 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 614 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_PK2_F16 OPR_VGPR 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 SRC2 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_DOT2_BF16_BF16 Compute the dot product of two packed 2-D BF16 float inputs, add the third input and store the result into a vector register. ENC_VOP3 default 615 VDST FMT_NUM_BF16 OPR_VGPR 16 SRC0 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC1 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC2 FMT_NUM_BF16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 615 VDST FMT_NUM_BF16 OPR_VGPR 16 SIMM32 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC1 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC2 FMT_NUM_BF16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 615 VDST FMT_NUM_BF16 OPR_VGPR 16 SRC0 FMT_NUM_PK2_BF16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC2 FMT_NUM_BF16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 615 VDST FMT_NUM_BF16 OPR_VGPR 16 SIMM32 FMT_NUM_PK2_BF16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC2 FMT_NUM_BF16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 615 VDST FMT_NUM_BF16 OPR_VGPR 16 SRC0 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC1 FMT_NUM_PK2_BF16 OPR_SRC 32 SIMM32 FMT_NUM_BF16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 615 VDST FMT_NUM_BF16 OPR_VGPR 16 SIMM32 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC1 FMT_NUM_PK2_BF16 OPR_SRC 32 SIMM32 FMT_NUM_BF16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 615 VDST FMT_NUM_BF16 OPR_VGPR 16 SRC0 FMT_NUM_PK2_BF16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_BF16 OPR_SRC 32 SIMM32 FMT_NUM_BF16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 615 VDST FMT_NUM_BF16 OPR_VGPR 16 SIMM32 FMT_NUM_PK2_BF16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_BF16 OPR_SRC 32 SIMM32 FMT_NUM_BF16 OPR_SRC 16 VOP3_VOP_DPP16 default 615 VDST FMT_NUM_BF16 OPR_VGPR 16 VSRC0 FMT_NUM_PK2_BF16 OPR_VGPR 32 SRC1 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC2 FMT_NUM_BF16 OPR_SRC 16 VOP3_VOP_DPP8 default 615 VDST FMT_NUM_BF16 OPR_VGPR 16 VSRC0 FMT_NUM_PK2_BF16 OPR_VGPR 32 SRC1 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC2 FMT_NUM_BF16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MINMAX_NUM_F32 V_MINMAX_F32 Select the IEEE minimumNumber() of the first two single-precision float inputs and then select the IEEE maximumNumber() of that result and third single-precision float input. Store the final result into a vector register. ENC_VOP3 default 616 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 616 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 616 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 616 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 616 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 616 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 616 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 616 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 616 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 616 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MAXMIN_NUM_F32 V_MAXMIN_F32 Select the IEEE maximumNumber() of the first two single-precision float inputs and then select the IEEE minimumNumber() of that result and third single-precision float input. Store the final result into a vector register. ENC_VOP3 default 617 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 617 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 617 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 617 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 617 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 617 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 617 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 617 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 617 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 617 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MINMAX_NUM_F16 V_MINMAX_F16 Select the IEEE minimumNumber() of the first two half-precision float inputs and then select the IEEE maximumNumber() of that result and third half-precision float input. Store the final result into a vector register. ENC_VOP3 default 618 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 618 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 618 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 618 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 618 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 618 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 618 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 618 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 618 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 618 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MAXMIN_NUM_F16 V_MAXMIN_F16 Select the IEEE maximumNumber() of the first two half-precision float inputs and then select the IEEE minimumNumber() of that result and third half-precision float input. Store the final result into a vector register. ENC_VOP3 default 619 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 619 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 619 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 619 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 619 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 619 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 619 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 619 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 619 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 619 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MINIMUMMAXIMUM_F32 Select the IEEE minimum() of the first two single-precision float inputs and then select the IEEE maximum() of that result and third single-precision float input. Store the final result into a vector register. ENC_VOP3 default 620 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 620 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 620 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 620 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 620 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 620 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 620 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 620 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 620 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 620 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MAXIMUMMINIMUM_F32 Select the IEEE maximum() of the first two single-precision float inputs and then select the IEEE minimum() of that result and third single-precision float input. Store the final result into a vector register. ENC_VOP3 default 621 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 621 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 621 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 621 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_2 621 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_2 621 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1_has_lit_2 621 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 621 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 621 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 621 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MINIMUMMAXIMUM_F16 Select the IEEE minimum() of the first two half-precision float inputs and then select the IEEE maximum() of that result and third half-precision float input. Store the final result into a vector register. ENC_VOP3 default 622 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 622 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 622 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 622 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 622 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 622 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 622 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 622 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 622 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 622 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MAXIMUMMINIMUM_F16 Select the IEEE maximum() of the first two half-precision float inputs and then select the IEEE minimum() of that result and third half-precision float input. Store the final result into a vector register. ENC_VOP3 default 623 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 623 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 623 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 623 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_2 623 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_2 623 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1_has_lit_2 623 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 623 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 623 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 623 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 SRC2 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_S_EXP_F32 Calculate 2 raised to the power of the single-precision float input and store the result into a scalar register. ENC_VOP3 default 640 VDST FMT_NUM_F32 OPR_SREG 32 SRC0 FMT_NUM_F32 OPR_SSRC 32 VOP3_INST_LITERAL has_lit_0 640 VDST FMT_NUM_F32 OPR_SREG 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_S_EXP_F16 Calculate 2 raised to the power of the half-precision float input and store the result into a scalar register. ENC_VOP3 default 641 VDST FMT_NUM_PK2_F16 OPR_SREG 32 SRC0 FMT_NUM_F16 OPR_SREG_LITERAL 16 VOP3_INST_LITERAL has_lit_0 641 VDST FMT_NUM_PK2_F16 OPR_SREG 32 SIMM32 FMT_NUM_F16 OPR_SREG_LITERAL 16 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_S_LOG_F32 Calculate the base 2 logarithm of the single-precision float input and store the result into a scalar register. ENC_VOP3 default 642 VDST FMT_NUM_F32 OPR_SREG 32 SRC0 FMT_NUM_F32 OPR_SSRC 32 VOP3_INST_LITERAL has_lit_0 642 VDST FMT_NUM_F32 OPR_SREG 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_S_LOG_F16 Calculate the base 2 logarithm of the half-precision float input and store the result into a scalar register. ENC_VOP3 default 643 VDST FMT_NUM_PK2_F16 OPR_SREG 32 SRC0 FMT_NUM_F16 OPR_SREG_LITERAL 16 VOP3_INST_LITERAL has_lit_0 643 VDST FMT_NUM_PK2_F16 OPR_SREG 32 SIMM32 FMT_NUM_F16 OPR_SREG_LITERAL 16 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_S_RCP_F32 Calculate the reciprocal of the single-precision float input using IEEE rules and store the result into a scalar register. ENC_VOP3 default 644 VDST FMT_NUM_F32 OPR_SREG 32 SRC0 FMT_NUM_F32 OPR_SSRC 32 VOP3_INST_LITERAL has_lit_0 644 VDST FMT_NUM_F32 OPR_SREG 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_S_RCP_F16 Calculate the reciprocal of the half-precision float input using IEEE rules and store the result into a scalar register. ENC_VOP3 default 645 VDST FMT_NUM_PK2_F16 OPR_SREG 32 SRC0 FMT_NUM_F16 OPR_SREG_LITERAL 16 VOP3_INST_LITERAL has_lit_0 645 VDST FMT_NUM_PK2_F16 OPR_SREG 32 SIMM32 FMT_NUM_F16 OPR_SREG_LITERAL 16 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_S_RSQ_F32 Calculate the reciprocal of the square root of the single-precision float input using IEEE rules and store the result into a scalar register. ENC_VOP3 default 646 VDST FMT_NUM_F32 OPR_SREG 32 SRC0 FMT_NUM_F32 OPR_SSRC 32 VOP3_INST_LITERAL has_lit_0 646 VDST FMT_NUM_F32 OPR_SREG 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_S_RSQ_F16 Calculate the reciprocal of the square root of the half-precision float input using IEEE rules and store the result into a scalar register. ENC_VOP3 default 647 VDST FMT_NUM_PK2_F16 OPR_SREG 32 SRC0 FMT_NUM_F16 OPR_SREG_LITERAL 16 VOP3_INST_LITERAL has_lit_0 647 VDST FMT_NUM_PK2_F16 OPR_SREG 32 SIMM32 FMT_NUM_F16 OPR_SREG_LITERAL 16 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_S_SQRT_F32 Calculate the square root of the single-precision float input using IEEE rules and store the result into a scalar register. ENC_VOP3 default 648 VDST FMT_NUM_F32 OPR_SREG 32 SRC0 FMT_NUM_F32 OPR_SSRC 32 VOP3_INST_LITERAL has_lit_0 648 VDST FMT_NUM_F32 OPR_SREG 32 SIMM32 FMT_NUM_F32 OPR_SSRC 32 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_S_SQRT_F16 Calculate the square root of the half-precision float input using IEEE rules and store the result into a scalar register. ENC_VOP3 default 649 VDST FMT_NUM_PK2_F16 OPR_SREG 32 SRC0 FMT_NUM_F16 OPR_SREG_LITERAL 16 VOP3_INST_LITERAL has_lit_0 649 VDST FMT_NUM_PK2_F16 OPR_SREG 32 SIMM32 FMT_NUM_F16 OPR_SREG_LITERAL 16 VALU TRANSCENDENTAL FALSE FALSE FALSE FALSE FALSE V_DIV_SCALE_F32 Given a single-precision float value to scale in the first input, a denominator in the second input and a numerator in the third input, scale the first input for division if required to avoid subnormal terms appearing during application of the Newton-Raphson correction method. Store the scaled result into a vector register and set the vector condition code iff post-scaling is required. VOP3_SDST_ENC default 764 VDST FMT_NUM_F32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_SDST_ENC_INST_LITERAL has_lit_0 764 VDST FMT_NUM_F32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_SDST_ENC_INST_LITERAL has_lit_1 764 VDST FMT_NUM_F32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_SDST_ENC_INST_LITERAL has_lit_0_has_lit_1 764 VDST FMT_NUM_F32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3_SDST_ENC_INST_LITERAL has_lit_2 764 VDST FMT_NUM_F32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_SDST_ENC_INST_LITERAL has_lit_0_has_lit_2 764 VDST FMT_NUM_F32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_SDST_ENC_INST_LITERAL has_lit_1_has_lit_2 764 VDST FMT_NUM_F32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_SDST_ENC_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 764 VDST FMT_NUM_F32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_DIV_SCALE_F64 Given a double-precision float value to scale in the first input, a denominator in the second input and a numerator in the third input, scale the first input for division if required to avoid subnormal terms appearing during application of the Newton-Raphson correction method. Store the scaled result into a vector register and set the vector condition code iff post-scaling is required. VOP3_SDST_ENC default 765 VDST FMT_NUM_F64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 SRC2 FMT_NUM_F64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_0 765 VDST FMT_NUM_F64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 SRC2 FMT_NUM_F64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_1 765 VDST FMT_NUM_F64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC2 FMT_NUM_F64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_0_has_lit_1 765 VDST FMT_NUM_F64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC2 FMT_NUM_F64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_2 765 VDST FMT_NUM_F64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_0_has_lit_2 765 VDST FMT_NUM_F64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_1_has_lit_2 765 VDST FMT_NUM_F64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 765 VDST FMT_NUM_F64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_MAD_CO_U64_U32 V_MAD_U64_U32 Multiply two unsigned integer inputs, add a third unsigned integer input, store the result into a 64-bit vector register and store the overflow/carryout into a scalar mask register. VOP3_SDST_ENC default 766 VDST FMT_NUM_U64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_0 766 VDST FMT_NUM_U64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_1 766 VDST FMT_NUM_U64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_0_has_lit_1 766 VDST FMT_NUM_U64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC2 FMT_NUM_U64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_2 766 VDST FMT_NUM_U64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_0_has_lit_2 766 VDST FMT_NUM_U64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_1_has_lit_2 766 VDST FMT_NUM_U64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 766 VDST FMT_NUM_U64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_MAD_CO_I64_I32 V_MAD_I64_I32 Multiply two signed integer inputs, add a third signed integer input, store the result into a 64-bit vector register and store the overflow/carryout into a scalar mask register. VOP3_SDST_ENC default 767 VDST FMT_NUM_I64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_0 767 VDST FMT_NUM_I64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_1 767 VDST FMT_NUM_I64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_0_has_lit_1 767 VDST FMT_NUM_I64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC2 FMT_NUM_I64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_2 767 VDST FMT_NUM_I64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_0_has_lit_2 767 VDST FMT_NUM_I64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_1_has_lit_2 767 VDST FMT_NUM_I64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I64 OPR_SRC 64 VOP3_SDST_ENC_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 767 VDST FMT_NUM_I64 OPR_VGPR 64 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_ADD_CO_U32 Add two unsigned 32-bit integer inputs, store the result into a vector register and store the carry-out mask into a scalar register. VOP3_SDST_ENC default 768 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_SDST_ENC_INST_LITERAL has_lit_0 768 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_SDST_ENC_INST_LITERAL has_lit_1 768 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_SDST_ENC_INST_LITERAL has_lit_0_has_lit_1 768 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_SDST_ENC_VOP_DPP16 default 768 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_SDST_ENC_VOP_DPP8 default 768 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_SUB_CO_U32 Subtract the second unsigned 32-bit integer input from the first input, store the result into a vector register and store the carry-out mask into a scalar register. VOP3_SDST_ENC default 769 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_SDST_ENC_INST_LITERAL has_lit_0 769 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_SDST_ENC_INST_LITERAL has_lit_1 769 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_SDST_ENC_INST_LITERAL has_lit_0_has_lit_1 769 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_SDST_ENC_VOP_DPP16 default 769 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_SDST_ENC_VOP_DPP8 default 769 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_SUBREV_CO_U32 Subtract the first unsigned 32-bit integer input from the second input, store the result into a vector register and store the carry-out mask into a scalar register. VOP3_SDST_ENC default 770 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_SDST_ENC_INST_LITERAL has_lit_0 770 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_SDST_ENC_INST_LITERAL has_lit_1 770 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_SDST_ENC_INST_LITERAL has_lit_0_has_lit_1 770 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_SDST_ENC_VOP_DPP16 default 770 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_VGPR 32 VOP3_SDST_ENC_VOP_DPP8 default 770 VDST FMT_NUM_U32 OPR_VGPR 32 SDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_ADD_NC_U16 Add two unsigned 16-bit integer inputs and store the result into a vector register. No carry-in or carry-out support. ENC_VOP3 default 771 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 771 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 771 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 771 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 771 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 771 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_SUB_NC_U16 Subtract the second unsigned 16-bit integer input from the first input and store the result into a vector register. No carry-in or carry-out support. ENC_VOP3 default 772 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 772 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 772 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 772 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 772 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 772 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MUL_LO_U16 Multiply two unsigned 16-bit integer inputs and store the low bits of the result into a vector register. ENC_VOP3 default 773 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 773 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 773 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 773 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 773 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 773 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_PK_I16_F32 Convert two single-precision float inputs into a packed signed 16-bit integer value and store the result into a vector register. ENC_VOP3 default 774 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 774 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 774 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 774 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 774 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 774 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_PK_U16_F32 Convert two single-precision float inputs into a packed unsigned 16-bit integer value and store the result into a vector register. ENC_VOP3 default 775 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 775 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 775 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 775 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 775 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 775 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MAX_U16 Select the maximum of two unsigned 16-bit integer inputs and store the selected value into a vector register. ENC_VOP3 default 777 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 777 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 777 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 777 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 777 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 777 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MAX_I16 Select the maximum of two signed 16-bit integer inputs and store the selected value into a vector register. ENC_VOP3 default 778 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 778 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 778 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 778 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP16 default 778 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP8 default 778 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MIN_U16 Select the minimum of two unsigned 16-bit integer inputs and store the selected value into a vector register. ENC_VOP3 default 779 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 779 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 779 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 779 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 779 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 779 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MIN_I16 Select the minimum of two signed 16-bit integer inputs and store the selected value into a vector register. ENC_VOP3 default 780 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 780 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 780 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 780 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP16 default 780 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP8 default 780 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_ADD_NC_I16 Add two signed 16-bit integer inputs and store the result into a vector register. No carry-in or carry-out support. ENC_VOP3 default 781 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 781 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 781 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 781 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP16 default 781 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP8 default 781 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_SUB_NC_I16 Subtract the second signed 16-bit integer input from the first input and store the result into a vector register. No carry-in or carry-out support. ENC_VOP3 default 782 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 782 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 782 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 782 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP16 default 782 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP8 default 782 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_PERMLANE16_VAR_B32 Perform arbitrary gather-style operation within a row (16 contiguous lanes). ENC_VOP3 default 783 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC_VGPR 32 VOP3_INST_LITERAL has_lit_0 783 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC_VGPR 32 VOP3_INST_LITERAL has_lit_1 783 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC_VGPR 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 783 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_PERMLANEX16_VAR_B32 Perform arbitrary gather-style operation across two rows (each row is 16 contiguous lanes). ENC_VOP3 default 784 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC_VGPR 32 VOP3_INST_LITERAL has_lit_0 784 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC_VGPR 32 VOP3_INST_LITERAL has_lit_1 784 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC_VGPR 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 784 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_PACK_B32_F16 Pack two half-precision float values into a single 32-bit value and store the result into a vector register. ENC_VOP3 default 785 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 785 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 785 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 785 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 785 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 785 VDST FMT_NUM_B32 OPR_VGPR 32 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_PK_NORM_I16_F16 V_CVT_PKNORM_I16_F16 Convert from two half-precision float inputs to a packed signed normalized short and store the result into a vector register. ENC_VOP3 default 786 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 786 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 786 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 786 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 786 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 786 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_PK_NORM_U16_F16 V_CVT_PKNORM_U16_F16 Convert from two half-precision float inputs to a packed unsigned normalized short and store the result into a vector register. ENC_VOP3 default 787 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 787 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 787 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 787 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 787 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 787 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_LDEXP_F32 Multiply the first input, a floating point value, by an integral power of 2 specified in the second input, a signed integer value, and store the floating point result into a vector register. ENC_VOP3 default 796 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 796 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 796 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 796 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 796 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 796 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_BFM_B32 Calculate a bitfield mask given a field offset and size and store the result into a vector register. ENC_VOP3 default 797 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 797 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 797 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 797 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 797 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 797 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_BCNT_U32_B32 Count the number of "1" bits in the vector input and store the result into a vector register. ENC_VOP3 default 798 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 798 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 798 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 798 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 798 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 798 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MBCNT_LO_U32_B32 For each lane 0 <= N < 32, examine the N least significant bits of the first input and count how many of those bits are "1". For each lane 32 <= N < 64, all "1" bits in the first input are counted. Add this count to the value in the second input and store the result into a vector register. ENC_VOP3 default 799 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 799 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 799 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 799 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 799 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 799 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MBCNT_HI_U32_B32 For each lane 32 <= N < 64, examine the N least significant bits of the first input and count how many of those bits are "1". For lane positions 0 <= N < 32 no bits are examined and the count is zero. Add this count to the value in the second input and store the result into a vector register. ENC_VOP3 default 800 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 800 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 800 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 800 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 800 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 800 VDST FMT_NUM_U32 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_PK_NORM_I16_F32 V_CVT_PKNORM_I16_F32 Convert from two single-precision float inputs to a packed signed normalized short and store the result into a vector register. ENC_VOP3 default 801 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 801 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 801 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 801 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 801 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 801 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_PK_NORM_U16_F32 V_CVT_PKNORM_U16_F32 Convert from two single-precision float inputs to a packed unsigned normalized short and store the result into a vector register. ENC_VOP3 default 802 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 802 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 802 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 802 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 802 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 802 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_PK_U16_U32 Convert from two unsigned 32-bit integer inputs to a packed unsigned 16-bit integer value and store the result into a vector register. ENC_VOP3 default 803 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 803 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 803 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 803 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 803 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 803 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_PK_I16_I32 Convert from two signed 32-bit integer inputs to a packed signed 16-bit integer value and store the result into a vector register. ENC_VOP3 default 804 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 804 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 804 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 804 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 804 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 804 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_SUB_NC_I32 Subtract the second signed 32-bit integer input from the first input and store the result into a vector register. No carry-in or carry-out support. ENC_VOP3 default 805 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 805 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 805 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 805 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 805 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 805 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_ADD_NC_I32 Add two signed 32-bit integer inputs and store the result into a vector register. No carry-in or carry-out support. ENC_VOP3 default 806 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 806 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 806 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 806 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 806 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 806 VDST FMT_NUM_I32 OPR_VGPR 32 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_LDEXP_F64 Multiply the first input, a floating point value, by an integral power of 2 specified in the second input, a signed integer value, and store the floating point result into a vector register. ENC_VOP3 default 811 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 811 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 811 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 811 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MUL_LO_U32 Multiply two unsigned 32-bit integer inputs and store the result into a vector register. ENC_VOP3 default 812 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 812 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 812 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 812 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MUL_HI_U32 Multiply two unsigned 32-bit integer inputs and store the high 32 bits of the result into a vector register. ENC_VOP3 default 813 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 813 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 813 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 813 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MUL_HI_I32 Multiply two signed 32-bit integer inputs and store the high 32 bits of the result into a vector register. ENC_VOP3 default 814 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 814 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 814 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 814 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_TRIG_PREOP_F64 Look up a 53-bit segment of 2/PI using an integer segment select in the second input. Scale the intermediate result by the exponent from the first double-precision float input and store the double-precision float result into a vector register. ENC_VOP3 default 815 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 815 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 815 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 815 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_B32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_LSHLREV_B16 Given a shift count in the first vector input, calculate the logical shift left of the second vector input and store the result into a vector register. ENC_VOP3 default 824 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 824 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 824 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 824 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 824 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 824 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_LSHRREV_B16 Given a shift count in the first vector input, calculate the logical shift right of the second vector input and store the result into a vector register. ENC_VOP3 default 825 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 825 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 825 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 825 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 825 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 825 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_ASHRREV_I16 Given a shift count in the first vector input, calculate the arithmetic shift right (preserving sign bit) of the second vector input and store the result into a vector register. ENC_VOP3 default 826 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 826 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 826 VDST FMT_NUM_I16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 826 VDST FMT_NUM_I16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP16 default 826 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP8 default 826 VDST FMT_NUM_I16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_LSHRREV_B64 Given a shift count in the first vector input, calculate the logical shift right of the second vector input and store the result into a vector register. ENC_VOP3 default 829 VDST FMT_NUM_U64 OPR_VGPR 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 829 VDST FMT_NUM_U64 OPR_VGPR 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 829 VDST FMT_NUM_U64 OPR_VGPR 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 829 VDST FMT_NUM_U64 OPR_VGPR 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_ASHRREV_I64 Given a shift count in the first vector input, calculate the arithmetic shift right (preserving sign bit) of the second vector input and store the result into a vector register. ENC_VOP3 default 830 VDST FMT_NUM_I64 OPR_VGPR 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 830 VDST FMT_NUM_I64 OPR_VGPR 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 830 VDST FMT_NUM_I64 OPR_VGPR 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 830 VDST FMT_NUM_I64 OPR_VGPR 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_I64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_MINIMUM_F64 Select the IEEE minimum() of two double-precision float inputs and store the result into a vector register. ENC_VOP3 default 833 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 833 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 833 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 833 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_MAXIMUM_F64 Select the IEEE maximum() of two double-precision float inputs and store the result into a vector register. ENC_VOP3 default 834 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 834 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 834 VDST FMT_NUM_F64 OPR_VGPR 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 834 VDST FMT_NUM_F64 OPR_VGPR 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VALU FALSE FALSE FALSE FALSE FALSE V_READLANE_B32 Read the scalar value in the specified lane of the first input where the lane select is in the second input. Store the result into a scalar register. ENC_VOP3 default 864 VDST FMT_NUM_B32 OPR_SREG 32 SRC0 FMT_NUM_B32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SSRC_LANESEL 32 VALU FALSE FALSE FALSE FALSE FALSE V_WRITELANE_B32 Write the scalar value in the first input into the specified lane of a vector register where the lane select is in the second input. ENC_VOP3 default 865 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SSRC 32 SRC1 FMT_NUM_B32 OPR_SSRC_LANESEL 32 VOP3_INST_LITERAL has_lit_0 865 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SRC1 FMT_NUM_B32 OPR_SSRC_LANESEL 32 VOP3_INST_LITERAL has_lit_1 865 VDST FMT_NUM_B32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SSRC 32 SIMM32 FMT_NUM_B32 OPR_SSRC_LANESEL 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 865 VDST FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SSRC 32 SIMM32 FMT_NUM_B32 OPR_SSRC_LANESEL 32 VALU FALSE FALSE FALSE FALSE FALSE V_AND_B16 Calculate bitwise AND on two vector inputs and store the result into a vector register. ENC_VOP3 default 866 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 866 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 866 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 866 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 866 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 866 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_OR_B16 Calculate bitwise OR on two vector inputs and store the result into a vector register. ENC_VOP3 default 867 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 867 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 867 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 867 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 867 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 867 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_XOR_B16 Calculate bitwise XOR on two vector inputs and store the result into a vector register. ENC_VOP3 default 868 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 868 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 868 VDST FMT_NUM_U16 OPR_VGPR 16 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 868 VDST FMT_NUM_U16 OPR_VGPR 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 868 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 868 VDST FMT_NUM_U16 OPR_VGPR 16 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MINIMUM_F32 Select the IEEE minimum() of two single-precision float inputs and store the result into a vector register. ENC_VOP3 default 869 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 869 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 869 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 869 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 869 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 869 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MAXIMUM_F32 Select the IEEE maximum() of two single-precision float inputs and store the result into a vector register. ENC_VOP3 default 870 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 870 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 870 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 870 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 870 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 870 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_MINIMUM_F16 Select the IEEE minimum() of two half-precision float inputs and store the result into a vector register. ENC_VOP3 default 871 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 871 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 871 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 871 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 871 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 871 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_MAXIMUM_F16 Select the IEEE maximum() of two half-precision float inputs and store the result into a vector register. ENC_VOP3 default 872 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 872 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 872 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 872 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 872 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 872 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_PK_FP8_F32 Convert from two single-precision float inputs to a packed FP8 float value using round to nearest even semantics and store the result into 16 bits of a vector register using OPSEL. ENC_VOP3 default 873 VDST FMT_NUM_PK2_FP8 OPR_VGPR 16 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 873 VDST FMT_NUM_PK2_FP8 OPR_VGPR 16 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 873 VDST FMT_NUM_PK2_FP8 OPR_VGPR 16 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 873 VDST FMT_NUM_PK2_FP8 OPR_VGPR 16 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 873 VDST FMT_NUM_PK2_FP8 OPR_VGPR 16 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 873 VDST FMT_NUM_PK2_FP8 OPR_VGPR 16 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_PK_BF8_F32 Convert from two single-precision float inputs to a packed BF8 float value using round to nearest even semantics and store the result into 16 bits of a vector register using OPSEL. ENC_VOP3 default 874 VDST FMT_NUM_PK2_BF8 OPR_VGPR 16 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 874 VDST FMT_NUM_PK2_BF8 OPR_VGPR 16 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 874 VDST FMT_NUM_PK2_BF8 OPR_VGPR 16 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 874 VDST FMT_NUM_PK2_BF8 OPR_VGPR 16 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 874 VDST FMT_NUM_PK2_BF8 OPR_VGPR 16 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 874 VDST FMT_NUM_PK2_BF8 OPR_VGPR 16 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_SR_FP8_F32 Convert from a single-precision float input to an FP8 float value with stochastic rounding using seed data from the second input. Store the result into 8 bits of a vector register using OPSEL to determine which byte of the destination to overwrite. ENC_VOP3 default 875 VDST FMT_NUM_FP8 OPR_VGPR 8 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 875 VDST FMT_NUM_FP8 OPR_VGPR 8 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 875 VDST FMT_NUM_FP8 OPR_VGPR 8 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 875 VDST FMT_NUM_FP8 OPR_VGPR 8 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 875 VDST FMT_NUM_FP8 OPR_VGPR 8 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 875 VDST FMT_NUM_FP8 OPR_VGPR 8 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_CVT_SR_BF8_F32 Convert from a single-precision float input to a BF8 float value with stochastic rounding using seed data from the second input. Store the result into 8 bits of a vector register using OPSEL to determine which byte of the destination to overwrite. ENC_VOP3 default 876 VDST FMT_NUM_BF8 OPR_VGPR 8 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 876 VDST FMT_NUM_BF8 OPR_VGPR 8 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 876 VDST FMT_NUM_BF8 OPR_VGPR 8 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 876 VDST FMT_NUM_BF8 OPR_VGPR 8 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 876 VDST FMT_NUM_BF8 OPR_VGPR 8 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 876 VDST FMT_NUM_BF8 OPR_VGPR 8 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_MAD_I16 Multiply two packed signed 16-bit integer inputs component-wise, add a packed signed 16-bit integer value from a third input component-wise, and store the result into a vector register. ENC_VOP3P default 0 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_I16 OPR_SRC 32 SRC1 FMT_NUM_PK2_I16 OPR_SRC 32 SRC2 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 0 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 SRC1 FMT_NUM_PK2_I16 OPR_SRC 32 SRC2 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 0 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_I16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 SRC2 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 0 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 SRC2 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_2 0 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_I16 OPR_SRC 32 SRC1 FMT_NUM_PK2_I16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_2 0 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 SRC1 FMT_NUM_PK2_I16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1_has_lit_2 0 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_I16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 0 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_MUL_LO_U16 Multiply two packed unsigned 16-bit integer inputs component-wise and store the low bits of each resulting component into a vector register. ENC_VOP3P default 1 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 1 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 1 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 1 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_ADD_I16 Add two packed signed 16-bit integer inputs component-wise and store the result into a vector register. No carry-in or carry-out support. ENC_VOP3P default 2 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_I16 OPR_SRC 32 SRC1 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 2 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 SRC1 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 2 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_I16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 2 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_SUB_I16 Subtract the second packed signed 16-bit integer input from the first input component-wise and store the result into a vector register. No carry-in or carry-out support. ENC_VOP3P default 3 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_I16 OPR_SRC 32 SRC1 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 3 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 SRC1 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 3 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_I16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 3 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_LSHLREV_B16 Given a packed shift count in the first vector input, calculate the component-wise logical shift left of the second packed vector input and store the result into a vector register. ENC_VOP3P default 4 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 4 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 4 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 4 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_LSHRREV_B16 Given a packed shift count in the first vector input, calculate the component-wise logical shift right of the second packed vector input and store the result into a vector register. ENC_VOP3P default 5 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 5 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 5 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 5 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_ASHRREV_I16 Given a packed shift count in the first vector input, calculate the component-wise arithmetic shift right (preserving sign bit) of the second packed vector input and store the result into a vector register. ENC_VOP3P default 6 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 6 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 6 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 6 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_MAX_I16 Select the component-wise maximum of two packed signed 16-bit integer inputs and store the selected values into a vector register. ENC_VOP3P default 7 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_I16 OPR_SRC 32 SRC1 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 7 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 SRC1 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 7 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_I16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 7 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_MIN_I16 Select the component-wise minimum of two packed signed 16-bit integer inputs and store the selected values into a vector register. ENC_VOP3P default 8 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_I16 OPR_SRC 32 SRC1 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 8 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 SRC1 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 8 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_I16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 8 VDST FMT_NUM_PK2_I16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_I16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_MAD_U16 Multiply two packed unsigned 16-bit integer inputs component-wise, add a packed unsigned 16-bit integer value from a third input component-wise, and store the result into a vector register. ENC_VOP3P default 9 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 SRC2 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 9 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 SRC2 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 9 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SRC2 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 9 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SRC2 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_2 9 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_2 9 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1_has_lit_2 9 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 9 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_ADD_U16 Add two packed unsigned 16-bit integer inputs component-wise and store the result into a vector register. No carry-in or carry-out support. ENC_VOP3P default 10 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 10 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 10 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 10 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_SUB_U16 Subtract the second packed unsigned 16-bit integer input from the first input component-wise and store the result into a vector register. No carry-in or carry-out support. ENC_VOP3P default 11 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 11 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 11 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 11 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_MAX_U16 Select the component-wise maximum of two packed unsigned 16-bit integer inputs and store the selected values into a vector register. ENC_VOP3P default 12 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 12 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 12 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 12 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_MIN_U16 Select the component-wise minimum of two packed unsigned 16-bit integer inputs and store the selected values into a vector register. ENC_VOP3P default 13 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 13 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SRC1 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 13 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 13 VDST FMT_NUM_PK2_U16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_U16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_FMA_F16 Multiply two packed half-precision float inputs component-wise and add a third input component-wise using fused multiply add, and store the result into a vector register. ENC_VOP3P default 14 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 SRC2 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 14 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 SRC2 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 14 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SRC2 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 14 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SRC2 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_2 14 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_2 14 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1_has_lit_2 14 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 14 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_ADD_F16 Add two packed half-precision float inputs component-wise and store the result into a vector register. No carry-in or carry-out support. ENC_VOP3P default 15 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 15 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 15 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 15 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_MUL_F16 Multiply two packed half-precision float inputs component-wise and store the result into a vector register. ENC_VOP3P default 16 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 16 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 16 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 16 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_DOT2_F32_F16 Compute the dot product of two packed 2-D half-precision float inputs in the single-precision float domain, add a single-precision float value from the third input and store the result into a vector register. ENC_VOP3P default 19 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 19 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 19 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 19 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_2 19 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_2 19 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1_has_lit_2 19 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 19 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_VOP_DPP16 default 19 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_PK2_F16 OPR_VGPR 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_VOP_DPP8 default 19 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_PK2_F16 OPR_VGPR 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_DOT4_I32_IU8 V_DOT4_I32_I8 Compute the dot product of two packed 4-D unsigned 8-bit integer inputs in the signed 32-bit integer domain, add a signed 32-bit integer value from the third input and store the result into a vector register. ENC_VOP3P default 22 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_IU8 OPR_SRC 32 SRC1 FMT_NUM_PK4_IU8 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 22 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_IU8 OPR_SRC 32 SRC1 FMT_NUM_PK4_IU8 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 22 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_IU8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_IU8 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 22 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_IU8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_IU8 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_2 22 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_IU8 OPR_SRC 32 SRC1 FMT_NUM_PK4_IU8 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_2 22 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_IU8 OPR_SRC 32 SRC1 FMT_NUM_PK4_IU8 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1_has_lit_2 22 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_IU8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_IU8 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 22 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_IU8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_IU8 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_DOT4_U32_U8 Compute the dot product of two packed 4-D unsigned 8-bit integer inputs in the unsigned 32-bit integer domain, add an unsigned 32-bit integer value from the third input and store the result into a vector register. ENC_VOP3P default 23 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_U8 OPR_SRC 32 SRC1 FMT_NUM_PK4_U8 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 23 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SRC1 FMT_NUM_PK4_U8 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 23 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 23 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_2 23 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_U8 OPR_SRC 32 SRC1 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_2 23 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SRC1 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1_has_lit_2 23 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 23 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_U8 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_DOT8_I32_IU4 V_DOT8_I32_I4 Compute the dot product of two packed 8-D unsigned 4-bit integer inputs in the signed 32-bit integer domain, add a signed 32-bit integer value from the third input and store the result into a vector register. ENC_VOP3P default 24 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_PK8_IU4 OPR_SRC 32 SRC1 FMT_NUM_PK8_IU4 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 24 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_PK8_IU4 OPR_SRC 32 SRC1 FMT_NUM_PK8_IU4 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 24 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_PK8_IU4 OPR_SRC 32 SIMM32 FMT_NUM_PK8_IU4 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 24 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_PK8_IU4 OPR_SRC 32 SIMM32 FMT_NUM_PK8_IU4 OPR_SRC 32 SRC2 FMT_NUM_I32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_2 24 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_PK8_IU4 OPR_SRC 32 SRC1 FMT_NUM_PK8_IU4 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_2 24 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_PK8_IU4 OPR_SRC 32 SRC1 FMT_NUM_PK8_IU4 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1_has_lit_2 24 VDST FMT_NUM_I32 OPR_VGPR 32 SRC0 FMT_NUM_PK8_IU4 OPR_SRC 32 SIMM32 FMT_NUM_PK8_IU4 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 24 VDST FMT_NUM_I32 OPR_VGPR 32 SIMM32 FMT_NUM_PK8_IU4 OPR_SRC 32 SIMM32 FMT_NUM_PK8_IU4 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_DOT8_U32_U4 Compute the dot product of two packed 8-D unsigned 4-bit integer inputs in the unsigned 32-bit integer domain, add an unsigned 32-bit integer value from the third input and store the result into a vector register. ENC_VOP3P default 25 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK8_U4 OPR_SRC 32 SRC1 FMT_NUM_PK8_U4 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 25 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK8_U4 OPR_SRC 32 SRC1 FMT_NUM_PK8_U4 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 25 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK8_U4 OPR_SRC 32 SIMM32 FMT_NUM_PK8_U4 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 25 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK8_U4 OPR_SRC 32 SIMM32 FMT_NUM_PK8_U4 OPR_SRC 32 SRC2 FMT_NUM_U32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_2 25 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK8_U4 OPR_SRC 32 SRC1 FMT_NUM_PK8_U4 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_2 25 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK8_U4 OPR_SRC 32 SRC1 FMT_NUM_PK8_U4 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1_has_lit_2 25 VDST FMT_NUM_U32 OPR_VGPR 32 SRC0 FMT_NUM_PK8_U4 OPR_SRC 32 SIMM32 FMT_NUM_PK8_U4 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 25 VDST FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_PK8_U4 OPR_SRC 32 SIMM32 FMT_NUM_PK8_U4 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_DOT2_F32_BF16 Compute the dot product of two packed 2-D BF16 float inputs in the single-precision float domain, add a single-precision float value from the third input and store the result into a vector register. ENC_VOP3P default 26 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC1 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 26 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC1 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 26 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK2_BF16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 26 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_BF16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_2 26 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC1 FMT_NUM_PK2_BF16 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_2 26 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC1 FMT_NUM_PK2_BF16 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1_has_lit_2 26 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK2_BF16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_BF16 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 26 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_BF16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_BF16 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_VOP_DPP16 default 26 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_PK2_BF16 OPR_VGPR 32 SRC1 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_VOP_DPP8 default 26 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_PK2_BF16 OPR_VGPR 32 SRC1 FMT_NUM_PK2_BF16 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_MIN_NUM_F16 V_PK_MIN_F16 Select the component-wise IEEE minimumNumber() of two half-precision float inputs and store the result into a vector register. ENC_VOP3P default 27 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 27 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 27 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 27 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_MAX_NUM_F16 V_PK_MAX_F16 Select the component-wise IEEE maximumNumber() of two half-precision float inputs and store the result into a vector register. ENC_VOP3P default 28 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 28 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 28 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 28 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_MINIMUM_F16 Select the component-wise IEEE minimum() of two half-precision float inputs and store the result into a vector register. ENC_VOP3P default 29 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 29 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 29 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 29 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_PK_MAXIMUM_F16 Select the component-wise IEEE maximum() of two half-precision float inputs and store the result into a vector register. ENC_VOP3P default 30 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 30 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SRC1 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 30 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SRC0 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 30 VDST FMT_NUM_PK2_F16 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_FMA_MIX_F32 Multiply two inputs and add a third input using fused multiply add where the inputs are a mix of half-precision float and single-precision float values. Store the result into a vector register. ENC_VOP3P default 32 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 32 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 32 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 32 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_2 32 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_2 32 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1_has_lit_2 32 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 32 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3P_VOP_DPP16 default 32 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3P_VOP_DPP8 default 32 VDST FMT_NUM_F32 OPR_VGPR 32 VSRC0 FMT_NUM_B32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_FMA_MIXLO_F16 Multiply two inputs and add a third input using fused multiply add where the inputs are a mix of half-precision float and single-precision float values. Convert the result to a half-precision float. Store the result into the low bits of a vector register. ENC_VOP3P default 33 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 33 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 33 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 33 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_2 33 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_2 33 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1_has_lit_2 33 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 33 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3P_VOP_DPP16 default 33 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_B32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3P_VOP_DPP8 default 33 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_B32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_FMA_MIXHI_F16 Multiply two inputs and add a third input using fused multiply add where the inputs are a mix of half-precision float and single-precision float values. Convert the result to a half-precision float. Store the result into the high bits of a vector register. ENC_VOP3P default 34 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 34 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 34 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 34 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_2 34 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_2 34 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_B32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1_has_lit_2 34 VDST FMT_NUM_F16 OPR_VGPR 16 SRC0 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 34 VDST FMT_NUM_F16 OPR_VGPR 16 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3P_VOP_DPP16 default 34 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_B32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VOP3P_VOP_DPP8 default 34 VDST FMT_NUM_F16 OPR_VGPR 16 VSRC0 FMT_NUM_B32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 SRC2 FMT_NUM_B32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_DOT4_F32_FP8_BF8 Compute the dot product of a packed 4-D FP8 float input and a packed 4-D BF8 float input in the single-precision float domain, add a single-precision float value from the third input and store the result into a vector register. ENC_VOP3P default 36 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_FP8 OPR_SRC 32 SRC1 FMT_NUM_PK4_BF8 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 36 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_FP8 OPR_SRC 32 SRC1 FMT_NUM_PK4_BF8 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 36 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_FP8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_BF8 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 36 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_FP8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_BF8 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_2 36 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_FP8 OPR_SRC 32 SRC1 FMT_NUM_PK4_BF8 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_2 36 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_FP8 OPR_SRC 32 SRC1 FMT_NUM_PK4_BF8 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1_has_lit_2 36 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_FP8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_BF8 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 36 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_FP8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_BF8 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_DOT4_F32_BF8_FP8 Compute the dot product of a packed 4-D BF8 float input and a packed 4-D FP8 float input in the single-precision float domain, add a single-precision float value from the third input and store the result into a vector register. ENC_VOP3P default 37 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_BF8 OPR_SRC 32 SRC1 FMT_NUM_PK4_FP8 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 37 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_BF8 OPR_SRC 32 SRC1 FMT_NUM_PK4_FP8 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 37 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_BF8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_FP8 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 37 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_BF8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_FP8 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_2 37 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_BF8 OPR_SRC 32 SRC1 FMT_NUM_PK4_FP8 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_2 37 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_BF8 OPR_SRC 32 SRC1 FMT_NUM_PK4_FP8 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1_has_lit_2 37 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_BF8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_FP8 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 37 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_BF8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_FP8 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_DOT4_F32_FP8_FP8 Compute the dot product of two packed 4-D FP8 float inputs in the single-precision float domain, add a single-precision float value from the third input and store the result into a vector register. ENC_VOP3P default 38 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_FP8 OPR_SRC 32 SRC1 FMT_NUM_PK4_FP8 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 38 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_FP8 OPR_SRC 32 SRC1 FMT_NUM_PK4_FP8 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 38 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_FP8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_FP8 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 38 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_FP8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_FP8 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_2 38 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_FP8 OPR_SRC 32 SRC1 FMT_NUM_PK4_FP8 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_2 38 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_FP8 OPR_SRC 32 SRC1 FMT_NUM_PK4_FP8 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1_has_lit_2 38 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_FP8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_FP8 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 38 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_FP8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_FP8 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_DOT4_F32_BF8_BF8 Compute the dot product of two packed 4-D BF8 float inputs in the single-precision float domain, add a single-precision float value from the third input and store the result into a vector register. ENC_VOP3P default 39 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_BF8 OPR_SRC 32 SRC1 FMT_NUM_PK4_BF8 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0 39 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_BF8 OPR_SRC 32 SRC1 FMT_NUM_PK4_BF8 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1 39 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_BF8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_BF8 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1 39 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_BF8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_BF8 OPR_SRC 32 SRC2 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_2 39 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_BF8 OPR_SRC 32 SRC1 FMT_NUM_PK4_BF8 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_2 39 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_BF8 OPR_SRC 32 SRC1 FMT_NUM_PK4_BF8 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_1_has_lit_2 39 VDST FMT_NUM_F32 OPR_VGPR 32 SRC0 FMT_NUM_PK4_BF8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_BF8 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3P_INST_LITERAL has_lit_0_has_lit_1_has_lit_2 39 VDST FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK4_BF8 OPR_SRC 32 SIMM32 FMT_NUM_PK4_BF8 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_WMMA_F32_16X16X16_F16 Multiply the 16x16 matrix in the first input by the 16x16 matrix in the second input and add the 16x16 matrix in the third input using fused multiply add. Store the resulting matrix into vector registers. ENC_VOP3P default 64 VDST FMT_WMMA_DC_16X16_F32 OPR_VGPR 256 SRC0 FMT_WMMA_AB_16X16_F16 OPR_SRC_VGPR 128 SRC1 FMT_WMMA_AB_16X16_F16 OPR_SRC_VGPR 128 SRC2 FMT_WMMA_DC_16X16_F32 OPR_SRC_VGPR_OR_INLINE 256 VALU WMMA FALSE FALSE FALSE FALSE FALSE V_WMMA_F32_16X16X16_BF16 Multiply the 16x16 matrix in the first input by the 16x16 matrix in the second input and add the 16x16 matrix in the third input using fused multiply add. Store the resulting matrix into vector registers. ENC_VOP3P default 65 VDST FMT_WMMA_DC_16X16_F32 OPR_VGPR 256 SRC0 FMT_WMMA_AB_16X16_BF16 OPR_SRC_VGPR 128 SRC1 FMT_WMMA_AB_16X16_BF16 OPR_SRC_VGPR 128 SRC2 FMT_WMMA_DC_16X16_F32 OPR_SRC_VGPR_OR_INLINE 256 VALU WMMA FALSE FALSE FALSE FALSE FALSE V_WMMA_F16_16X16X16_F16 Multiply the 16x16 matrix in the first input by the 16x16 matrix in the second input and add the 16x16 matrix in the third input using fused multiply add. Store the resulting matrix into vector registers. ENC_VOP3P default 66 VDST FMT_WMMA_DC_16X16_F16 OPR_VGPR 128 SRC0 FMT_WMMA_AB_16X16_F16 OPR_SRC_VGPR 128 SRC1 FMT_WMMA_AB_16X16_F16 OPR_SRC_VGPR 128 SRC2 FMT_WMMA_DC_16X16_F16 OPR_SRC_VGPR_OR_INLINE 128 VALU WMMA FALSE FALSE FALSE FALSE FALSE V_WMMA_BF16_16X16X16_BF16 Multiply the 16x16 matrix in the first input by the 16x16 matrix in the second input and add the 16x16 matrix in the third input using fused multiply add. Store the resulting matrix into vector registers. ENC_VOP3P default 67 VDST FMT_WMMA_DC_16X16_BF16 OPR_VGPR 128 SRC0 FMT_WMMA_AB_16X16_BF16 OPR_SRC_VGPR 128 SRC1 FMT_WMMA_AB_16X16_BF16 OPR_SRC_VGPR 128 SRC2 FMT_WMMA_DC_16X16_BF16 OPR_SRC_VGPR_OR_INLINE 128 VALU WMMA FALSE FALSE FALSE FALSE FALSE V_WMMA_I32_16X16X16_IU8 Multiply the 16x16 matrix in the first input by the 16x16 matrix in the second input and add the 16x16 matrix in the third input using fused multiply add. Store the resulting matrix into vector registers. ENC_VOP3P default 68 VDST FMT_WMMA_DC_16X16_I32 OPR_VGPR 256 SRC0 FMT_WMMA_AB_16X16_IU8 OPR_SRC_VGPR 64 SRC1 FMT_WMMA_AB_16X16_IU8 OPR_SRC_VGPR 64 SRC2 FMT_WMMA_DC_16X16_I32 OPR_SRC_VGPR_OR_INLINE 256 VALU WMMA FALSE FALSE FALSE FALSE FALSE V_WMMA_I32_16X16X16_IU4 Multiply the 16x16 matrix in the first input by the 16x16 matrix in the second input and add the 16x16 matrix in the third input using fused multiply add. Store the resulting matrix into vector registers. ENC_VOP3P default 69 VDST FMT_WMMA_DC_16X16_I32 OPR_VGPR 256 SRC0 FMT_WMMA_AB_16X16_IU4 OPR_SRC_VGPR 32 SRC1 FMT_WMMA_AB_16X16_IU4 OPR_SRC_VGPR 32 SRC2 FMT_WMMA_DC_16X16_I32 OPR_SRC_VGPR_OR_INLINE 256 VALU WMMA FALSE FALSE FALSE FALSE FALSE V_WMMA_F32_16X16X16_FP8_FP8 WMMA matrix multiplication with 8-bit floating point inputs and 32-bit floating point result. Multiply the 16x16 matrix in the first input by the 16x16 matrix in the second input and add the 16x16 matrix in the third input using fused multiply add. Store the resulting matrix into vector registers. ENC_VOP3P default 70 VDST FMT_WMMA_DC_16X16_F32 OPR_VGPR 256 SRC0 FMT_WMMA_AB_16X16_FP8 OPR_SRC_VGPR 64 SRC1 FMT_WMMA_AB_16X16_FP8 OPR_SRC_VGPR 64 SRC2 FMT_WMMA_DC_16X16_F32 OPR_SRC_VGPR_OR_INLINE 256 VALU WMMA FALSE FALSE FALSE FALSE FALSE V_WMMA_F32_16X16X16_FP8_BF8 WMMA matrix multiplication with 8-bit floating point inputs and 32-bit floating point result. Multiply the 16x16 matrix in the first input by the 16x16 matrix in the second input and add the 16x16 matrix in the third input using fused multiply add. Store the resulting matrix into vector registers. ENC_VOP3P default 71 VDST FMT_WMMA_DC_16X16_F32 OPR_VGPR 256 SRC0 FMT_WMMA_AB_16X16_FP8 OPR_SRC_VGPR 64 SRC1 FMT_WMMA_AB_16X16_BF8 OPR_SRC_VGPR 64 SRC2 FMT_WMMA_DC_16X16_F32 OPR_SRC_VGPR_OR_INLINE 256 VALU WMMA FALSE FALSE FALSE FALSE FALSE V_WMMA_F32_16X16X16_BF8_FP8 WMMA matrix multiplication with 8-bit floating point inputs and 32-bit floating point result. Multiply the 16x16 matrix in the first input by the 16x16 matrix in the second input and add the 16x16 matrix in the third input using fused multiply add. Store the resulting matrix into vector registers. ENC_VOP3P default 72 VDST FMT_WMMA_DC_16X16_F32 OPR_VGPR 256 SRC0 FMT_WMMA_AB_16X16_BF8 OPR_SRC_VGPR 64 SRC1 FMT_WMMA_AB_16X16_FP8 OPR_SRC_VGPR 64 SRC2 FMT_WMMA_DC_16X16_F32 OPR_SRC_VGPR_OR_INLINE 256 VALU WMMA FALSE FALSE FALSE FALSE FALSE V_WMMA_F32_16X16X16_BF8_BF8 WMMA matrix multiplication with 8-bit floating point inputs and 32-bit floating point result. Multiply the 16x16 matrix in the first input by the 16x16 matrix in the second input and add the 16x16 matrix in the third input using fused multiply add. Store the resulting matrix into vector registers. ENC_VOP3P default 73 VDST FMT_WMMA_DC_16X16_F32 OPR_VGPR 256 SRC0 FMT_WMMA_AB_16X16_BF8 OPR_SRC_VGPR 64 SRC1 FMT_WMMA_AB_16X16_BF8 OPR_SRC_VGPR 64 SRC2 FMT_WMMA_DC_16X16_F32 OPR_SRC_VGPR_OR_INLINE 256 VALU WMMA FALSE FALSE FALSE FALSE FALSE V_WMMA_I32_16X16X32_IU4 Multiply the 16x32 matrix in the first input by the 32x16 matrix in the second input and add the 16x16 matrix in the third input using fused multiply add. Store the resulting matrix into vector registers. ENC_VOP3P default 74 VDST FMT_WMMA_DC_16X16_I32 OPR_VGPR 256 SRC0 FMT_WMMA_AB_16X32_IU4 OPR_SRC_VGPR 64 SRC1 FMT_WMMA_AB_16X32_IU4 OPR_SRC_VGPR 64 SRC2 FMT_WMMA_DC_16X16_I32 OPR_SRC_VGPR_OR_INLINE 256 VALU WMMA FALSE FALSE FALSE FALSE FALSE V_SWMMAC_F32_16X16X32_F16 Multiply the 16x32 matrix in the first input by the 32x16 matrix in the second input and accumulate the result into the 16x16 matrix in the destination registers using fused multiply add. Sparse indexes for the first matrix are given in the third input. ENC_VOP3P default 80 VDST FMT_WMMA_DC_16X16_F32 OPR_VGPR 256 SRC0 FMT_WMMA_AB_16X16_F16 OPR_SRC_VGPR 128 SRC1 FMT_WMMA_AB_16X32_F16 OPR_SRC_VGPR 256 SRC2 FMT_WMMA_INDEX_SET OPR_SRC_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_SWMMAC_F32_16X16X32_BF16 Multiply the 16x32 matrix in the first input by the 32x16 matrix in the second input and accumulate the result into the 16x16 matrix in the destination registers using fused multiply add. Sparse indexes for the first matrix are given in the third input. ENC_VOP3P default 81 VDST FMT_WMMA_DC_16X16_F32 OPR_VGPR 256 SRC0 FMT_WMMA_AB_16X16_BF16 OPR_SRC_VGPR 128 SRC1 FMT_WMMA_AB_16X32_BF16 OPR_SRC_VGPR 256 SRC2 FMT_WMMA_INDEX_SET OPR_SRC_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_SWMMAC_F16_16X16X32_F16 Multiply the 16x32 matrix in the first input by the 32x16 matrix in the second input and accumulate the result into the 16x16 matrix in the destination registers using fused multiply add. Sparse indexes for the first matrix are given in the third input. ENC_VOP3P default 82 VDST FMT_WMMA_DC_16X16_F16 OPR_VGPR 128 SRC0 FMT_WMMA_AB_16X16_F16 OPR_SRC_VGPR 128 SRC1 FMT_WMMA_AB_16X32_F16 OPR_SRC_VGPR 256 SRC2 FMT_WMMA_INDEX_SET OPR_SRC_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_SWMMAC_BF16_16X16X32_BF16 Multiply the 16x32 matrix in the first input by the 32x16 matrix in the second input and accumulate the result into the 16x16 matrix in the destination registers using fused multiply add. Sparse indexes for the first matrix are given in the third input. ENC_VOP3P default 83 VDST FMT_WMMA_DC_16X16_BF16 OPR_VGPR 128 SRC0 FMT_WMMA_AB_16X16_BF16 OPR_SRC_VGPR 128 SRC1 FMT_WMMA_AB_16X32_BF16 OPR_SRC_VGPR 256 SRC2 FMT_WMMA_INDEX_SET OPR_SRC_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_SWMMAC_I32_16X16X32_IU8 Multiply the 16x32 matrix in the first input by the 32x16 matrix in the second input and accumulate the result into the 16x16 matrix in the destination registers using fused multiply add. Sparse indexes for the first matrix are given in the third input. ENC_VOP3P default 84 VDST FMT_WMMA_DC_16X16_I32 OPR_VGPR 256 SRC0 FMT_WMMA_AB_16X16_IU8 OPR_SRC_VGPR 64 SRC1 FMT_WMMA_AB_16X32_IU8 OPR_SRC_VGPR 128 SRC2 FMT_WMMA_INDEX_SET OPR_SRC_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_SWMMAC_I32_16X16X32_IU4 Multiply the 16x32 matrix in the first input by the 32x16 matrix in the second input and accumulate the result into the 16x16 matrix in the destination registers using fused multiply add. Sparse indexes for the first matrix are given in the third input. ENC_VOP3P default 85 VDST FMT_WMMA_DC_16X16_I32 OPR_VGPR 256 SRC0 FMT_WMMA_AB_16X16_IU4 OPR_SRC_VGPR 32 SRC1 FMT_WMMA_AB_16X32_IU4 OPR_SRC_VGPR 64 SRC2 FMT_WMMA_INDEX_SET OPR_SRC_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_SWMMAC_I32_16X16X64_IU4 Multiply the 16x64 matrix in the first input by the 64x16 matrix in the second input and accumulate the result into the 16x16 matrix in the destination registers using fused multiply add. Sparse indexes for the first matrix are given in the third input. ENC_VOP3P default 86 VDST FMT_WMMA_DC_16X16_I32 OPR_VGPR 256 SRC0 FMT_WMMA_AB_16X32_IU4 OPR_SRC_VGPR 64 SRC1 FMT_WMMA_AB_16X64_IU4 OPR_SRC_VGPR 128 SRC2 FMT_WMMA_INDEX_SET OPR_SRC_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_SWMMAC_F32_16X16X32_FP8_FP8 Multiply the 16x32 matrix in the first input by the 32x16 matrix in the second input and accumulate the result into the 16x16 matrix in the destination registers using fused multiply add. Sparse indexes for the first matrix are given in the third input. ENC_VOP3P default 87 VDST FMT_WMMA_DC_16X16_F32 OPR_VGPR 256 SRC0 FMT_WMMA_AB_16X16_FP8 OPR_SRC_VGPR 64 SRC1 FMT_WMMA_AB_16X32_FP8 OPR_SRC_VGPR 128 SRC2 FMT_WMMA_INDEX_SET OPR_SRC_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_SWMMAC_F32_16X16X32_FP8_BF8 Multiply the 16x32 matrix in the first input by the 32x16 matrix in the second input and accumulate the result into the 16x16 matrix in the destination registers using fused multiply add. Sparse indexes for the first matrix are given in the third input. ENC_VOP3P default 88 VDST FMT_WMMA_DC_16X16_F32 OPR_VGPR 256 SRC0 FMT_WMMA_AB_16X16_FP8 OPR_SRC_VGPR 64 SRC1 FMT_WMMA_AB_16X32_BF8 OPR_SRC_VGPR 128 SRC2 FMT_WMMA_INDEX_SET OPR_SRC_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_SWMMAC_F32_16X16X32_BF8_FP8 Multiply the 16x32 matrix in the first input by the 32x16 matrix in the second input and accumulate the result into the 16x16 matrix in the destination registers using fused multiply add. Sparse indexes for the first matrix are given in the third input. ENC_VOP3P default 89 VDST FMT_WMMA_DC_16X16_F32 OPR_VGPR 256 SRC0 FMT_WMMA_AB_16X16_BF8 OPR_SRC_VGPR 64 SRC1 FMT_WMMA_AB_16X32_FP8 OPR_SRC_VGPR 128 SRC2 FMT_WMMA_INDEX_SET OPR_SRC_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_SWMMAC_F32_16X16X32_BF8_BF8 Multiply the 16x32 matrix in the first input by the 32x16 matrix in the second input and accumulate the result into the 16x16 matrix in the destination registers using fused multiply add. Sparse indexes for the first matrix are given in the third input. ENC_VOP3P default 90 VDST FMT_WMMA_DC_16X16_F32 OPR_VGPR 256 SRC0 FMT_WMMA_AB_16X16_BF8 OPR_SRC_VGPR 64 SRC1 FMT_WMMA_AB_16X32_BF8 OPR_SRC_VGPR 128 SRC2 FMT_WMMA_INDEX_SET OPR_SRC_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LT_F16 Set the per-lane condition code to 1 iff the first input is less than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 1 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_VOP_DPP8 default 1 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 1 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 1 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 1 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 1 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 1 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 1 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 ENC_VOPC default 1 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 1 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_EQ_F16 Set the per-lane condition code to 1 iff the first input is equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 2 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_VOP_DPP8 default 2 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 2 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 2 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 2 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 2 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 2 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 2 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 ENC_VOPC default 2 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 2 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LE_F16 Set the per-lane condition code to 1 iff the first input is less than or equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 3 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_VOP_DPP8 default 3 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 3 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 3 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 3 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 3 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 3 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 3 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 ENC_VOPC default 3 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 3 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_GT_F16 Set the per-lane condition code to 1 iff the first input is greater than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 4 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_VOP_DPP8 default 4 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 4 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 4 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 4 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 4 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 4 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 4 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 ENC_VOPC default 4 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 4 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LG_F16 Set the per-lane condition code to 1 iff the first input is less than or greater than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 5 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_VOP_DPP8 default 5 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 5 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 5 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 5 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 5 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 5 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 5 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 ENC_VOPC default 5 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 5 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_GE_F16 Set the per-lane condition code to 1 iff the first input is greater than or equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 6 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_VOP_DPP8 default 6 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 6 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 6 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 6 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 6 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 6 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 6 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 ENC_VOPC default 6 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 6 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_O_F16 Set the per-lane condition code to 1 iff the first input is orderable to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 7 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_VOP_DPP8 default 7 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 7 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 7 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 7 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 7 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 7 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 7 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 ENC_VOPC default 7 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 7 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_U_F16 Set the per-lane condition code to 1 iff the first input is not orderable to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 8 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_VOP_DPP8 default 8 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 8 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 8 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 8 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 8 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 8 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 8 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 ENC_VOPC default 8 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 8 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NGE_F16 Set the per-lane condition code to 1 iff the first input is not greater than or equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 9 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_VOP_DPP8 default 9 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 9 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 9 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 9 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 9 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 9 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 9 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 ENC_VOPC default 9 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 9 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NLG_F16 Set the per-lane condition code to 1 iff the first input is not less than or greater than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 10 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_VOP_DPP8 default 10 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 10 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 10 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 10 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 10 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 10 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 10 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 ENC_VOPC default 10 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 10 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NGT_F16 Set the per-lane condition code to 1 iff the first input is not greater than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 11 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_VOP_DPP8 default 11 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 11 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 11 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 11 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 11 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 11 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 11 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 ENC_VOPC default 11 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 11 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NLE_F16 Set the per-lane condition code to 1 iff the first input is not less than or equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 12 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_VOP_DPP8 default 12 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 12 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 12 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 12 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 12 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 12 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 12 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 ENC_VOPC default 12 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 12 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NEQ_F16 Set the per-lane condition code to 1 iff the first input is not equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 13 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_VOP_DPP8 default 13 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 13 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 13 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 13 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 13 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 13 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 13 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 ENC_VOPC default 13 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 13 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NLT_F16 Set the per-lane condition code to 1 iff the first input is not less than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 14 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_VOP_DPP8 default 14 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 14 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 14 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 14 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 14 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 14 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 14 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 ENC_VOPC default 14 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 14 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LT_F32 Set the per-lane condition code to 1 iff the first input is less than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 17 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_VOP_DPP8 default 17 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 17 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 17 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 17 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 17 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 17 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 17 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 ENC_VOPC default 17 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 17 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_EQ_F32 Set the per-lane condition code to 1 iff the first input is equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 18 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_VOP_DPP8 default 18 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 18 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 18 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 18 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 18 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 18 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 18 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 ENC_VOPC default 18 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 18 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LE_F32 Set the per-lane condition code to 1 iff the first input is less than or equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 19 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_VOP_DPP8 default 19 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 19 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 19 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 19 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 19 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 19 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 19 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 ENC_VOPC default 19 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 19 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_GT_F32 Set the per-lane condition code to 1 iff the first input is greater than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 20 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_VOP_DPP8 default 20 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 20 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 20 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 20 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 20 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 20 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 20 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 ENC_VOPC default 20 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 20 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LG_F32 Set the per-lane condition code to 1 iff the first input is less than or greater than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 21 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_VOP_DPP8 default 21 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 21 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 21 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 21 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 21 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 21 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 21 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 ENC_VOPC default 21 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 21 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_GE_F32 Set the per-lane condition code to 1 iff the first input is greater than or equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 22 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_VOP_DPP8 default 22 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 22 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 22 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 22 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 22 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 22 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 22 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 ENC_VOPC default 22 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 22 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_O_F32 Set the per-lane condition code to 1 iff the first input is orderable to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 23 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_VOP_DPP8 default 23 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 23 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 23 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 23 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 23 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 23 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 23 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 ENC_VOPC default 23 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 23 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_U_F32 Set the per-lane condition code to 1 iff the first input is not orderable to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 24 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_VOP_DPP8 default 24 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 24 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 24 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 24 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 24 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 24 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 24 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 ENC_VOPC default 24 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 24 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NGE_F32 Set the per-lane condition code to 1 iff the first input is not greater than or equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 25 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_VOP_DPP8 default 25 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 25 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 25 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 25 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 25 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 25 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 25 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 ENC_VOPC default 25 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 25 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NLG_F32 Set the per-lane condition code to 1 iff the first input is not less than or greater than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 26 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_VOP_DPP8 default 26 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 26 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 26 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 26 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 26 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 26 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 26 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 ENC_VOPC default 26 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 26 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NGT_F32 Set the per-lane condition code to 1 iff the first input is not greater than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 27 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_VOP_DPP8 default 27 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 27 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 27 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 27 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 27 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 27 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 27 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 ENC_VOPC default 27 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 27 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NLE_F32 Set the per-lane condition code to 1 iff the first input is not less than or equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 28 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_VOP_DPP8 default 28 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 28 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 28 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 28 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 28 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 28 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 28 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 ENC_VOPC default 28 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 28 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NEQ_F32 Set the per-lane condition code to 1 iff the first input is not equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 29 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_VOP_DPP8 default 29 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 29 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 29 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 29 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 29 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 29 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 29 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 ENC_VOPC default 29 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 29 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NLT_F32 Set the per-lane condition code to 1 iff the first input is not less than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 30 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_VOP_DPP8 default 30 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 ENC_VOP3 default 30 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 30 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 30 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 30 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP16 default 30 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 VOP3_VOP_DPP8 default 30 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 ENC_VOPC default 30 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 30 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LT_F64 Set the per-lane condition code to 1 iff the first input is less than the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 33 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 33 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 33 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 33 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOPC default 33 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 33 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_EQ_F64 Set the per-lane condition code to 1 iff the first input is equal to the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 34 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 34 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 34 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 34 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOPC default 34 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 34 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LE_F64 Set the per-lane condition code to 1 iff the first input is less than or equal to the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 35 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 35 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 35 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 35 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOPC default 35 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 35 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_GT_F64 Set the per-lane condition code to 1 iff the first input is greater than the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 36 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 36 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 36 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 36 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOPC default 36 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 36 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LG_F64 Set the per-lane condition code to 1 iff the first input is less than or greater than the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 37 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 37 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 37 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 37 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOPC default 37 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 37 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_GE_F64 Set the per-lane condition code to 1 iff the first input is greater than or equal to the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 38 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 38 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 38 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 38 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOPC default 38 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 38 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_O_F64 Set the per-lane condition code to 1 iff the first input is orderable to the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 39 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 39 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 39 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 39 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOPC default 39 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 39 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_U_F64 Set the per-lane condition code to 1 iff the first input is not orderable to the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 40 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 40 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 40 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 40 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOPC default 40 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 40 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NGE_F64 Set the per-lane condition code to 1 iff the first input is not greater than or equal to the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 41 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 41 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 41 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 41 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOPC default 41 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 41 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NLG_F64 Set the per-lane condition code to 1 iff the first input is not less than or greater than the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 42 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 42 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 42 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 42 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOPC default 42 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 42 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NGT_F64 Set the per-lane condition code to 1 iff the first input is not greater than the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 43 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 43 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 43 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 43 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOPC default 43 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 43 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NLE_F64 Set the per-lane condition code to 1 iff the first input is not less than or equal to the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 44 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 44 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 44 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 44 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOPC default 44 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 44 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NEQ_F64 Set the per-lane condition code to 1 iff the first input is not equal to the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 45 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 45 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 45 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 45 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOPC default 45 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 45 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NLT_F64 Set the per-lane condition code to 1 iff the first input is not less than the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 46 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 46 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 46 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 46 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 ENC_VOPC default 46 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 46 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LT_I16 Set the per-lane condition code to 1 iff the first input is less than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 49 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 VOPC_VOP_DPP8 default 49 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 ENC_VOP3 default 49 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 49 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 49 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 49 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP16 default 49 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP8 default 49 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 ENC_VOPC default 49 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 49 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_EQ_I16 Set the per-lane condition code to 1 iff the first input is equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 50 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 VOPC_VOP_DPP8 default 50 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 ENC_VOP3 default 50 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 50 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 50 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 50 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP16 default 50 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP8 default 50 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 ENC_VOPC default 50 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 50 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LE_I16 Set the per-lane condition code to 1 iff the first input is less than or equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 51 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 VOPC_VOP_DPP8 default 51 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 ENC_VOP3 default 51 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 51 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 51 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 51 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP16 default 51 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP8 default 51 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 ENC_VOPC default 51 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 51 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_GT_I16 Set the per-lane condition code to 1 iff the first input is greater than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 52 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 VOPC_VOP_DPP8 default 52 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 ENC_VOP3 default 52 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 52 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 52 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 52 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP16 default 52 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP8 default 52 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 ENC_VOPC default 52 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 52 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NE_I16 Set the per-lane condition code to 1 iff the first input is not equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 53 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 VOPC_VOP_DPP8 default 53 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 ENC_VOP3 default 53 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 53 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 53 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 53 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP16 default 53 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP8 default 53 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 ENC_VOPC default 53 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 53 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_GE_I16 Set the per-lane condition code to 1 iff the first input is greater than or equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 54 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 VOPC_VOP_DPP8 default 54 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 ENC_VOP3 default 54 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 54 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 54 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 54 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP16 default 54 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 VOP3_VOP_DPP8 default 54 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 ENC_VOPC default 54 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 54 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LT_U16 Set the per-lane condition code to 1 iff the first input is less than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 57 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 VOPC_VOP_DPP8 default 57 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 ENC_VOP3 default 57 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 57 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 57 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 57 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 57 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 57 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 ENC_VOPC default 57 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 57 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_EQ_U16 Set the per-lane condition code to 1 iff the first input is equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 58 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 VOPC_VOP_DPP8 default 58 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 ENC_VOP3 default 58 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 58 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 58 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 58 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 58 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 58 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 ENC_VOPC default 58 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 58 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LE_U16 Set the per-lane condition code to 1 iff the first input is less than or equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 59 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 VOPC_VOP_DPP8 default 59 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 ENC_VOP3 default 59 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 59 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 59 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 59 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 59 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 59 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 ENC_VOPC default 59 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 59 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_GT_U16 Set the per-lane condition code to 1 iff the first input is greater than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 60 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 VOPC_VOP_DPP8 default 60 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 ENC_VOP3 default 60 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 60 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 60 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 60 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 60 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 60 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 ENC_VOPC default 60 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 60 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NE_U16 Set the per-lane condition code to 1 iff the first input is not equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 61 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 VOPC_VOP_DPP8 default 61 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 ENC_VOP3 default 61 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 61 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 61 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 61 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 61 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 61 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 ENC_VOPC default 61 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 61 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_GE_U16 Set the per-lane condition code to 1 iff the first input is greater than or equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 62 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 VOPC_VOP_DPP8 default 62 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 ENC_VOP3 default 62 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 62 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 62 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 62 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP16 default 62 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 VOP3_VOP_DPP8 default 62 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 ENC_VOPC default 62 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 62 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LT_I32 Set the per-lane condition code to 1 iff the first input is less than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 65 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOPC_VOP_DPP8 default 65 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 ENC_VOP3 default 65 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 65 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 65 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 65 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 65 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 65 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 ENC_VOPC default 65 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 65 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_EQ_I32 Set the per-lane condition code to 1 iff the first input is equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 66 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOPC_VOP_DPP8 default 66 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 ENC_VOP3 default 66 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 66 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 66 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 66 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 66 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 66 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 ENC_VOPC default 66 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 66 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LE_I32 Set the per-lane condition code to 1 iff the first input is less than or equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 67 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOPC_VOP_DPP8 default 67 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 ENC_VOP3 default 67 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 67 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 67 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 67 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 67 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 67 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 ENC_VOPC default 67 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 67 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_GT_I32 Set the per-lane condition code to 1 iff the first input is greater than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 68 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOPC_VOP_DPP8 default 68 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 ENC_VOP3 default 68 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 68 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 68 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 68 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 68 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 68 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 ENC_VOPC default 68 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 68 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NE_I32 Set the per-lane condition code to 1 iff the first input is not equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 69 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOPC_VOP_DPP8 default 69 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 ENC_VOP3 default 69 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 69 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 69 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 69 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 69 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 69 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 ENC_VOPC default 69 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 69 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_GE_I32 Set the per-lane condition code to 1 iff the first input is greater than or equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 70 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOPC_VOP_DPP8 default 70 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 ENC_VOP3 default 70 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 70 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 70 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 70 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP16 default 70 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 VOP3_VOP_DPP8 default 70 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 ENC_VOPC default 70 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 70 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LT_U32 Set the per-lane condition code to 1 iff the first input is less than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 73 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOPC_VOP_DPP8 default 73 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP3 default 73 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 73 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 73 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 73 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 73 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 73 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 ENC_VOPC default 73 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 73 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_EQ_U32 Set the per-lane condition code to 1 iff the first input is equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 74 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOPC_VOP_DPP8 default 74 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP3 default 74 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 74 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 74 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 74 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 74 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 74 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 ENC_VOPC default 74 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 74 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LE_U32 Set the per-lane condition code to 1 iff the first input is less than or equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 75 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOPC_VOP_DPP8 default 75 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP3 default 75 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 75 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 75 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 75 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 75 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 75 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 ENC_VOPC default 75 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 75 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_GT_U32 Set the per-lane condition code to 1 iff the first input is greater than the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 76 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOPC_VOP_DPP8 default 76 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP3 default 76 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 76 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 76 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 76 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 76 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 76 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 ENC_VOPC default 76 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 76 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NE_U32 Set the per-lane condition code to 1 iff the first input is not equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 77 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOPC_VOP_DPP8 default 77 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP3 default 77 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 77 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 77 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 77 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 77 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 77 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 ENC_VOPC default 77 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 77 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_GE_U32 Set the per-lane condition code to 1 iff the first input is greater than or equal to the second input. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 78 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOPC_VOP_DPP8 default 78 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 ENC_VOP3 default 78 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 78 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 78 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 78 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP16 default 78 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 VOP3_VOP_DPP8 default 78 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 ENC_VOPC default 78 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 78 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LT_I64 Set the per-lane condition code to 1 iff the first input is less than the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 81 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 81 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 81 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 81 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 ENC_VOPC default 81 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 81 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_EQ_I64 Set the per-lane condition code to 1 iff the first input is equal to the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 82 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 82 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 82 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 82 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 ENC_VOPC default 82 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 82 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LE_I64 Set the per-lane condition code to 1 iff the first input is less than or equal to the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 83 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 83 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 83 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 83 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 ENC_VOPC default 83 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 83 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_GT_I64 Set the per-lane condition code to 1 iff the first input is greater than the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 84 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 84 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 84 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 84 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 ENC_VOPC default 84 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 84 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NE_I64 Set the per-lane condition code to 1 iff the first input is not equal to the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 85 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 85 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 85 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 85 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 ENC_VOPC default 85 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 85 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_GE_I64 Set the per-lane condition code to 1 iff the first input is greater than or equal to the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 86 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 86 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 86 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 86 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 ENC_VOPC default 86 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 86 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LT_U64 Set the per-lane condition code to 1 iff the first input is less than the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 89 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 89 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 89 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 89 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 ENC_VOPC default 89 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 89 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_EQ_U64 Set the per-lane condition code to 1 iff the first input is equal to the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 90 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 90 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 90 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 90 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 ENC_VOPC default 90 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 90 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_LE_U64 Set the per-lane condition code to 1 iff the first input is less than or equal to the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 91 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 91 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 91 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 91 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 ENC_VOPC default 91 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 91 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_GT_U64 Set the per-lane condition code to 1 iff the first input is greater than the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 92 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 92 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 92 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 92 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 ENC_VOPC default 92 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 92 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_NE_U64 Set the per-lane condition code to 1 iff the first input is not equal to the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 93 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 93 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 93 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 93 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 ENC_VOPC default 93 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 93 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_GE_U64 Set the per-lane condition code to 1 iff the first input is greater than or equal to the second input. Store the result into VCC or a scalar register. ENC_VOP3 default 94 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0 94 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_1 94 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 94 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 ENC_VOPC default 94 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 VOPC_INST_LITERAL has_lit 94 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_CLASS_F16 Evaluate the IEEE numeric class function specified as a 10 bit mask in the second input on the first input, a half-precision float, and set the per-lane condition code to the result. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 125 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_VOP_DPP8 default 125 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 ENC_VOP3 default 125 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0 125 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_1 125 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_INST_LITERAL has_lit_0_has_lit_1 125 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP16 default 125 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 VOP3_VOP_DPP8 default 125 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 ENC_VOPC default 125 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VOPC_INST_LITERAL has_lit 125 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_CLASS_F32 Evaluate the IEEE numeric class function specified as a 10 bit mask in the second input on the first input, a single-precision float, and set the per-lane condition code to the result. Store the result into VCC or a scalar register. VOPC_VOP_DPP16 default 126 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_B32 OPR_VGPR 32 VOPC_VOP_DPP8 default 126 FMT_NUM_M64 OPR_VCC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_B32 OPR_VGPR 32 ENC_VOP3 default 126 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 126 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 126 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 126 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_VOP_DPP16 default 126 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 VOP3_VOP_DPP8 default 126 VDST FMT_NUM_M64 OPR_SREG 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 ENC_VOPC default 126 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_B32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 126 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_B32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMP_CLASS_F64 Evaluate the IEEE numeric class function specified as a 10 bit mask in the second input on the first input, a double-precision float, and set the per-lane condition code to the result. Store the result into VCC or a scalar register. ENC_VOP3 default 127 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0 127 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_1 127 VDST FMT_NUM_M64 OPR_SREG 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOP3_INST_LITERAL has_lit_0_has_lit_1 127 VDST FMT_NUM_M64 OPR_SREG 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_B32 OPR_SRC 32 ENC_VOPC default 127 FMT_NUM_M64 OPR_VCC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_B32 OPR_VGPR 32 VOPC_INST_LITERAL has_lit 127 FMT_NUM_M64 OPR_VCC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_B32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LT_F16 Set the per-lane condition code to 1 iff the first input is less than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 129 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 129 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 129 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 129 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 129 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 129 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 129 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 129 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 129 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 129 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_EQ_F16 Set the per-lane condition code to 1 iff the first input is equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 130 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 130 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 130 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 130 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 130 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 130 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 130 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 130 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 130 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 130 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LE_F16 Set the per-lane condition code to 1 iff the first input is less than or equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 131 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 131 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 131 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 131 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 131 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 131 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 131 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 131 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 131 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 131 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_GT_F16 Set the per-lane condition code to 1 iff the first input is greater than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 132 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 132 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 132 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 132 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 132 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 132 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 132 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 132 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 132 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 132 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LG_F16 Set the per-lane condition code to 1 iff the first input is less than or greater than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 133 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 133 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 133 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 133 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 133 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 133 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 133 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 133 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 133 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 133 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_GE_F16 Set the per-lane condition code to 1 iff the first input is greater than or equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 134 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 134 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 134 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 134 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 134 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 134 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 134 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 134 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 134 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 134 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_O_F16 Set the per-lane condition code to 1 iff the first input is orderable to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 135 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 135 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 135 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 135 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 135 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 135 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 135 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 135 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 135 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 135 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_U_F16 Set the per-lane condition code to 1 iff the first input is not orderable to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 136 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 136 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 136 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 136 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 136 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 136 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 136 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 136 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 136 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 136 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NGE_F16 Set the per-lane condition code to 1 iff the first input is not greater than or equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 137 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 137 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 137 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 137 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 137 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 137 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 137 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 137 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 137 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 137 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NLG_F16 Set the per-lane condition code to 1 iff the first input is not less than or greater than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 138 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 138 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 138 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 138 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 138 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 138 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 138 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 138 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 138 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 138 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NGT_F16 Set the per-lane condition code to 1 iff the first input is not greater than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 139 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 139 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 139 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 139 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 139 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 139 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 139 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 139 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 139 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 139 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NLE_F16 Set the per-lane condition code to 1 iff the first input is not less than or equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 140 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 140 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 140 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 140 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 140 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 140 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 140 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 140 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 140 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 140 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NEQ_F16 Set the per-lane condition code to 1 iff the first input is not equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 141 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 141 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 141 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 141 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 141 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 141 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 141 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 141 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 141 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 141 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NLT_F16 Set the per-lane condition code to 1 iff the first input is not less than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 142 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 142 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 142 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 142 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 142 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 142 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 142 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 142 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 142 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 142 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LT_F32 Set the per-lane condition code to 1 iff the first input is less than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 145 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 145 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 145 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 145 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 145 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 145 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 145 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 145 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 145 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 145 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_EQ_F32 Set the per-lane condition code to 1 iff the first input is equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 146 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 146 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 146 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 146 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 146 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 146 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 146 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 146 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 146 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 146 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LE_F32 Set the per-lane condition code to 1 iff the first input is less than or equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 147 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 147 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 147 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 147 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 147 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 147 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 147 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 147 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 147 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 147 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_GT_F32 Set the per-lane condition code to 1 iff the first input is greater than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 148 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 148 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 148 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 148 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 148 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 148 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 148 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 148 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 148 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 148 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LG_F32 Set the per-lane condition code to 1 iff the first input is less than or greater than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 149 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 149 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 149 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 149 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 149 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 149 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 149 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 149 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 149 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 149 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_GE_F32 Set the per-lane condition code to 1 iff the first input is greater than or equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 150 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 150 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 150 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 150 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 150 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 150 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 150 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 150 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 150 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 150 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_O_F32 Set the per-lane condition code to 1 iff the first input is orderable to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 151 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 151 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 151 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 151 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 151 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 151 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 151 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 151 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 151 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 151 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_U_F32 Set the per-lane condition code to 1 iff the first input is not orderable to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 152 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 152 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 152 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 152 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 152 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 152 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 152 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 152 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 152 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 152 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NGE_F32 Set the per-lane condition code to 1 iff the first input is not greater than or equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 153 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 153 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 153 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 153 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 153 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 153 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 153 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 153 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 153 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 153 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NLG_F32 Set the per-lane condition code to 1 iff the first input is not less than or greater than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 154 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 154 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 154 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 154 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 154 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 154 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 154 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 154 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 154 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 154 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NGT_F32 Set the per-lane condition code to 1 iff the first input is not greater than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 155 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 155 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 155 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 155 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 155 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 155 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 155 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 155 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 155 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 155 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NLE_F32 Set the per-lane condition code to 1 iff the first input is not less than or equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 156 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 156 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 156 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 156 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 156 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 156 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 156 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 156 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 156 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 156 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NEQ_F32 Set the per-lane condition code to 1 iff the first input is not equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 157 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 157 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 157 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 157 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 157 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 157 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 157 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 157 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 157 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 157 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NLT_F32 Set the per-lane condition code to 1 iff the first input is not less than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 158 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 158 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 158 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 158 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 158 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 158 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 158 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 158 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 158 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 158 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LT_F64 Set the per-lane condition code to 1 iff the first input is less than the second input. Store the result into the EXEC mask. ENC_VOP3 default 161 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 161 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 161 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 161 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 161 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 161 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_EQ_F64 Set the per-lane condition code to 1 iff the first input is equal to the second input. Store the result into the EXEC mask. ENC_VOP3 default 162 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 162 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 162 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 162 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 162 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 162 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LE_F64 Set the per-lane condition code to 1 iff the first input is less than or equal to the second input. Store the result into the EXEC mask. ENC_VOP3 default 163 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 163 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 163 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 163 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 163 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 163 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_GT_F64 Set the per-lane condition code to 1 iff the first input is greater than the second input. Store the result into the EXEC mask. ENC_VOP3 default 164 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 164 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 164 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 164 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 164 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 164 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LG_F64 Set the per-lane condition code to 1 iff the first input is less than or greater than the second input. Store the result into the EXEC mask. ENC_VOP3 default 165 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 165 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 165 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 165 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 165 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 165 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_GE_F64 Set the per-lane condition code to 1 iff the first input is greater than or equal to the second input. Store the result into the EXEC mask. ENC_VOP3 default 166 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 166 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 166 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 166 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 166 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 166 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_O_F64 Set the per-lane condition code to 1 iff the first input is orderable to the second input. Store the result into the EXEC mask. ENC_VOP3 default 167 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 167 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 167 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 167 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 167 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 167 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_U_F64 Set the per-lane condition code to 1 iff the first input is not orderable to the second input. Store the result into the EXEC mask. ENC_VOP3 default 168 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 168 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 168 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 168 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 168 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 168 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NGE_F64 Set the per-lane condition code to 1 iff the first input is not greater than or equal to the second input. Store the result into the EXEC mask. ENC_VOP3 default 169 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 169 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 169 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 169 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 169 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 169 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NLG_F64 Set the per-lane condition code to 1 iff the first input is not less than or greater than the second input. Store the result into the EXEC mask. ENC_VOP3 default 170 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 170 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 170 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 170 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 170 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 170 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NGT_F64 Set the per-lane condition code to 1 iff the first input is not greater than the second input. Store the result into the EXEC mask. ENC_VOP3 default 171 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 171 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 171 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 171 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 171 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 171 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NLE_F64 Set the per-lane condition code to 1 iff the first input is not less than or equal to the second input. Store the result into the EXEC mask. ENC_VOP3 default 172 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 172 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 172 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 172 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 172 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 172 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NEQ_F64 Set the per-lane condition code to 1 iff the first input is not equal to the second input. Store the result into the EXEC mask. ENC_VOP3 default 173 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 173 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 173 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 173 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 173 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 173 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NLT_F64 Set the per-lane condition code to 1 iff the first input is not less than the second input. Store the result into the EXEC mask. ENC_VOP3 default 174 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 174 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 174 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 174 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 174 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 174 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_F64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LT_I16 Set the per-lane condition code to 1 iff the first input is less than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 177 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 177 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 177 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 177 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 177 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 177 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 177 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 177 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 177 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 177 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_EQ_I16 Set the per-lane condition code to 1 iff the first input is equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 178 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 178 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 178 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 178 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 178 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 178 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 178 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 178 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 178 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 178 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LE_I16 Set the per-lane condition code to 1 iff the first input is less than or equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 179 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 179 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 179 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 179 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 179 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 179 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 179 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 179 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 179 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 179 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_GT_I16 Set the per-lane condition code to 1 iff the first input is greater than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 180 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 180 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 180 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 180 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 180 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 180 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 180 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 180 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 180 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 180 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NE_I16 Set the per-lane condition code to 1 iff the first input is not equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 181 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 181 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 181 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 181 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 181 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 181 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 181 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 181 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 181 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 181 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_GE_I16 Set the per-lane condition code to 1 iff the first input is greater than or equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 182 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 182 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 182 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 182 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 182 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 182 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 SIMM32 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 182 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 182 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I16 OPR_VGPR 16 SRC1 FMT_NUM_I16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 182 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 182 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I16 OPR_SRC 16 VSRC1 FMT_NUM_I16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LT_U16 Set the per-lane condition code to 1 iff the first input is less than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 185 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 185 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 185 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 185 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 185 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 185 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 185 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 185 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 185 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 185 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_EQ_U16 Set the per-lane condition code to 1 iff the first input is equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 186 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 186 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 186 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 186 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 186 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 186 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 186 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 186 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 186 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 186 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LE_U16 Set the per-lane condition code to 1 iff the first input is less than or equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 187 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 187 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 187 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 187 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 187 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 187 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 187 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 187 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 187 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 187 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_GT_U16 Set the per-lane condition code to 1 iff the first input is greater than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 188 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 188 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 188 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 188 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 188 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 188 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 188 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 188 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 188 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 188 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NE_U16 Set the per-lane condition code to 1 iff the first input is not equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 189 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 189 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 189 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 189 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 189 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 189 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 189 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 189 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 189 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 189 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_GE_U16 Set the per-lane condition code to 1 iff the first input is greater than or equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 190 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 190 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 190 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 190 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 190 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 190 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 SIMM32 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 190 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 190 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U16 OPR_VGPR 16 SRC1 FMT_NUM_U16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 190 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 190 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U16 OPR_SRC 16 VSRC1 FMT_NUM_U16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LT_I32 Set the per-lane condition code to 1 iff the first input is less than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 193 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 193 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 193 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 193 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 193 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 193 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 193 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 193 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 193 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 193 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_EQ_I32 Set the per-lane condition code to 1 iff the first input is equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 194 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 194 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 194 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 194 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 194 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 194 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 194 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 194 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 194 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 194 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LE_I32 Set the per-lane condition code to 1 iff the first input is less than or equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 195 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 195 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 195 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 195 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 195 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 195 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 195 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 195 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 195 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 195 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_GT_I32 Set the per-lane condition code to 1 iff the first input is greater than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 196 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 196 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 196 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 196 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 196 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 196 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 196 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 196 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 196 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 196 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NE_I32 Set the per-lane condition code to 1 iff the first input is not equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 197 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 197 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 197 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 197 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 197 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 197 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 197 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 197 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 197 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 197 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_GE_I32 Set the per-lane condition code to 1 iff the first input is greater than or equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 198 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 198 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 198 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 198 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 198 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 198 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 SIMM32 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 198 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 198 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_I32 OPR_VGPR 32 SRC1 FMT_NUM_I32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 198 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 198 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I32 OPR_SRC 32 VSRC1 FMT_NUM_I32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LT_U32 Set the per-lane condition code to 1 iff the first input is less than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 201 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 201 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 201 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 201 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 201 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 201 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 201 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 201 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 201 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 201 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_EQ_U32 Set the per-lane condition code to 1 iff the first input is equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 202 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 202 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 202 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 202 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 202 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 202 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 202 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 202 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 202 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 202 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LE_U32 Set the per-lane condition code to 1 iff the first input is less than or equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 203 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 203 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 203 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 203 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 203 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 203 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 203 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 203 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 203 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 203 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_GT_U32 Set the per-lane condition code to 1 iff the first input is greater than the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 204 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 204 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 204 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 204 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 204 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 204 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 204 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 204 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 204 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 204 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NE_U32 Set the per-lane condition code to 1 iff the first input is not equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 205 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 205 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 205 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 205 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 205 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 205 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 205 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 205 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 205 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 205 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_GE_U32 Set the per-lane condition code to 1 iff the first input is greater than or equal to the second input. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 206 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 206 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 206 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 206 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 206 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 206 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 206 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 206 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_U32 OPR_VGPR 32 SRC1 FMT_NUM_U32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 206 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 206 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRC1 FMT_NUM_U32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LT_I64 Set the per-lane condition code to 1 iff the first input is less than the second input. Store the result into the EXEC mask. ENC_VOP3 default 209 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 209 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 209 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 209 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 209 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 209 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_EQ_I64 Set the per-lane condition code to 1 iff the first input is equal to the second input. Store the result into the EXEC mask. ENC_VOP3 default 210 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 210 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 210 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 210 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 210 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 210 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LE_I64 Set the per-lane condition code to 1 iff the first input is less than or equal to the second input. Store the result into the EXEC mask. ENC_VOP3 default 211 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 211 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 211 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 211 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 211 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 211 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_GT_I64 Set the per-lane condition code to 1 iff the first input is greater than the second input. Store the result into the EXEC mask. ENC_VOP3 default 212 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 212 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 212 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 212 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 212 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 212 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NE_I64 Set the per-lane condition code to 1 iff the first input is not equal to the second input. Store the result into the EXEC mask. ENC_VOP3 default 213 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 213 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 213 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 213 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 213 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 213 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_GE_I64 Set the per-lane condition code to 1 iff the first input is greater than or equal to the second input. Store the result into the EXEC mask. ENC_VOP3 default 214 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 214 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SRC1 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 214 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 214 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 214 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 214 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_I64 OPR_SRC 64 VSRC1 FMT_NUM_I64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LT_U64 Set the per-lane condition code to 1 iff the first input is less than the second input. Store the result into the EXEC mask. ENC_VOP3 default 217 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 217 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 217 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 217 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 217 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 217 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_EQ_U64 Set the per-lane condition code to 1 iff the first input is equal to the second input. Store the result into the EXEC mask. ENC_VOP3 default 218 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 218 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 218 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 218 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 218 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 218 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_LE_U64 Set the per-lane condition code to 1 iff the first input is less than or equal to the second input. Store the result into the EXEC mask. ENC_VOP3 default 219 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 219 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 219 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 219 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 219 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 219 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_GT_U64 Set the per-lane condition code to 1 iff the first input is greater than the second input. Store the result into the EXEC mask. ENC_VOP3 default 220 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 220 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 220 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 220 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 220 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 220 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_NE_U64 Set the per-lane condition code to 1 iff the first input is not equal to the second input. Store the result into the EXEC mask. ENC_VOP3 default 221 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 221 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 221 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 221 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 221 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 221 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_GE_U64 Set the per-lane condition code to 1 iff the first input is greater than or equal to the second input. Store the result into the EXEC mask. ENC_VOP3 default 222 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 222 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SRC1 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 222 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 222 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 222 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 222 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_U64 OPR_SRC 64 VSRC1 FMT_NUM_U64 OPR_VGPR 64 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_CLASS_F16 Evaluate the IEEE numeric class function specified as a 10 bit mask in the second input on the first input, a half-precision float, and set the per-lane condition code to the result. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 253 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 253 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 253 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 253 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 253 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 253 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 SIMM32 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 253 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 253 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F16 OPR_VGPR 16 SRC1 FMT_NUM_F16 OPR_SRC 16 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 253 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 253 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F16 OPR_SRC 16 VSRC1 FMT_NUM_F16 OPR_VGPR 16 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_CLASS_F32 Evaluate the IEEE numeric class function specified as a 10 bit mask in the second input on the first input, a single-precision float, and set the per-lane condition code to the result. Store the result into the EXEC mask. VOPC_VOP_DPP16 default 254 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_VOP_DPP8 default 254 FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 VSRC1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOP3 default 254 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 254 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SRC1 FMT_NUM_B32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 254 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 254 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP16 default 254 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_VOP_DPP8 default 254 VDST FMT_NUM_M64 OPR_EXEC 64 VSRC0 FMT_NUM_F32 OPR_VGPR 32 SRC1 FMT_NUM_B32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 254 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 254 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRC1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_CMPX_CLASS_F64 Evaluate the IEEE numeric class function specified as a 10 bit mask in the second input on the first input, a double-precision float, and set the per-lane condition code to the result. Store the result into the EXEC mask. ENC_VOP3 default 255 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_B32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0 255 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SRC1 FMT_NUM_B32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_1 255 VDST FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_B32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOP3_INST_LITERAL has_lit_0_has_lit_1 255 VDST FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 SIMM32 FMT_NUM_B32 OPR_SRC 32 FMT_NUM_M64 OPR_SDST_EXEC 64 ENC_VOPC default 255 FMT_NUM_M64 OPR_EXEC 64 SRC0 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VOPC_INST_LITERAL has_lit 255 FMT_NUM_M64 OPR_EXEC 64 SIMM32 FMT_NUM_F64 OPR_SRC 64 VSRC1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_M64 OPR_SDST_EXEC 64 VALU FALSE FALSE FALSE FALSE FALSE V_DUAL_FMAC_F32 Multiply two floating point inputs and accumulate the result into the destination register using fused multiply add. VOPDXY default 0 VDSTX FMT_NUM_F32 OPR_VGPR 32 SRCX0 FMT_NUM_F32 OPR_SRC 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 0 VDSTX FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY default 0 VDSTY FMT_NUM_F32 OPR_VGPR 32 SRCY0 FMT_NUM_F32 OPR_SRC 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 0 VDSTY FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_DUAL_FMAAK_F32 Multiply two single-precision float inputs and add a literal constant using fused multiply add, and store the result into a vector register. VOPDXY_INST_LITERAL default 1 VDSTX FMT_NUM_F32 OPR_VGPR 32 SRCX0 FMT_NUM_F32 OPR_SRC 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_F32 OPR_SIMM32 32 VOPDXY_INST_LITERAL has_lit 1 VDSTX FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_F32 OPR_SIMM32 32 VOPDXY_INST_LITERAL default 1 VDSTY FMT_NUM_F32 OPR_VGPR 32 SRCY0 FMT_NUM_F32 OPR_SRC 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_F32 OPR_SIMM32 32 VOPDXY_INST_LITERAL has_lit 1 VDSTY FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 FMT_NUM_F32 OPR_SIMM32 32 VALU FALSE FALSE FALSE FALSE FALSE V_DUAL_FMAMK_F32 Multiply a single-precision float input with a literal constant and add a second single-precision float input using fused multiply add, and store the result into a vector register. VOPDXY_INST_LITERAL default 2 VDSTX FMT_NUM_F32 OPR_VGPR 32 SRCX0 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_F32 OPR_SIMM32 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 2 VDSTX FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_F32 OPR_SIMM32 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY_INST_LITERAL default 2 VDSTY FMT_NUM_F32 OPR_VGPR 32 SRCY0 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_F32 OPR_SIMM32 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 2 VDSTY FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 FMT_NUM_F32 OPR_SIMM32 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_DUAL_MUL_F32 Multiply two floating point inputs and store the result into a vector register. VOPDXY default 3 VDSTX FMT_NUM_F32 OPR_VGPR 32 SRCX0 FMT_NUM_F32 OPR_SRC 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 3 VDSTX FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY default 3 VDSTY FMT_NUM_F32 OPR_VGPR 32 SRCY0 FMT_NUM_F32 OPR_SRC 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 3 VDSTY FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_DUAL_ADD_F32 Add two floating point inputs and store the result into a vector register. VOPDXY default 4 VDSTX FMT_NUM_F32 OPR_VGPR 32 SRCX0 FMT_NUM_F32 OPR_SRC 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 4 VDSTX FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY default 4 VDSTY FMT_NUM_F32 OPR_VGPR 32 SRCY0 FMT_NUM_F32 OPR_SRC 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 4 VDSTY FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_DUAL_SUB_F32 Subtract the second floating point input from the first input and store the result into a vector register. VOPDXY default 5 VDSTX FMT_NUM_F32 OPR_VGPR 32 SRCX0 FMT_NUM_F32 OPR_SRC 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 5 VDSTX FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY default 5 VDSTY FMT_NUM_F32 OPR_VGPR 32 SRCY0 FMT_NUM_F32 OPR_SRC 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 5 VDSTY FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_DUAL_SUBREV_F32 Subtract the first floating point input from the second input and store the result into a vector register. VOPDXY default 6 VDSTX FMT_NUM_F32 OPR_VGPR 32 SRCX0 FMT_NUM_F32 OPR_SRC 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 6 VDSTX FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY default 6 VDSTY FMT_NUM_F32 OPR_VGPR 32 SRCY0 FMT_NUM_F32 OPR_SRC 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 6 VDSTY FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_DUAL_MUL_DX9_ZERO_F32 Multiply two floating point inputs and store the result into a vector register. Follows DX9 rules where 0.0 times anything produces 0.0 (this differs from other APIs when the other input is infinity or NaN). VOPDXY default 7 VDSTX FMT_NUM_F32 OPR_VGPR 32 SRCX0 FMT_NUM_F32 OPR_SRC 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 7 VDSTX FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY default 7 VDSTY FMT_NUM_F32 OPR_VGPR 32 SRCY0 FMT_NUM_F32 OPR_SRC 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 7 VDSTY FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_DUAL_MOV_B32 Move 32-bit data from a vector input into a vector register. VOPDXY default 8 VDSTX FMT_NUM_B32 OPR_VGPR 32 SRCX0 FMT_NUM_B32 OPR_SRC 32 VOPDXY_INST_LITERAL has_lit 8 VDSTX FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VOPDXY default 8 VDSTY FMT_NUM_B32 OPR_VGPR 32 SRCY0 FMT_NUM_B32 OPR_SRC 32 VOPDXY_INST_LITERAL has_lit 8 VDSTY FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VALU FALSE FALSE FALSE FALSE FALSE V_DUAL_CNDMASK_B32 Copy data from one of two inputs based on the per-lane condition code and store the result into a vector register. VOPDXY default 9 VDSTX FMT_NUM_B32 OPR_VGPR 32 SRCX0 FMT_NUM_B32 OPR_SRC 32 VSRCX1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VOPDXY_INST_LITERAL has_lit 9 VDSTX FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VSRCX1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VOPDXY default 9 VDSTY FMT_NUM_B32 OPR_VGPR 32 SRCY0 FMT_NUM_B32 OPR_SRC 32 VSRCY1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VOPDXY_INST_LITERAL has_lit 9 VDSTY FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VSRCY1 FMT_NUM_B32 OPR_VGPR 32 FMT_NUM_M64 OPR_VCC 64 VALU FALSE FALSE FALSE FALSE FALSE V_DUAL_MAX_NUM_F32 Select the IEEE maximumNumber() of two single-precision float inputs and store the result into a vector register. VOPDXY default 10 VDSTX FMT_NUM_F32 OPR_VGPR 32 SRCX0 FMT_NUM_F32 OPR_SRC 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 10 VDSTX FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY default 10 VDSTY FMT_NUM_F32 OPR_VGPR 32 SRCY0 FMT_NUM_F32 OPR_SRC 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 10 VDSTY FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_DUAL_MIN_NUM_F32 Select the IEEE minimumNumber() of two single-precision float inputs and store the result into a vector register. VOPDXY default 11 VDSTX FMT_NUM_F32 OPR_VGPR 32 SRCX0 FMT_NUM_F32 OPR_SRC 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 11 VDSTX FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRCX1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY default 11 VDSTY FMT_NUM_F32 OPR_VGPR 32 SRCY0 FMT_NUM_F32 OPR_SRC 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 11 VDSTY FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_F32 OPR_SRC 32 VSRCY1 FMT_NUM_F32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_DUAL_DOT2ACC_F32_F16 Compute the dot product of two packed 2-D half-precision float inputs in the single-precision float domain and accumulate the resulting single-precision float value into the destination vector register. The initial value in D is used as S2. VOPDXY default 12 VDSTX FMT_NUM_F32 OPR_VGPR 32 SRCX0 FMT_NUM_PK2_F16 OPR_SRC 32 VSRCX1 FMT_NUM_PK2_F16 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 12 VDSTX FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 VSRCX1 FMT_NUM_PK2_F16 OPR_VGPR 32 VOPDXY default 12 VDSTY FMT_NUM_F32 OPR_VGPR 32 SRCY0 FMT_NUM_PK2_F16 OPR_SRC 32 VSRCY1 FMT_NUM_PK2_F16 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 12 VDSTY FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_F16 OPR_SRC 32 VSRCY1 FMT_NUM_PK2_F16 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_DUAL_DOT2ACC_F32_BF16 Dot product of packed brain-float values, accumulate with destination. The initial value in D is used as S2. VOPDXY default 13 VDSTX FMT_NUM_F32 OPR_VGPR 32 SRCX0 FMT_NUM_PK2_BF16 OPR_SRC 32 VSRCX1 FMT_NUM_PK2_BF16 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 13 VDSTX FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_BF16 OPR_SRC 32 VSRCX1 FMT_NUM_PK2_BF16 OPR_VGPR 32 VOPDXY default 13 VDSTY FMT_NUM_F32 OPR_VGPR 32 SRCY0 FMT_NUM_PK2_BF16 OPR_SRC 32 VSRCY1 FMT_NUM_PK2_BF16 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 13 VDSTY FMT_NUM_F32 OPR_VGPR 32 SIMM32 FMT_NUM_PK2_BF16 OPR_SRC 32 VSRCY1 FMT_NUM_PK2_BF16 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_DUAL_ADD_NC_U32 Add two unsigned 32-bit integer inputs and store the result into a vector register. No carry-in or carry-out support. VOPDXY default 16 VDSTY FMT_NUM_U32 OPR_VGPR 32 SRCY0 FMT_NUM_U32 OPR_SRC 32 VSRCY1 FMT_NUM_U32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 16 VDSTY FMT_NUM_U32 OPR_VGPR 32 SIMM32 FMT_NUM_U32 OPR_SRC 32 VSRCY1 FMT_NUM_U32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_DUAL_LSHLREV_B32 Given a shift count in the first vector input, calculate the logical shift left of the second vector input and store the result into a vector register. VOPDXY default 17 VDSTY FMT_NUM_B32 OPR_VGPR 32 SRCY0 FMT_NUM_B32 OPR_SRC 32 VSRCY1 FMT_NUM_B32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 17 VDSTY FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VSRCY1 FMT_NUM_B32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE V_DUAL_AND_B32 Calculate bitwise AND on two vector inputs and store the result into a vector register. VOPDXY default 18 VDSTY FMT_NUM_B32 OPR_VGPR 32 SRCY0 FMT_NUM_B32 OPR_SRC 32 VSRCY1 FMT_NUM_B32 OPR_VGPR 32 VOPDXY_INST_LITERAL has_lit 18 VDSTY FMT_NUM_B32 OPR_VGPR 32 SIMM32 FMT_NUM_B32 OPR_SRC 32 VSRCY1 FMT_NUM_B32 OPR_VGPR 32 VALU FALSE FALSE FALSE FALSE FALSE IMAGE_MSAA_LOAD Load up to 4 samples of 1 component from an MSAA resource with a user-specified fragment ID. No sampling is performed. ENC_VSAMPLE default 24 VDATA FMT_ANY OPR_VGPR 128 FMT_ANY OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. ENC_VSAMPLE default 27 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 96 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_D Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for user derivatives are provided by the address registers. ENC_VSAMPLE default 28 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 288 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_L Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for LOD are provided by the address registers. ENC_VSAMPLE default 29 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_B Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for LOD bias are provided by the address registers. ENC_VSAMPLE default 30 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_LZ Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for are provided by the address registers. Mipmap level is set to zero. ENC_VSAMPLE default 31 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 96 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF are provided by the address registers. ENC_VSAMPLE default 32 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C_D Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF, user derivatives are provided by the address registers. ENC_VSAMPLE default 33 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 320 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C_L Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF, LOD are provided by the address registers. ENC_VSAMPLE default 34 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 160 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C_B Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF, LOD bias are provided by the address registers. ENC_VSAMPLE default 35 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 160 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C_LZ Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF are provided by the address registers. Mipmap level is set to zero. ENC_VSAMPLE default 36 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_O Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for user offsets are provided by the address registers. ENC_VSAMPLE default 37 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_D_O Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for user derivatives, user offsets are provided by the address registers. ENC_VSAMPLE default 38 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 320 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_L_O Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for LOD, user offsets are provided by the address registers. ENC_VSAMPLE default 39 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 160 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_B_O Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for LOD bias, user offsets are provided by the address registers. ENC_VSAMPLE default 40 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 160 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_LZ_O Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for user offsets are provided by the address registers. Mipmap level is set to zero. ENC_VSAMPLE default 41 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C_O Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF, user offsets are provided by the address registers. ENC_VSAMPLE default 42 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 160 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C_D_O Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF, user derivatives, user offsets are provided by the address registers. ENC_VSAMPLE default 43 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 352 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C_L_O Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF, LOD, user offsets are provided by the address registers. ENC_VSAMPLE default 44 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 192 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C_B_O Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF, LOD bias, user offsets are provided by the address registers. ENC_VSAMPLE default 45 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 192 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C_LZ_O Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF, user offsets are provided by the address registers. Mipmap level is set to zero. ENC_VSAMPLE default 46 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 160 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_GATHER4 Gather 4 single-component texels from a 2x2 matrix on an image surface. Store the result into vector registers. The DMASK selects which channel to read from (R, G, B, A) and must only have one bit set to 1. ENC_VSAMPLE default 47 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 96 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_GATHER4_L Gather 4 single-component texels from a 2x2 matrix on an image surface. Store the result into vector registers. The DMASK selects which channel to read from (R, G, B, A) and must only have one bit set to 1. Additional data for LOD are provided by the address registers. ENC_VSAMPLE default 48 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_GATHER4_B Gather 4 single-component texels from a 2x2 matrix on an image surface. Store the result into vector registers. The DMASK selects which channel to read from (R, G, B, A) and must only have one bit set to 1. Additional data for LOD bias are provided by the address registers. ENC_VSAMPLE default 49 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_GATHER4_LZ Gather 4 single-component texels from a 2x2 matrix on an image surface. Store the result into vector registers. The DMASK selects which channel to read from (R, G, B, A) and must only have one bit set to 1. Additional data for are provided by the address registers. Mipmap level is set to zero. ENC_VSAMPLE default 50 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 96 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_GATHER4_C Gather 4 single-component texels from a 2x2 matrix on an image surface. Store the result into vector registers. The DMASK selects which channel to read from (R, G, B, A) and must only have one bit set to 1. Additional data for PCF are provided by the address registers. ENC_VSAMPLE default 51 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_GATHER4_C_LZ Gather 4 single-component texels from a 2x2 matrix on an image surface. Store the result into vector registers. The DMASK selects which channel to read from (R, G, B, A) and must only have one bit set to 1. Additional data for PCF are provided by the address registers. Mipmap level is set to zero. ENC_VSAMPLE default 52 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_GATHER4_O Gather 4 single-component texels from a 2x2 matrix on an image surface. Store the result into vector registers. The DMASK selects which channel to read from (R, G, B, A) and must only have one bit set to 1. Additional data for user offsets are provided by the address registers. ENC_VSAMPLE default 53 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_GATHER4_LZ_O Gather 4 single-component texels from a 2x2 matrix on an image surface. Store the result into vector registers. The DMASK selects which channel to read from (R, G, B, A) and must only have one bit set to 1. Additional data for user offsets are provided by the address registers. Mipmap level is set to zero. ENC_VSAMPLE default 54 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_GATHER4_C_LZ_O Gather 4 single-component texels from a 2x2 matrix on an image surface. Store the result into vector registers. The DMASK selects which channel to read from (R, G, B, A) and must only have one bit set to 1. Additional data for PCF, user offsets are provided by the address registers. Mipmap level is set to zero. ENC_VSAMPLE default 55 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 160 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_GET_LOD Return the calculated level of detail (LOD) for the provided input as two single-precision float values. No memory access is performed. ENC_VSAMPLE default 56 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 96 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_D_G16 Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for 16-bit derivatives are provided by the address registers. ENC_VSAMPLE default 57 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 224 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C_D_G16 Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF, 16-bit derivatives are provided by the address registers. ENC_VSAMPLE default 58 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 256 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_D_O_G16 Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for user offsets, 16-bit derivatives are provided by the address registers. ENC_VSAMPLE default 59 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 256 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C_D_O_G16 Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF, user offsets, 16-bit derivatives are provided by the address registers. ENC_VSAMPLE default 60 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 288 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_CL Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for LOD clamp are provided by the address registers. ENC_VSAMPLE default 64 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_D_CL Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for user derivatives, LOD clamp are provided by the address registers. ENC_VSAMPLE default 65 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 320 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_B_CL Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for LOD bias, LOD clamp are provided by the address registers. ENC_VSAMPLE default 66 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 160 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C_CL Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF, LOD clamp are provided by the address registers. ENC_VSAMPLE default 67 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 160 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C_D_CL Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF, user derivatives, LOD clamp are provided by the address registers. ENC_VSAMPLE default 68 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 352 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C_B_CL Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF, LOD bias, LOD clamp are provided by the address registers. ENC_VSAMPLE default 69 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 192 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_CL_O Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for LOD clamp, user offsets are provided by the address registers. ENC_VSAMPLE default 70 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 160 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_D_CL_O Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for user derivatives, LOD clamp, user offsets are provided by the address registers. ENC_VSAMPLE default 71 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 352 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_B_CL_O Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for LOD bias, LOD clamp, user offsets are provided by the address registers. ENC_VSAMPLE default 72 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 192 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C_CL_O Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF, LOD clamp, user offsets are provided by the address registers. ENC_VSAMPLE default 73 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 192 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C_D_CL_O Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF, user derivatives, LOD clamp, user offsets are provided by the address registers. ENC_VSAMPLE default 74 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 384 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C_B_CL_O Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF, LOD bias, LOD clamp, user offsets are provided by the address registers. ENC_VSAMPLE default 75 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 224 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C_D_CL_G16 Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF, LOD clamp, 16-bit derivatives are provided by the address registers. ENC_VSAMPLE default 84 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 288 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_D_CL_O_G16 Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for LOD clamp, user offsets, 16-bit derivatives are provided by the address registers. ENC_VSAMPLE default 85 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 288 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_C_D_CL_O_G16 Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for PCF, LOD clamp, user offsets, 16-bit derivatives are provided by the address registers. ENC_VSAMPLE default 86 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 320 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_SAMPLE_D_CL_G16 Sample texels from an image surface using texel coordinates provided by the address input registers and store the result into vector registers. Additional data for LOD clamp, 16-bit derivatives are provided by the address registers. ENC_VSAMPLE default 95 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 256 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_GATHER4_CL Gather 4 single-component texels from a 2x2 matrix on an image surface. Store the result into vector registers. The DMASK selects which channel to read from (R, G, B, A) and must only have one bit set to 1. Additional data for LOD clamp are provided by the address registers. ENC_VSAMPLE default 96 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 128 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_GATHER4_B_CL Gather 4 single-component texels from a 2x2 matrix on an image surface. Store the result into vector registers. The DMASK selects which channel to read from (R, G, B, A) and must only have one bit set to 1. Additional data for LOD bias, LOD clamp are provided by the address registers. ENC_VSAMPLE default 97 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 160 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_GATHER4_C_CL Gather 4 single-component texels from a 2x2 matrix on an image surface. Store the result into vector registers. The DMASK selects which channel to read from (R, G, B, A) and must only have one bit set to 1. Additional data for PCF, LOD clamp are provided by the address registers. ENC_VSAMPLE default 98 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 160 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_GATHER4_C_L Gather 4 single-component texels from a 2x2 matrix on an image surface. Store the result into vector registers. The DMASK selects which channel to read from (R, G, B, A) and must only have one bit set to 1. Additional data for PCF, LOD are provided by the address registers. ENC_VSAMPLE default 99 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 160 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_GATHER4_C_B Gather 4 single-component texels from a 2x2 matrix on an image surface. Store the result into vector registers. The DMASK selects which channel to read from (R, G, B, A) and must only have one bit set to 1. Additional data for PCF, LOD bias are provided by the address registers. ENC_VSAMPLE default 100 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 160 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_GATHER4_C_B_CL Gather 4 single-component texels from a 2x2 matrix on an image surface. Store the result into vector registers. The DMASK selects which channel to read from (R, G, B, A) and must only have one bit set to 1. Additional data for PCF, LOD bias, LOD clamp are provided by the address registers. ENC_VSAMPLE default 101 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 192 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE IMAGE_GATHER4H Gather 4 single-component texels from a 4x1 row vector on an image surface. Store the result into vector registers. The DMASK selects which channel to read from (R, G, B, A) and must only have one bit set to 1. ENC_VSAMPLE default 144 VDATA FMT_ANY OPR_VGPR 128 FMT_NUM_F32 OPR_VGPR 96 RSRC FMT_IMG OPR_SREG 256 SAMP FMT_SAMP OPR_SREG 128 VMEM SAMPLE FALSE FALSE FALSE FALSE FALSE SCRATCH_LOAD_U8 SCRATCH_LOAD_UBYTE Load 8 bits of unsigned data from the scratch aperture, zero extend to 32 bits and store the result into a vector register. ENC_VSCRATCH default 16 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_U8 OPR_GPUMEM 8 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_LOAD_I8 SCRATCH_LOAD_SBYTE Load 8 bits of signed data from the scratch aperture, sign extend to 32 bits and store the result into a vector register. ENC_VSCRATCH default 17 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_I8 OPR_GPUMEM 8 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_LOAD_U16 SCRATCH_LOAD_USHORT Load 16 bits of unsigned data from the scratch aperture, zero extend to 32 bits and store the result into a vector register. ENC_VSCRATCH default 18 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_U16 OPR_GPUMEM 16 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_LOAD_I16 SCRATCH_LOAD_SSHORT Load 16 bits of signed data from the scratch aperture, sign extend to 32 bits and store the result into a vector register. ENC_VSCRATCH default 19 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_I16 OPR_GPUMEM 16 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_LOAD_B32 SCRATCH_LOAD_DWORD Load 32 bits of data from the scratch aperture into a vector register. ENC_VSCRATCH default 20 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_LOAD_B64 SCRATCH_LOAD_DWORDX2 Load 64 bits of data from the scratch aperture into a vector register. ENC_VSCRATCH default 21 VDST FMT_ANY OPR_VGPR 64 VADDR FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_B64 OPR_GPUMEM 64 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_LOAD_B96 SCRATCH_LOAD_DWORDX3 Load 96 bits of data from the scratch aperture into a vector register. ENC_VSCRATCH default 22 VDST FMT_ANY OPR_VGPR 96 VADDR FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_B96 OPR_GPUMEM 96 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_LOAD_B128 SCRATCH_LOAD_DWORDX4 Load 128 bits of data from the scratch aperture into a vector register. ENC_VSCRATCH default 23 VDST FMT_ANY OPR_VGPR 128 VADDR FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_B128 OPR_GPUMEM 128 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_STORE_B8 SCRATCH_STORE_BYTE Store 8 bits of data from a vector register into the scratch aperture. ENC_VSCRATCH default 24 VADDR FMT_ANY OPR_VGPR 32 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_B8 OPR_GPUMEM 8 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_STORE_B16 SCRATCH_STORE_SHORT Store 16 bits of data from a vector register into the scratch aperture. ENC_VSCRATCH default 25 VADDR FMT_ANY OPR_VGPR 32 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_B16 OPR_GPUMEM 16 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_STORE_B32 SCRATCH_STORE_DWORD Store 32 bits of data from vector input registers into the scratch aperture. ENC_VSCRATCH default 26 VADDR FMT_ANY OPR_VGPR 32 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_B32 OPR_GPUMEM 32 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_STORE_B64 SCRATCH_STORE_DWORDX2 Store 64 bits of data from vector input registers into the scratch aperture. ENC_VSCRATCH default 27 VADDR FMT_ANY OPR_VGPR 32 VSRC FMT_ANY OPR_VGPR 64 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_B64 OPR_GPUMEM 64 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_STORE_B96 SCRATCH_STORE_DWORDX3 Store 96 bits of data from vector input registers into the scratch aperture. ENC_VSCRATCH default 28 VADDR FMT_ANY OPR_VGPR 32 VSRC FMT_ANY OPR_VGPR 96 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_B96 OPR_GPUMEM 96 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_STORE_B128 SCRATCH_STORE_DWORDX4 Store 128 bits of data from vector input registers into the scratch aperture. ENC_VSCRATCH default 29 VADDR FMT_ANY OPR_VGPR 32 VSRC FMT_ANY OPR_VGPR 128 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_B128 OPR_GPUMEM 128 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_LOAD_D16_U8 SCRATCH_LOAD_UBYTE_D16 Load 8 bits of unsigned data from the scratch aperture, zero extend to 16 bits and store the result into the low 16 bits of a 32-bit vector register. ENC_VSCRATCH default 30 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_U8 OPR_GPUMEM 8 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_LOAD_D16_I8 SCRATCH_LOAD_SBYTE_D16 Load 8 bits of signed data from the scratch aperture, sign extend to 16 bits and store the result into the low 16 bits of a 32-bit vector register. ENC_VSCRATCH default 31 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_I8 OPR_GPUMEM 8 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_LOAD_D16_B16 SCRATCH_LOAD_SHORT_D16 Load 16 bits of unsigned data from the scratch aperture and store the result into the low 16 bits of a 32-bit vector register. ENC_VSCRATCH default 32 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_B16 OPR_GPUMEM 16 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_LOAD_D16_HI_U8 SCRATCH_LOAD_UBYTE_D16_HI Load 8 bits of unsigned data from the scratch aperture, zero extend to 16 bits and store the result into the high 16 bits of a 32-bit vector register. ENC_VSCRATCH default 33 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_U8 OPR_GPUMEM 8 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_LOAD_D16_HI_I8 SCRATCH_LOAD_SBYTE_D16_HI Load 8 bits of signed data from the scratch aperture, sign extend to 16 bits and store the result into the high 16 bits of a 32-bit vector register. ENC_VSCRATCH default 34 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_I8 OPR_GPUMEM 8 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_LOAD_D16_HI_B16 SCRATCH_LOAD_SHORT_D16_HI Load 16 bits of unsigned data from the scratch aperture and store the result into the high 16 bits of a 32-bit vector register. ENC_VSCRATCH default 35 VDST FMT_ANY OPR_VGPR 32 VADDR FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_B16 OPR_GPUMEM 16 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_STORE_D16_HI_B8 SCRATCH_STORE_BYTE_D16_HI Store 8 bits of data from the high 16 bits of a 32-bit vector register into the scratch aperture. ENC_VSCRATCH default 36 VADDR FMT_ANY OPR_VGPR 32 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_B8 OPR_GPUMEM 8 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_STORE_D16_HI_B16 SCRATCH_STORE_SHORT_D16_HI Store 16 bits of data from the high 16 bits of a 32-bit vector register into the scratch aperture. ENC_VSCRATCH default 37 VADDR FMT_ANY OPR_VGPR 32 VSRC FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_B16 OPR_GPUMEM 16 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_LOAD_BLOCK Load a block of data from the scratch aperture. ENC_VSCRATCH default 83 VDST FMT_ANY OPR_VGPR 1024 VADDR FMT_ANY OPR_VGPR 32 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_B1024 OPR_GPUMEM 1024 FMT_NUM_B32 OPR_SDST_M0 32 VMEM TEXTURE FALSE FALSE FALSE FALSE FALSE SCRATCH_STORE_BLOCK Store a block of data to the scratch aperture. ENC_VSCRATCH default 84 VADDR FMT_ANY OPR_VGPR 32 VSRC FMT_ANY OPR_VGPR 1024 SADDR FMT_ANY OPR_SREG 32 FMT_NUM_B1024 OPR_GPUMEM 1024 FMT_NUM_B32 OPR_SDST_M0 32 VMEM TEXTURE FMT_ANY Value can be anything. bits 1 1 Bits 1 0 FMT_BUF 64-bit virtual memory address. descriptor 64 1 Descriptor 64 0 FMT_IMG 256-bit image resource constant. descriptor 256 1 Descriptor 256 0 FMT_IMG_BVH 128-bit BVH resource constant. descriptor 128 1 Descriptor 128 0 FMT_NUM_B1 1 bit of data. bits 1 1 Bits 1 0 FMT_NUM_B1024 1024 bits of arbitrary data. bits 1024 1 Bits 1024 0 FMT_NUM_B128 128 bits of arbitrary data. bits 128 1 Bits 128 0 FMT_NUM_B16 16 bits of arbitrary data. bits 16 1 Bits 16 0 FMT_NUM_B256 256 bits of arbitrary data. bits 256 1 Bits 256 0 FMT_NUM_B32 32 bits of arbitrary data. bits 32 1 Bits 32 0 FMT_NUM_B512 512 bits of arbitrary data. bits 512 1 Bits 512 0 FMT_NUM_B64 64 bits of arbitrary data. bits 64 1 Bits 64 0 FMT_NUM_B8 8 bits of arbitrary data. bits 8 1 Bits 8 0 FMT_NUM_B96 96 bits of arbitrary data. bits 96 1 Bits 96 0 FMT_NUM_BF16 16-bit brain floating point value. float 16 1 Mantissa 7 0 Exponent 8 7 Sign 1 15 FMT_NUM_BF8 float 8 1 Mantissa 2 0 Exponent 5 2 Sign 1 7 FMT_NUM_F16 16-bit half-precision floating point value. float 16 1 Mantissa 10 0 Exponent 5 10 Sign 1 15 FMT_NUM_F32 32-bit single-precision floating point value. float 32 1 Mantissa 23 0 Exponent 8 23 Sign 1 31 FMT_NUM_F64 64-bit double-precision floating point value. float 64 1 Mantissa 52 0 Exponent 11 52 Sign 1 63 FMT_NUM_FP8 float 8 1 Mantissa 3 0 Exponent 4 3 Sign 1 7 FMT_NUM_I16 Signed 16-bit integral data. integer 16 1 Integer 16 0 FMT_NUM_I24 Signed 24-bit integral data. integer 24 1 Integer 24 0 FMT_NUM_I32 Signed 32-bit integral data. integer 32 1 Integer 32 0 FMT_NUM_I4 Signed 4-bit integral data. integer 4 1 Integer 4 0 FMT_NUM_I64 Signed 64-bit integral data. integer 64 1 Integer 64 0 FMT_NUM_I8 Signed 8-bit integral data. integer 8 1 Integer 8 0 FMT_NUM_IU4 Signed or unsigned 4-bit integral data; signedness is determined by opcode modifier. integer 4 1 Integer 4 0 FMT_NUM_IU8 Signed or unsigned 8-bit integral data; signedness is determined by opcode modifier. integer 8 1 Integer 8 0 FMT_NUM_M64 64-bit mask value, read or written to a consecutive run of GPRs. The first GPR contains the least significant dword of the data. This format is generally used for lane masks and may be truncated to 32 bits when running in wave32 mode. Scalar operands with the M64 format do not have an alignment restriction when running in wave32 mode. integer 64 1 Integer 64 0 FMT_NUM_PK2_B16 16 bits of arbitrary data. Packed 2-component value. bits 32 2 Bits 16 0 Bits 16 16 FMT_NUM_PK2_B32 32 bits of arbitrary data. Packed 2-component value. bits 64 2 Bits 32 0 Bits 32 32 FMT_NUM_PK2_B64 64 bits of arbitrary data. Packed 2-component value. bits 128 2 Bits 64 0 Bits 64 64 FMT_NUM_PK2_BF16 16-bit brain floating point value. Packed 2-component value. float 32 2 Mantissa 7 0 Exponent 8 7 Sign 1 15 Mantissa 7 16 Exponent 8 23 Sign 1 31 FMT_NUM_PK2_BF8 Packed 2-component value. float 16 2 Mantissa 2 0 Exponent 5 2 Sign 1 7 Mantissa 2 8 Exponent 5 10 Sign 1 15 FMT_NUM_PK2_F16 16-bit half-precision floating point value. Packed 2-component value. float 32 2 Mantissa 10 0 Exponent 5 10 Sign 1 15 Mantissa 10 16 Exponent 5 26 Sign 1 31 FMT_NUM_PK2_F32 32-bit single-precision floating point value. Packed 2-component value. float 64 2 Mantissa 23 0 Exponent 8 23 Sign 1 31 Mantissa 23 32 Exponent 8 55 Sign 1 63 FMT_NUM_PK2_FP8 Packed 2-component value. float 16 2 Mantissa 3 0 Exponent 4 3 Sign 1 7 Mantissa 3 8 Exponent 4 11 Sign 1 15 FMT_NUM_PK2_I16 Signed 16-bit integral data. Packed 2-component value. integer 32 2 Integer 16 0 Integer 16 16 FMT_NUM_PK2_I8 Signed 8-bit integral data. Packed 2-component value. integer 16 2 Integer 8 0 Integer 8 8 FMT_NUM_PK2_U16 Unsigned 16-bit integral data. Packed 2-component value. integer 32 2 Integer 16 0 Integer 16 16 FMT_NUM_PK2_U8 Unsigned 8-bit integral data. Packed 2-component value. integer 16 2 Integer 8 0 Integer 8 8 FMT_NUM_PK4_B8 8 bits of arbitrary data. Packed 4-component value. bits 32 4 Bits 8 0 Bits 8 8 Bits 8 16 Bits 8 24 FMT_NUM_PK4_BF8 Packed 4-component value. float 32 4 Mantissa 2 0 Exponent 5 2 Sign 1 7 Mantissa 2 8 Exponent 5 10 Sign 1 15 Mantissa 2 16 Exponent 5 18 Sign 1 23 Mantissa 2 24 Exponent 5 26 Sign 1 31 FMT_NUM_PK4_FP8 Packed 4-component value. float 32 4 Mantissa 3 0 Exponent 4 3 Sign 1 7 Mantissa 3 8 Exponent 4 11 Sign 1 15 Mantissa 3 16 Exponent 4 19 Sign 1 23 Mantissa 3 24 Exponent 4 27 Sign 1 31 FMT_NUM_PK4_I8 Signed 8-bit integral data. Packed 4-component value. integer 32 4 Integer 8 0 Integer 8 8 Integer 8 16 Integer 8 24 FMT_NUM_PK4_IU8 Signed or unsigned 8-bit integral data; signedness is determined by opcode modifier. Packed 4-component value. integer 32 4 Integer 8 0 Integer 8 8 Integer 8 16 Integer 8 24 FMT_NUM_PK4_U8 Unsigned 8-bit integral data. Packed 4-component value. integer 32 4 Integer 8 0 Integer 8 8 Integer 8 16 Integer 8 24 FMT_NUM_PK8_I4 Signed 4-bit integral data. Packed 8-component value. integer 32 8 Integer 4 0 Integer 4 4 Integer 4 8 Integer 4 12 Integer 4 16 Integer 4 20 Integer 4 24 Integer 4 28 FMT_NUM_PK8_IU4 Signed or unsigned 4-bit integral data; signedness is determined by opcode modifier. Packed 8-component value. integer 32 8 Integer 4 0 Integer 4 4 Integer 4 8 Integer 4 12 Integer 4 16 Integer 4 20 Integer 4 24 Integer 4 28 FMT_NUM_PK8_U4 Unsigned 4-bit integral data. Packed 8-component value. integer 32 8 Integer 4 0 Integer 4 4 Integer 4 8 Integer 4 12 Integer 4 16 Integer 4 20 Integer 4 24 Integer 4 28 FMT_NUM_U16 Unsigned 16-bit integral data. integer 16 1 Integer 16 0 FMT_NUM_U24 Unsigned 24-bit integral data. integer 24 1 Integer 24 0 FMT_NUM_U32 Unsigned 32-bit integral data. integer 32 1 Integer 32 0 FMT_NUM_U4 Unsigned 4-bit integral data. integer 4 1 Integer 4 0 FMT_NUM_U64 Unsigned 64-bit integral data. integer 64 1 Integer 64 0 FMT_NUM_U8 Unsigned 8-bit integral data. integer 8 1 Integer 8 0 FMT_RSRC 128-bit buffer resource constant. descriptor 128 1 Descriptor 128 0 FMT_RSRC_SCALAR 128-bit buffer resource constant. descriptor 128 1 Descriptor 128 0 FMT_RSRC_SCRATCH 128-bit buffer resource constant. descriptor 128 1 Descriptor 128 0 FMT_RSRC_SCRATCH_BYTE 128-bit buffer resource constant. descriptor 128 1 Descriptor 128 0 FMT_RSRC_SCRATCH_STRIDE 128-bit buffer resource constant. descriptor 128 1 Descriptor 128 0 FMT_RSRC_TYPED 128-bit buffer resource constant. descriptor 128 1 Descriptor 128 0 FMT_RSRC_TYPED_BYTE 128-bit buffer resource constant. descriptor 128 1 Descriptor 128 0 FMT_RSRC_TYPED_STRIDE 128-bit buffer resource constant. descriptor 128 1 Descriptor 128 0 FMT_RSRC_VECTOR 128-bit buffer resource constant. descriptor 128 1 Descriptor 128 0 FMT_RSRC_VECTOR_BYTE 128-bit buffer resource constant. descriptor 128 1 Descriptor 128 0 FMT_RSRC_VECTOR_STRIDE 128-bit buffer resource constant. descriptor 128 1 Descriptor 128 0 FMT_SAMP 128-bit image sampler constant. descriptor 128 1 Descriptor 128 0 FMT_WMMA_AB_16X16_BF16 WMMA 16x16 multiplicand matrix (A or B) with 16-bit brain float values. WMMA AB matrices are 16x16 (or larger) and the individual elements are packed; each VGPR lane can contain 2 16-bit elements, 4 8-bit elements or 8 4-bit elements. float 128 8 Mantissa 7 0 Exponent 8 7 Sign 1 15 Mantissa 7 16 Exponent 8 23 Sign 1 31 Mantissa 7 32 Exponent 8 39 Sign 1 47 Mantissa 7 48 Exponent 8 55 Sign 1 63 Mantissa 7 64 Exponent 8 71 Sign 1 79 Mantissa 7 80 Exponent 8 87 Sign 1 95 Mantissa 7 96 Exponent 8 103 Sign 1 111 Mantissa 7 112 Exponent 8 119 Sign 1 127 FMT_WMMA_AB_16X16_BF8 WMMA 16x16 multiplicand matrix (A or B) with 8-bit brain floats. WMMA AB matrices are 16x16 (or larger) and the individual elements are packed; each VGPR lane can contain 2 16-bit elements, 4 8-bit elements or 8 4-bit elements. float 64 8 Mantissa 2 0 Exponent 5 2 Sign 1 7 Mantissa 2 8 Exponent 5 10 Sign 1 15 Mantissa 2 16 Exponent 5 18 Sign 1 23 Mantissa 2 24 Exponent 5 26 Sign 1 31 Mantissa 2 32 Exponent 5 34 Sign 1 39 Mantissa 2 40 Exponent 5 42 Sign 1 47 Mantissa 2 48 Exponent 5 50 Sign 1 55 Mantissa 2 56 Exponent 5 58 Sign 1 63 FMT_WMMA_AB_16X16_F16 WMMA 16x16 multiplicand matrix (A or B) with 16-bit float values. WMMA AB matrices are 16x16 (or larger) and the individual elements are packed; each VGPR lane can contain 2 16-bit elements, 4 8-bit elements or 8 4-bit elements. float 128 8 Mantissa 10 0 Exponent 5 10 Sign 1 15 Mantissa 10 16 Exponent 5 26 Sign 1 31 Mantissa 10 32 Exponent 5 42 Sign 1 47 Mantissa 10 48 Exponent 5 58 Sign 1 63 Mantissa 10 64 Exponent 5 74 Sign 1 79 Mantissa 10 80 Exponent 5 90 Sign 1 95 Mantissa 10 96 Exponent 5 106 Sign 1 111 Mantissa 10 112 Exponent 5 122 Sign 1 127 FMT_WMMA_AB_16X16_FP8 WMMA 16x16 multiplicand matrix (A or B) with 8-bit floats. WMMA AB matrices are 16x16 (or larger) and the individual elements are packed; each VGPR lane can contain 2 16-bit elements, 4 8-bit elements or 8 4-bit elements. float 64 8 Mantissa 3 0 Exponent 4 3 Sign 1 7 Mantissa 3 8 Exponent 4 11 Sign 1 15 Mantissa 3 16 Exponent 4 19 Sign 1 23 Mantissa 3 24 Exponent 4 27 Sign 1 31 Mantissa 3 32 Exponent 4 35 Sign 1 39 Mantissa 3 40 Exponent 4 43 Sign 1 47 Mantissa 3 48 Exponent 4 51 Sign 1 55 Mantissa 3 56 Exponent 4 59 Sign 1 63 FMT_WMMA_AB_16X16_IU4 WMMA 16x16 multiplicand matrix (A or B) with 4-bit integers whose signedness is determined by the NEG modifier. In WAVE64 mode lanes 32+ are ignored. WMMA AB matrices are 16x16 (or larger) and the individual elements are packed; each VGPR lane can contain 2 16-bit elements, 4 8-bit elements or 8 4-bit elements. integer 32 8 Integer 4 0 Integer 4 4 Integer 4 8 Integer 4 12 Integer 4 16 Integer 4 20 Integer 4 24 Integer 4 28 FMT_WMMA_AB_16X16_IU8 WMMA 16x16 multiplicand matrix (A or B) with 8-bit integers whose signedness is determined by the NEG modifier. WMMA AB matrices are 16x16 (or larger) and the individual elements are packed; each VGPR lane can contain 2 16-bit elements, 4 8-bit elements or 8 4-bit elements. integer 64 8 Integer 8 0 Integer 8 8 Integer 8 16 Integer 8 24 Integer 8 32 Integer 8 40 Integer 8 48 Integer 8 56 FMT_WMMA_AB_16X32_BF16 WMMA 16x32 multiplicand matrix (A or B) with 16-bit brain float values. WMMA AB matrices are 16x16 (or larger) and the individual elements are packed; each VGPR lane can contain 2 16-bit elements, 4 8-bit elements or 8 4-bit elements. float 256 16 Mantissa 7 0 Exponent 8 7 Sign 1 15 Mantissa 7 16 Exponent 8 23 Sign 1 31 Mantissa 7 32 Exponent 8 39 Sign 1 47 Mantissa 7 48 Exponent 8 55 Sign 1 63 Mantissa 7 64 Exponent 8 71 Sign 1 79 Mantissa 7 80 Exponent 8 87 Sign 1 95 Mantissa 7 96 Exponent 8 103 Sign 1 111 Mantissa 7 112 Exponent 8 119 Sign 1 127 Mantissa 7 128 Exponent 8 135 Sign 1 143 Mantissa 7 144 Exponent 8 151 Sign 1 159 Mantissa 7 160 Exponent 8 167 Sign 1 175 Mantissa 7 176 Exponent 8 183 Sign 1 191 Mantissa 7 192 Exponent 8 199 Sign 1 207 Mantissa 7 208 Exponent 8 215 Sign 1 223 Mantissa 7 224 Exponent 8 231 Sign 1 239 Mantissa 7 240 Exponent 8 247 Sign 1 255 FMT_WMMA_AB_16X32_BF8 WMMA 16x32 multiplicand matrix (A or B) with 8-bit brain floats. WMMA AB matrices are 16x16 (or larger) and the individual elements are packed; each VGPR lane can contain 2 16-bit elements, 4 8-bit elements or 8 4-bit elements. float 128 16 Mantissa 2 0 Exponent 5 2 Sign 1 7 Mantissa 2 8 Exponent 5 10 Sign 1 15 Mantissa 2 16 Exponent 5 18 Sign 1 23 Mantissa 2 24 Exponent 5 26 Sign 1 31 Mantissa 2 32 Exponent 5 34 Sign 1 39 Mantissa 2 40 Exponent 5 42 Sign 1 47 Mantissa 2 48 Exponent 5 50 Sign 1 55 Mantissa 2 56 Exponent 5 58 Sign 1 63 Mantissa 2 64 Exponent 5 66 Sign 1 71 Mantissa 2 72 Exponent 5 74 Sign 1 79 Mantissa 2 80 Exponent 5 82 Sign 1 87 Mantissa 2 88 Exponent 5 90 Sign 1 95 Mantissa 2 96 Exponent 5 98 Sign 1 103 Mantissa 2 104 Exponent 5 106 Sign 1 111 Mantissa 2 112 Exponent 5 114 Sign 1 119 Mantissa 2 120 Exponent 5 122 Sign 1 127 FMT_WMMA_AB_16X32_F16 WMMA 16x32 multiplicand matrix (A or B) with 16-bit float values. WMMA AB matrices are 16x16 (or larger) and the individual elements are packed; each VGPR lane can contain 2 16-bit elements, 4 8-bit elements or 8 4-bit elements. float 256 16 Mantissa 10 0 Exponent 5 10 Sign 1 15 Mantissa 10 16 Exponent 5 26 Sign 1 31 Mantissa 10 32 Exponent 5 42 Sign 1 47 Mantissa 10 48 Exponent 5 58 Sign 1 63 Mantissa 10 64 Exponent 5 74 Sign 1 79 Mantissa 10 80 Exponent 5 90 Sign 1 95 Mantissa 10 96 Exponent 5 106 Sign 1 111 Mantissa 10 112 Exponent 5 122 Sign 1 127 Mantissa 10 128 Exponent 5 138 Sign 1 143 Mantissa 10 144 Exponent 5 154 Sign 1 159 Mantissa 10 160 Exponent 5 170 Sign 1 175 Mantissa 10 176 Exponent 5 186 Sign 1 191 Mantissa 10 192 Exponent 5 202 Sign 1 207 Mantissa 10 208 Exponent 5 218 Sign 1 223 Mantissa 10 224 Exponent 5 234 Sign 1 239 Mantissa 10 240 Exponent 5 250 Sign 1 255 FMT_WMMA_AB_16X32_FP8 WMMA 16x32 multiplicand matrix (A or B) with 8-bit floats. WMMA AB matrices are 16x16 (or larger) and the individual elements are packed; each VGPR lane can contain 2 16-bit elements, 4 8-bit elements or 8 4-bit elements. float 128 16 Mantissa 3 0 Exponent 4 3 Sign 1 7 Mantissa 3 8 Exponent 4 11 Sign 1 15 Mantissa 3 16 Exponent 4 19 Sign 1 23 Mantissa 3 24 Exponent 4 27 Sign 1 31 Mantissa 3 32 Exponent 4 35 Sign 1 39 Mantissa 3 40 Exponent 4 43 Sign 1 47 Mantissa 3 48 Exponent 4 51 Sign 1 55 Mantissa 3 56 Exponent 4 59 Sign 1 63 Mantissa 3 64 Exponent 4 67 Sign 1 71 Mantissa 3 72 Exponent 4 75 Sign 1 79 Mantissa 3 80 Exponent 4 83 Sign 1 87 Mantissa 3 88 Exponent 4 91 Sign 1 95 Mantissa 3 96 Exponent 4 99 Sign 1 103 Mantissa 3 104 Exponent 4 107 Sign 1 111 Mantissa 3 112 Exponent 4 115 Sign 1 119 Mantissa 3 120 Exponent 4 123 Sign 1 127 FMT_WMMA_AB_16X32_IU4 WMMA 16x32 multiplicand matrix (A or B) with 4-bit integers whose signedness is determined by the NEG modifier. WMMA AB matrices are 16x16 (or larger) and the individual elements are packed; each VGPR lane can contain 2 16-bit elements, 4 8-bit elements or 8 4-bit elements. integer 64 16 Integer 4 0 Integer 4 4 Integer 4 8 Integer 4 12 Integer 4 16 Integer 4 20 Integer 4 24 Integer 4 28 Integer 4 32 Integer 4 36 Integer 4 40 Integer 4 44 Integer 4 48 Integer 4 52 Integer 4 56 Integer 4 60 FMT_WMMA_AB_16X32_IU8 WMMA 16x32 multiplicand matrix (A or B) with 8-bit integers whose signedness is determined by the NEG modifier. WMMA AB matrices are 16x16 (or larger) and the individual elements are packed; each VGPR lane can contain 2 16-bit elements, 4 8-bit elements or 8 4-bit elements. integer 128 16 Integer 8 0 Integer 8 8 Integer 8 16 Integer 8 24 Integer 8 32 Integer 8 40 Integer 8 48 Integer 8 56 Integer 8 64 Integer 8 72 Integer 8 80 Integer 8 88 Integer 8 96 Integer 8 104 Integer 8 112 Integer 8 120 FMT_WMMA_AB_16X64_IU4 WMMA 16x64 multiplicand matrix (A or B) with 4-bit integers whose signedness is determined by the NEG modifier. WMMA AB matrices are 16x16 (or larger) and the individual elements are packed; each VGPR lane can contain 2 16-bit elements, 4 8-bit elements or 8 4-bit elements. integer 128 32 Integer 4 0 Integer 4 4 Integer 4 8 Integer 4 12 Integer 4 16 Integer 4 20 Integer 4 24 Integer 4 28 Integer 4 32 Integer 4 36 Integer 4 40 Integer 4 44 Integer 4 48 Integer 4 52 Integer 4 56 Integer 4 60 Integer 4 64 Integer 4 68 Integer 4 72 Integer 4 76 Integer 4 80 Integer 4 84 Integer 4 88 Integer 4 92 Integer 4 96 Integer 4 100 Integer 4 104 Integer 4 108 Integer 4 112 Integer 4 116 Integer 4 120 Integer 4 124 FMT_WMMA_DC_16X16_BF16 WMMA destination (D) or accumulator (C) 16x16 matrix with 16-bit brain float values. WMMA DC matrices are 16x16 (or larger) and the individual elements are packed. The total number of VGPRs required depends on the wave size. float 128 8 Mantissa 7 0 Exponent 8 7 Sign 1 15 Mantissa 7 16 Exponent 8 23 Sign 1 31 Mantissa 7 32 Exponent 8 39 Sign 1 47 Mantissa 7 48 Exponent 8 55 Sign 1 63 Mantissa 7 64 Exponent 8 71 Sign 1 79 Mantissa 7 80 Exponent 8 87 Sign 1 95 Mantissa 7 96 Exponent 8 103 Sign 1 111 Mantissa 7 112 Exponent 8 119 Sign 1 127 FMT_WMMA_DC_16X16_F16 WMMA destination (D) or accumulator (C) 16x16 matrix with 16-bit float values. WMMA DC matrices are 16x16 (or larger) and the individual elements are packed. The total number of VGPRs required depends on the wave size. float 128 8 Mantissa 10 0 Exponent 5 10 Sign 1 15 Mantissa 10 16 Exponent 5 26 Sign 1 31 Mantissa 10 32 Exponent 5 42 Sign 1 47 Mantissa 10 48 Exponent 5 58 Sign 1 63 Mantissa 10 64 Exponent 5 74 Sign 1 79 Mantissa 10 80 Exponent 5 90 Sign 1 95 Mantissa 10 96 Exponent 5 106 Sign 1 111 Mantissa 10 112 Exponent 5 122 Sign 1 127 FMT_WMMA_DC_16X16_F32 WMMA destination (D) or accumulator (C) 16x16 matrix with single precision float values. WMMA DC matrices are 16x16 (or larger) and the individual elements are packed. The total number of VGPRs required depends on the wave size. float 256 8 Mantissa 23 0 Exponent 8 23 Sign 1 31 Mantissa 23 32 Exponent 8 55 Sign 1 63 Mantissa 23 64 Exponent 8 87 Sign 1 95 Mantissa 23 96 Exponent 8 119 Sign 1 127 Mantissa 23 128 Exponent 8 151 Sign 1 159 Mantissa 23 160 Exponent 8 183 Sign 1 191 Mantissa 23 192 Exponent 8 215 Sign 1 223 Mantissa 23 224 Exponent 8 247 Sign 1 255 FMT_WMMA_DC_16X16_I32 WMMA destination (D) or accumulator (C) 16x16 matrix with 32-bit signed integer values. WMMA DC matrices are 16x16 (or larger) and the individual elements are packed. The total number of VGPRs required depends on the wave size. integer 256 8 Integer 32 0 Integer 32 32 Integer 32 64 Integer 32 96 Integer 32 128 Integer 32 160 Integer 32 192 Integer 32 224 FMT_WMMA_INDEX_SET Index set for SWMMAC instructions (see OPF_SWMMAC). Consists of 2-bit indices with the constraint that for each pair, index0 < index1. bits 32 16 Bits 2 0 Bits 2 2 Bits 2 4 Bits 2 6 Bits 2 8 Bits 2 10 Bits 2 12 Bits 2 14 Bits 2 16 Bits 2 18 Bits 2 20 Bits 2 22 Bits 2 24 Bits 2 26 Bits 2 28 Bits 2 30 OPR_ATTR Attribute parameter for interpolation. attr0 First interpolation attribute. Increment from here for additional attributes. There are SQ_NUM_ATTR attributes in total. 0 attr1 <p>See above.</p> 1 attr2 <p>See above.</p> 2 attr3 <p>See above.</p> 3 attr4 <p>See above.</p> 4 attr5 <p>See above.</p> 5 attr6 <p>See above.</p> 6 attr7 <p>See above.</p> 7 attr8 <p>See above.</p> 8 attr9 <p>See above.</p> 9 attr10 <p>See above.</p> 10 attr11 <p>See above.</p> 11 attr12 <p>See above.</p> 12 attr13 <p>See above.</p> 13 attr14 <p>See above.</p> 14 attr15 <p>See above.</p> 15 attr16 <p>See above.</p> 16 attr17 <p>See above.</p> 17 attr18 <p>See above.</p> 18 attr19 <p>See above.</p> 19 attr20 <p>See above.</p> 20 attr21 <p>See above.</p> 21 attr22 <p>See above.</p> 22 attr23 <p>See above.</p> 23 attr24 <p>See above.</p> 24 attr25 <p>See above.</p> 25 attr26 <p>See above.</p> 26 attr27 <p>See above.</p> 27 attr28 <p>See above.</p> 28 attr29 <p>See above.</p> 29 attr30 <p>See above.</p> 30 attr31 <p>See above.</p> 31 attr32 <p>See above.</p> 32 OPR_CLAUSE Description of a clause following this instruction. BREAK_SPAN This field is set to the number of instructions to issue before each clause break. If set to zero then there are no clause breaks. If set to nonzero value then the maximum number of instructions between clause breaks is 15. 4 8 LENGTH This field is set to the logical number of instructions in the clause, minus 1 (e.g. if a clause has 4 instructions, program this field to 3). The minimum number of instructions required for a clause is 2 and the maximum number of instructions is 63, therefore this field must be programmed in the range [1, 62] inclusive. 6 0 OPR_DELAY Specify one or more delay conditions to apply to the next VALU instructions in the instruction stream. INSTID0 Hazard to delay for with the next VALU instruction. 4 0 instid_no_dep No dependency on any prior instruction. 0 instid_valu_dep_1 Dependent on previous VALU instruction, 1 instruction(s) back. 1 instid_valu_dep_2 Dependent on previous VALU instruction, 2 instruction(s) back. 2 instid_valu_dep_3 Dependent on previous VALU instruction, 3 instruction(s) back. 3 instid_valu_dep_4 Dependent on previous VALU instruction, 4 instruction(s) back. 4 instid_trans32_dep_1 Dependent on previous TRANS32 instruction, 1 instruction(s) back. 5 instid_trans32_dep_2 Dependent on previous TRANS32 instruction, 2 instruction(s) back. 6 instid_trans32_dep_3 Dependent on previous TRANS32 instruction, 3 instruction(s) back. 7 instid_fma_accum_cycle_1 Single cycle penalty for FMA accumulation (reserved). 8 instid_salu_cycle_1 1 cycle penalty for a prior SALU instruction. 9 instid_salu_cycle_2 2 cycle penalty for a prior SALU instruction. 10 instid_salu_cycle_3 3 cycle penalty for a prior SALU instruction. 11 INSTID1 Hazard to delay for with the VALU instruction identified by INSTSKIP. 4 7 instid_no_dep No dependency on any prior instruction. 0 instid_valu_dep_1 Dependent on previous VALU instruction, 1 instruction(s) back. 1 instid_valu_dep_2 Dependent on previous VALU instruction, 2 instruction(s) back. 2 instid_valu_dep_3 Dependent on previous VALU instruction, 3 instruction(s) back. 3 instid_valu_dep_4 Dependent on previous VALU instruction, 4 instruction(s) back. 4 instid_trans32_dep_1 Dependent on previous TRANS32 instruction, 1 instruction(s) back. 5 instid_trans32_dep_2 Dependent on previous TRANS32 instruction, 2 instruction(s) back. 6 instid_trans32_dep_3 Dependent on previous TRANS32 instruction, 3 instruction(s) back. 7 instid_fma_accum_cycle_1 Single cycle penalty for FMA accumulation (reserved). 8 instid_salu_cycle_1 1 cycle penalty for a prior SALU instruction. 9 instid_salu_cycle_2 2 cycle penalty for a prior SALU instruction. 10 instid_salu_cycle_3 3 cycle penalty for a prior SALU instruction. 11 INSTSKIP Identify the VALU instruction that the second delay condition applies to. 3 4 instskip_same Apply second dependency to same instruction (2 dependencies on one instruction). 0 instskip_next Apply second dependency to next instruction (no skip). 1 instskip_skip_1 Skip 1 instruction(s) then apply dependency. 2 instskip_skip_2 Skip 2 instruction(s) then apply dependency. 3 instskip_skip_3 Skip 3 instruction(s) then apply dependency. 4 instskip_skip_4 Skip 4 instruction(s) then apply dependency. 5 OPR_DSMEM Describes an operation that reads or writes DS memory. dsmem I/O to DS memory 0 OPR_EXEC 64-bit vector execute mask. exec EXEC[63:0] 126 OPR_GPUMEM Describes an operation that reads or writes GPU memory. gpumem I/O to GPU memory 0 OPR_HWREG Reference to a hardware register that is manipulated by getreg or setreg. This operand encodes the ID of the register to access and allows access to a subset of the full register by specifying an offset and size in bits. ID ID of hardware register to access. 6 0 hw_reg_wave_mode Wave mode. Readable and writable by all. 1 hw_reg_wave_status Wave Status. Read-only by all. 2 hw_reg_wave_state_priv Wave status. Privileged state that may need to be modified during context switch. Readable by all, writable only when PRIV=1. 4 hw_reg_wave_gpr_alloc Wave GPR allocation. 5 hw_reg_wave_lds_alloc Wave LDS allocation. 6 hw_reg_perf_snapshot_data SQ state on a performance snapshot capture. 10 hw_reg_perf_snapshot_pc_lo 32 LSBs of PC of wave when snapshot is taken. 11 hw_reg_perf_snapshot_pc_hi MSBs of PC when wave was snapshot. 12 hw_reg_perf_snapshot_data1 SQ performance counter snapshot data1 15 hw_reg_perf_snapshot_data2 SQ performance counter snapshot data2: Holds Memory counter state 16 hw_reg_wave_excp_flag_priv Wave exception flags. Readable by all, writable only when PRIV=1. 17 hw_reg_wave_excp_flag_user Wave exception flags. Readable and writable by all. 18 hw_reg_wave_trap_ctrl Wave trap control. Readable by all, writable only when PRIV=1. 19 hw_reg_wave_scratch_base_lo Wave base address of scratch memory address (lsbs). 20 hw_reg_wave_scratch_base_hi Wave base address of scratch memory address (msbs). 21 hw_reg_wave_hw_id1 Wave Hardware ID1. Read-only. 23 hw_reg_wave_hw_id2 Wave Hardware ID2. Read-only. 24 hw_reg_shader_cycles_lo Low bits of shader cycles counter. 60 bits. One instance per SQ, automatically counts once per cycle when SQ is busy, automatically wraps around when it reaches its maximum value (expected to occur approximately once every 9 years at 4GHz). Readable and writable by the host. When accessed by the host, SHADER_CYCLES_LO must be accessed first, followed by SHADER_CYCLES_HI. The counter is frozen on the SHADER_CYCLES_LO access and unfrozen on the SHADER_CYCLES_HI access to prevent wraparound. Read-only within the shader via S_GETREG_B32. When read, it is recommended to read HI, then LO, then HI again; if the two HI reads differ then repeat the process. This will detect any possible wraparound condition. 29 hw_reg_shader_cycles_hi High bits of shader cycles counter. 60 bits. One instance per SQ, automatically counts once per cycle when SQ is busy, automatically wraps around when it reaches its maximum value (expected to occur approximately once every 9 years at 4GHz). See SHADER_CYCLES_LO for restrictions on reading and writing this value. 30 OFFSET LSB offset of register bits to access. 5 6 SIZE Size of register bits to access, minus 1. Set this field to 31 to read/write all bits of the hardware register. 5 11 OPR_LABEL A relative jump offset. TARGET Singed jump target relative to PC+4, in dwords. 16 0 OPR_PC Program counter implied operand. pc pc[63:0] 0 OPR_SDST All scalar destination operands. OPR_SDST_EXEC OPR_SDST_M0 OPR_SREG s0 Scalar GPR 0. Increment from here for additional GPRs. There are NUM_SGPR SGPRs in total. Alignment restriction: unless otherwise noted, 64-bit SGPR values must be even-aligned (first SGPR address is even); 128-bit and wider SGPR values must be aligned to a multiple of 4. 0 s1 <p>See above.</p> 1 s2 <p>See above.</p> 2 s3 <p>See above.</p> 3 s4 <p>See above.</p> 4 s5 <p>See above.</p> 5 s6 <p>See above.</p> 6 s7 <p>See above.</p> 7 s8 <p>See above.</p> 8 s9 <p>See above.</p> 9 s10 <p>See above.</p> 10 s11 <p>See above.</p> 11 s12 <p>See above.</p> 12 s13 <p>See above.</p> 13 s14 <p>See above.</p> 14 s15 <p>See above.</p> 15 s16 <p>See above.</p> 16 s17 <p>See above.</p> 17 s18 <p>See above.</p> 18 s19 <p>See above.</p> 19 s20 <p>See above.</p> 20 s21 <p>See above.</p> 21 s22 <p>See above.</p> 22 s23 <p>See above.</p> 23 s24 <p>See above.</p> 24 s25 <p>See above.</p> 25 s26 <p>See above.</p> 26 s27 <p>See above.</p> 27 s28 <p>See above.</p> 28 s29 <p>See above.</p> 29 s30 <p>See above.</p> 30 s31 <p>See above.</p> 31 s32 <p>See above.</p> 32 s33 <p>See above.</p> 33 s34 <p>See above.</p> 34 s35 <p>See above.</p> 35 s36 <p>See above.</p> 36 s37 <p>See above.</p> 37 s38 <p>See above.</p> 38 s39 <p>See above.</p> 39 s40 <p>See above.</p> 40 s41 <p>See above.</p> 41 s42 <p>See above.</p> 42 s43 <p>See above.</p> 43 s44 <p>See above.</p> 44 s45 <p>See above.</p> 45 s46 <p>See above.</p> 46 s47 <p>See above.</p> 47 s48 <p>See above.</p> 48 s49 <p>See above.</p> 49 s50 <p>See above.</p> 50 s51 <p>See above.</p> 51 s52 <p>See above.</p> 52 s53 <p>See above.</p> 53 s54 <p>See above.</p> 54 s55 <p>See above.</p> 55 s56 <p>See above.</p> 56 s57 <p>See above.</p> 57 s58 <p>See above.</p> 58 s59 <p>See above.</p> 59 s60 <p>See above.</p> 60 s61 <p>See above.</p> 61 s62 <p>See above.</p> 62 s63 <p>See above.</p> 63 s64 <p>See above.</p> 64 s65 <p>See above.</p> 65 s66 <p>See above.</p> 66 s67 <p>See above.</p> 67 s68 <p>See above.</p> 68 s69 <p>See above.</p> 69 s70 <p>See above.</p> 70 s71 <p>See above.</p> 71 s72 <p>See above.</p> 72 s73 <p>See above.</p> 73 s74 <p>See above.</p> 74 s75 <p>See above.</p> 75 s76 <p>See above.</p> 76 s77 <p>See above.</p> 77 s78 <p>See above.</p> 78 s79 <p>See above.</p> 79 s80 <p>See above.</p> 80 s81 <p>See above.</p> 81 s82 <p>See above.</p> 82 s83 <p>See above.</p> 83 s84 <p>See above.</p> 84 s85 <p>See above.</p> 85 s86 <p>See above.</p> 86 s87 <p>See above.</p> 87 s88 <p>See above.</p> 88 s89 <p>See above.</p> 89 s90 <p>See above.</p> 90 s91 <p>See above.</p> 91 s92 <p>See above.</p> 92 s93 <p>See above.</p> 93 s94 <p>See above.</p> 94 s95 <p>See above.</p> 95 s96 <p>See above.</p> 96 s97 <p>See above.</p> 97 s98 <p>See above.</p> 98 s99 <p>See above.</p> 99 s100 <p>See above.</p> 100 s101 <p>See above.</p> 101 s102 <p>See above.</p> 102 s103 <p>See above.</p> 103 s104 <p>See above.</p> 104 s105 <p>See above.</p> 105 ttmp0 Trap handler temps (privileged). Increment from here for additional TTMPs. There are NUM_TTMP TTMPs in total. {TTMP1,TTMP0} = PC_save{hi,lo}. Alignment restriction: unless otherwise noted, 64-bit SGPR values must be even-aligned (first SGPR address is even); 128-bit and wider SGPR values must be aligned to a multiple of 4. 108 ttmp1 Trap handler temps (privileged). 109 ttmp2 Trap handler temps (privileged). 110 ttmp3 Trap handler temps (privileged). 111 ttmp4 Trap handler temps (privileged). 112 ttmp5 Trap handler temps (privileged). 113 ttmp6 Trap handler temps (privileged). 114 ttmp7 Trap handler temps (privileged). 115 ttmp8 Trap handler temps (privileged). 116 ttmp9 Trap handler temps (privileged). 117 ttmp10 Trap handler temps (privileged). 118 ttmp11 Trap handler temps (privileged). 119 ttmp12 Trap handler temps (privileged). 120 ttmp13 Trap handler temps (privileged). 121 ttmp14 Trap handler temps (privileged). 122 ttmp15 Trap handler temps (privileged). 123 vcc_lo VCC[31:0] 106 vcc_hi VCC[63:32] 107 null Do not write destination. For scalar memory, do not use SOFFSET in address calculation. For scratch addressing, do not use an SGPR in the addressing calculation. Otherwise read back zero. 124 m0 Special register used to hold DS addresses, relative indices, and send-message values. 125 exec_lo exec[31:0] 126 exec_hi exec[63:32] 127 OPR_SDST_EXEC The execute mask. exec_lo exec[31:0] 126 exec_hi exec[63:32] 127 OPR_SDST_M0 Internal hardware register that is used to pass additional control information into certain instructions. m0 Special register used to hold DS addresses, relative indices, and send-message values. 125 OPR_SENDMSG Send a message back to SPI. HAS_RTN Set to false. 1 7 MSG Message type. 7 0 msg_interrupt Send a software-generated interrupt to the host. M0[23:0] carries user data. IDs are also sent to the host as is the case with all interrupts. 1 msg_hs_tessfactor Indicates HS tessellation factor is all zero or one for all patches in this HS workgroup. Data from M0[0]: 1 = "all are zero or one". This message is optional; if sent it should only be sent once and should not be sent by a non-HS shader stage. 2 msg_dealloc_vgprs Deallocate all VGPRs for this wave, allowing another wave to allocate these VGPRs before this wave ends. Use only when next instruction is S_ENDPGM. Typically used when a shader is waiting memory-write-acknowledgments before ending. Hardware waits for all previous instructions to complete except stores (all dependency and instructions are zero except VSCNT before this is issued). After issuing this the wave has no VGPRs or scratch memory. Any previous writes to scratch memory must have completed before issuing this deallocation. Wave do not respond to context-save after deallocating VGPRs (since it can no longer write to memory). Also, all vector instructions are ignored. This message is not compatible with Dynamic VGPRs mode and its use is illegal in any wavefront where Dynamic VGPRs are enabled (whether or not S_ALLOC_VGPR has been executed). 3 msg_gs_alloc_req Request GS space in the parameter cache. M0[9:0] = number of vertices, M0[22:12] = number of primitives. In response, a GS-allocation response is broadcast to the workgroup for non-zero requests. 9 SYSTEM Must be set to zero in all shader programs. 1 8 OPR_SENDMSG_RTN Send a message back to SPI, with a return value expected. HAS_RTN Set to true. 1 7 MSG Message type. 7 0 msg_rtn_get_doorbell Get the doorbell ID associated with this wave. Does not exist for ME0 -- return 0x0bad in this case. Also returns 0x0bad for invalid pipeID or queueID. 128 msg_rtn_get_ddid Get the draw ID or dispatch ID associated with this wave. 129 msg_rtn_get_tma Get the Trap Memory Address: [31:0] or [63:0] depending on the request size. 130 msg_rtn_get_realtime Get the value of the constant frequency (REFCLK) time counter: [31:0] or [63:0] depending on the request size. 131 msg_rtn_save_wave Used in context switching in indicate this wave is ready to be context saved. Only the trap handler can send this message (user shaders have this converted to MSG_RTN_ILLEGAL_MSG). 132 msg_rtn_get_tba Gets the Trap Base Address: [31:0] or [63:0] depending on the request size. 133 msg_rtn_get_tba_to_pc Internal hardware use only. Do not use in shader programs. 134 msg_rtn_get_se_hw_id Get the SE_ID and AID_ID. 135 msg_rtn_illegal_msg Illegal message with data return to wave. 255 SYSTEM Must be set to zero in all shader programs. 1 8 OPR_SIMM16 16-bit integer constant. VALUE Integer constant value. 16 0 OPR_SIMM24 24-bit integer constant. VALUE Integer constant value. 24 0 OPR_SIMM32 32-bit integer constant. VALUE Integer constant value. 32 0 OPR_SIMM5 5-bit integer constant. VALUE Integer constant value. 5 0 OPR_SIMM8 8-bit integer constant. VALUE Integer constant value. 8 0 OPR_SLEEP SLEEP operand specifying duration and sleep-forever mode. DURATION Determines the sleep duration. The wave sleeps for (64*(SIMM16[6:0]-1) .. 64*SIMM16[6:0]) clocks. The exact amount of delay is approximate. Compare with S_NOP. When set to zero, no sleep occurs. 7 0 SLEEP_FOREVER If set to 1, enables "sleep forever" mode. The wave sleeps until woken up by one of the following: S_WAKEUP S_WAKEUP_BARRIER An exception Host trap Other trap events Wave KILL 1 15 OPR_SMEM_OFFSET Operand is a scalar memory offset, which may come from a scalar GPR, from M0 or from a 24-bit immediate constant. OPR_SDST_M0 OPR_SIMM24 OPR_SMEM_OFFSET_NOK OPR_SREG s0 Scalar GPR 0. Increment from here for additional GPRs. There are NUM_SGPR SGPRs in total. Alignment restriction: unless otherwise noted, 64-bit SGPR values must be even-aligned (first SGPR address is even); 128-bit and wider SGPR values must be aligned to a multiple of 4. 0 s1 <p>See above.</p> 1 s2 <p>See above.</p> 2 s3 <p>See above.</p> 3 s4 <p>See above.</p> 4 s5 <p>See above.</p> 5 s6 <p>See above.</p> 6 s7 <p>See above.</p> 7 s8 <p>See above.</p> 8 s9 <p>See above.</p> 9 s10 <p>See above.</p> 10 s11 <p>See above.</p> 11 s12 <p>See above.</p> 12 s13 <p>See above.</p> 13 s14 <p>See above.</p> 14 s15 <p>See above.</p> 15 s16 <p>See above.</p> 16 s17 <p>See above.</p> 17 s18 <p>See above.</p> 18 s19 <p>See above.</p> 19 s20 <p>See above.</p> 20 s21 <p>See above.</p> 21 s22 <p>See above.</p> 22 s23 <p>See above.</p> 23 s24 <p>See above.</p> 24 s25 <p>See above.</p> 25 s26 <p>See above.</p> 26 s27 <p>See above.</p> 27 s28 <p>See above.</p> 28 s29 <p>See above.</p> 29 s30 <p>See above.</p> 30 s31 <p>See above.</p> 31 s32 <p>See above.</p> 32 s33 <p>See above.</p> 33 s34 <p>See above.</p> 34 s35 <p>See above.</p> 35 s36 <p>See above.</p> 36 s37 <p>See above.</p> 37 s38 <p>See above.</p> 38 s39 <p>See above.</p> 39 s40 <p>See above.</p> 40 s41 <p>See above.</p> 41 s42 <p>See above.</p> 42 s43 <p>See above.</p> 43 s44 <p>See above.</p> 44 s45 <p>See above.</p> 45 s46 <p>See above.</p> 46 s47 <p>See above.</p> 47 s48 <p>See above.</p> 48 s49 <p>See above.</p> 49 s50 <p>See above.</p> 50 s51 <p>See above.</p> 51 s52 <p>See above.</p> 52 s53 <p>See above.</p> 53 s54 <p>See above.</p> 54 s55 <p>See above.</p> 55 s56 <p>See above.</p> 56 s57 <p>See above.</p> 57 s58 <p>See above.</p> 58 s59 <p>See above.</p> 59 s60 <p>See above.</p> 60 s61 <p>See above.</p> 61 s62 <p>See above.</p> 62 s63 <p>See above.</p> 63 s64 <p>See above.</p> 64 s65 <p>See above.</p> 65 s66 <p>See above.</p> 66 s67 <p>See above.</p> 67 s68 <p>See above.</p> 68 s69 <p>See above.</p> 69 s70 <p>See above.</p> 70 s71 <p>See above.</p> 71 s72 <p>See above.</p> 72 s73 <p>See above.</p> 73 s74 <p>See above.</p> 74 s75 <p>See above.</p> 75 s76 <p>See above.</p> 76 s77 <p>See above.</p> 77 s78 <p>See above.</p> 78 s79 <p>See above.</p> 79 s80 <p>See above.</p> 80 s81 <p>See above.</p> 81 s82 <p>See above.</p> 82 s83 <p>See above.</p> 83 s84 <p>See above.</p> 84 s85 <p>See above.</p> 85 s86 <p>See above.</p> 86 s87 <p>See above.</p> 87 s88 <p>See above.</p> 88 s89 <p>See above.</p> 89 s90 <p>See above.</p> 90 s91 <p>See above.</p> 91 s92 <p>See above.</p> 92 s93 <p>See above.</p> 93 s94 <p>See above.</p> 94 s95 <p>See above.</p> 95 s96 <p>See above.</p> 96 s97 <p>See above.</p> 97 s98 <p>See above.</p> 98 s99 <p>See above.</p> 99 s100 <p>See above.</p> 100 s101 <p>See above.</p> 101 s102 <p>See above.</p> 102 s103 <p>See above.</p> 103 s104 <p>See above.</p> 104 s105 <p>See above.</p> 105 ttmp0 Trap handler temps (privileged). Increment from here for additional TTMPs. There are NUM_TTMP TTMPs in total. {TTMP1,TTMP0} = PC_save{hi,lo}. Alignment restriction: unless otherwise noted, 64-bit SGPR values must be even-aligned (first SGPR address is even); 128-bit and wider SGPR values must be aligned to a multiple of 4. 108 ttmp1 Trap handler temps (privileged). 109 ttmp2 Trap handler temps (privileged). 110 ttmp3 Trap handler temps (privileged). 111 ttmp4 Trap handler temps (privileged). 112 ttmp5 Trap handler temps (privileged). 113 ttmp6 Trap handler temps (privileged). 114 ttmp7 Trap handler temps (privileged). 115 ttmp8 Trap handler temps (privileged). 116 ttmp9 Trap handler temps (privileged). 117 ttmp10 Trap handler temps (privileged). 118 ttmp11 Trap handler temps (privileged). 119 ttmp12 Trap handler temps (privileged). 120 ttmp13 Trap handler temps (privileged). 121 ttmp14 Trap handler temps (privileged). 122 ttmp15 Trap handler temps (privileged). 123 vcc_lo VCC[31:0] 106 vcc_hi VCC[63:32] 107 null Do not write destination. For scalar memory, do not use SOFFSET in address calculation. For scratch addressing, do not use an SGPR in the addressing calculation. Otherwise read back zero. 124 m0 Special register used to hold DS addresses, relative indices, and send-message values. 125 m0 Special register used to hold DS addresses, relative indices, and send-message values. 125 VALUE Integer constant value. 24 0 OPR_SMEM_OFFSET_NOK Operand is a scalar memory offset, which may come from a scalar GPR or M0. OPR_SDST_M0 OPR_SREG s0 Scalar GPR 0. Increment from here for additional GPRs. There are NUM_SGPR SGPRs in total. Alignment restriction: unless otherwise noted, 64-bit SGPR values must be even-aligned (first SGPR address is even); 128-bit and wider SGPR values must be aligned to a multiple of 4. 0 s1 <p>See above.</p> 1 s2 <p>See above.</p> 2 s3 <p>See above.</p> 3 s4 <p>See above.</p> 4 s5 <p>See above.</p> 5 s6 <p>See above.</p> 6 s7 <p>See above.</p> 7 s8 <p>See above.</p> 8 s9 <p>See above.</p> 9 s10 <p>See above.</p> 10 s11 <p>See above.</p> 11 s12 <p>See above.</p> 12 s13 <p>See above.</p> 13 s14 <p>See above.</p> 14 s15 <p>See above.</p> 15 s16 <p>See above.</p> 16 s17 <p>See above.</p> 17 s18 <p>See above.</p> 18 s19 <p>See above.</p> 19 s20 <p>See above.</p> 20 s21 <p>See above.</p> 21 s22 <p>See above.</p> 22 s23 <p>See above.</p> 23 s24 <p>See above.</p> 24 s25 <p>See above.</p> 25 s26 <p>See above.</p> 26 s27 <p>See above.</p> 27 s28 <p>See above.</p> 28 s29 <p>See above.</p> 29 s30 <p>See above.</p> 30 s31 <p>See above.</p> 31 s32 <p>See above.</p> 32 s33 <p>See above.</p> 33 s34 <p>See above.</p> 34 s35 <p>See above.</p> 35 s36 <p>See above.</p> 36 s37 <p>See above.</p> 37 s38 <p>See above.</p> 38 s39 <p>See above.</p> 39 s40 <p>See above.</p> 40 s41 <p>See above.</p> 41 s42 <p>See above.</p> 42 s43 <p>See above.</p> 43 s44 <p>See above.</p> 44 s45 <p>See above.</p> 45 s46 <p>See above.</p> 46 s47 <p>See above.</p> 47 s48 <p>See above.</p> 48 s49 <p>See above.</p> 49 s50 <p>See above.</p> 50 s51 <p>See above.</p> 51 s52 <p>See above.</p> 52 s53 <p>See above.</p> 53 s54 <p>See above.</p> 54 s55 <p>See above.</p> 55 s56 <p>See above.</p> 56 s57 <p>See above.</p> 57 s58 <p>See above.</p> 58 s59 <p>See above.</p> 59 s60 <p>See above.</p> 60 s61 <p>See above.</p> 61 s62 <p>See above.</p> 62 s63 <p>See above.</p> 63 s64 <p>See above.</p> 64 s65 <p>See above.</p> 65 s66 <p>See above.</p> 66 s67 <p>See above.</p> 67 s68 <p>See above.</p> 68 s69 <p>See above.</p> 69 s70 <p>See above.</p> 70 s71 <p>See above.</p> 71 s72 <p>See above.</p> 72 s73 <p>See above.</p> 73 s74 <p>See above.</p> 74 s75 <p>See above.</p> 75 s76 <p>See above.</p> 76 s77 <p>See above.</p> 77 s78 <p>See above.</p> 78 s79 <p>See above.</p> 79 s80 <p>See above.</p> 80 s81 <p>See above.</p> 81 s82 <p>See above.</p> 82 s83 <p>See above.</p> 83 s84 <p>See above.</p> 84 s85 <p>See above.</p> 85 s86 <p>See above.</p> 86 s87 <p>See above.</p> 87 s88 <p>See above.</p> 88 s89 <p>See above.</p> 89 s90 <p>See above.</p> 90 s91 <p>See above.</p> 91 s92 <p>See above.</p> 92 s93 <p>See above.</p> 93 s94 <p>See above.</p> 94 s95 <p>See above.</p> 95 s96 <p>See above.</p> 96 s97 <p>See above.</p> 97 s98 <p>See above.</p> 98 s99 <p>See above.</p> 99 s100 <p>See above.</p> 100 s101 <p>See above.</p> 101 s102 <p>See above.</p> 102 s103 <p>See above.</p> 103 s104 <p>See above.</p> 104 s105 <p>See above.</p> 105 ttmp0 Trap handler temps (privileged). Increment from here for additional TTMPs. There are NUM_TTMP TTMPs in total. {TTMP1,TTMP0} = PC_save{hi,lo}. Alignment restriction: unless otherwise noted, 64-bit SGPR values must be even-aligned (first SGPR address is even); 128-bit and wider SGPR values must be aligned to a multiple of 4. 108 ttmp1 Trap handler temps (privileged). 109 ttmp2 Trap handler temps (privileged). 110 ttmp3 Trap handler temps (privileged). 111 ttmp4 Trap handler temps (privileged). 112 ttmp5 Trap handler temps (privileged). 113 ttmp6 Trap handler temps (privileged). 114 ttmp7 Trap handler temps (privileged). 115 ttmp8 Trap handler temps (privileged). 116 ttmp9 Trap handler temps (privileged). 117 ttmp10 Trap handler temps (privileged). 118 ttmp11 Trap handler temps (privileged). 119 ttmp12 Trap handler temps (privileged). 120 ttmp13 Trap handler temps (privileged). 121 ttmp14 Trap handler temps (privileged). 122 ttmp15 Trap handler temps (privileged). 123 vcc_lo VCC[31:0] 106 vcc_hi VCC[63:32] 107 null Do not write destination. For scalar memory, do not use SOFFSET in address calculation. For scratch addressing, do not use an SGPR in the addressing calculation. Otherwise read back zero. 124 m0 Special register used to hold DS addresses, relative indices, and send-message values. 125 OPR_SRC All operands. Covers all operands that are allowed as scalar or vector sources. OPR_SDST OPR_SDST_EXEC OPR_SDST_M0 OPR_SRC_VGPR OPR_SREG OPR_SSRC_SPECIAL_SCC s0 Scalar GPR 0. Increment from here for additional GPRs. There are NUM_SGPR SGPRs in total. Alignment restriction: unless otherwise noted, 64-bit SGPR values must be even-aligned (first SGPR address is even); 128-bit and wider SGPR values must be aligned to a multiple of 4. 0 s1 <p>See above.</p> 1 s2 <p>See above.</p> 2 s3 <p>See above.</p> 3 s4 <p>See above.</p> 4 s5 <p>See above.</p> 5 s6 <p>See above.</p> 6 s7 <p>See above.</p> 7 s8 <p>See above.</p> 8 s9 <p>See above.</p> 9 s10 <p>See above.</p> 10 s11 <p>See above.</p> 11 s12 <p>See above.</p> 12 s13 <p>See above.</p> 13 s14 <p>See above.</p> 14 s15 <p>See above.</p> 15 s16 <p>See above.</p> 16 s17 <p>See above.</p> 17 s18 <p>See above.</p> 18 s19 <p>See above.</p> 19 s20 <p>See above.</p> 20 s21 <p>See above.</p> 21 s22 <p>See above.</p> 22 s23 <p>See above.</p> 23 s24 <p>See above.</p> 24 s25 <p>See above.</p> 25 s26 <p>See above.</p> 26 s27 <p>See above.</p> 27 s28 <p>See above.</p> 28 s29 <p>See above.</p> 29 s30 <p>See above.</p> 30 s31 <p>See above.</p> 31 s32 <p>See above.</p> 32 s33 <p>See above.</p> 33 s34 <p>See above.</p> 34 s35 <p>See above.</p> 35 s36 <p>See above.</p> 36 s37 <p>See above.</p> 37 s38 <p>See above.</p> 38 s39 <p>See above.</p> 39 s40 <p>See above.</p> 40 s41 <p>See above.</p> 41 s42 <p>See above.</p> 42 s43 <p>See above.</p> 43 s44 <p>See above.</p> 44 s45 <p>See above.</p> 45 s46 <p>See above.</p> 46 s47 <p>See above.</p> 47 s48 <p>See above.</p> 48 s49 <p>See above.</p> 49 s50 <p>See above.</p> 50 s51 <p>See above.</p> 51 s52 <p>See above.</p> 52 s53 <p>See above.</p> 53 s54 <p>See above.</p> 54 s55 <p>See above.</p> 55 s56 <p>See above.</p> 56 s57 <p>See above.</p> 57 s58 <p>See above.</p> 58 s59 <p>See above.</p> 59 s60 <p>See above.</p> 60 s61 <p>See above.</p> 61 s62 <p>See above.</p> 62 s63 <p>See above.</p> 63 s64 <p>See above.</p> 64 s65 <p>See above.</p> 65 s66 <p>See above.</p> 66 s67 <p>See above.</p> 67 s68 <p>See above.</p> 68 s69 <p>See above.</p> 69 s70 <p>See above.</p> 70 s71 <p>See above.</p> 71 s72 <p>See above.</p> 72 s73 <p>See above.</p> 73 s74 <p>See above.</p> 74 s75 <p>See above.</p> 75 s76 <p>See above.</p> 76 s77 <p>See above.</p> 77 s78 <p>See above.</p> 78 s79 <p>See above.</p> 79 s80 <p>See above.</p> 80 s81 <p>See above.</p> 81 s82 <p>See above.</p> 82 s83 <p>See above.</p> 83 s84 <p>See above.</p> 84 s85 <p>See above.</p> 85 s86 <p>See above.</p> 86 s87 <p>See above.</p> 87 s88 <p>See above.</p> 88 s89 <p>See above.</p> 89 s90 <p>See above.</p> 90 s91 <p>See above.</p> 91 s92 <p>See above.</p> 92 s93 <p>See above.</p> 93 s94 <p>See above.</p> 94 s95 <p>See above.</p> 95 s96 <p>See above.</p> 96 s97 <p>See above.</p> 97 s98 <p>See above.</p> 98 s99 <p>See above.</p> 99 s100 <p>See above.</p> 100 s101 <p>See above.</p> 101 s102 <p>See above.</p> 102 s103 <p>See above.</p> 103 s104 <p>See above.</p> 104 s105 <p>See above.</p> 105 ttmp0 Trap handler temps (privileged). Increment from here for additional TTMPs. There are NUM_TTMP TTMPs in total. {TTMP1,TTMP0} = PC_save{hi,lo}. Alignment restriction: unless otherwise noted, 64-bit SGPR values must be even-aligned (first SGPR address is even); 128-bit and wider SGPR values must be aligned to a multiple of 4. 108 ttmp1 Trap handler temps (privileged). 109 ttmp2 Trap handler temps (privileged). 110 ttmp3 Trap handler temps (privileged). 111 ttmp4 Trap handler temps (privileged). 112 ttmp5 Trap handler temps (privileged). 113 ttmp6 Trap handler temps (privileged). 114 ttmp7 Trap handler temps (privileged). 115 ttmp8 Trap handler temps (privileged). 116 ttmp9 Trap handler temps (privileged). 117 ttmp10 Trap handler temps (privileged). 118 ttmp11 Trap handler temps (privileged). 119 ttmp12 Trap handler temps (privileged). 120 ttmp13 Trap handler temps (privileged). 121 ttmp14 Trap handler temps (privileged). 122 ttmp15 Trap handler temps (privileged). 123 vcc_lo VCC[31:0] 106 vcc_hi VCC[63:32] 107 null Do not write destination. For scalar memory, do not use SOFFSET in address calculation. For scratch addressing, do not use an SGPR in the addressing calculation. Otherwise read back zero. 124 m0 Special register used to hold DS addresses, relative indices, and send-message values. 125 exec_lo exec[31:0] 126 exec_hi exec[63:32] 127 0 0 128 1 1 (integer) 129 2 2 (integer) 130 3 3 (integer) 131 4 4 (integer) 132 5 5 (integer) 133 6 6 (integer) 134 7 7 (integer) 135 8 8 (integer) 136 9 9 (integer) 137 10 10 (integer) 138 11 11 (integer) 139 12 12 (integer) 140 13 13 (integer) 141 14 14 (integer) 142 15 15 (integer) 143 16 16 (integer) 144 17 17 (integer) 145 18 18 (integer) 146 19 19 (integer) 147 20 20 (integer) 148 21 21 (integer) 149 22 22 (integer) 150 23 23 (integer) 151 24 24 (integer) 152 25 25 (integer) 153 26 26 (integer) 154 27 27 (integer) 155 28 28 (integer) 156 29 29 (integer) 157 30 30 (integer) 158 31 31 (integer) 159 32 32 (integer) 160 33 33 (integer) 161 34 34 (integer) 162 35 35 (integer) 163 36 36 (integer) 164 37 37 (integer) 165 38 38 (integer) 166 39 39 (integer) 167 40 40 (integer) 168 41 41 (integer) 169 42 42 (integer) 170 43 43 (integer) 171 44 44 (integer) 172 45 45 (integer) 173 46 46 (integer) 174 47 47 (integer) 175 48 48 (integer) 176 49 49 (integer) 177 50 50 (integer) 178 51 51 (integer) 179 52 52 (integer) 180 53 53 (integer) 181 54 54 (integer) 182 55 55 (integer) 183 56 56 (integer) 184 57 57 (integer) 185 58 58 (integer) 186 59 59 (integer) 187 60 60 (integer) 188 61 61 (integer) 189 62 62 (integer) 190 63 63 (integer) 191 64 64 (integer) 192 -1 -1 (integer) 193 -2 -2 (integer) 194 -3 -3 (integer) 195 -4 -4 (integer) 196 -5 -5 (integer) 197 -6 -6 (integer) 198 -7 -7 (integer) 199 -8 -8 (integer) 200 -9 -9 (integer) 201 -10 -10 (integer) 202 -11 -11 (integer) 203 -12 -12 (integer) 204 -13 -13 (integer) 205 -14 -14 (integer) 206 -15 -15 (integer) 207 -16 -16 (integer) 208 0.5 0.5 240 -0.5 -0.5 241 1.0 1.0 242 -1.0 -1.0 243 2.0 2.0 244 -2.0 -2.0 245 4.0 4.0 246 -4.0 -4.0 247 0.15915494 1 / (2 * PI). For F16 and PK2_F16 formats: 0x3118. For BF16 and PK2_BF16 formats: 0x3e22 (truncated from 32-bit format with no rounding). For all 64-bit formats (F64, B64, I64 etc.): 0x3fc45f30_6dc9c882. For all other cases (including 32-bit, B16, I16): 0x3e22f983. 248 src_scc Scalar condition code. 253 src_shared_base 235 src_shared_limit 236 src_private_base 237 src_private_limit 238 v0 Vector GPR 0. Increment from here for additional GPRs. There are NUM_VGPR VGPRs in total. The constant SQ_SRC_VGPR_BIT to set or clear the high order bit for vector GPRs in this operand. 256 v1 <p>See above.</p> 257 v2 <p>See above.</p> 258 v3 <p>See above.</p> 259 v4 <p>See above.</p> 260 v5 <p>See above.</p> 261 v6 <p>See above.</p> 262 v7 <p>See above.</p> 263 v8 <p>See above.</p> 264 v9 <p>See above.</p> 265 v10 <p>See above.</p> 266 v11 <p>See above.</p> 267 v12 <p>See above.</p> 268 v13 <p>See above.</p> 269 v14 <p>See above.</p> 270 v15 <p>See above.</p> 271 v16 <p>See above.</p> 272 v17 <p>See above.</p> 273 v18 <p>See above.</p> 274 v19 <p>See above.</p> 275 v20 <p>See above.</p> 276 v21 <p>See above.</p> 277 v22 <p>See above.</p> 278 v23 <p>See above.</p> 279 v24 <p>See above.</p> 280 v25 <p>See above.</p> 281 v26 <p>See above.</p> 282 v27 <p>See above.</p> 283 v28 <p>See above.</p> 284 v29 <p>See above.</p> 285 v30 <p>See above.</p> 286 v31 <p>See above.</p> 287 v32 <p>See above.</p> 288 v33 <p>See above.</p> 289 v34 <p>See above.</p> 290 v35 <p>See above.</p> 291 v36 <p>See above.</p> 292 v37 <p>See above.</p> 293 v38 <p>See above.</p> 294 v39 <p>See above.</p> 295 v40 <p>See above.</p> 296 v41 <p>See above.</p> 297 v42 <p>See above.</p> 298 v43 <p>See above.</p> 299 v44 <p>See above.</p> 300 v45 <p>See above.</p> 301 v46 <p>See above.</p> 302 v47 <p>See above.</p> 303 v48 <p>See above.</p> 304 v49 <p>See above.</p> 305 v50 <p>See above.</p> 306 v51 <p>See above.</p> 307 v52 <p>See above.</p> 308 v53 <p>See above.</p> 309 v54 <p>See above.</p> 310 v55 <p>See above.</p> 311 v56 <p>See above.</p> 312 v57 <p>See above.</p> 313 v58 <p>See above.</p> 314 v59 <p>See above.</p> 315 v60 <p>See above.</p> 316 v61 <p>See above.</p> 317 v62 <p>See above.</p> 318 v63 <p>See above.</p> 319 v64 <p>See above.</p> 320 v65 <p>See above.</p> 321 v66 <p>See above.</p> 322 v67 <p>See above.</p> 323 v68 <p>See above.</p> 324 v69 <p>See above.</p> 325 v70 <p>See above.</p> 326 v71 <p>See above.</p> 327 v72 <p>See above.</p> 328 v73 <p>See above.</p> 329 v74 <p>See above.</p> 330 v75 <p>See above.</p> 331 v76 <p>See above.</p> 332 v77 <p>See above.</p> 333 v78 <p>See above.</p> 334 v79 <p>See above.</p> 335 v80 <p>See above.</p> 336 v81 <p>See above.</p> 337 v82 <p>See above.</p> 338 v83 <p>See above.</p> 339 v84 <p>See above.</p> 340 v85 <p>See above.</p> 341 v86 <p>See above.</p> 342 v87 <p>See above.</p> 343 v88 <p>See above.</p> 344 v89 <p>See above.</p> 345 v90 <p>See above.</p> 346 v91 <p>See above.</p> 347 v92 <p>See above.</p> 348 v93 <p>See above.</p> 349 v94 <p>See above.</p> 350 v95 <p>See above.</p> 351 v96 <p>See above.</p> 352 v97 <p>See above.</p> 353 v98 <p>See above.</p> 354 v99 <p>See above.</p> 355 v100 <p>See above.</p> 356 v101 <p>See above.</p> 357 v102 <p>See above.</p> 358 v103 <p>See above.</p> 359 v104 <p>See above.</p> 360 v105 <p>See above.</p> 361 v106 <p>See above.</p> 362 v107 <p>See above.</p> 363 v108 <p>See above.</p> 364 v109 <p>See above.</p> 365 v110 <p>See above.</p> 366 v111 <p>See above.</p> 367 v112 <p>See above.</p> 368 v113 <p>See above.</p> 369 v114 <p>See above.</p> 370 v115 <p>See above.</p> 371 v116 <p>See above.</p> 372 v117 <p>See above.</p> 373 v118 <p>See above.</p> 374 v119 <p>See above.</p> 375 v120 <p>See above.</p> 376 v121 <p>See above.</p> 377 v122 <p>See above.</p> 378 v123 <p>See above.</p> 379 v124 <p>See above.</p> 380 v125 <p>See above.</p> 381 v126 <p>See above.</p> 382 v127 <p>See above.</p> 383 v128 <p>See above.</p> 384 v129 <p>See above.</p> 385 v130 <p>See above.</p> 386 v131 <p>See above.</p> 387 v132 <p>See above.</p> 388 v133 <p>See above.</p> 389 v134 <p>See above.</p> 390 v135 <p>See above.</p> 391 v136 <p>See above.</p> 392 v137 <p>See above.</p> 393 v138 <p>See above.</p> 394 v139 <p>See above.</p> 395 v140 <p>See above.</p> 396 v141 <p>See above.</p> 397 v142 <p>See above.</p> 398 v143 <p>See above.</p> 399 v144 <p>See above.</p> 400 v145 <p>See above.</p> 401 v146 <p>See above.</p> 402 v147 <p>See above.</p> 403 v148 <p>See above.</p> 404 v149 <p>See above.</p> 405 v150 <p>See above.</p> 406 v151 <p>See above.</p> 407 v152 <p>See above.</p> 408 v153 <p>See above.</p> 409 v154 <p>See above.</p> 410 v155 <p>See above.</p> 411 v156 <p>See above.</p> 412 v157 <p>See above.</p> 413 v158 <p>See above.</p> 414 v159 <p>See above.</p> 415 v160 <p>See above.</p> 416 v161 <p>See above.</p> 417 v162 <p>See above.</p> 418 v163 <p>See above.</p> 419 v164 <p>See above.</p> 420 v165 <p>See above.</p> 421 v166 <p>See above.</p> 422 v167 <p>See above.</p> 423 v168 <p>See above.</p> 424 v169 <p>See above.</p> 425 v170 <p>See above.</p> 426 v171 <p>See above.</p> 427 v172 <p>See above.</p> 428 v173 <p>See above.</p> 429 v174 <p>See above.</p> 430 v175 <p>See above.</p> 431 v176 <p>See above.</p> 432 v177 <p>See above.</p> 433 v178 <p>See above.</p> 434 v179 <p>See above.</p> 435 v180 <p>See above.</p> 436 v181 <p>See above.</p> 437 v182 <p>See above.</p> 438 v183 <p>See above.</p> 439 v184 <p>See above.</p> 440 v185 <p>See above.</p> 441 v186 <p>See above.</p> 442 v187 <p>See above.</p> 443 v188 <p>See above.</p> 444 v189 <p>See above.</p> 445 v190 <p>See above.</p> 446 v191 <p>See above.</p> 447 v192 <p>See above.</p> 448 v193 <p>See above.</p> 449 v194 <p>See above.</p> 450 v195 <p>See above.</p> 451 v196 <p>See above.</p> 452 v197 <p>See above.</p> 453 v198 <p>See above.</p> 454 v199 <p>See above.</p> 455 v200 <p>See above.</p> 456 v201 <p>See above.</p> 457 v202 <p>See above.</p> 458 v203 <p>See above.</p> 459 v204 <p>See above.</p> 460 v205 <p>See above.</p> 461 v206 <p>See above.</p> 462 v207 <p>See above.</p> 463 v208 <p>See above.</p> 464 v209 <p>See above.</p> 465 v210 <p>See above.</p> 466 v211 <p>See above.</p> 467 v212 <p>See above.</p> 468 v213 <p>See above.</p> 469 v214 <p>See above.</p> 470 v215 <p>See above.</p> 471 v216 <p>See above.</p> 472 v217 <p>See above.</p> 473 v218 <p>See above.</p> 474 v219 <p>See above.</p> 475 v220 <p>See above.</p> 476 v221 <p>See above.</p> 477 v222 <p>See above.</p> 478 v223 <p>See above.</p> 479 v224 <p>See above.</p> 480 v225 <p>See above.</p> 481 v226 <p>See above.</p> 482 v227 <p>See above.</p> 483 v228 <p>See above.</p> 484 v229 <p>See above.</p> 485 v230 <p>See above.</p> 486 v231 <p>See above.</p> 487 v232 <p>See above.</p> 488 v233 <p>See above.</p> 489 v234 <p>See above.</p> 490 v235 <p>See above.</p> 491 v236 <p>See above.</p> 492 v237 <p>See above.</p> 493 v238 <p>See above.</p> 494 v239 <p>See above.</p> 495 v240 <p>See above.</p> 496 v241 <p>See above.</p> 497 v242 <p>See above.</p> 498 v243 <p>See above.</p> 499 v244 <p>See above.</p> 500 v245 <p>See above.</p> 501 v246 <p>See above.</p> 502 v247 <p>See above.</p> 503 v248 <p>See above.</p> 504 v249 <p>See above.</p> 505 v250 <p>See above.</p> 506 v251 <p>See above.</p> 507 v252 <p>See above.</p> 508 v253 <p>See above.</p> 509 v254 <p>See above.</p> 510 v255 <p>See above.</p> 511 src_literal 32-bit literal constant follows this instruction. 16-bit operands use only the 16 LSBs of this constant. 64-bit unsigned integer operands (U64, B64) zero-extend the constant. 64-bit signed integer operands (I64) sign-extend the constant. 64-bit floating-point operands use the constant as the high 32 MSBs of a double-precision floating point value (1 sign bit, 11 exponent bits and 20 mantissa bits). 255 OPR_SRC_VGPR Operand must be a vector GPR. Uses a 9-bit operand field. v0 Vector GPR 0. Increment from here for additional GPRs. There are NUM_VGPR VGPRs in total. The constant SQ_SRC_VGPR_BIT to set or clear the high order bit for vector GPRs in this operand. 256 v1 <p>See above.</p> 257 v2 <p>See above.</p> 258 v3 <p>See above.</p> 259 v4 <p>See above.</p> 260 v5 <p>See above.</p> 261 v6 <p>See above.</p> 262 v7 <p>See above.</p> 263 v8 <p>See above.</p> 264 v9 <p>See above.</p> 265 v10 <p>See above.</p> 266 v11 <p>See above.</p> 267 v12 <p>See above.</p> 268 v13 <p>See above.</p> 269 v14 <p>See above.</p> 270 v15 <p>See above.</p> 271 v16 <p>See above.</p> 272 v17 <p>See above.</p> 273 v18 <p>See above.</p> 274 v19 <p>See above.</p> 275 v20 <p>See above.</p> 276 v21 <p>See above.</p> 277 v22 <p>See above.</p> 278 v23 <p>See above.</p> 279 v24 <p>See above.</p> 280 v25 <p>See above.</p> 281 v26 <p>See above.</p> 282 v27 <p>See above.</p> 283 v28 <p>See above.</p> 284 v29 <p>See above.</p> 285 v30 <p>See above.</p> 286 v31 <p>See above.</p> 287 v32 <p>See above.</p> 288 v33 <p>See above.</p> 289 v34 <p>See above.</p> 290 v35 <p>See above.</p> 291 v36 <p>See above.</p> 292 v37 <p>See above.</p> 293 v38 <p>See above.</p> 294 v39 <p>See above.</p> 295 v40 <p>See above.</p> 296 v41 <p>See above.</p> 297 v42 <p>See above.</p> 298 v43 <p>See above.</p> 299 v44 <p>See above.</p> 300 v45 <p>See above.</p> 301 v46 <p>See above.</p> 302 v47 <p>See above.</p> 303 v48 <p>See above.</p> 304 v49 <p>See above.</p> 305 v50 <p>See above.</p> 306 v51 <p>See above.</p> 307 v52 <p>See above.</p> 308 v53 <p>See above.</p> 309 v54 <p>See above.</p> 310 v55 <p>See above.</p> 311 v56 <p>See above.</p> 312 v57 <p>See above.</p> 313 v58 <p>See above.</p> 314 v59 <p>See above.</p> 315 v60 <p>See above.</p> 316 v61 <p>See above.</p> 317 v62 <p>See above.</p> 318 v63 <p>See above.</p> 319 v64 <p>See above.</p> 320 v65 <p>See above.</p> 321 v66 <p>See above.</p> 322 v67 <p>See above.</p> 323 v68 <p>See above.</p> 324 v69 <p>See above.</p> 325 v70 <p>See above.</p> 326 v71 <p>See above.</p> 327 v72 <p>See above.</p> 328 v73 <p>See above.</p> 329 v74 <p>See above.</p> 330 v75 <p>See above.</p> 331 v76 <p>See above.</p> 332 v77 <p>See above.</p> 333 v78 <p>See above.</p> 334 v79 <p>See above.</p> 335 v80 <p>See above.</p> 336 v81 <p>See above.</p> 337 v82 <p>See above.</p> 338 v83 <p>See above.</p> 339 v84 <p>See above.</p> 340 v85 <p>See above.</p> 341 v86 <p>See above.</p> 342 v87 <p>See above.</p> 343 v88 <p>See above.</p> 344 v89 <p>See above.</p> 345 v90 <p>See above.</p> 346 v91 <p>See above.</p> 347 v92 <p>See above.</p> 348 v93 <p>See above.</p> 349 v94 <p>See above.</p> 350 v95 <p>See above.</p> 351 v96 <p>See above.</p> 352 v97 <p>See above.</p> 353 v98 <p>See above.</p> 354 v99 <p>See above.</p> 355 v100 <p>See above.</p> 356 v101 <p>See above.</p> 357 v102 <p>See above.</p> 358 v103 <p>See above.</p> 359 v104 <p>See above.</p> 360 v105 <p>See above.</p> 361 v106 <p>See above.</p> 362 v107 <p>See above.</p> 363 v108 <p>See above.</p> 364 v109 <p>See above.</p> 365 v110 <p>See above.</p> 366 v111 <p>See above.</p> 367 v112 <p>See above.</p> 368 v113 <p>See above.</p> 369 v114 <p>See above.</p> 370 v115 <p>See above.</p> 371 v116 <p>See above.</p> 372 v117 <p>See above.</p> 373 v118 <p>See above.</p> 374 v119 <p>See above.</p> 375 v120 <p>See above.</p> 376 v121 <p>See above.</p> 377 v122 <p>See above.</p> 378 v123 <p>See above.</p> 379 v124 <p>See above.</p> 380 v125 <p>See above.</p> 381 v126 <p>See above.</p> 382 v127 <p>See above.</p> 383 v128 <p>See above.</p> 384 v129 <p>See above.</p> 385 v130 <p>See above.</p> 386 v131 <p>See above.</p> 387 v132 <p>See above.</p> 388 v133 <p>See above.</p> 389 v134 <p>See above.</p> 390 v135 <p>See above.</p> 391 v136 <p>See above.</p> 392 v137 <p>See above.</p> 393 v138 <p>See above.</p> 394 v139 <p>See above.</p> 395 v140 <p>See above.</p> 396 v141 <p>See above.</p> 397 v142 <p>See above.</p> 398 v143 <p>See above.</p> 399 v144 <p>See above.</p> 400 v145 <p>See above.</p> 401 v146 <p>See above.</p> 402 v147 <p>See above.</p> 403 v148 <p>See above.</p> 404 v149 <p>See above.</p> 405 v150 <p>See above.</p> 406 v151 <p>See above.</p> 407 v152 <p>See above.</p> 408 v153 <p>See above.</p> 409 v154 <p>See above.</p> 410 v155 <p>See above.</p> 411 v156 <p>See above.</p> 412 v157 <p>See above.</p> 413 v158 <p>See above.</p> 414 v159 <p>See above.</p> 415 v160 <p>See above.</p> 416 v161 <p>See above.</p> 417 v162 <p>See above.</p> 418 v163 <p>See above.</p> 419 v164 <p>See above.</p> 420 v165 <p>See above.</p> 421 v166 <p>See above.</p> 422 v167 <p>See above.</p> 423 v168 <p>See above.</p> 424 v169 <p>See above.</p> 425 v170 <p>See above.</p> 426 v171 <p>See above.</p> 427 v172 <p>See above.</p> 428 v173 <p>See above.</p> 429 v174 <p>See above.</p> 430 v175 <p>See above.</p> 431 v176 <p>See above.</p> 432 v177 <p>See above.</p> 433 v178 <p>See above.</p> 434 v179 <p>See above.</p> 435 v180 <p>See above.</p> 436 v181 <p>See above.</p> 437 v182 <p>See above.</p> 438 v183 <p>See above.</p> 439 v184 <p>See above.</p> 440 v185 <p>See above.</p> 441 v186 <p>See above.</p> 442 v187 <p>See above.</p> 443 v188 <p>See above.</p> 444 v189 <p>See above.</p> 445 v190 <p>See above.</p> 446 v191 <p>See above.</p> 447 v192 <p>See above.</p> 448 v193 <p>See above.</p> 449 v194 <p>See above.</p> 450 v195 <p>See above.</p> 451 v196 <p>See above.</p> 452 v197 <p>See above.</p> 453 v198 <p>See above.</p> 454 v199 <p>See above.</p> 455 v200 <p>See above.</p> 456 v201 <p>See above.</p> 457 v202 <p>See above.</p> 458 v203 <p>See above.</p> 459 v204 <p>See above.</p> 460 v205 <p>See above.</p> 461 v206 <p>See above.</p> 462 v207 <p>See above.</p> 463 v208 <p>See above.</p> 464 v209 <p>See above.</p> 465 v210 <p>See above.</p> 466 v211 <p>See above.</p> 467 v212 <p>See above.</p> 468 v213 <p>See above.</p> 469 v214 <p>See above.</p> 470 v215 <p>See above.</p> 471 v216 <p>See above.</p> 472 v217 <p>See above.</p> 473 v218 <p>See above.</p> 474 v219 <p>See above.</p> 475 v220 <p>See above.</p> 476 v221 <p>See above.</p> 477 v222 <p>See above.</p> 478 v223 <p>See above.</p> 479 v224 <p>See above.</p> 480 v225 <p>See above.</p> 481 v226 <p>See above.</p> 482 v227 <p>See above.</p> 483 v228 <p>See above.</p> 484 v229 <p>See above.</p> 485 v230 <p>See above.</p> 486 v231 <p>See above.</p> 487 v232 <p>See above.</p> 488 v233 <p>See above.</p> 489 v234 <p>See above.</p> 490 v235 <p>See above.</p> 491 v236 <p>See above.</p> 492 v237 <p>See above.</p> 493 v238 <p>See above.</p> 494 v239 <p>See above.</p> 495 v240 <p>See above.</p> 496 v241 <p>See above.</p> 497 v242 <p>See above.</p> 498 v243 <p>See above.</p> 499 v244 <p>See above.</p> 500 v245 <p>See above.</p> 501 v246 <p>See above.</p> 502 v247 <p>See above.</p> 503 v248 <p>See above.</p> 504 v249 <p>See above.</p> 505 v250 <p>See above.</p> 506 v251 <p>See above.</p> 507 v252 <p>See above.</p> 508 v253 <p>See above.</p> 509 v254 <p>See above.</p> 510 v255 <p>See above.</p> 511 OPR_SRC_VGPR_OR_INLINE Operand must be a vector GPR or an inline constant (no literal constants or scalar GPRs). Uses a 9-bit operand field. OPR_SRC_VGPR 0 0 128 1 1 (integer) 129 2 2 (integer) 130 3 3 (integer) 131 4 4 (integer) 132 5 5 (integer) 133 6 6 (integer) 134 7 7 (integer) 135 8 8 (integer) 136 9 9 (integer) 137 10 10 (integer) 138 11 11 (integer) 139 12 12 (integer) 140 13 13 (integer) 141 14 14 (integer) 142 15 15 (integer) 143 16 16 (integer) 144 17 17 (integer) 145 18 18 (integer) 146 19 19 (integer) 147 20 20 (integer) 148 21 21 (integer) 149 22 22 (integer) 150 23 23 (integer) 151 24 24 (integer) 152 25 25 (integer) 153 26 26 (integer) 154 27 27 (integer) 155 28 28 (integer) 156 29 29 (integer) 157 30 30 (integer) 158 31 31 (integer) 159 32 32 (integer) 160 33 33 (integer) 161 34 34 (integer) 162 35 35 (integer) 163 36 36 (integer) 164 37 37 (integer) 165 38 38 (integer) 166 39 39 (integer) 167 40 40 (integer) 168 41 41 (integer) 169 42 42 (integer) 170 43 43 (integer) 171 44 44 (integer) 172 45 45 (integer) 173 46 46 (integer) 174 47 47 (integer) 175 48 48 (integer) 176 49 49 (integer) 177 50 50 (integer) 178 51 51 (integer) 179 52 52 (integer) 180 53 53 (integer) 181 54 54 (integer) 182 55 55 (integer) 183 56 56 (integer) 184 57 57 (integer) 185 58 58 (integer) 186 59 59 (integer) 187 60 60 (integer) 188 61 61 (integer) 189 62 62 (integer) 190 63 63 (integer) 191 64 64 (integer) 192 -1 -1 (integer) 193 -2 -2 (integer) 194 -3 -3 (integer) 195 -4 -4 (integer) 196 -5 -5 (integer) 197 -6 -6 (integer) 198 -7 -7 (integer) 199 -8 -8 (integer) 200 -9 -9 (integer) 201 -10 -10 (integer) 202 -11 -11 (integer) 203 -12 -12 (integer) 204 -13 -13 (integer) 205 -14 -14 (integer) 206 -15 -15 (integer) 207 -16 -16 (integer) 208 0.5 0.5 240 -0.5 -0.5 241 1.0 1.0 242 -1.0 -1.0 243 2.0 2.0 244 -2.0 -2.0 245 4.0 4.0 246 -4.0 -4.0 247 0.15915494 1 / (2 * PI). For F16 and PK2_F16 formats: 0x3118. For BF16 and PK2_BF16 formats: 0x3e22 (truncated from 32-bit format with no rounding). For all 64-bit formats (F64, B64, I64 etc.): 0x3fc45f30_6dc9c882. For all other cases (including 32-bit, B16, I16): 0x3e22f983. 248 v0 Vector GPR 0. Increment from here for additional GPRs. There are NUM_VGPR VGPRs in total. The constant SQ_SRC_VGPR_BIT to set or clear the high order bit for vector GPRs in this operand. 256 v1 <p>See above.</p> 257 v2 <p>See above.</p> 258 v3 <p>See above.</p> 259 v4 <p>See above.</p> 260 v5 <p>See above.</p> 261 v6 <p>See above.</p> 262 v7 <p>See above.</p> 263 v8 <p>See above.</p> 264 v9 <p>See above.</p> 265 v10 <p>See above.</p> 266 v11 <p>See above.</p> 267 v12 <p>See above.</p> 268 v13 <p>See above.</p> 269 v14 <p>See above.</p> 270 v15 <p>See above.</p> 271 v16 <p>See above.</p> 272 v17 <p>See above.</p> 273 v18 <p>See above.</p> 274 v19 <p>See above.</p> 275 v20 <p>See above.</p> 276 v21 <p>See above.</p> 277 v22 <p>See above.</p> 278 v23 <p>See above.</p> 279 v24 <p>See above.</p> 280 v25 <p>See above.</p> 281 v26 <p>See above.</p> 282 v27 <p>See above.</p> 283 v28 <p>See above.</p> 284 v29 <p>See above.</p> 285 v30 <p>See above.</p> 286 v31 <p>See above.</p> 287 v32 <p>See above.</p> 288 v33 <p>See above.</p> 289 v34 <p>See above.</p> 290 v35 <p>See above.</p> 291 v36 <p>See above.</p> 292 v37 <p>See above.</p> 293 v38 <p>See above.</p> 294 v39 <p>See above.</p> 295 v40 <p>See above.</p> 296 v41 <p>See above.</p> 297 v42 <p>See above.</p> 298 v43 <p>See above.</p> 299 v44 <p>See above.</p> 300 v45 <p>See above.</p> 301 v46 <p>See above.</p> 302 v47 <p>See above.</p> 303 v48 <p>See above.</p> 304 v49 <p>See above.</p> 305 v50 <p>See above.</p> 306 v51 <p>See above.</p> 307 v52 <p>See above.</p> 308 v53 <p>See above.</p> 309 v54 <p>See above.</p> 310 v55 <p>See above.</p> 311 v56 <p>See above.</p> 312 v57 <p>See above.</p> 313 v58 <p>See above.</p> 314 v59 <p>See above.</p> 315 v60 <p>See above.</p> 316 v61 <p>See above.</p> 317 v62 <p>See above.</p> 318 v63 <p>See above.</p> 319 v64 <p>See above.</p> 320 v65 <p>See above.</p> 321 v66 <p>See above.</p> 322 v67 <p>See above.</p> 323 v68 <p>See above.</p> 324 v69 <p>See above.</p> 325 v70 <p>See above.</p> 326 v71 <p>See above.</p> 327 v72 <p>See above.</p> 328 v73 <p>See above.</p> 329 v74 <p>See above.</p> 330 v75 <p>See above.</p> 331 v76 <p>See above.</p> 332 v77 <p>See above.</p> 333 v78 <p>See above.</p> 334 v79 <p>See above.</p> 335 v80 <p>See above.</p> 336 v81 <p>See above.</p> 337 v82 <p>See above.</p> 338 v83 <p>See above.</p> 339 v84 <p>See above.</p> 340 v85 <p>See above.</p> 341 v86 <p>See above.</p> 342 v87 <p>See above.</p> 343 v88 <p>See above.</p> 344 v89 <p>See above.</p> 345 v90 <p>See above.</p> 346 v91 <p>See above.</p> 347 v92 <p>See above.</p> 348 v93 <p>See above.</p> 349 v94 <p>See above.</p> 350 v95 <p>See above.</p> 351 v96 <p>See above.</p> 352 v97 <p>See above.</p> 353 v98 <p>See above.</p> 354 v99 <p>See above.</p> 355 v100 <p>See above.</p> 356 v101 <p>See above.</p> 357 v102 <p>See above.</p> 358 v103 <p>See above.</p> 359 v104 <p>See above.</p> 360 v105 <p>See above.</p> 361 v106 <p>See above.</p> 362 v107 <p>See above.</p> 363 v108 <p>See above.</p> 364 v109 <p>See above.</p> 365 v110 <p>See above.</p> 366 v111 <p>See above.</p> 367 v112 <p>See above.</p> 368 v113 <p>See above.</p> 369 v114 <p>See above.</p> 370 v115 <p>See above.</p> 371 v116 <p>See above.</p> 372 v117 <p>See above.</p> 373 v118 <p>See above.</p> 374 v119 <p>See above.</p> 375 v120 <p>See above.</p> 376 v121 <p>See above.</p> 377 v122 <p>See above.</p> 378 v123 <p>See above.</p> 379 v124 <p>See above.</p> 380 v125 <p>See above.</p> 381 v126 <p>See above.</p> 382 v127 <p>See above.</p> 383 v128 <p>See above.</p> 384 v129 <p>See above.</p> 385 v130 <p>See above.</p> 386 v131 <p>See above.</p> 387 v132 <p>See above.</p> 388 v133 <p>See above.</p> 389 v134 <p>See above.</p> 390 v135 <p>See above.</p> 391 v136 <p>See above.</p> 392 v137 <p>See above.</p> 393 v138 <p>See above.</p> 394 v139 <p>See above.</p> 395 v140 <p>See above.</p> 396 v141 <p>See above.</p> 397 v142 <p>See above.</p> 398 v143 <p>See above.</p> 399 v144 <p>See above.</p> 400 v145 <p>See above.</p> 401 v146 <p>See above.</p> 402 v147 <p>See above.</p> 403 v148 <p>See above.</p> 404 v149 <p>See above.</p> 405 v150 <p>See above.</p> 406 v151 <p>See above.</p> 407 v152 <p>See above.</p> 408 v153 <p>See above.</p> 409 v154 <p>See above.</p> 410 v155 <p>See above.</p> 411 v156 <p>See above.</p> 412 v157 <p>See above.</p> 413 v158 <p>See above.</p> 414 v159 <p>See above.</p> 415 v160 <p>See above.</p> 416 v161 <p>See above.</p> 417 v162 <p>See above.</p> 418 v163 <p>See above.</p> 419 v164 <p>See above.</p> 420 v165 <p>See above.</p> 421 v166 <p>See above.</p> 422 v167 <p>See above.</p> 423 v168 <p>See above.</p> 424 v169 <p>See above.</p> 425 v170 <p>See above.</p> 426 v171 <p>See above.</p> 427 v172 <p>See above.</p> 428 v173 <p>See above.</p> 429 v174 <p>See above.</p> 430 v175 <p>See above.</p> 431 v176 <p>See above.</p> 432 v177 <p>See above.</p> 433 v178 <p>See above.</p> 434 v179 <p>See above.</p> 435 v180 <p>See above.</p> 436 v181 <p>See above.</p> 437 v182 <p>See above.</p> 438 v183 <p>See above.</p> 439 v184 <p>See above.</p> 440 v185 <p>See above.</p> 441 v186 <p>See above.</p> 442 v187 <p>See above.</p> 443 v188 <p>See above.</p> 444 v189 <p>See above.</p> 445 v190 <p>See above.</p> 446 v191 <p>See above.</p> 447 v192 <p>See above.</p> 448 v193 <p>See above.</p> 449 v194 <p>See above.</p> 450 v195 <p>See above.</p> 451 v196 <p>See above.</p> 452 v197 <p>See above.</p> 453 v198 <p>See above.</p> 454 v199 <p>See above.</p> 455 v200 <p>See above.</p> 456 v201 <p>See above.</p> 457 v202 <p>See above.</p> 458 v203 <p>See above.</p> 459 v204 <p>See above.</p> 460 v205 <p>See above.</p> 461 v206 <p>See above.</p> 462 v207 <p>See above.</p> 463 v208 <p>See above.</p> 464 v209 <p>See above.</p> 465 v210 <p>See above.</p> 466 v211 <p>See above.</p> 467 v212 <p>See above.</p> 468 v213 <p>See above.</p> 469 v214 <p>See above.</p> 470 v215 <p>See above.</p> 471 v216 <p>See above.</p> 472 v217 <p>See above.</p> 473 v218 <p>See above.</p> 474 v219 <p>See above.</p> 475 v220 <p>See above.</p> 476 v221 <p>See above.</p> 477 v222 <p>See above.</p> 478 v223 <p>See above.</p> 479 v224 <p>See above.</p> 480 v225 <p>See above.</p> 481 v226 <p>See above.</p> 482 v227 <p>See above.</p> 483 v228 <p>See above.</p> 484 v229 <p>See above.</p> 485 v230 <p>See above.</p> 486 v231 <p>See above.</p> 487 v232 <p>See above.</p> 488 v233 <p>See above.</p> 489 v234 <p>See above.</p> 490 v235 <p>See above.</p> 491 v236 <p>See above.</p> 492 v237 <p>See above.</p> 493 v238 <p>See above.</p> 494 v239 <p>See above.</p> 495 v240 <p>See above.</p> 496 v241 <p>See above.</p> 497 v242 <p>See above.</p> 498 v243 <p>See above.</p> 499 v244 <p>See above.</p> 500 v245 <p>See above.</p> 501 v246 <p>See above.</p> 502 v247 <p>See above.</p> 503 v248 <p>See above.</p> 504 v249 <p>See above.</p> 505 v250 <p>See above.</p> 506 v251 <p>See above.</p> 507 v252 <p>See above.</p> 508 v253 <p>See above.</p> 509 v254 <p>See above.</p> 510 v255 <p>See above.</p> 511 OPR_SREG Any scalar GPR operand including VCC and NULL. s0 Scalar GPR 0. Increment from here for additional GPRs. There are NUM_SGPR SGPRs in total. Alignment restriction: unless otherwise noted, 64-bit SGPR values must be even-aligned (first SGPR address is even); 128-bit and wider SGPR values must be aligned to a multiple of 4. 0 s1 <p>See above.</p> 1 s2 <p>See above.</p> 2 s3 <p>See above.</p> 3 s4 <p>See above.</p> 4 s5 <p>See above.</p> 5 s6 <p>See above.</p> 6 s7 <p>See above.</p> 7 s8 <p>See above.</p> 8 s9 <p>See above.</p> 9 s10 <p>See above.</p> 10 s11 <p>See above.</p> 11 s12 <p>See above.</p> 12 s13 <p>See above.</p> 13 s14 <p>See above.</p> 14 s15 <p>See above.</p> 15 s16 <p>See above.</p> 16 s17 <p>See above.</p> 17 s18 <p>See above.</p> 18 s19 <p>See above.</p> 19 s20 <p>See above.</p> 20 s21 <p>See above.</p> 21 s22 <p>See above.</p> 22 s23 <p>See above.</p> 23 s24 <p>See above.</p> 24 s25 <p>See above.</p> 25 s26 <p>See above.</p> 26 s27 <p>See above.</p> 27 s28 <p>See above.</p> 28 s29 <p>See above.</p> 29 s30 <p>See above.</p> 30 s31 <p>See above.</p> 31 s32 <p>See above.</p> 32 s33 <p>See above.</p> 33 s34 <p>See above.</p> 34 s35 <p>See above.</p> 35 s36 <p>See above.</p> 36 s37 <p>See above.</p> 37 s38 <p>See above.</p> 38 s39 <p>See above.</p> 39 s40 <p>See above.</p> 40 s41 <p>See above.</p> 41 s42 <p>See above.</p> 42 s43 <p>See above.</p> 43 s44 <p>See above.</p> 44 s45 <p>See above.</p> 45 s46 <p>See above.</p> 46 s47 <p>See above.</p> 47 s48 <p>See above.</p> 48 s49 <p>See above.</p> 49 s50 <p>See above.</p> 50 s51 <p>See above.</p> 51 s52 <p>See above.</p> 52 s53 <p>See above.</p> 53 s54 <p>See above.</p> 54 s55 <p>See above.</p> 55 s56 <p>See above.</p> 56 s57 <p>See above.</p> 57 s58 <p>See above.</p> 58 s59 <p>See above.</p> 59 s60 <p>See above.</p> 60 s61 <p>See above.</p> 61 s62 <p>See above.</p> 62 s63 <p>See above.</p> 63 s64 <p>See above.</p> 64 s65 <p>See above.</p> 65 s66 <p>See above.</p> 66 s67 <p>See above.</p> 67 s68 <p>See above.</p> 68 s69 <p>See above.</p> 69 s70 <p>See above.</p> 70 s71 <p>See above.</p> 71 s72 <p>See above.</p> 72 s73 <p>See above.</p> 73 s74 <p>See above.</p> 74 s75 <p>See above.</p> 75 s76 <p>See above.</p> 76 s77 <p>See above.</p> 77 s78 <p>See above.</p> 78 s79 <p>See above.</p> 79 s80 <p>See above.</p> 80 s81 <p>See above.</p> 81 s82 <p>See above.</p> 82 s83 <p>See above.</p> 83 s84 <p>See above.</p> 84 s85 <p>See above.</p> 85 s86 <p>See above.</p> 86 s87 <p>See above.</p> 87 s88 <p>See above.</p> 88 s89 <p>See above.</p> 89 s90 <p>See above.</p> 90 s91 <p>See above.</p> 91 s92 <p>See above.</p> 92 s93 <p>See above.</p> 93 s94 <p>See above.</p> 94 s95 <p>See above.</p> 95 s96 <p>See above.</p> 96 s97 <p>See above.</p> 97 s98 <p>See above.</p> 98 s99 <p>See above.</p> 99 s100 <p>See above.</p> 100 s101 <p>See above.</p> 101 s102 <p>See above.</p> 102 s103 <p>See above.</p> 103 s104 <p>See above.</p> 104 s105 <p>See above.</p> 105 ttmp0 Trap handler temps (privileged). Increment from here for additional TTMPs. There are NUM_TTMP TTMPs in total. {TTMP1,TTMP0} = PC_save{hi,lo}. Alignment restriction: unless otherwise noted, 64-bit SGPR values must be even-aligned (first SGPR address is even); 128-bit and wider SGPR values must be aligned to a multiple of 4. 108 ttmp1 Trap handler temps (privileged). 109 ttmp2 Trap handler temps (privileged). 110 ttmp3 Trap handler temps (privileged). 111 ttmp4 Trap handler temps (privileged). 112 ttmp5 Trap handler temps (privileged). 113 ttmp6 Trap handler temps (privileged). 114 ttmp7 Trap handler temps (privileged). 115 ttmp8 Trap handler temps (privileged). 116 ttmp9 Trap handler temps (privileged). 117 ttmp10 Trap handler temps (privileged). 118 ttmp11 Trap handler temps (privileged). 119 ttmp12 Trap handler temps (privileged). 120 ttmp13 Trap handler temps (privileged). 121 ttmp14 Trap handler temps (privileged). 122 ttmp15 Trap handler temps (privileged). 123 vcc_lo VCC[31:0] 106 vcc_hi VCC[63:32] 107 null Do not write destination. For scalar memory, do not use SOFFSET in address calculation. For scratch addressing, do not use an SGPR in the addressing calculation. Otherwise read back zero. 124 OPR_SREG_LITERAL Vector memory operand that may be a scalar register or literal (but not inline). OPR_SREG s0 Scalar GPR 0. Increment from here for additional GPRs. There are NUM_SGPR SGPRs in total. Alignment restriction: unless otherwise noted, 64-bit SGPR values must be even-aligned (first SGPR address is even); 128-bit and wider SGPR values must be aligned to a multiple of 4. 0 s1 <p>See above.</p> 1 s2 <p>See above.</p> 2 s3 <p>See above.</p> 3 s4 <p>See above.</p> 4 s5 <p>See above.</p> 5 s6 <p>See above.</p> 6 s7 <p>See above.</p> 7 s8 <p>See above.</p> 8 s9 <p>See above.</p> 9 s10 <p>See above.</p> 10 s11 <p>See above.</p> 11 s12 <p>See above.</p> 12 s13 <p>See above.</p> 13 s14 <p>See above.</p> 14 s15 <p>See above.</p> 15 s16 <p>See above.</p> 16 s17 <p>See above.</p> 17 s18 <p>See above.</p> 18 s19 <p>See above.</p> 19 s20 <p>See above.</p> 20 s21 <p>See above.</p> 21 s22 <p>See above.</p> 22 s23 <p>See above.</p> 23 s24 <p>See above.</p> 24 s25 <p>See above.</p> 25 s26 <p>See above.</p> 26 s27 <p>See above.</p> 27 s28 <p>See above.</p> 28 s29 <p>See above.</p> 29 s30 <p>See above.</p> 30 s31 <p>See above.</p> 31 s32 <p>See above.</p> 32 s33 <p>See above.</p> 33 s34 <p>See above.</p> 34 s35 <p>See above.</p> 35 s36 <p>See above.</p> 36 s37 <p>See above.</p> 37 s38 <p>See above.</p> 38 s39 <p>See above.</p> 39 s40 <p>See above.</p> 40 s41 <p>See above.</p> 41 s42 <p>See above.</p> 42 s43 <p>See above.</p> 43 s44 <p>See above.</p> 44 s45 <p>See above.</p> 45 s46 <p>See above.</p> 46 s47 <p>See above.</p> 47 s48 <p>See above.</p> 48 s49 <p>See above.</p> 49 s50 <p>See above.</p> 50 s51 <p>See above.</p> 51 s52 <p>See above.</p> 52 s53 <p>See above.</p> 53 s54 <p>See above.</p> 54 s55 <p>See above.</p> 55 s56 <p>See above.</p> 56 s57 <p>See above.</p> 57 s58 <p>See above.</p> 58 s59 <p>See above.</p> 59 s60 <p>See above.</p> 60 s61 <p>See above.</p> 61 s62 <p>See above.</p> 62 s63 <p>See above.</p> 63 s64 <p>See above.</p> 64 s65 <p>See above.</p> 65 s66 <p>See above.</p> 66 s67 <p>See above.</p> 67 s68 <p>See above.</p> 68 s69 <p>See above.</p> 69 s70 <p>See above.</p> 70 s71 <p>See above.</p> 71 s72 <p>See above.</p> 72 s73 <p>See above.</p> 73 s74 <p>See above.</p> 74 s75 <p>See above.</p> 75 s76 <p>See above.</p> 76 s77 <p>See above.</p> 77 s78 <p>See above.</p> 78 s79 <p>See above.</p> 79 s80 <p>See above.</p> 80 s81 <p>See above.</p> 81 s82 <p>See above.</p> 82 s83 <p>See above.</p> 83 s84 <p>See above.</p> 84 s85 <p>See above.</p> 85 s86 <p>See above.</p> 86 s87 <p>See above.</p> 87 s88 <p>See above.</p> 88 s89 <p>See above.</p> 89 s90 <p>See above.</p> 90 s91 <p>See above.</p> 91 s92 <p>See above.</p> 92 s93 <p>See above.</p> 93 s94 <p>See above.</p> 94 s95 <p>See above.</p> 95 s96 <p>See above.</p> 96 s97 <p>See above.</p> 97 s98 <p>See above.</p> 98 s99 <p>See above.</p> 99 s100 <p>See above.</p> 100 s101 <p>See above.</p> 101 s102 <p>See above.</p> 102 s103 <p>See above.</p> 103 s104 <p>See above.</p> 104 s105 <p>See above.</p> 105 ttmp0 Trap handler temps (privileged). Increment from here for additional TTMPs. There are NUM_TTMP TTMPs in total. {TTMP1,TTMP0} = PC_save{hi,lo}. Alignment restriction: unless otherwise noted, 64-bit SGPR values must be even-aligned (first SGPR address is even); 128-bit and wider SGPR values must be aligned to a multiple of 4. 108 ttmp1 Trap handler temps (privileged). 109 ttmp2 Trap handler temps (privileged). 110 ttmp3 Trap handler temps (privileged). 111 ttmp4 Trap handler temps (privileged). 112 ttmp5 Trap handler temps (privileged). 113 ttmp6 Trap handler temps (privileged). 114 ttmp7 Trap handler temps (privileged). 115 ttmp8 Trap handler temps (privileged). 116 ttmp9 Trap handler temps (privileged). 117 ttmp10 Trap handler temps (privileged). 118 ttmp11 Trap handler temps (privileged). 119 ttmp12 Trap handler temps (privileged). 120 ttmp13 Trap handler temps (privileged). 121 ttmp14 Trap handler temps (privileged). 122 ttmp15 Trap handler temps (privileged). 123 vcc_lo VCC[31:0] 106 vcc_hi VCC[63:32] 107 null Do not write destination. For scalar memory, do not use SOFFSET in address calculation. For scratch addressing, do not use an SGPR in the addressing calculation. Otherwise read back zero. 124 src_literal 32-bit literal constant follows this instruction. 16-bit operands use only the 16 LSBs of this constant. 64-bit unsigned integer operands (U64, B64) zero-extend the constant. 64-bit signed integer operands (I64) sign-extend the constant. 64-bit floating-point operands use the constant as the high 32 MSBs of a double-precision floating point value (1 sign bit, 11 exponent bits and 20 mantissa bits). 255 OPR_SREG_M0 Vector memory operand that may be a scalar register or m0. OPR_SDST_M0 OPR_SREG s0 Scalar GPR 0. Increment from here for additional GPRs. There are NUM_SGPR SGPRs in total. Alignment restriction: unless otherwise noted, 64-bit SGPR values must be even-aligned (first SGPR address is even); 128-bit and wider SGPR values must be aligned to a multiple of 4. 0 s1 <p>See above.</p> 1 s2 <p>See above.</p> 2 s3 <p>See above.</p> 3 s4 <p>See above.</p> 4 s5 <p>See above.</p> 5 s6 <p>See above.</p> 6 s7 <p>See above.</p> 7 s8 <p>See above.</p> 8 s9 <p>See above.</p> 9 s10 <p>See above.</p> 10 s11 <p>See above.</p> 11 s12 <p>See above.</p> 12 s13 <p>See above.</p> 13 s14 <p>See above.</p> 14 s15 <p>See above.</p> 15 s16 <p>See above.</p> 16 s17 <p>See above.</p> 17 s18 <p>See above.</p> 18 s19 <p>See above.</p> 19 s20 <p>See above.</p> 20 s21 <p>See above.</p> 21 s22 <p>See above.</p> 22 s23 <p>See above.</p> 23 s24 <p>See above.</p> 24 s25 <p>See above.</p> 25 s26 <p>See above.</p> 26 s27 <p>See above.</p> 27 s28 <p>See above.</p> 28 s29 <p>See above.</p> 29 s30 <p>See above.</p> 30 s31 <p>See above.</p> 31 s32 <p>See above.</p> 32 s33 <p>See above.</p> 33 s34 <p>See above.</p> 34 s35 <p>See above.</p> 35 s36 <p>See above.</p> 36 s37 <p>See above.</p> 37 s38 <p>See above.</p> 38 s39 <p>See above.</p> 39 s40 <p>See above.</p> 40 s41 <p>See above.</p> 41 s42 <p>See above.</p> 42 s43 <p>See above.</p> 43 s44 <p>See above.</p> 44 s45 <p>See above.</p> 45 s46 <p>See above.</p> 46 s47 <p>See above.</p> 47 s48 <p>See above.</p> 48 s49 <p>See above.</p> 49 s50 <p>See above.</p> 50 s51 <p>See above.</p> 51 s52 <p>See above.</p> 52 s53 <p>See above.</p> 53 s54 <p>See above.</p> 54 s55 <p>See above.</p> 55 s56 <p>See above.</p> 56 s57 <p>See above.</p> 57 s58 <p>See above.</p> 58 s59 <p>See above.</p> 59 s60 <p>See above.</p> 60 s61 <p>See above.</p> 61 s62 <p>See above.</p> 62 s63 <p>See above.</p> 63 s64 <p>See above.</p> 64 s65 <p>See above.</p> 65 s66 <p>See above.</p> 66 s67 <p>See above.</p> 67 s68 <p>See above.</p> 68 s69 <p>See above.</p> 69 s70 <p>See above.</p> 70 s71 <p>See above.</p> 71 s72 <p>See above.</p> 72 s73 <p>See above.</p> 73 s74 <p>See above.</p> 74 s75 <p>See above.</p> 75 s76 <p>See above.</p> 76 s77 <p>See above.</p> 77 s78 <p>See above.</p> 78 s79 <p>See above.</p> 79 s80 <p>See above.</p> 80 s81 <p>See above.</p> 81 s82 <p>See above.</p> 82 s83 <p>See above.</p> 83 s84 <p>See above.</p> 84 s85 <p>See above.</p> 85 s86 <p>See above.</p> 86 s87 <p>See above.</p> 87 s88 <p>See above.</p> 88 s89 <p>See above.</p> 89 s90 <p>See above.</p> 90 s91 <p>See above.</p> 91 s92 <p>See above.</p> 92 s93 <p>See above.</p> 93 s94 <p>See above.</p> 94 s95 <p>See above.</p> 95 s96 <p>See above.</p> 96 s97 <p>See above.</p> 97 s98 <p>See above.</p> 98 s99 <p>See above.</p> 99 s100 <p>See above.</p> 100 s101 <p>See above.</p> 101 s102 <p>See above.</p> 102 s103 <p>See above.</p> 103 s104 <p>See above.</p> 104 s105 <p>See above.</p> 105 ttmp0 Trap handler temps (privileged). Increment from here for additional TTMPs. There are NUM_TTMP TTMPs in total. {TTMP1,TTMP0} = PC_save{hi,lo}. Alignment restriction: unless otherwise noted, 64-bit SGPR values must be even-aligned (first SGPR address is even); 128-bit and wider SGPR values must be aligned to a multiple of 4. 108 ttmp1 Trap handler temps (privileged). 109 ttmp2 Trap handler temps (privileged). 110 ttmp3 Trap handler temps (privileged). 111 ttmp4 Trap handler temps (privileged). 112 ttmp5 Trap handler temps (privileged). 113 ttmp6 Trap handler temps (privileged). 114 ttmp7 Trap handler temps (privileged). 115 ttmp8 Trap handler temps (privileged). 116 ttmp9 Trap handler temps (privileged). 117 ttmp10 Trap handler temps (privileged). 118 ttmp11 Trap handler temps (privileged). 119 ttmp12 Trap handler temps (privileged). 120 ttmp13 Trap handler temps (privileged). 121 ttmp14 Trap handler temps (privileged). 122 ttmp15 Trap handler temps (privileged). 123 vcc_lo VCC[31:0] 106 vcc_hi VCC[63:32] 107 null Do not write destination. For scalar memory, do not use SOFFSET in address calculation. For scratch addressing, do not use an SGPR in the addressing calculation. Otherwise read back zero. 124 m0 Special register used to hold DS addresses, relative indices, and send-message values. 125 OPR_SSRC All scalar operands. Covers all operands that are allowed as scalar sources. OPR_SDST OPR_SDST_EXEC OPR_SDST_M0 OPR_SREG OPR_SSRC_SPECIAL_SCC s0 Scalar GPR 0. Increment from here for additional GPRs. There are NUM_SGPR SGPRs in total. Alignment restriction: unless otherwise noted, 64-bit SGPR values must be even-aligned (first SGPR address is even); 128-bit and wider SGPR values must be aligned to a multiple of 4. 0 s1 <p>See above.</p> 1 s2 <p>See above.</p> 2 s3 <p>See above.</p> 3 s4 <p>See above.</p> 4 s5 <p>See above.</p> 5 s6 <p>See above.</p> 6 s7 <p>See above.</p> 7 s8 <p>See above.</p> 8 s9 <p>See above.</p> 9 s10 <p>See above.</p> 10 s11 <p>See above.</p> 11 s12 <p>See above.</p> 12 s13 <p>See above.</p> 13 s14 <p>See above.</p> 14 s15 <p>See above.</p> 15 s16 <p>See above.</p> 16 s17 <p>See above.</p> 17 s18 <p>See above.</p> 18 s19 <p>See above.</p> 19 s20 <p>See above.</p> 20 s21 <p>See above.</p> 21 s22 <p>See above.</p> 22 s23 <p>See above.</p> 23 s24 <p>See above.</p> 24 s25 <p>See above.</p> 25 s26 <p>See above.</p> 26 s27 <p>See above.</p> 27 s28 <p>See above.</p> 28 s29 <p>See above.</p> 29 s30 <p>See above.</p> 30 s31 <p>See above.</p> 31 s32 <p>See above.</p> 32 s33 <p>See above.</p> 33 s34 <p>See above.</p> 34 s35 <p>See above.</p> 35 s36 <p>See above.</p> 36 s37 <p>See above.</p> 37 s38 <p>See above.</p> 38 s39 <p>See above.</p> 39 s40 <p>See above.</p> 40 s41 <p>See above.</p> 41 s42 <p>See above.</p> 42 s43 <p>See above.</p> 43 s44 <p>See above.</p> 44 s45 <p>See above.</p> 45 s46 <p>See above.</p> 46 s47 <p>See above.</p> 47 s48 <p>See above.</p> 48 s49 <p>See above.</p> 49 s50 <p>See above.</p> 50 s51 <p>See above.</p> 51 s52 <p>See above.</p> 52 s53 <p>See above.</p> 53 s54 <p>See above.</p> 54 s55 <p>See above.</p> 55 s56 <p>See above.</p> 56 s57 <p>See above.</p> 57 s58 <p>See above.</p> 58 s59 <p>See above.</p> 59 s60 <p>See above.</p> 60 s61 <p>See above.</p> 61 s62 <p>See above.</p> 62 s63 <p>See above.</p> 63 s64 <p>See above.</p> 64 s65 <p>See above.</p> 65 s66 <p>See above.</p> 66 s67 <p>See above.</p> 67 s68 <p>See above.</p> 68 s69 <p>See above.</p> 69 s70 <p>See above.</p> 70 s71 <p>See above.</p> 71 s72 <p>See above.</p> 72 s73 <p>See above.</p> 73 s74 <p>See above.</p> 74 s75 <p>See above.</p> 75 s76 <p>See above.</p> 76 s77 <p>See above.</p> 77 s78 <p>See above.</p> 78 s79 <p>See above.</p> 79 s80 <p>See above.</p> 80 s81 <p>See above.</p> 81 s82 <p>See above.</p> 82 s83 <p>See above.</p> 83 s84 <p>See above.</p> 84 s85 <p>See above.</p> 85 s86 <p>See above.</p> 86 s87 <p>See above.</p> 87 s88 <p>See above.</p> 88 s89 <p>See above.</p> 89 s90 <p>See above.</p> 90 s91 <p>See above.</p> 91 s92 <p>See above.</p> 92 s93 <p>See above.</p> 93 s94 <p>See above.</p> 94 s95 <p>See above.</p> 95 s96 <p>See above.</p> 96 s97 <p>See above.</p> 97 s98 <p>See above.</p> 98 s99 <p>See above.</p> 99 s100 <p>See above.</p> 100 s101 <p>See above.</p> 101 s102 <p>See above.</p> 102 s103 <p>See above.</p> 103 s104 <p>See above.</p> 104 s105 <p>See above.</p> 105 ttmp0 Trap handler temps (privileged). Increment from here for additional TTMPs. There are NUM_TTMP TTMPs in total. {TTMP1,TTMP0} = PC_save{hi,lo}. Alignment restriction: unless otherwise noted, 64-bit SGPR values must be even-aligned (first SGPR address is even); 128-bit and wider SGPR values must be aligned to a multiple of 4. 108 ttmp1 Trap handler temps (privileged). 109 ttmp2 Trap handler temps (privileged). 110 ttmp3 Trap handler temps (privileged). 111 ttmp4 Trap handler temps (privileged). 112 ttmp5 Trap handler temps (privileged). 113 ttmp6 Trap handler temps (privileged). 114 ttmp7 Trap handler temps (privileged). 115 ttmp8 Trap handler temps (privileged). 116 ttmp9 Trap handler temps (privileged). 117 ttmp10 Trap handler temps (privileged). 118 ttmp11 Trap handler temps (privileged). 119 ttmp12 Trap handler temps (privileged). 120 ttmp13 Trap handler temps (privileged). 121 ttmp14 Trap handler temps (privileged). 122 ttmp15 Trap handler temps (privileged). 123 vcc_lo VCC[31:0] 106 vcc_hi VCC[63:32] 107 null Do not write destination. For scalar memory, do not use SOFFSET in address calculation. For scratch addressing, do not use an SGPR in the addressing calculation. Otherwise read back zero. 124 m0 Special register used to hold DS addresses, relative indices, and send-message values. 125 exec_lo exec[31:0] 126 exec_hi exec[63:32] 127 0 0 128 1 1 (integer) 129 2 2 (integer) 130 3 3 (integer) 131 4 4 (integer) 132 5 5 (integer) 133 6 6 (integer) 134 7 7 (integer) 135 8 8 (integer) 136 9 9 (integer) 137 10 10 (integer) 138 11 11 (integer) 139 12 12 (integer) 140 13 13 (integer) 141 14 14 (integer) 142 15 15 (integer) 143 16 16 (integer) 144 17 17 (integer) 145 18 18 (integer) 146 19 19 (integer) 147 20 20 (integer) 148 21 21 (integer) 149 22 22 (integer) 150 23 23 (integer) 151 24 24 (integer) 152 25 25 (integer) 153 26 26 (integer) 154 27 27 (integer) 155 28 28 (integer) 156 29 29 (integer) 157 30 30 (integer) 158 31 31 (integer) 159 32 32 (integer) 160 33 33 (integer) 161 34 34 (integer) 162 35 35 (integer) 163 36 36 (integer) 164 37 37 (integer) 165 38 38 (integer) 166 39 39 (integer) 167 40 40 (integer) 168 41 41 (integer) 169 42 42 (integer) 170 43 43 (integer) 171 44 44 (integer) 172 45 45 (integer) 173 46 46 (integer) 174 47 47 (integer) 175 48 48 (integer) 176 49 49 (integer) 177 50 50 (integer) 178 51 51 (integer) 179 52 52 (integer) 180 53 53 (integer) 181 54 54 (integer) 182 55 55 (integer) 183 56 56 (integer) 184 57 57 (integer) 185 58 58 (integer) 186 59 59 (integer) 187 60 60 (integer) 188 61 61 (integer) 189 62 62 (integer) 190 63 63 (integer) 191 64 64 (integer) 192 -1 -1 (integer) 193 -2 -2 (integer) 194 -3 -3 (integer) 195 -4 -4 (integer) 196 -5 -5 (integer) 197 -6 -6 (integer) 198 -7 -7 (integer) 199 -8 -8 (integer) 200 -9 -9 (integer) 201 -10 -10 (integer) 202 -11 -11 (integer) 203 -12 -12 (integer) 204 -13 -13 (integer) 205 -14 -14 (integer) 206 -15 -15 (integer) 207 -16 -16 (integer) 208 0.5 0.5 240 -0.5 -0.5 241 1.0 1.0 242 -1.0 -1.0 243 2.0 2.0 244 -2.0 -2.0 245 4.0 4.0 246 -4.0 -4.0 247 0.15915494 1 / (2 * PI). For F16 and PK2_F16 formats: 0x3118. For BF16 and PK2_BF16 formats: 0x3e22 (truncated from 32-bit format with no rounding). For all 64-bit formats (F64, B64, I64 etc.): 0x3fc45f30_6dc9c882. For all other cases (including 32-bit, B16, I16): 0x3e22f983. 248 src_scc Scalar condition code. 253 src_shared_base 235 src_shared_limit 236 src_private_base 237 src_private_limit 238 src_literal 32-bit literal constant follows this instruction. 16-bit operands use only the 16 LSBs of this constant. 64-bit unsigned integer operands (U64, B64) zero-extend the constant. 64-bit signed integer operands (I64) sign-extend the constant. 64-bit floating-point operands use the constant as the high 32 MSBs of a double-precision floating point value (1 sign bit, 11 exponent bits and 20 mantissa bits). 255 OPR_SSRC_BARRIER_ID OPR_SDST_M0 -1 -1 (integer) 193 -2 -2 (integer) 194 0 0 128 1 1 (integer) 129 2 2 (integer) 130 3 3 (integer) 131 4 4 (integer) 132 5 5 (integer) 133 6 6 (integer) 134 7 7 (integer) 135 8 8 (integer) 136 9 9 (integer) 137 10 10 (integer) 138 11 11 (integer) 139 12 12 (integer) 140 13 13 (integer) 141 14 14 (integer) 142 15 15 (integer) 143 16 16 (integer) 144 17 17 (integer) 145 18 18 (integer) 146 19 19 (integer) 147 20 20 (integer) 148 21 21 (integer) 149 22 22 (integer) 150 23 23 (integer) 151 24 24 (integer) 152 25 25 (integer) 153 26 26 (integer) 154 27 27 (integer) 155 28 28 (integer) 156 29 29 (integer) 157 30 30 (integer) 158 31 31 (integer) 159 m0 Special register used to hold DS addresses, relative indices, and send-message values. 125 OPR_SSRC_LANESEL Operand is legal for a lane-select operand, generally scalar integer values in the range 0..63. OPR_SDST_M0 OPR_SREG s0 Scalar GPR 0. Increment from here for additional GPRs. There are NUM_SGPR SGPRs in total. Alignment restriction: unless otherwise noted, 64-bit SGPR values must be even-aligned (first SGPR address is even); 128-bit and wider SGPR values must be aligned to a multiple of 4. 0 s1 <p>See above.</p> 1 s2 <p>See above.</p> 2 s3 <p>See above.</p> 3 s4 <p>See above.</p> 4 s5 <p>See above.</p> 5 s6 <p>See above.</p> 6 s7 <p>See above.</p> 7 s8 <p>See above.</p> 8 s9 <p>See above.</p> 9 s10 <p>See above.</p> 10 s11 <p>See above.</p> 11 s12 <p>See above.</p> 12 s13 <p>See above.</p> 13 s14 <p>See above.</p> 14 s15 <p>See above.</p> 15 s16 <p>See above.</p> 16 s17 <p>See above.</p> 17 s18 <p>See above.</p> 18 s19 <p>See above.</p> 19 s20 <p>See above.</p> 20 s21 <p>See above.</p> 21 s22 <p>See above.</p> 22 s23 <p>See above.</p> 23 s24 <p>See above.</p> 24 s25 <p>See above.</p> 25 s26 <p>See above.</p> 26 s27 <p>See above.</p> 27 s28 <p>See above.</p> 28 s29 <p>See above.</p> 29 s30 <p>See above.</p> 30 s31 <p>See above.</p> 31 s32 <p>See above.</p> 32 s33 <p>See above.</p> 33 s34 <p>See above.</p> 34 s35 <p>See above.</p> 35 s36 <p>See above.</p> 36 s37 <p>See above.</p> 37 s38 <p>See above.</p> 38 s39 <p>See above.</p> 39 s40 <p>See above.</p> 40 s41 <p>See above.</p> 41 s42 <p>See above.</p> 42 s43 <p>See above.</p> 43 s44 <p>See above.</p> 44 s45 <p>See above.</p> 45 s46 <p>See above.</p> 46 s47 <p>See above.</p> 47 s48 <p>See above.</p> 48 s49 <p>See above.</p> 49 s50 <p>See above.</p> 50 s51 <p>See above.</p> 51 s52 <p>See above.</p> 52 s53 <p>See above.</p> 53 s54 <p>See above.</p> 54 s55 <p>See above.</p> 55 s56 <p>See above.</p> 56 s57 <p>See above.</p> 57 s58 <p>See above.</p> 58 s59 <p>See above.</p> 59 s60 <p>See above.</p> 60 s61 <p>See above.</p> 61 s62 <p>See above.</p> 62 s63 <p>See above.</p> 63 s64 <p>See above.</p> 64 s65 <p>See above.</p> 65 s66 <p>See above.</p> 66 s67 <p>See above.</p> 67 s68 <p>See above.</p> 68 s69 <p>See above.</p> 69 s70 <p>See above.</p> 70 s71 <p>See above.</p> 71 s72 <p>See above.</p> 72 s73 <p>See above.</p> 73 s74 <p>See above.</p> 74 s75 <p>See above.</p> 75 s76 <p>See above.</p> 76 s77 <p>See above.</p> 77 s78 <p>See above.</p> 78 s79 <p>See above.</p> 79 s80 <p>See above.</p> 80 s81 <p>See above.</p> 81 s82 <p>See above.</p> 82 s83 <p>See above.</p> 83 s84 <p>See above.</p> 84 s85 <p>See above.</p> 85 s86 <p>See above.</p> 86 s87 <p>See above.</p> 87 s88 <p>See above.</p> 88 s89 <p>See above.</p> 89 s90 <p>See above.</p> 90 s91 <p>See above.</p> 91 s92 <p>See above.</p> 92 s93 <p>See above.</p> 93 s94 <p>See above.</p> 94 s95 <p>See above.</p> 95 s96 <p>See above.</p> 96 s97 <p>See above.</p> 97 s98 <p>See above.</p> 98 s99 <p>See above.</p> 99 s100 <p>See above.</p> 100 s101 <p>See above.</p> 101 s102 <p>See above.</p> 102 s103 <p>See above.</p> 103 s104 <p>See above.</p> 104 s105 <p>See above.</p> 105 ttmp0 Trap handler temps (privileged). Increment from here for additional TTMPs. There are NUM_TTMP TTMPs in total. {TTMP1,TTMP0} = PC_save{hi,lo}. Alignment restriction: unless otherwise noted, 64-bit SGPR values must be even-aligned (first SGPR address is even); 128-bit and wider SGPR values must be aligned to a multiple of 4. 108 ttmp1 Trap handler temps (privileged). 109 ttmp2 Trap handler temps (privileged). 110 ttmp3 Trap handler temps (privileged). 111 ttmp4 Trap handler temps (privileged). 112 ttmp5 Trap handler temps (privileged). 113 ttmp6 Trap handler temps (privileged). 114 ttmp7 Trap handler temps (privileged). 115 ttmp8 Trap handler temps (privileged). 116 ttmp9 Trap handler temps (privileged). 117 ttmp10 Trap handler temps (privileged). 118 ttmp11 Trap handler temps (privileged). 119 ttmp12 Trap handler temps (privileged). 120 ttmp13 Trap handler temps (privileged). 121 ttmp14 Trap handler temps (privileged). 122 ttmp15 Trap handler temps (privileged). 123 vcc_lo VCC[31:0] 106 vcc_hi VCC[63:32] 107 null Do not write destination. For scalar memory, do not use SOFFSET in address calculation. For scratch addressing, do not use an SGPR in the addressing calculation. Otherwise read back zero. 124 m0 Special register used to hold DS addresses, relative indices, and send-message values. 125 0 0 128 1 1 (integer) 129 2 2 (integer) 130 3 3 (integer) 131 4 4 (integer) 132 5 5 (integer) 133 6 6 (integer) 134 7 7 (integer) 135 8 8 (integer) 136 9 9 (integer) 137 10 10 (integer) 138 11 11 (integer) 139 12 12 (integer) 140 13 13 (integer) 141 14 14 (integer) 142 15 15 (integer) 143 16 16 (integer) 144 17 17 (integer) 145 18 18 (integer) 146 19 19 (integer) 147 20 20 (integer) 148 21 21 (integer) 149 22 22 (integer) 150 23 23 (integer) 151 24 24 (integer) 152 25 25 (integer) 153 26 26 (integer) 154 27 27 (integer) 155 28 28 (integer) 156 29 29 (integer) 157 30 30 (integer) 158 31 31 (integer) 159 32 32 (integer) 160 33 33 (integer) 161 34 34 (integer) 162 35 35 (integer) 163 36 36 (integer) 164 37 37 (integer) 165 38 38 (integer) 166 39 39 (integer) 167 40 40 (integer) 168 41 41 (integer) 169 42 42 (integer) 170 43 43 (integer) 171 44 44 (integer) 172 45 45 (integer) 173 46 46 (integer) 174 47 47 (integer) 175 48 48 (integer) 176 49 49 (integer) 177 50 50 (integer) 178 51 51 (integer) 179 52 52 (integer) 180 53 53 (integer) 181 54 54 (integer) 182 55 55 (integer) 183 56 56 (integer) 184 57 57 (integer) 185 58 58 (integer) 186 59 59 (integer) 187 60 60 (integer) 188 61 61 (integer) 189 62 62 (integer) 190 63 63 (integer) 191 OPR_SSRC_SPECIAL_SCC src_scc Scalar condition code. 253 OPR_TGT Target to export shader data to. mrt0 Output to colour MRT 0. Increment from here for additional MRTs. There are EXP_NUM_MRT MRTs in total. 0 mrt1 <p>See above.</p> 1 mrt2 <p>See above.</p> 2 mrt3 <p>See above.</p> 3 mrt4 <p>See above.</p> 4 mrt5 <p>See above.</p> 5 mrt6 <p>See above.</p> 6 mrt7 <p>See above.</p> 7 mrtz Output to Z. 8 pos0 Output to position 0. Increment from here for additional positions. There are EXP_NUM_POS positions in total. POS4 (the last one) is for stereo rendering. 12 pos1 <p>See above.</p> 13 pos2 <p>See above.</p> 14 pos3 <p>See above.</p> 15 pos4 <p>See above.</p> 16 prim Output NGG primitive (connectivity) data. 20 dual_src_blend0 Dual source blend left/right. There are EXP_NUM_DUAL_SRC_BLEND exports in total. 21 dual_src_blend1 <p>See above.</p> 22 OPR_VCC 64-bit vector condition code. vcc VCC[63:0] 106 OPR_VERSION Microcode version header. MDP 1 15 VERSION Version number. 8 0 uc_version_gfx7 Baseline microcode for GFX7.0. 0 uc_version_gfx10 Baseline microcode for GFX10.0. 4 uc_version_gfx11 Baseline microcode for GFX11. 6 uc_version_gfx12 Baseline microcode for GFX12. 9 W32 1 14 W64 1 13 OPR_VGPR Operand must be a vector GPR. Uses an 8-bit operand field. v0 Vector GPR 0. Increment from here for additional GPRs. There are NUM_VGPR VGPRs in total. 0 v1 <p>See above.</p> 1 v2 <p>See above.</p> 2 v3 <p>See above.</p> 3 v4 <p>See above.</p> 4 v5 <p>See above.</p> 5 v6 <p>See above.</p> 6 v7 <p>See above.</p> 7 v8 <p>See above.</p> 8 v9 <p>See above.</p> 9 v10 <p>See above.</p> 10 v11 <p>See above.</p> 11 v12 <p>See above.</p> 12 v13 <p>See above.</p> 13 v14 <p>See above.</p> 14 v15 <p>See above.</p> 15 v16 <p>See above.</p> 16 v17 <p>See above.</p> 17 v18 <p>See above.</p> 18 v19 <p>See above.</p> 19 v20 <p>See above.</p> 20 v21 <p>See above.</p> 21 v22 <p>See above.</p> 22 v23 <p>See above.</p> 23 v24 <p>See above.</p> 24 v25 <p>See above.</p> 25 v26 <p>See above.</p> 26 v27 <p>See above.</p> 27 v28 <p>See above.</p> 28 v29 <p>See above.</p> 29 v30 <p>See above.</p> 30 v31 <p>See above.</p> 31 v32 <p>See above.</p> 32 v33 <p>See above.</p> 33 v34 <p>See above.</p> 34 v35 <p>See above.</p> 35 v36 <p>See above.</p> 36 v37 <p>See above.</p> 37 v38 <p>See above.</p> 38 v39 <p>See above.</p> 39 v40 <p>See above.</p> 40 v41 <p>See above.</p> 41 v42 <p>See above.</p> 42 v43 <p>See above.</p> 43 v44 <p>See above.</p> 44 v45 <p>See above.</p> 45 v46 <p>See above.</p> 46 v47 <p>See above.</p> 47 v48 <p>See above.</p> 48 v49 <p>See above.</p> 49 v50 <p>See above.</p> 50 v51 <p>See above.</p> 51 v52 <p>See above.</p> 52 v53 <p>See above.</p> 53 v54 <p>See above.</p> 54 v55 <p>See above.</p> 55 v56 <p>See above.</p> 56 v57 <p>See above.</p> 57 v58 <p>See above.</p> 58 v59 <p>See above.</p> 59 v60 <p>See above.</p> 60 v61 <p>See above.</p> 61 v62 <p>See above.</p> 62 v63 <p>See above.</p> 63 v64 <p>See above.</p> 64 v65 <p>See above.</p> 65 v66 <p>See above.</p> 66 v67 <p>See above.</p> 67 v68 <p>See above.</p> 68 v69 <p>See above.</p> 69 v70 <p>See above.</p> 70 v71 <p>See above.</p> 71 v72 <p>See above.</p> 72 v73 <p>See above.</p> 73 v74 <p>See above.</p> 74 v75 <p>See above.</p> 75 v76 <p>See above.</p> 76 v77 <p>See above.</p> 77 v78 <p>See above.</p> 78 v79 <p>See above.</p> 79 v80 <p>See above.</p> 80 v81 <p>See above.</p> 81 v82 <p>See above.</p> 82 v83 <p>See above.</p> 83 v84 <p>See above.</p> 84 v85 <p>See above.</p> 85 v86 <p>See above.</p> 86 v87 <p>See above.</p> 87 v88 <p>See above.</p> 88 v89 <p>See above.</p> 89 v90 <p>See above.</p> 90 v91 <p>See above.</p> 91 v92 <p>See above.</p> 92 v93 <p>See above.</p> 93 v94 <p>See above.</p> 94 v95 <p>See above.</p> 95 v96 <p>See above.</p> 96 v97 <p>See above.</p> 97 v98 <p>See above.</p> 98 v99 <p>See above.</p> 99 v100 <p>See above.</p> 100 v101 <p>See above.</p> 101 v102 <p>See above.</p> 102 v103 <p>See above.</p> 103 v104 <p>See above.</p> 104 v105 <p>See above.</p> 105 v106 <p>See above.</p> 106 v107 <p>See above.</p> 107 v108 <p>See above.</p> 108 v109 <p>See above.</p> 109 v110 <p>See above.</p> 110 v111 <p>See above.</p> 111 v112 <p>See above.</p> 112 v113 <p>See above.</p> 113 v114 <p>See above.</p> 114 v115 <p>See above.</p> 115 v116 <p>See above.</p> 116 v117 <p>See above.</p> 117 v118 <p>See above.</p> 118 v119 <p>See above.</p> 119 v120 <p>See above.</p> 120 v121 <p>See above.</p> 121 v122 <p>See above.</p> 122 v123 <p>See above.</p> 123 v124 <p>See above.</p> 124 v125 <p>See above.</p> 125 v126 <p>See above.</p> 126 v127 <p>See above.</p> 127 v128 <p>See above.</p> 128 v129 <p>See above.</p> 129 v130 <p>See above.</p> 130 v131 <p>See above.</p> 131 v132 <p>See above.</p> 132 v133 <p>See above.</p> 133 v134 <p>See above.</p> 134 v135 <p>See above.</p> 135 v136 <p>See above.</p> 136 v137 <p>See above.</p> 137 v138 <p>See above.</p> 138 v139 <p>See above.</p> 139 v140 <p>See above.</p> 140 v141 <p>See above.</p> 141 v142 <p>See above.</p> 142 v143 <p>See above.</p> 143 v144 <p>See above.</p> 144 v145 <p>See above.</p> 145 v146 <p>See above.</p> 146 v147 <p>See above.</p> 147 v148 <p>See above.</p> 148 v149 <p>See above.</p> 149 v150 <p>See above.</p> 150 v151 <p>See above.</p> 151 v152 <p>See above.</p> 152 v153 <p>See above.</p> 153 v154 <p>See above.</p> 154 v155 <p>See above.</p> 155 v156 <p>See above.</p> 156 v157 <p>See above.</p> 157 v158 <p>See above.</p> 158 v159 <p>See above.</p> 159 v160 <p>See above.</p> 160 v161 <p>See above.</p> 161 v162 <p>See above.</p> 162 v163 <p>See above.</p> 163 v164 <p>See above.</p> 164 v165 <p>See above.</p> 165 v166 <p>See above.</p> 166 v167 <p>See above.</p> 167 v168 <p>See above.</p> 168 v169 <p>See above.</p> 169 v170 <p>See above.</p> 170 v171 <p>See above.</p> 171 v172 <p>See above.</p> 172 v173 <p>See above.</p> 173 v174 <p>See above.</p> 174 v175 <p>See above.</p> 175 v176 <p>See above.</p> 176 v177 <p>See above.</p> 177 v178 <p>See above.</p> 178 v179 <p>See above.</p> 179 v180 <p>See above.</p> 180 v181 <p>See above.</p> 181 v182 <p>See above.</p> 182 v183 <p>See above.</p> 183 v184 <p>See above.</p> 184 v185 <p>See above.</p> 185 v186 <p>See above.</p> 186 v187 <p>See above.</p> 187 v188 <p>See above.</p> 188 v189 <p>See above.</p> 189 v190 <p>See above.</p> 190 v191 <p>See above.</p> 191 v192 <p>See above.</p> 192 v193 <p>See above.</p> 193 v194 <p>See above.</p> 194 v195 <p>See above.</p> 195 v196 <p>See above.</p> 196 v197 <p>See above.</p> 197 v198 <p>See above.</p> 198 v199 <p>See above.</p> 199 v200 <p>See above.</p> 200 v201 <p>See above.</p> 201 v202 <p>See above.</p> 202 v203 <p>See above.</p> 203 v204 <p>See above.</p> 204 v205 <p>See above.</p> 205 v206 <p>See above.</p> 206 v207 <p>See above.</p> 207 v208 <p>See above.</p> 208 v209 <p>See above.</p> 209 v210 <p>See above.</p> 210 v211 <p>See above.</p> 211 v212 <p>See above.</p> 212 v213 <p>See above.</p> 213 v214 <p>See above.</p> 214 v215 <p>See above.</p> 215 v216 <p>See above.</p> 216 v217 <p>See above.</p> 217 v218 <p>See above.</p> 218 v219 <p>See above.</p> 219 v220 <p>See above.</p> 220 v221 <p>See above.</p> 221 v222 <p>See above.</p> 222 v223 <p>See above.</p> 223 v224 <p>See above.</p> 224 v225 <p>See above.</p> 225 v226 <p>See above.</p> 226 v227 <p>See above.</p> 227 v228 <p>See above.</p> 228 v229 <p>See above.</p> 229 v230 <p>See above.</p> 230 v231 <p>See above.</p> 231 v232 <p>See above.</p> 232 v233 <p>See above.</p> 233 v234 <p>See above.</p> 234 v235 <p>See above.</p> 235 v236 <p>See above.</p> 236 v237 <p>See above.</p> 237 v238 <p>See above.</p> 238 v239 <p>See above.</p> 239 v240 <p>See above.</p> 240 v241 <p>See above.</p> 241 v242 <p>See above.</p> 242 v243 <p>See above.</p> 243 v244 <p>See above.</p> 244 v245 <p>See above.</p> 245 v246 <p>See above.</p> 246 v247 <p>See above.</p> 247 v248 <p>See above.</p> 248 v249 <p>See above.</p> 249 v250 <p>See above.</p> 250 v251 <p>See above.</p> 251 v252 <p>See above.</p> 252 v253 <p>See above.</p> 253 v254 <p>See above.</p> 254 v255 <p>See above.</p> 255 OPR_WAITCNT A waitcnt operand specifying expcnt, lgkmcnt and vmcnt. EXP Export wait count. 0x7 means do not wait on EXPCNT. 3 0 LGKM LGKM wait count. 0x3f means do not wait on LGKMCNT. 6 4 VM VM wait count. 0x3f means do not wait on VMCNT. 6 10 OPR_WAIT_ALU A waitcnt operand for ALU and miscellaneous dependency counters. HOLD_CNT Wait for HOLD_CNT <= N (N = 1 disables check). 1 7 SA_SDST Wait for SA_SDST <= N (N = 1 disables check). 1 0 VA_SDST Wait for VA_SDST <= N (N = 7 disables check). 3 9 VA_SSRC Wait for VA_SSRC <= N (N = 1 disables check). 1 8 VA_VCC Wait for VA_VCC <= N (N = 1 disables check). 1 1 VA_VDST Wait for VA_VDST <= N (N = 15 disables check). 4 12 VM_VSRC Wait for VM_VSRC <= N (N = 7 disables check). 3 2 OPR_WAIT_EVENT Specify one or more events to wait for. EXPORT_READY If this value is ONE then sleep until the export_ready bit is 1. If the export_ready bit is already 1, no sleep occurs. Effect is the same as the export_ready check performed before issuing an export instruction. No wait occurs if this value is ZERO. 1 1 OPR_WAIT_MEM_DS A waitcnt operand for combined memory (LOAD or STORE) and DS counter. DS Wait for DSCNT <= N. 6 0 MEM Wait for either LOADCNT <= N or STORECNT <= N (depending on instruction). 6 8 SALU Scalar ALU (SALU) instructions operate on values that are common to all work-items in the wave. These operations consist of 32-bit integer or floating-point arithmetic, and 32 or 64-bit bitwise operations, and they store results in scalar registers. SMEM Scalar Memory (SMEM) instructions allow a shader program to load data from memory into scalar registers, making the data accessible to all work-items in the wave. VALU Vector ALU (VALU) instructions perform arithmetic or logical operations on data for each of the wave's work-item and write results back to vector registers, scalar registers or the EXEC mask. VMEM Vector-memory (VMEM) instructions transfer data to and from vector registers, making the data uniquely available for each work-item in the wave. EXPORT Export instructions copy data from a VGPR to one of the export buffers (position, color, or Z). BRANCH Branch instructions allow programs to jump - direct or indirect - to an instruction to be executed next, conditionally or unconditionally. MESSAGE Message instructions send a message to fixed function hardware or the host and can be used to request that a value be returned to the wave. WAVE_CONTROL Wave control instructions control a wave's execution these include dependency, delay, scheduling, and wave termination operations. TRAP Trap instructions transition into and out of the wave's trap handler FLOATING_POINT BUFFER TEXTURE LOAD STORE SAMPLE BVH ATOMIC FLAT DATA_SHARE STATIC MFMA WMMA TRANSCENDENTAL