Files
derek.holloway 7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
CodeQL / Analyze (python) (push) Canceled after 0s
pre-commit / pre-commit (push) Canceled after 0s
Multi-Arch CI / setup (push) Canceled after 0s
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Setup (Multi-Arch) / setup (push) Canceled after 0s
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Multi-Arch CI / CI Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
Flatten subprojects into a single base monorepo
2026-09-02 16:34:47 -07:00

320 lines
9.1 KiB
LLVM

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5
; RUN: llc -mtriple=riscv32 -global-isel -global-isel-abort=1 < %s 2>&1 | FileCheck %s --check-prefixes=RV32
; RUN: llc -mtriple=riscv64 -global-isel -global-isel-abort=1 < %s 2>&1 | FileCheck %s --check-prefixes=RV64
define i2 @bitreverse_i2(i2 %x) {
; RV32-LABEL: bitreverse_i2:
; RV32: # %bb.0:
; RV32-NEXT: slli a1, a0, 1
; RV32-NEXT: andi a0, a0, 3
; RV32-NEXT: andi a1, a1, 2
; RV32-NEXT: srli a0, a0, 1
; RV32-NEXT: or a0, a1, a0
; RV32-NEXT: ret
;
; RV64-LABEL: bitreverse_i2:
; RV64: # %bb.0:
; RV64-NEXT: slli a1, a0, 1
; RV64-NEXT: andi a0, a0, 3
; RV64-NEXT: andi a1, a1, 2
; RV64-NEXT: srli a0, a0, 1
; RV64-NEXT: or a0, a1, a0
; RV64-NEXT: ret
%rev = call i2 @llvm.bitreverse.i2(i2 %x)
ret i2 %rev
}
define i3 @bitreverse_i3(i3 %x) {
; RV32-LABEL: bitreverse_i3:
; RV32: # %bb.0:
; RV32-NEXT: slli a1, a0, 2
; RV32-NEXT: andi a0, a0, 7
; RV32-NEXT: andi a1, a1, 4
; RV32-NEXT: andi a2, a0, 2
; RV32-NEXT: or a1, a1, a2
; RV32-NEXT: srli a0, a0, 2
; RV32-NEXT: or a0, a1, a0
; RV32-NEXT: ret
;
; RV64-LABEL: bitreverse_i3:
; RV64: # %bb.0:
; RV64-NEXT: slli a1, a0, 2
; RV64-NEXT: andi a0, a0, 7
; RV64-NEXT: andi a1, a1, 4
; RV64-NEXT: andi a2, a0, 2
; RV64-NEXT: or a1, a1, a2
; RV64-NEXT: srli a0, a0, 2
; RV64-NEXT: or a0, a1, a0
; RV64-NEXT: ret
%rev = call i3 @llvm.bitreverse.i3(i3 %x)
ret i3 %rev
}
define i4 @bitreverse_i4(i4 %x) {
; RV32-LABEL: bitreverse_i4:
; RV32: # %bb.0:
; RV32-NEXT: slli a1, a0, 3
; RV32-NEXT: slli a2, a0, 1
; RV32-NEXT: andi a0, a0, 15
; RV32-NEXT: andi a1, a1, 8
; RV32-NEXT: andi a2, a2, 4
; RV32-NEXT: or a1, a1, a2
; RV32-NEXT: srli a2, a0, 1
; RV32-NEXT: andi a2, a2, 2
; RV32-NEXT: or a1, a1, a2
; RV32-NEXT: srli a0, a0, 3
; RV32-NEXT: or a0, a1, a0
; RV32-NEXT: ret
;
; RV64-LABEL: bitreverse_i4:
; RV64: # %bb.0:
; RV64-NEXT: slli a1, a0, 3
; RV64-NEXT: slli a2, a0, 1
; RV64-NEXT: andi a0, a0, 15
; RV64-NEXT: andi a1, a1, 8
; RV64-NEXT: andi a2, a2, 4
; RV64-NEXT: or a1, a1, a2
; RV64-NEXT: srli a2, a0, 1
; RV64-NEXT: andi a2, a2, 2
; RV64-NEXT: or a1, a1, a2
; RV64-NEXT: srli a0, a0, 3
; RV64-NEXT: or a0, a1, a0
; RV64-NEXT: ret
%rev = call i4 @llvm.bitreverse.i4(i4 %x)
ret i4 %rev
}
define i7 @bitreverse_i7(i7 %x) {
; RV32-LABEL: bitreverse_i7:
; RV32: # %bb.0:
; RV32-NEXT: slli a1, a0, 6
; RV32-NEXT: slli a2, a0, 4
; RV32-NEXT: slli a3, a0, 2
; RV32-NEXT: andi a0, a0, 127
; RV32-NEXT: andi a1, a1, 64
; RV32-NEXT: andi a2, a2, 32
; RV32-NEXT: andi a3, a3, 16
; RV32-NEXT: or a1, a1, a2
; RV32-NEXT: andi a2, a0, 8
; RV32-NEXT: or a2, a3, a2
; RV32-NEXT: srli a3, a0, 2
; RV32-NEXT: or a1, a1, a2
; RV32-NEXT: srli a2, a0, 4
; RV32-NEXT: andi a3, a3, 4
; RV32-NEXT: andi a2, a2, 2
; RV32-NEXT: or a2, a3, a2
; RV32-NEXT: or a1, a1, a2
; RV32-NEXT: srli a0, a0, 6
; RV32-NEXT: or a0, a1, a0
; RV32-NEXT: ret
;
; RV64-LABEL: bitreverse_i7:
; RV64: # %bb.0:
; RV64-NEXT: slli a1, a0, 6
; RV64-NEXT: slli a2, a0, 4
; RV64-NEXT: slli a3, a0, 2
; RV64-NEXT: andi a0, a0, 127
; RV64-NEXT: andi a1, a1, 64
; RV64-NEXT: andi a2, a2, 32
; RV64-NEXT: andi a3, a3, 16
; RV64-NEXT: or a1, a1, a2
; RV64-NEXT: andi a2, a0, 8
; RV64-NEXT: or a2, a3, a2
; RV64-NEXT: srli a3, a0, 2
; RV64-NEXT: or a1, a1, a2
; RV64-NEXT: srli a2, a0, 4
; RV64-NEXT: andi a3, a3, 4
; RV64-NEXT: andi a2, a2, 2
; RV64-NEXT: or a2, a3, a2
; RV64-NEXT: or a1, a1, a2
; RV64-NEXT: srli a0, a0, 6
; RV64-NEXT: or a0, a1, a0
; RV64-NEXT: ret
%rev = call i7 @llvm.bitreverse.i7(i7 %x)
ret i7 %rev
}
define i24 @bitreverse_i24(i24 %x) {
; RV32-LABEL: bitreverse_i24:
; RV32: # %bb.0:
; RV32-NEXT: slli a1, a0, 16
; RV32-NEXT: lui a2, 4096
; RV32-NEXT: lui a3, 1048335
; RV32-NEXT: addi a2, a2, -1
; RV32-NEXT: addi a3, a3, 240
; RV32-NEXT: and a0, a0, a2
; RV32-NEXT: srli a0, a0, 16
; RV32-NEXT: or a0, a0, a1
; RV32-NEXT: and a1, a3, a2
; RV32-NEXT: and a1, a0, a1
; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: and a0, a0, a3
; RV32-NEXT: lui a3, 1047757
; RV32-NEXT: addi a3, a3, -820
; RV32-NEXT: srli a1, a1, 4
; RV32-NEXT: or a0, a1, a0
; RV32-NEXT: and a1, a3, a2
; RV32-NEXT: and a1, a0, a1
; RV32-NEXT: slli a0, a0, 2
; RV32-NEXT: and a0, a0, a3
; RV32-NEXT: lui a3, 1047211
; RV32-NEXT: addi a3, a3, -1366
; RV32-NEXT: and a2, a3, a2
; RV32-NEXT: srli a1, a1, 2
; RV32-NEXT: or a0, a1, a0
; RV32-NEXT: and a2, a0, a2
; RV32-NEXT: slli a0, a0, 1
; RV32-NEXT: srli a2, a2, 1
; RV32-NEXT: and a0, a0, a3
; RV32-NEXT: or a0, a2, a0
; RV32-NEXT: ret
;
; RV64-LABEL: bitreverse_i24:
; RV64: # %bb.0:
; RV64-NEXT: slli a1, a0, 16
; RV64-NEXT: lui a2, 4096
; RV64-NEXT: lui a3, 1048335
; RV64-NEXT: addi a2, a2, -1
; RV64-NEXT: addi a3, a3, 240
; RV64-NEXT: and a0, a0, a2
; RV64-NEXT: srli a0, a0, 16
; RV64-NEXT: or a0, a0, a1
; RV64-NEXT: and a1, a3, a2
; RV64-NEXT: and a1, a0, a1
; RV64-NEXT: slli a0, a0, 4
; RV64-NEXT: and a0, a0, a3
; RV64-NEXT: lui a3, 1047757
; RV64-NEXT: addi a3, a3, -820
; RV64-NEXT: srli a1, a1, 4
; RV64-NEXT: or a0, a1, a0
; RV64-NEXT: and a1, a3, a2
; RV64-NEXT: and a1, a0, a1
; RV64-NEXT: slli a0, a0, 2
; RV64-NEXT: and a0, a0, a3
; RV64-NEXT: lui a3, 1047211
; RV64-NEXT: addi a3, a3, -1366
; RV64-NEXT: and a2, a3, a2
; RV64-NEXT: srli a1, a1, 2
; RV64-NEXT: or a0, a1, a0
; RV64-NEXT: and a2, a0, a2
; RV64-NEXT: slli a0, a0, 1
; RV64-NEXT: srli a2, a2, 1
; RV64-NEXT: and a0, a0, a3
; RV64-NEXT: or a0, a2, a0
; RV64-NEXT: ret
%rev = call i24 @llvm.bitreverse.i24(i24 %x)
ret i24 %rev
}
define i2 @test_ctpop_i2(i2 %a) {
; RV32-LABEL: test_ctpop_i2:
; RV32: # %bb.0:
; RV32-NEXT: addi sp, sp, -16
; RV32-NEXT: .cfi_def_cfa_offset 16
; RV32-NEXT: sw ra, 12(sp) # 4-byte Folded Spill
; RV32-NEXT: .cfi_offset ra, -4
; RV32-NEXT: andi a0, a0, 3
; RV32-NEXT: srli a1, a0, 1
; RV32-NEXT: sub a0, a0, a1
; RV32-NEXT: zext.b a1, a0
; RV32-NEXT: srli a1, a1, 2
; RV32-NEXT: andi a1, a1, 51
; RV32-NEXT: andi a0, a0, 51
; RV32-NEXT: add a0, a1, a0
; RV32-NEXT: srli a1, a0, 4
; RV32-NEXT: add a0, a1, a0
; RV32-NEXT: andi a0, a0, 15
; RV32-NEXT: li a1, 1
; RV32-NEXT: call __mulsi3
; RV32-NEXT: zext.b a0, a0
; RV32-NEXT: lw ra, 12(sp) # 4-byte Folded Reload
; RV32-NEXT: .cfi_restore ra
; RV32-NEXT: addi sp, sp, 16
; RV32-NEXT: .cfi_def_cfa_offset 0
; RV32-NEXT: ret
;
; RV64-LABEL: test_ctpop_i2:
; RV64: # %bb.0:
; RV64-NEXT: addi sp, sp, -16
; RV64-NEXT: .cfi_def_cfa_offset 16
; RV64-NEXT: sd ra, 8(sp) # 8-byte Folded Spill
; RV64-NEXT: .cfi_offset ra, -8
; RV64-NEXT: andi a0, a0, 3
; RV64-NEXT: srli a1, a0, 1
; RV64-NEXT: sub a0, a0, a1
; RV64-NEXT: zext.b a1, a0
; RV64-NEXT: srli a1, a1, 2
; RV64-NEXT: andi a1, a1, 51
; RV64-NEXT: andi a0, a0, 51
; RV64-NEXT: add a0, a1, a0
; RV64-NEXT: srli a1, a0, 4
; RV64-NEXT: add a0, a1, a0
; RV64-NEXT: andi a0, a0, 15
; RV64-NEXT: li a1, 1
; RV64-NEXT: call __muldi3
; RV64-NEXT: zext.b a0, a0
; RV64-NEXT: ld ra, 8(sp) # 8-byte Folded Reload
; RV64-NEXT: .cfi_restore ra
; RV64-NEXT: addi sp, sp, 16
; RV64-NEXT: .cfi_def_cfa_offset 0
; RV64-NEXT: ret
%1 = call i2 @llvm.ctpop.i2(i2 %a)
ret i2 %1
}
define i11 @test_ctpop_i11(i11 %a) {
; RV32-LABEL: test_ctpop_i11:
; RV32: # %bb.0:
; RV32-NEXT: andi a0, a0, 2047
; RV32-NEXT: lui a1, 5
; RV32-NEXT: srli a2, a0, 1
; RV32-NEXT: addi a1, a1, 1365
; RV32-NEXT: and a1, a2, a1
; RV32-NEXT: lui a2, 3
; RV32-NEXT: addi a2, a2, 819
; RV32-NEXT: sub a0, a0, a1
; RV32-NEXT: slli a1, a0, 16
; RV32-NEXT: and a0, a0, a2
; RV32-NEXT: srli a1, a1, 16
; RV32-NEXT: srli a1, a1, 2
; RV32-NEXT: and a1, a1, a2
; RV32-NEXT: lui a2, 1
; RV32-NEXT: add a0, a1, a0
; RV32-NEXT: srli a1, a0, 4
; RV32-NEXT: add a0, a1, a0
; RV32-NEXT: addi a1, a2, -241
; RV32-NEXT: and a0, a0, a1
; RV32-NEXT: srli a1, a0, 8
; RV32-NEXT: add a0, a0, a1
; RV32-NEXT: zext.b a0, a0
; RV32-NEXT: ret
;
; RV64-LABEL: test_ctpop_i11:
; RV64: # %bb.0:
; RV64-NEXT: andi a0, a0, 2047
; RV64-NEXT: lui a1, 5
; RV64-NEXT: srli a2, a0, 1
; RV64-NEXT: addi a1, a1, 1365
; RV64-NEXT: and a1, a2, a1
; RV64-NEXT: lui a2, 3
; RV64-NEXT: addi a2, a2, 819
; RV64-NEXT: sub a0, a0, a1
; RV64-NEXT: slli a1, a0, 48
; RV64-NEXT: and a0, a0, a2
; RV64-NEXT: srli a1, a1, 48
; RV64-NEXT: srli a1, a1, 2
; RV64-NEXT: and a1, a1, a2
; RV64-NEXT: lui a2, 1
; RV64-NEXT: add a0, a1, a0
; RV64-NEXT: srli a1, a0, 4
; RV64-NEXT: add a0, a1, a0
; RV64-NEXT: addi a1, a2, -241
; RV64-NEXT: and a0, a0, a1
; RV64-NEXT: srli a1, a0, 8
; RV64-NEXT: add a0, a0, a1
; RV64-NEXT: zext.b a0, a0
; RV64-NEXT: ret
%1 = call i11 @llvm.ctpop.i11(i11 %a)
ret i11 %1
}