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
70 lines
2.4 KiB
LLVM
70 lines
2.4 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc < %s -mtriple=x86_64-unknown -mattr=+bmi2 | FileCheck %s
|
|
; RUN: llc < %s -mtriple=x86_64-unknown -mcpu=core-avx2 | FileCheck %s
|
|
;
|
|
; When a MULX operand already lives in RDX (the implicit source register) at
|
|
; function entry, no register-copy preamble should be emitted before the MULX.
|
|
; In SysV x86-64 ABI the 3rd integer argument arrives in %rdx.
|
|
|
|
; Pure 64x64->128 multiply: arg 'a' is the 3rd argument, arriving in %rdx.
|
|
; The mul i128 node is emitted with operands in (b, a) order (as clang does);
|
|
define void @mul64_u128_a_in_rdx(ptr %hi, ptr %lo, i64 %a, i64 %b) {
|
|
; CHECK-LABEL: mul64_u128_a_in_rdx:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: mulxq %rcx, %rcx, %rax
|
|
; CHECK-NEXT: movq %rcx, (%rsi)
|
|
; CHECK-NEXT: movq %rax, (%rdi)
|
|
; CHECK-NEXT: retq
|
|
%za = zext i64 %a to i128
|
|
%zb = zext i64 %b to i128
|
|
%r = mul nuw i128 %zb, %za
|
|
%lo_val = trunc i128 %r to i64
|
|
store i64 %lo_val, ptr %lo
|
|
%hi_shr = lshr i128 %r, 64
|
|
%hi_val = trunc nuw i128 %hi_shr to i64
|
|
store i64 %hi_val, ptr %hi
|
|
ret void
|
|
}
|
|
|
|
; Same multiply with operands in natural (a, b) order.
|
|
define void @mul64_u128_a_in_rdx_natural_order(ptr %hi, ptr %lo, i64 %a, i64 %b) {
|
|
; CHECK-LABEL: mul64_u128_a_in_rdx_natural_order:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: mulxq %rcx, %rcx, %rax
|
|
; CHECK-NEXT: movq %rcx, (%rsi)
|
|
; CHECK-NEXT: movq %rax, (%rdi)
|
|
; CHECK-NEXT: retq
|
|
%za = zext i64 %a to i128
|
|
%zb = zext i64 %b to i128
|
|
%r = mul nuw i128 %za, %zb
|
|
%lo_val = trunc i128 %r to i64
|
|
store i64 %lo_val, ptr %lo
|
|
%hi_shr = lshr i128 %r, 64
|
|
%hi_val = trunc nuw i128 %hi_shr to i64
|
|
store i64 %hi_val, ptr %hi
|
|
ret void
|
|
}
|
|
|
|
; Multiply-add: hi:lo = a*b + c. 'a' is the 3rd arg in %rdx, 'c' is in %r8.
|
|
define void @muladd64_a_in_rdx(ptr %hi, ptr %lo, i64 %a, i64 %b, i64 %c) {
|
|
; CHECK-LABEL: muladd64_a_in_rdx:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: mulxq %rcx, %rax, %rcx
|
|
; CHECK-NEXT: addq %r8, %rax
|
|
; CHECK-NEXT: adcq $0, %rcx
|
|
; CHECK-NEXT: movq %rax, (%rsi)
|
|
; CHECK-NEXT: movq %rcx, (%rdi)
|
|
; CHECK-NEXT: retq
|
|
%za = zext i64 %a to i128
|
|
%zb = zext i64 %b to i128
|
|
%r = mul nuw i128 %zb, %za
|
|
%zc = zext i64 %c to i128
|
|
%r2 = add nuw i128 %r, %zc
|
|
%lo_val = trunc i128 %r2 to i64
|
|
store i64 %lo_val, ptr %lo
|
|
%hi_shr = lshr i128 %r2, 64
|
|
%hi_val = trunc nuw i128 %hi_shr to i64
|
|
store i64 %hi_val, ptr %hi
|
|
ret void
|
|
}
|