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

108 lines
7.8 KiB
LLVM

; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt -S -passes=argpromotion -mtriple=powerpc64le-unknown-linux-gnu < %s | FileCheck %s
; Test to check that we do not promote arguments when the
; type size is greater than 128 bits.
define internal fastcc void @print_acc(ptr nocapture readonly %a) nounwind {
; CHECK-LABEL: @print_acc(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[TMP0:%.*]] = load <512 x i1>, ptr [[A:%.*]], align 64
; CHECK-NEXT: [[TMP1:%.*]] = tail call { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } @llvm.ppc.mma.disassemble.acc(<512 x i1> [[TMP0]])
; CHECK-NEXT: [[TMP2:%.*]] = extractvalue { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } [[TMP1]], 0
; CHECK-NEXT: ret void
;
entry:
%0 = load <512 x i1>, ptr %a, align 64
%1 = tail call { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } @llvm.ppc.mma.disassemble.acc(<512 x i1> %0)
%2 = extractvalue { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } %1, 0
ret void
}
declare { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } @llvm.ppc.mma.disassemble.acc(<512 x i1>) nounwind
define dso_local void @test(ptr nocapture %a, <16 x i8> %ac) {
; CHECK-LABEL: @test(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[TMP0:%.*]] = tail call <512 x i1> @llvm.ppc.mma.xvf32ger(<16 x i8> [[AC:%.*]], <16 x i8> [[AC]])
; CHECK-NEXT: store <512 x i1> [[TMP0]], ptr [[A:%.*]], align 64
; CHECK-NEXT: tail call fastcc void @print_acc(ptr nonnull [[A]])
; CHECK-NEXT: ret void
;
entry:
%0 = tail call <512 x i1> @llvm.ppc.mma.xvf32ger(<16 x i8> %ac, <16 x i8> %ac)
store <512 x i1> %0, ptr %a, align 64
tail call fastcc void @print_acc(ptr nonnull %a)
ret void
}
declare <512 x i1> @llvm.ppc.mma.xvf32ger(<16 x i8>, <16 x i8>) nounwind
@.str = private unnamed_addr constant [11 x i8] c"Vector: { \00", align 1
@.str.1 = private unnamed_addr constant [5 x i8] c"%d, \00", align 1
@.str.2 = private unnamed_addr constant [6 x i8] c"%d }\0A\00", align 1
define internal fastcc void @printWideVec(<16 x i32> %ptr.val) nounwind {
; CHECK-LABEL: @printWideVec(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[CALL:%.*]] = tail call signext i32 (ptr, ...) @printf(ptr nonnull dereferenceable(1) @.str)
; CHECK-NEXT: [[VECEXT:%.*]] = extractelement <16 x i32> [[PTR_VAL:%.*]], i32 0
; CHECK-NEXT: [[CALL1:%.*]] = tail call signext i32 (ptr, ...) @printf(ptr nonnull dereferenceable(1) @.str.1, i32 signext [[VECEXT]])
; CHECK-NEXT: [[VECEXT_1:%.*]] = extractelement <16 x i32> [[PTR_VAL]], i32 1
; CHECK-NEXT: [[CALL1_1:%.*]] = tail call signext i32 (ptr, ...) @printf(ptr nonnull dereferenceable(1) @.str.1, i32 signext [[VECEXT_1]])
; CHECK-NEXT: [[VECEXT_2:%.*]] = extractelement <16 x i32> [[PTR_VAL]], i32 2
; CHECK-NEXT: [[CALL1_2:%.*]] = tail call signext i32 (ptr, ...) @printf(ptr nonnull dereferenceable(1) @.str.1, i32 signext [[VECEXT_2]])
; CHECK-NEXT: [[VECEXT_3:%.*]] = extractelement <16 x i32> [[PTR_VAL]], i32 3
; CHECK-NEXT: [[CALL1_3:%.*]] = tail call signext i32 (ptr, ...) @printf(ptr nonnull dereferenceable(1) @.str.1, i32 signext [[VECEXT_3]])
; CHECK-NEXT: [[VECEXT_4:%.*]] = extractelement <16 x i32> [[PTR_VAL]], i32 4
; CHECK-NEXT: [[CALL1_4:%.*]] = tail call signext i32 (ptr, ...) @printf(ptr nonnull dereferenceable(1) @.str.1, i32 signext [[VECEXT_4]])
; CHECK-NEXT: [[VECEXT_5:%.*]] = extractelement <16 x i32> [[PTR_VAL]], i32 5
; CHECK-NEXT: [[CALL1_5:%.*]] = tail call signext i32 (ptr, ...) @printf(ptr nonnull dereferenceable(1) @.str.1, i32 signext [[VECEXT_5]])
; CHECK-NEXT: [[VECEXT_6:%.*]] = extractelement <16 x i32> [[PTR_VAL]], i32 6
; CHECK-NEXT: [[CALL1_6:%.*]] = tail call signext i32 (ptr, ...) @printf(ptr nonnull dereferenceable(1) @.str.1, i32 signext [[VECEXT_6]])
; CHECK-NEXT: [[VECEXT2:%.*]] = extractelement <16 x i32> [[PTR_VAL]], i32 7
; CHECK-NEXT: [[CALL3:%.*]] = tail call signext i32 (ptr, ...) @printf(ptr nonnull dereferenceable(1) @.str.2, i32 signext [[VECEXT2]])
; CHECK-NEXT: ret void
;
entry:
%call = tail call signext i32 (ptr, ...) @printf(ptr nonnull dereferenceable(1) @.str)
%vecext = extractelement <16 x i32> %ptr.val, i32 0
%call1 = tail call signext i32 (ptr, ...) @printf(ptr nonnull dereferenceable(1) @.str.1, i32 signext %vecext)
%vecext.1 = extractelement <16 x i32> %ptr.val, i32 1
%call1.1 = tail call signext i32 (ptr, ...) @printf(ptr nonnull dereferenceable(1) @.str.1, i32 signext %vecext.1)
%vecext.2 = extractelement <16 x i32> %ptr.val, i32 2
%call1.2 = tail call signext i32 (ptr, ...) @printf(ptr nonnull dereferenceable(1) @.str.1, i32 signext %vecext.2)
%vecext.3 = extractelement <16 x i32> %ptr.val, i32 3
%call1.3 = tail call signext i32 (ptr, ...) @printf(ptr nonnull dereferenceable(1) @.str.1, i32 signext %vecext.3)
%vecext.4 = extractelement <16 x i32> %ptr.val, i32 4
%call1.4 = tail call signext i32 (ptr, ...) @printf(ptr nonnull dereferenceable(1) @.str.1, i32 signext %vecext.4)
%vecext.5 = extractelement <16 x i32> %ptr.val, i32 5
%call1.5 = tail call signext i32 (ptr, ...) @printf(ptr nonnull dereferenceable(1) @.str.1, i32 signext %vecext.5)
%vecext.6 = extractelement <16 x i32> %ptr.val, i32 6
%call1.6 = tail call signext i32 (ptr, ...) @printf(ptr nonnull dereferenceable(1) @.str.1, i32 signext %vecext.6)
%vecext2 = extractelement <16 x i32> %ptr.val, i32 7
%call3 = tail call signext i32 (ptr, ...) @printf(ptr nonnull dereferenceable(1) @.str.2, i32 signext %vecext2)
ret void
}
declare noundef signext i32 @printf(ptr nocapture noundef readonly, ...) nounwind
define dso_local void @test1(<4 x i32> %a, <4 x i32> %b) nounwind {
; CHECK-LABEL: @test1(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[TMP0:%.*]] = shufflevector <4 x i32> [[A:%.*]], <4 x i32> undef, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <4 x i32> [[B:%.*]], <4 x i32> undef, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
; CHECK-NEXT: [[TMP2:%.*]] = shufflevector <16 x i32> [[TMP0]], <16 x i32> <i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31>
; CHECK-NEXT: [[VECINIT22:%.*]] = shufflevector <16 x i32> [[TMP2]], <16 x i32> [[TMP1]], <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 16, i32 17, i32 18, i32 19, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
; CHECK-NEXT: tail call fastcc void @printWideVec(<16 x i32> [[VECINIT22]])
; CHECK-NEXT: ret void
;
entry:
%0 = shufflevector <4 x i32> %a, <4 x i32> undef, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
%1 = shufflevector <4 x i32> %b, <4 x i32> undef, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
%2 = shufflevector <16 x i32> %0, <16 x i32> <i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31>
%vecinit22 = shufflevector <16 x i32> %2, <16 x i32> %1, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 16, i32 17, i32 18, i32 19, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
tail call fastcc void @printWideVec(<16 x i32> %vecinit22)
ret void
}