From 49a5dd3dac285ba12f3fcaa55cacbea5968f5a37 Mon Sep 17 00:00:00 2001 From: Shilei Tian Date: Thu, 1 May 2025 14:10:31 -0400 Subject: [PATCH] [NFC][AMDGPU] Correct the check line update script for `llvm/test/CodeGen/AMDGPU/attributor-flatscratchinit-undefined-behavior.ll` --- ...utor-flatscratchinit-undefined-behavior.ll | 100 +++++++++++++++++- 1 file changed, 96 insertions(+), 4 deletions(-) diff --git a/llvm/test/CodeGen/AMDGPU/attributor-flatscratchinit-undefined-behavior.ll b/llvm/test/CodeGen/AMDGPU/attributor-flatscratchinit-undefined-behavior.ll index 1b422252573d..93ebeafe0dee 100644 --- a/llvm/test/CodeGen/AMDGPU/attributor-flatscratchinit-undefined-behavior.ll +++ b/llvm/test/CodeGen/AMDGPU/attributor-flatscratchinit-undefined-behavior.ll @@ -1,4 +1,4 @@ -; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5 +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-attributes --check-globals all --version 5 ; RUN: opt -S -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 -passes=amdgpu-attributor < %s | FileCheck -check-prefixes=GFX9 %s ; RUN: opt -S -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -passes=amdgpu-attributor < %s | FileCheck -check-prefixes=GFX10 %s @@ -10,23 +10,67 @@ ;; tests of addrspacecast define void @with_private_to_flat_addrspacecast(ptr addrspace(5) %ptr) #0 { +; GFX9-LABEL: define void @with_private_to_flat_addrspacecast( +; GFX9-SAME: ptr addrspace(5) [[PTR:%.*]]) #[[ATTR0:[0-9]+]] { +; GFX9-NEXT: [[STOF:%.*]] = addrspacecast ptr addrspace(5) [[PTR]] to ptr +; GFX9-NEXT: store volatile i32 0, ptr [[STOF]], align 4 +; GFX9-NEXT: ret void +; +; GFX10-LABEL: define void @with_private_to_flat_addrspacecast( +; GFX10-SAME: ptr addrspace(5) [[PTR:%.*]]) #[[ATTR0:[0-9]+]] { +; GFX10-NEXT: [[STOF:%.*]] = addrspacecast ptr addrspace(5) [[PTR]] to ptr +; GFX10-NEXT: store volatile i32 0, ptr [[STOF]], align 4 +; GFX10-NEXT: ret void +; %stof = addrspacecast ptr addrspace(5) %ptr to ptr store volatile i32 0, ptr %stof ret void } define amdgpu_kernel void @with_private_to_flat_addrspacecast_cc_kernel(ptr addrspace(5) %ptr) #0 { +; GFX9-LABEL: define amdgpu_kernel void @with_private_to_flat_addrspacecast_cc_kernel( +; GFX9-SAME: ptr addrspace(5) [[PTR:%.*]]) #[[ATTR0]] { +; GFX9-NEXT: [[STOF:%.*]] = addrspacecast ptr addrspace(5) [[PTR]] to ptr +; GFX9-NEXT: store volatile i32 0, ptr [[STOF]], align 4 +; GFX9-NEXT: ret void +; +; GFX10-LABEL: define amdgpu_kernel void @with_private_to_flat_addrspacecast_cc_kernel( +; GFX10-SAME: ptr addrspace(5) [[PTR:%.*]]) #[[ATTR0]] { +; GFX10-NEXT: [[STOF:%.*]] = addrspacecast ptr addrspace(5) [[PTR]] to ptr +; GFX10-NEXT: store volatile i32 0, ptr [[STOF]], align 4 +; GFX10-NEXT: ret void +; %stof = addrspacecast ptr addrspace(5) %ptr to ptr store volatile i32 0, ptr %stof ret void } define void @call_with_private_to_flat_addrspacecast(ptr addrspace(5) %ptr) #0 { +; GFX9-LABEL: define void @call_with_private_to_flat_addrspacecast( +; GFX9-SAME: ptr addrspace(5) [[PTR:%.*]]) #[[ATTR0]] { +; GFX9-NEXT: call void @with_private_to_flat_addrspacecast(ptr addrspace(5) [[PTR]]) +; GFX9-NEXT: ret void +; +; GFX10-LABEL: define void @call_with_private_to_flat_addrspacecast( +; GFX10-SAME: ptr addrspace(5) [[PTR:%.*]]) #[[ATTR0]] { +; GFX10-NEXT: call void @with_private_to_flat_addrspacecast(ptr addrspace(5) [[PTR]]) +; GFX10-NEXT: ret void +; call void @with_private_to_flat_addrspacecast(ptr addrspace(5) %ptr) ret void } define amdgpu_kernel void @call_with_private_to_flat_addrspacecast_cc_kernel(ptr addrspace(5) %ptr) #0 { +; GFX9-LABEL: define amdgpu_kernel void @call_with_private_to_flat_addrspacecast_cc_kernel( +; GFX9-SAME: ptr addrspace(5) [[PTR:%.*]]) #[[ATTR0]] { +; GFX9-NEXT: call void @with_private_to_flat_addrspacecast(ptr addrspace(5) [[PTR]]) +; GFX9-NEXT: ret void +; +; GFX10-LABEL: define amdgpu_kernel void @call_with_private_to_flat_addrspacecast_cc_kernel( +; GFX10-SAME: ptr addrspace(5) [[PTR:%.*]]) #[[ATTR0]] { +; GFX10-NEXT: call void @with_private_to_flat_addrspacecast(ptr addrspace(5) [[PTR]]) +; GFX10-NEXT: ret void +; call void @with_private_to_flat_addrspacecast(ptr addrspace(5) %ptr) ret void } @@ -34,6 +78,16 @@ define amdgpu_kernel void @call_with_private_to_flat_addrspacecast_cc_kernel(ptr ;; tests of addrspacecast in a constant define amdgpu_kernel void @private_constant_expression_use(ptr addrspace(1) nocapture %out) #0 { +; GFX9-LABEL: define amdgpu_kernel void @private_constant_expression_use( +; GFX9-SAME: ptr addrspace(1) captures(none) [[OUT:%.*]]) #[[ATTR0]] { +; GFX9-NEXT: store volatile ptr addrspacecast (ptr addrspace(5) inttoptr (i32 123 to ptr addrspace(5)) to ptr), ptr addrspace(1) [[OUT]], align 8 +; GFX9-NEXT: ret void +; +; GFX10-LABEL: define amdgpu_kernel void @private_constant_expression_use( +; GFX10-SAME: ptr addrspace(1) captures(none) [[OUT:%.*]]) #[[ATTR0]] { +; GFX10-NEXT: store volatile ptr addrspacecast (ptr addrspace(5) inttoptr (i32 123 to ptr addrspace(5)) to ptr), ptr addrspace(1) [[OUT]], align 8 +; GFX10-NEXT: ret void +; store volatile ptr addrspacecast (ptr addrspace(5) inttoptr (i32 123 to ptr addrspace(5)) to ptr), ptr addrspace(1) %out, align 8 ret void } @@ -41,23 +95,61 @@ define amdgpu_kernel void @private_constant_expression_use(ptr addrspace(1) noca ;; tests of intrinsics define amdgpu_kernel void @calls_intrin_ascast_cc_kernel(ptr addrspace(3) %ptr) #0 { +; GFX9-LABEL: define amdgpu_kernel void @calls_intrin_ascast_cc_kernel( +; GFX9-SAME: ptr addrspace(3) [[PTR:%.*]]) #[[ATTR0]] { +; GFX9-NEXT: [[TMP1:%.*]] = call ptr @llvm.amdgcn.addrspacecast.nonnull.p0.p3(ptr addrspace(3) [[PTR]]) +; GFX9-NEXT: store volatile i32 7, ptr [[TMP1]], align 4 +; GFX9-NEXT: ret void +; +; GFX10-LABEL: define amdgpu_kernel void @calls_intrin_ascast_cc_kernel( +; GFX10-SAME: ptr addrspace(3) [[PTR:%.*]]) #[[ATTR0]] { +; GFX10-NEXT: [[TMP1:%.*]] = call ptr @llvm.amdgcn.addrspacecast.nonnull.p0.p3(ptr addrspace(3) [[PTR]]) +; GFX10-NEXT: store volatile i32 7, ptr [[TMP1]], align 4 +; GFX10-NEXT: ret void +; %1 = call ptr @llvm.amdgcn.addrspacecast.nonnull.p0.p3(ptr addrspace(3) %ptr) store volatile i32 7, ptr %1, align 4 ret void } define void @calls_intrin_ascast(ptr addrspace(3) %ptr) #0 { +; GFX9-LABEL: define void @calls_intrin_ascast( +; GFX9-SAME: ptr addrspace(3) [[PTR:%.*]]) #[[ATTR0]] { +; GFX9-NEXT: [[TMP1:%.*]] = call ptr @llvm.amdgcn.addrspacecast.nonnull.p0.p3(ptr addrspace(3) [[PTR]]) +; GFX9-NEXT: store volatile i32 7, ptr [[TMP1]], align 4 +; GFX9-NEXT: ret void +; +; GFX10-LABEL: define void @calls_intrin_ascast( +; GFX10-SAME: ptr addrspace(3) [[PTR:%.*]]) #[[ATTR0]] { +; GFX10-NEXT: [[TMP1:%.*]] = call ptr @llvm.amdgcn.addrspacecast.nonnull.p0.p3(ptr addrspace(3) [[PTR]]) +; GFX10-NEXT: store volatile i32 7, ptr [[TMP1]], align 4 +; GFX10-NEXT: ret void +; %1 = call ptr @llvm.amdgcn.addrspacecast.nonnull.p0.p3(ptr addrspace(3) %ptr) store volatile i32 7, ptr %1, align 4 ret void } define amdgpu_kernel void @call_calls_intrin_ascast_cc_kernel(ptr addrspace(3) %ptr) #0 { +; GFX9-LABEL: define amdgpu_kernel void @call_calls_intrin_ascast_cc_kernel( +; GFX9-SAME: ptr addrspace(3) [[PTR:%.*]]) #[[ATTR0]] { +; GFX9-NEXT: call void @calls_intrin_ascast(ptr addrspace(3) [[PTR]]) +; GFX9-NEXT: ret void +; +; GFX10-LABEL: define amdgpu_kernel void @call_calls_intrin_ascast_cc_kernel( +; GFX10-SAME: ptr addrspace(3) [[PTR:%.*]]) #[[ATTR0]] { +; GFX10-NEXT: call void @calls_intrin_ascast(ptr addrspace(3) [[PTR]]) +; GFX10-NEXT: ret void +; call void @calls_intrin_ascast(ptr addrspace(3) %ptr) ret void } attributes #0 = { "amdgpu-no-flat-scratch-init" } - -; GFX9: attributes #0 = { "amdgpu-agpr-alloc"="0" "amdgpu-no-completion-action" "amdgpu-no-default-queue" "amdgpu-no-dispatch-id" "amdgpu-no-dispatch-ptr" "amdgpu-no-flat-scratch-init" "amdgpu-no-heap-ptr" "amdgpu-no-hostcall-ptr" "amdgpu-no-implicitarg-ptr" "amdgpu-no-lds-kernel-id" "amdgpu-no-multigrid-sync-arg" "amdgpu-no-queue-ptr" "amdgpu-no-workgroup-id-x" "amdgpu-no-workgroup-id-y" "amdgpu-no-workgroup-id-z" "amdgpu-no-workitem-id-x" "amdgpu-no-workitem-id-y" "amdgpu-no-workitem-id-z" "target-cpu"="gfx900" "uniform-work-group-size"="false" } -; GFX10: attributes #0 = { "amdgpu-agpr-alloc"="0" "amdgpu-no-completion-action" "amdgpu-no-default-queue" "amdgpu-no-dispatch-id" "amdgpu-no-dispatch-ptr" "amdgpu-no-flat-scratch-init" "amdgpu-no-heap-ptr" "amdgpu-no-hostcall-ptr" "amdgpu-no-implicitarg-ptr" "amdgpu-no-lds-kernel-id" "amdgpu-no-multigrid-sync-arg" "amdgpu-no-queue-ptr" "amdgpu-no-workgroup-id-x" "amdgpu-no-workgroup-id-y" "amdgpu-no-workgroup-id-z" "amdgpu-no-workitem-id-x" "amdgpu-no-workitem-id-y" "amdgpu-no-workitem-id-z" "target-cpu"="gfx1010" "uniform-work-group-size"="false" } +;. +; GFX9: attributes #[[ATTR0]] = { "amdgpu-agpr-alloc"="0" "amdgpu-no-completion-action" "amdgpu-no-default-queue" "amdgpu-no-dispatch-id" "amdgpu-no-dispatch-ptr" "amdgpu-no-flat-scratch-init" "amdgpu-no-heap-ptr" "amdgpu-no-hostcall-ptr" "amdgpu-no-implicitarg-ptr" "amdgpu-no-lds-kernel-id" "amdgpu-no-multigrid-sync-arg" "amdgpu-no-queue-ptr" "amdgpu-no-workgroup-id-x" "amdgpu-no-workgroup-id-y" "amdgpu-no-workgroup-id-z" "amdgpu-no-workitem-id-x" "amdgpu-no-workitem-id-y" "amdgpu-no-workitem-id-z" "amdgpu-waves-per-eu"="4,10" "target-cpu"="gfx900" "uniform-work-group-size"="false" } +; GFX9: attributes #[[ATTR1:[0-9]+]] = { nocallback nofree nosync nounwind speculatable willreturn memory(none) "target-cpu"="gfx900" } +;. +; GFX10: attributes #[[ATTR0]] = { "amdgpu-agpr-alloc"="0" "amdgpu-no-completion-action" "amdgpu-no-default-queue" "amdgpu-no-dispatch-id" "amdgpu-no-dispatch-ptr" "amdgpu-no-flat-scratch-init" "amdgpu-no-heap-ptr" "amdgpu-no-hostcall-ptr" "amdgpu-no-implicitarg-ptr" "amdgpu-no-lds-kernel-id" "amdgpu-no-multigrid-sync-arg" "amdgpu-no-queue-ptr" "amdgpu-no-workgroup-id-x" "amdgpu-no-workgroup-id-y" "amdgpu-no-workgroup-id-z" "amdgpu-no-workitem-id-x" "amdgpu-no-workitem-id-y" "amdgpu-no-workitem-id-z" "amdgpu-waves-per-eu"="8,20" "target-cpu"="gfx1010" "uniform-work-group-size"="false" } +; GFX10: attributes #[[ATTR1:[0-9]+]] = { nocallback nofree nosync nounwind speculatable willreturn memory(none) "target-cpu"="gfx1010" } +;.