304f1d59ca
This switches everything to use the memory attribute proposed in https://discourse.llvm.org/t/rfc-unify-memory-effect-attributes/65579. The old argmemonly, inaccessiblememonly and inaccessiblemem_or_argmemonly attributes are dropped. The readnone, readonly and writeonly attributes are restricted to parameters only. The old attributes are auto-upgraded both in bitcode and IR. The bitcode upgrade is a policy requirement that has to be retained indefinitely. The IR upgrade is mainly there so it's not necessary to update all tests using memory attributes in this patch, which is already large enough. We could drop that part after migrating tests, or retain it longer term, to make it easier to import IR from older LLVM versions. High-level Function/CallBase APIs like doesNotAccessMemory() or setDoesNotAccessMemory() are mapped transparently to the memory attribute. Code that directly manipulates attributes (e.g. via AttributeList) on the other hand needs to switch to working with the memory attribute instead. Differential Revision: https://reviews.llvm.org/D135780
170 lines
9.6 KiB
LLVM
170 lines
9.6 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --function-signature --check-globals
|
|
; RUN: opt -S -passes=openmp-opt < %s | FileCheck %s
|
|
|
|
; void foo(double x) {
|
|
; #pragma omp target map(from:x)
|
|
; {
|
|
; x = sin(M_PI);
|
|
; #pragma omp parallel
|
|
; { }
|
|
; }
|
|
; }
|
|
|
|
target triple = "nvptx64"
|
|
|
|
%struct.ident_t = type { i32, i32, i32, i32, i8* }
|
|
|
|
@0 = private unnamed_addr constant [23 x i8] c";unknown;unknown;0;0;;\00", align 1
|
|
@1 = private unnamed_addr constant %struct.ident_t { i32 0, i32 2, i32 0, i32 0, i8* getelementptr inbounds ([23 x i8], [23 x i8]* @0, i32 0, i32 0) }, align 8
|
|
@__omp_offloading_fd02_404433c2_main_l5_exec_mode = weak constant i8 1
|
|
@llvm.compiler.used = appending global [1 x i8*] [i8* @__omp_offloading_fd02_404433c2_main_l5_exec_mode], section "llvm.metadata"
|
|
|
|
; Function Attrs: alwaysinline convergent norecurse nounwind
|
|
;.
|
|
; CHECK: @[[GLOB0:[0-9]+]] = private unnamed_addr constant [23 x i8] c"
|
|
; CHECK: @[[GLOB1:[0-9]+]] = private unnamed_addr constant [[STRUCT_IDENT_T:%.*]] { i32 0, i32 2, i32 0, i32 0, i8* getelementptr inbounds ([23 x i8], [23 x i8]* @[[GLOB0]], i32 0, i32 0) }, align 8
|
|
; CHECK: @[[__OMP_OFFLOADING_FD02_404433C2_MAIN_L5_EXEC_MODE:[a-zA-Z0-9_$"\\.-]+]] = weak constant i8 3
|
|
; CHECK: @[[LLVM_COMPILER_USED:[a-zA-Z0-9_$"\\.-]+]] = appending global [1 x i8*] [i8* @__omp_offloading_fd02_404433c2_main_l5_exec_mode], section "llvm.metadata"
|
|
; CHECK: @[[GLOB2:[0-9]+]] = private unnamed_addr constant [[STRUCT_IDENT_T:%.*]] { i32 0, i32 2, i32 0, i32 22, i8* getelementptr inbounds ([23 x i8], [23 x i8]* @[[GLOB0]], i32 0, i32 0) }, align 8
|
|
;.
|
|
define weak void @__omp_offloading_fd02_404433c2_main_l5(double* nonnull align 8 dereferenceable(8) %x) local_unnamed_addr #0 {
|
|
; CHECK-LABEL: define {{[^@]+}}@__omp_offloading_fd02_404433c2_main_l5
|
|
; CHECK-SAME: (double* nonnull align 8 dereferenceable(8) [[X:%.*]]) local_unnamed_addr #[[ATTR0:[0-9]+]] {
|
|
; CHECK-NEXT: entry:
|
|
; CHECK-NEXT: [[CAPTURED_VARS_ADDRS:%.*]] = alloca [0 x i8*], align 8
|
|
; CHECK-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_target_init(%struct.ident_t* nonnull @[[GLOB1]], i8 2, i1 false, i1 false) #[[ATTR3:[0-9]+]]
|
|
; CHECK-NEXT: [[EXEC_USER_CODE:%.*]] = icmp eq i32 [[TMP0]], -1
|
|
; CHECK-NEXT: br i1 [[EXEC_USER_CODE]], label [[USER_CODE_ENTRY:%.*]], label [[COMMON_RET:%.*]]
|
|
; CHECK: common.ret:
|
|
; CHECK-NEXT: ret void
|
|
; CHECK: user_code.entry:
|
|
; CHECK-NEXT: [[TMP1:%.*]] = call i32 @__kmpc_global_thread_num(%struct.ident_t* nonnull @[[GLOB1]]) #[[ATTR3]]
|
|
; CHECK-NEXT: [[CALL_I:%.*]] = call double @__nv_sin(double 0x400921FB54442D18) #[[ATTR7:[0-9]+]]
|
|
; CHECK-NEXT: br label [[REGION_CHECK_TID:%.*]]
|
|
; CHECK: region.check.tid:
|
|
; CHECK-NEXT: [[TMP2:%.*]] = call i32 @__kmpc_get_hardware_thread_id_in_block()
|
|
; CHECK-NEXT: [[TMP3:%.*]] = icmp eq i32 [[TMP2]], 0
|
|
; CHECK-NEXT: br i1 [[TMP3]], label [[REGION_GUARDED:%.*]], label [[REGION_BARRIER:%.*]]
|
|
; CHECK: region.guarded:
|
|
; CHECK-NEXT: store double [[CALL_I]], double* [[X]], align 8, !tbaa [[TBAA8:![0-9]+]]
|
|
; CHECK-NEXT: br label [[REGION_GUARDED_END:%.*]]
|
|
; CHECK: region.guarded.end:
|
|
; CHECK-NEXT: br label [[REGION_BARRIER]]
|
|
; CHECK: region.barrier:
|
|
; CHECK-NEXT: call void @__kmpc_barrier_simple_spmd(%struct.ident_t* @[[GLOB2]], i32 [[TMP2]])
|
|
; CHECK-NEXT: br label [[REGION_EXIT:%.*]]
|
|
; CHECK: region.exit:
|
|
; CHECK-NEXT: [[TMP4:%.*]] = getelementptr inbounds [0 x i8*], [0 x i8*]* [[CAPTURED_VARS_ADDRS]], i64 0, i64 0
|
|
; CHECK-NEXT: call void @__kmpc_parallel_51(%struct.ident_t* nonnull @[[GLOB1]], i32 [[TMP1]], i32 1, i32 -1, i32 -1, i8* bitcast (void (i32*, i32*)* @__omp_outlined__ to i8*), i8* bitcast (void (i16, i32)* @__omp_outlined___wrapper to i8*), i8** nonnull [[TMP4]], i64 0) #[[ATTR3]]
|
|
; CHECK-NEXT: call void @__kmpc_target_deinit(%struct.ident_t* nonnull @[[GLOB1]], i8 2, i1 false) #[[ATTR3]]
|
|
; CHECK-NEXT: br label [[COMMON_RET]]
|
|
;
|
|
entry:
|
|
%captured_vars_addrs = alloca [0 x i8*], align 8
|
|
%0 = call i32 @__kmpc_target_init(%struct.ident_t* nonnull @1, i8 1, i1 true, i1 true) #3
|
|
%exec_user_code = icmp eq i32 %0, -1
|
|
br i1 %exec_user_code, label %user_code.entry, label %common.ret
|
|
|
|
common.ret: ; preds = %entry, %user_code.entry
|
|
ret void
|
|
|
|
user_code.entry: ; preds = %entry
|
|
%1 = call i32 @__kmpc_global_thread_num(%struct.ident_t* nonnull @1)
|
|
%call.i = call double @__nv_sin(double 0x400921FB54442D18) #6
|
|
store double %call.i, double* %x, align 8, !tbaa !8
|
|
%2 = getelementptr inbounds [0 x i8*], [0 x i8*]* %captured_vars_addrs, i64 0, i64 0
|
|
call void @__kmpc_parallel_51(%struct.ident_t* nonnull @1, i32 %1, i32 1, i32 -1, i32 -1, i8* bitcast (void (i32*, i32*)* @__omp_outlined__ to i8*), i8* bitcast (void (i16, i32)* @__omp_outlined___wrapper to i8*), i8** nonnull %2, i64 0) #3
|
|
call void @__kmpc_target_deinit(%struct.ident_t* nonnull @1, i8 1, i1 true) #3
|
|
br label %common.ret
|
|
}
|
|
|
|
declare i32 @__kmpc_target_init(%struct.ident_t*, i8, i1, i1) local_unnamed_addr
|
|
|
|
; Function Attrs: alwaysinline mustprogress nofree norecurse nosync nounwind readnone willreturn
|
|
define internal void @__omp_outlined__(i32* noalias nocapture %.global_tid., i32* noalias nocapture %.bound_tid.) #1 {
|
|
; CHECK-LABEL: define {{[^@]+}}@__omp_outlined__
|
|
; CHECK-SAME: (i32* noalias nocapture [[DOTGLOBAL_TID_:%.*]], i32* noalias nocapture [[DOTBOUND_TID_:%.*]]) #[[ATTR1:[0-9]+]] {
|
|
; CHECK-NEXT: entry:
|
|
; CHECK-NEXT: ret void
|
|
;
|
|
entry:
|
|
ret void
|
|
}
|
|
|
|
; Function Attrs: norecurse nounwind
|
|
define internal void @__omp_outlined___wrapper(i16 zeroext %0, i32 %1) #2 {
|
|
; CHECK-LABEL: define {{[^@]+}}@__omp_outlined___wrapper
|
|
; CHECK-SAME: (i16 zeroext [[TMP0:%.*]], i32 [[TMP1:%.*]]) #[[ATTR2:[0-9]+]] {
|
|
; CHECK-NEXT: entry:
|
|
; CHECK-NEXT: [[GLOBAL_ARGS:%.*]] = alloca i8**, align 8
|
|
; CHECK-NEXT: call void @__kmpc_get_shared_variables(i8*** nonnull [[GLOBAL_ARGS]]) #[[ATTR3]]
|
|
; CHECK-NEXT: ret void
|
|
;
|
|
entry:
|
|
%global_args = alloca i8**, align 8
|
|
call void @__kmpc_get_shared_variables(i8*** nonnull %global_args) #3
|
|
ret void
|
|
}
|
|
|
|
declare void @__kmpc_get_shared_variables(i8***) local_unnamed_addr
|
|
|
|
; Function Attrs: nounwind
|
|
declare i32 @__kmpc_global_thread_num(%struct.ident_t*) local_unnamed_addr #3
|
|
|
|
; Function Attrs: alwaysinline
|
|
declare void @__kmpc_parallel_51(%struct.ident_t*, i32, i32, i32, i32, i8*, i8*, i8**, i64) local_unnamed_addr #4
|
|
|
|
declare void @__kmpc_target_deinit(%struct.ident_t*, i8, i1) local_unnamed_addr
|
|
|
|
; Function Attrs: convergent
|
|
declare double @__nv_sin(double) local_unnamed_addr #5
|
|
|
|
attributes #0 = { alwaysinline convergent norecurse nounwind "frame-pointer"="all" "min-legal-vector-width"="0" "no-trapping-math"="true" "stack-protector-buffer-size"="8" }
|
|
attributes #1 = { alwaysinline mustprogress nofree norecurse nosync nounwind readnone willreturn "frame-pointer"="all" "min-legal-vector-width"="0" "no-trapping-math"="true" "stack-protector-buffer-size"="8" }
|
|
attributes #2 = { norecurse nounwind "frame-pointer"="all" "min-legal-vector-width"="0" "no-trapping-math"="true" "stack-protector-buffer-size"="8" }
|
|
attributes #3 = { nounwind }
|
|
attributes #4 = { alwaysinline }
|
|
attributes #5 = { convergent "frame-pointer"="all" "no-trapping-math"="true" "stack-protector-buffer-size"="8" }
|
|
attributes #6 = { convergent nounwind "llvm.assume"="ompx_spmd_amenable" }
|
|
|
|
!omp_offload.info = !{!0}
|
|
!nvvm.annotations = !{!1}
|
|
!llvm.module.flags = !{!2, !3, !4, !5, !6}
|
|
!llvm.ident = !{!7}
|
|
|
|
!0 = !{i32 0, i32 64770, i32 1078211522, !"main", i32 5, i32 0}
|
|
!1 = !{void (double*)* @__omp_offloading_fd02_404433c2_main_l5, !"kernel", i32 1}
|
|
!2 = !{i32 1, !"wchar_size", i32 4}
|
|
!3 = !{i32 7, !"openmp", i32 50}
|
|
!4 = !{i32 7, !"openmp-device", i32 50}
|
|
!5 = !{i32 8, !"PIC Level", i32 2}
|
|
!6 = !{i32 7, !"frame-pointer", i32 2}
|
|
!7 = !{!"clang version 14.0.0"}
|
|
!8 = !{!9, !9, i64 0}
|
|
!9 = !{!"double", !10, i64 0}
|
|
!10 = !{!"omnipotent char", !11, i64 0}
|
|
!11 = !{!"Simple C/C++ TBAA"}
|
|
;.
|
|
; CHECK: attributes #[[ATTR0]] = { alwaysinline convergent norecurse nounwind "frame-pointer"="all" "min-legal-vector-width"="0" "no-trapping-math"="true" "stack-protector-buffer-size"="8" }
|
|
; CHECK: attributes #[[ATTR1]] = { alwaysinline mustprogress nofree norecurse nosync nounwind willreturn memory(none) "frame-pointer"="all" "min-legal-vector-width"="0" "no-trapping-math"="true" "stack-protector-buffer-size"="8" }
|
|
; CHECK: attributes #[[ATTR2]] = { norecurse nounwind "frame-pointer"="all" "min-legal-vector-width"="0" "no-trapping-math"="true" "stack-protector-buffer-size"="8" }
|
|
; CHECK: attributes #[[ATTR3]] = { nounwind }
|
|
; CHECK: attributes #[[ATTR4:[0-9]+]] = { alwaysinline }
|
|
; CHECK: attributes #[[ATTR5:[0-9]+]] = { convergent "frame-pointer"="all" "no-trapping-math"="true" "stack-protector-buffer-size"="8" }
|
|
; CHECK: attributes #[[ATTR6:[0-9]+]] = { convergent nounwind }
|
|
; CHECK: attributes #[[ATTR7]] = { convergent nounwind "llvm.assume"="ompx_spmd_amenable" }
|
|
;.
|
|
; CHECK: [[META0:![0-9]+]] = !{i32 0, i32 64770, i32 1078211522, !"main", i32 5, i32 0}
|
|
; CHECK: [[META1:![0-9]+]] = !{void (double*)* @__omp_offloading_fd02_404433c2_main_l5, !"kernel", i32 1}
|
|
; CHECK: [[META2:![0-9]+]] = !{i32 1, !"wchar_size", i32 4}
|
|
; CHECK: [[META3:![0-9]+]] = !{i32 7, !"openmp", i32 50}
|
|
; CHECK: [[META4:![0-9]+]] = !{i32 7, !"openmp-device", i32 50}
|
|
; CHECK: [[META5:![0-9]+]] = !{i32 8, !"PIC Level", i32 2}
|
|
; CHECK: [[META6:![0-9]+]] = !{i32 7, !"frame-pointer", i32 2}
|
|
; CHECK: [[META7:![0-9]+]] = !{!"clang version 14.0.0"}
|
|
; CHECK: [[TBAA8]] = !{!9, !9, i64 0}
|
|
; CHECK: [[META9:![0-9]+]] = !{!"double", !10, i64 0}
|
|
; CHECK: [[META10:![0-9]+]] = !{!"omnipotent char", !11, i64 0}
|
|
; CHECK: [[META11:![0-9]+]] = !{!"Simple C/C++ TBAA"}
|
|
;.
|