[llvm] 6400918 - [OpenMP] AlwaysInline __kmpc_parallel_51 to improve inlining hueristics

[via llvm-commits]

Author: Joseph Huber
Date: 2021-08-10T14:41:43-04:00
New Revision: 640091884f81b377737bb73a3f076c7069666607

URL: https://github.com/llvm/llvm-project/commit/640091884f81b377737bb73a3f076c7069666607
DIFF: https://github.com/llvm/llvm-project/commit/640091884f81b377737bb73a3f076c7069666607.diff

LOG: [OpenMP] AlwaysInline __kmpc_parallel_51 to improve inlining hueristics

This patch adds the `AlwaysInline` attribute to the `__kmpc_parallel_51`
device runtime call. This improves inlining heuristics which encourages
the indirect function pointer arguemnt to also be inlined. This greatly
improves performance for a few applications whose outlined regions were
not inlined otherwise.

Reviewed By: jdoerfert

Differential Revision: https://reviews.llvm.org/D107839

Added: 
    

Modified: 
    llvm/include/llvm/Frontend/OpenMP/OMPKinds.def
    llvm/test/Transforms/OpenMP/parallel_level_fold.ll

Removed: 
    


################################################################################
diff  --git a/llvm/include/llvm/Frontend/OpenMP/OMPKinds.def b/llvm/include/llvm/Frontend/OpenMP/OMPKinds.def
index 6c07a58906568..26141c9ce249e 100644
--- a/llvm/include/llvm/Frontend/OpenMP/OMPKinds.def
+++ b/llvm/include/llvm/Frontend/OpenMP/OMPKinds.def
@@ -510,6 +510,11 @@ __OMP_ATTRS_SET(NoCaptureAttrs,
                     ? AttributeSet(EnumAttr(NoCapture))
                     : AttributeSet(EnumAttr(NoCapture)))
 
+__OMP_ATTRS_SET(AlwaysInlineAttrs,
+                OptimisticAttributes
+                    ? AttributeSet(EnumAttr(AlwaysInline))
+                    : AttributeSet(EnumAttr(AlwaysInline)))
+
 #if 0
 __OMP_ATTRS_SET(InaccessibleOnlyAttrs,
                 OptimisticAttributes
@@ -897,6 +902,9 @@ __OMP_RTL_ATTRS(__tgt_push_mapper_component, ForkAttrs, AttributeSet(),
 __OMP_RTL_ATTRS(__kmpc_task_allow_completion_event, DefaultAttrs,
                 ReturnPtrAttrs, ParamAttrs(ReadOnlyPtrAttrs))
 
+__OMP_RTL_ATTRS(__kmpc_parallel_51, AlwaysInlineAttrs, AttributeSet(),
+                ParamAttrs())
+
 #undef __OMP_RTL_ATTRS
 #undef OMP_RTL_ATTRS
 #undef AttributeSet

diff  --git a/llvm/test/Transforms/OpenMP/parallel_level_fold.ll b/llvm/test/Transforms/OpenMP/parallel_level_fold.ll
index 29afba97ab56e..75530a38d9764 100644
--- a/llvm/test/Transforms/OpenMP/parallel_level_fold.ll
+++ b/llvm/test/Transforms/OpenMP/parallel_level_fold.ll
@@ -108,7 +108,7 @@ define internal void @spmd_helper() {
 
 define internal void @__kmpc_parallel_51(%struct.ident_t*, i32, i32, i32, i32, i8*, i8*, i8**, i64) {
 ; CHECK-LABEL: define {{[^@]+}}@__kmpc_parallel_51
-; CHECK-SAME: (%struct.ident_t* noalias nocapture nofree readnone align 536870912 [[TMP0:%.*]], i32 [[TMP1:%.*]], i32 [[TMP2:%.*]], i32 [[TMP3:%.*]], i32 [[TMP4:%.*]], i8* noalias nocapture nofree readnone align 536870912 [[TMP5:%.*]], i8* noalias nocapture nofree readnone align 536870912 [[TMP6:%.*]], i8** noalias nocapture nofree readnone align 536870912 [[TMP7:%.*]], i64 [[TMP8:%.*]]) {
+; CHECK-SAME: (%struct.ident_t* noalias nocapture nofree readnone align 536870912 [[TMP0:%.*]], i32 [[TMP1:%.*]], i32 [[TMP2:%.*]], i32 [[TMP3:%.*]], i32 [[TMP4:%.*]], i8* noalias nocapture nofree readnone align 536870912 [[TMP5:%.*]], i8* noalias nocapture nofree readnone align 536870912 [[TMP6:%.*]], i8** noalias nocapture nofree readnone align 536870912 [[TMP7:%.*]], i64 [[TMP8:%.*]]) #[[ATTR0:[0-9]+]] {
 ; CHECK-NEXT:    call void @parallel_helper()
 ; CHECK-NEXT:    ret void
 ;
@@ -142,6 +142,8 @@ declare void @__kmpc_target_deinit(%struct.ident_t* nocapture readnone, i1 zeroe
 !3 = !{void ()* @spmd, !"kernel", i32 1}
 !4 = !{void ()* @parallel, !"kernel", i32 1}
 ;.
+; CHECK: attributes #[[ATTR0]] = { alwaysinline }
+;.
 ; CHECK: [[META0:![0-9]+]] = !{i32 7, !"openmp", i32 50}
 ; CHECK: [[META1:![0-9]+]] = !{i32 7, !"openmp-device", i32 50}
 ; CHECK: [[META2:![0-9]+]] = !{void ()* @none_spmd, !"kernel", i32 1}