[Openmp-commits] [PATCH] D88474: [nfc][libomptarget] Drop parameter to named_sync
Jon Chesterfield via Phabricator via Openmp-commits
openmp-commits at lists.llvm.org
Tue Sep 29 03:21:12 PDT 2020
JonChesterfield created this revision.
JonChesterfield added reviewers: jdoerfert, ABataev, grokos, jhuber6.
Herald added subscribers: openmp-commits, aaron.ballman, tpr, jvesely.
Herald added a project: OpenMP.
JonChesterfield requested review of this revision.
Herald added a subscriber: sstefan1.
[nfc][libomptarget] Drop parameter to named_sync
named_sync has one call site (in sync.cu) where it always passed L1_BARRIER.
Folding this into the call site and dropping the macro is a simplification.
amdgpu doesn't have ptx' bar.sync instruction. A correct implementation of
__kmpc_impl_named_sync in terms of shared memory is much easier if it can
assume that the barrier argument is this constant. Said implementation is left
for a second patch.
Repository:
rG LLVM Github Monorepo
https://reviews.llvm.org/D88474
Files:
openmp/libomptarget/deviceRTLs/amdgcn/src/target_impl.h
openmp/libomptarget/deviceRTLs/common/src/sync.cu
openmp/libomptarget/deviceRTLs/nvptx/src/target_impl.h
Index: openmp/libomptarget/deviceRTLs/nvptx/src/target_impl.h
===================================================================
--- openmp/libomptarget/deviceRTLs/nvptx/src/target_impl.h
+++ openmp/libomptarget/deviceRTLs/nvptx/src/target_impl.h
@@ -37,10 +37,6 @@
#define WARPSIZE 32
-// The named barrier for active parallel threads of a team in an L1 parallel
-// region to synchronize with each other.
-#define L1_BARRIER (1)
-
// Maximum number of preallocated arguments to an outlined parallel/simd function.
// Anything more requires dynamic memory allocation.
#define MAX_SHARED_ARGS 20
@@ -187,7 +183,10 @@
#endif // CUDA_VERSION
}
-INLINE void __kmpc_impl_named_sync(int barrier, uint32_t num_threads) {
+INLINE void __kmpc_impl_named_sync(uint32_t num_threads) {
+ // The named barrier for active parallel threads of a team in an L1 parallel
+ // region to synchronize with each other.
+ int barrier = 1;
asm volatile("bar.sync %0, %1;"
:
: "r"(barrier), "r"(num_threads)
Index: openmp/libomptarget/deviceRTLs/common/src/sync.cu
===================================================================
--- openmp/libomptarget/deviceRTLs/common/src/sync.cu
+++ openmp/libomptarget/deviceRTLs/common/src/sync.cu
@@ -60,8 +60,7 @@
PRINT(LD_SYNC,
"call kmpc_barrier with %d omp threads, sync parameter %d\n",
(int)numberOfActiveOMPThreads, (int)threads);
- // Barrier #1 is for synchronization among active threads.
- __kmpc_impl_named_sync(L1_BARRIER, threads);
+ __kmpc_impl_named_sync(threads);
}
} else {
// Still need to flush the memory per the standard.
Index: openmp/libomptarget/deviceRTLs/amdgcn/src/target_impl.h
===================================================================
--- openmp/libomptarget/deviceRTLs/amdgcn/src/target_impl.h
+++ openmp/libomptarget/deviceRTLs/amdgcn/src/target_impl.h
@@ -42,10 +42,6 @@
#define WARPSIZE 64
-// The named barrier for active parallel threads of a team in an L1 parallel
-// region to synchronize with each other.
-#define L1_BARRIER (1)
-
// Maximum number of preallocated arguments to an outlined parallel/simd
// function. Anything more requires dynamic memory allocation.
#define MAX_SHARED_ARGS 20
@@ -113,10 +109,9 @@
// AMDGCN doesn't need to sync threads in a warp
}
-INLINE void __kmpc_impl_named_sync(int barrier, uint32_t num_threads) {
- // we have protected the master warp from releasing from its barrier
- // due to a full workgroup barrier in the middle of a work function.
- // So it is ok to issue a full workgroup barrier here.
+INLINE void __kmpc_impl_named_sync(uint32_t num_threads) {
+ (void)num_threads;
+ // TODO: Implement on top of __SHARED__
__builtin_amdgcn_s_barrier();
}
-------------- next part --------------
A non-text attachment was scrubbed...
Name: D88474.294921.patch
Type: text/x-patch
Size: 2818 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/openmp-commits/attachments/20200929/ee45ca1f/attachment.bin>
More information about the Openmp-commits
mailing list