[libc-commits] [PATCH] D146725: [libc] Implement memory fences on NVPTX

[Joseph Huber via Phabricator via libc-commits]

jhuber6 created this revision.
jhuber6 added reviewers: JonChesterfield, jdoerfert, tianshilei1992, sivachandra, tra.
Herald added subscribers: libc-commits, mattd, gchakrabarti, asavonic, ecnelises, tschuett.
Herald added projects: libc-project, All.
jhuber6 requested review of this revision.

Memory fences are not handled by the NVPTX backend. We need to replace
them with a memory barrier intrinsic function. This doesn't include the
ordering, but should perform the necessary functionality, albeit slower.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D146725

Files:
  libc/src/__support/CPP/atomic.h


Index: libc/src/__support/CPP/atomic.h
===================================================================
--- libc/src/__support/CPP/atomic.h
+++ libc/src/__support/CPP/atomic.h
@@ -10,6 +10,7 @@
 #define LLVM_LIBC_SRC_SUPPORT_CPP_ATOMIC_H
 
 #include "src/__support/macros/attributes.h"
+#include "src/__support/macros/properties/architectures.h"
 
 #include "type_traits.h"
 
@@ -96,7 +97,14 @@
 
 // Issue a thread fence with the given memory ordering.
 LIBC_INLINE void atomic_thread_fence(MemoryOrder mem_ord) {
+// The NVPTX backend currently does not support atomic thread fences so we use a
+// full system fence instead.
+#ifdef LIBC_TARGET_ARCH_IS_NVPTX
+  (void)mem_ord;
+  __nvvm_membar_sys();
+#else
   __atomic_thread_fence(int(mem_ord));
+#endif
 }
 
 } // namespace cpp


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D146725.507741.patch
Type: text/x-patch
Size: 787 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/libc-commits/attachments/20230323/1c038c31/attachment.bin>