diff options
| author | Jakub Chlanda <jakub@codeplay.com> | 2023-09-27 08:51:26 +0200 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2023-09-27 08:51:26 +0200 |
| commit | dfab31b41b4988b6dc8129840eba68f0c36c0f13 (patch) | |
| tree | 38cc342bf229fc35e8edadd229f7ccaf86fd1dcf /clang/lib/Parse/ParseOpenMP.cpp | |
| parent | d1653c8e9ba9fa23add726e1c772419cb968d0fa (diff) | |
| download | llvm-dfab31b41b4988b6dc8129840eba68f0c36c0f13.zip llvm-dfab31b41b4988b6dc8129840eba68f0c36c0f13.tar.gz llvm-dfab31b41b4988b6dc8129840eba68f0c36c0f13.tar.bz2 | |
[NVPTX] Add support for maxclusterrank in launch_bounds (#66496)
Since SM_90 CUDA supports specifying additional argument to the
launch_bounds attribute: maxBlocksPerCluster, to express the maximum
number of CTAs that can be part of the cluster. See:
https://docs.nvidia.com/cuda/parallel-thread-execution/index.html#cluster-dimension-directives-maxclusterrank
and
https://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#launch-bounds
for details.
Diffstat (limited to 'clang/lib/Parse/ParseOpenMP.cpp')
| -rw-r--r-- | clang/lib/Parse/ParseOpenMP.cpp | 3 |
1 files changed, 2 insertions, 1 deletions
diff --git a/clang/lib/Parse/ParseOpenMP.cpp b/clang/lib/Parse/ParseOpenMP.cpp index 605b976..8a8a126 100644 --- a/clang/lib/Parse/ParseOpenMP.cpp +++ b/clang/lib/Parse/ParseOpenMP.cpp @@ -3739,7 +3739,8 @@ OMPClause *Parser::ParseOpenMPOMPXAttributesClause(bool ParseOnly) { continue; if (auto *A = Actions.CreateLaunchBoundsAttr( PA, PA.getArgAsExpr(0), - PA.getNumArgs() > 1 ? PA.getArgAsExpr(1) : nullptr)) + PA.getNumArgs() > 1 ? PA.getArgAsExpr(1) : nullptr, + PA.getNumArgs() > 2 ? PA.getArgAsExpr(2) : nullptr)) Attrs.push_back(A); continue; default: |