[Openmp-commits] [PATCH] D97195: [OpenMP][NVPTX] Fixed a compilation error in deviceRTLs caused by unsupported feature in release verion of LLVM

[Shilei Tian via Phabricator via Openmp-commits]

tianshilei1992 created this revision.
tianshilei1992 added reviewers: jdoerfert, JonChesterfield, protze.joachim.
Herald added subscribers: guansong, yaxunl, mgorny.
tianshilei1992 requested review of this revision.
Herald added subscribers: openmp-commits, sstefan1.
Herald added a project: OpenMP.

`ptx71` is not supported in release version of LLVM yet. As a result,
the support of CUDA 11.2 and CUDA 11.1 caused a compilation error as mentioned
in D97004 <https://reviews.llvm.org/D97004>. Since the support in D97004 <https://reviews.llvm.org/D97004> is just a WA for releease, and we'll not
use it in the near future, using `ptx70` for CUDA 11 is feasible.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D97195

Files:
  openmp/libomptarget/deviceRTLs/nvptx/CMakeLists.txt


Index: openmp/libomptarget/deviceRTLs/nvptx/CMakeLists.txt
===================================================================
--- openmp/libomptarget/deviceRTLs/nvptx/CMakeLists.txt
+++ openmp/libomptarget/deviceRTLs/nvptx/CMakeLists.txt
@@ -153,7 +153,7 @@
 # This map is from clang/lib/Driver/ToolChains/Cuda.cpp.
 # The last element is the default case.
 set(cuda_version_list 112 111 110 102 101 100 92 91 90 80)
-set(ptx_feature_list 71 71 70 65 64 63 61 61 60 42)
+set(ptx_feature_list 70 70 70 65 64 63 61 61 60 42)
 # The following two lines of ugly code is not needed when the minimal CMake
 # version requirement is 3.17+.
 list(LENGTH cuda_version_list num_version_supported)


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D97195.325447.patch
Type: text/x-patch
Size: 688 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/openmp-commits/attachments/20210222/4a0fc184/attachment.bin>