[Openmp-commits] [PATCH] D97198: [OpenMP][Clang][NVPTX] Only build one bitcode library for each SM

[Shilei Tian via Phabricator via Openmp-commits]

tianshilei1992 created this revision.
tianshilei1992 added reviewers: jdoerfert, JonChesterfield, ABataev, grokos, ye-luo.
Herald added subscribers: guansong, yaxunl, mgorny.
tianshilei1992 requested review of this revision.
Herald added subscribers: openmp-commits, cfe-commits, sstefan1.
Herald added projects: clang, OpenMP.

In D97003 <https://reviews.llvm.org/D97003>, CUDA 9.2 is the minimum requirement for OpenMP offloading on
NVPTX target. We don't need to have macros in source code to select right functions
based on CUDA version. we don't need to compile multiple bitcode libraries of
different CUDA versions for each SM. We don't need to worry about future
compatibility with newer CUDA version.

`-target-feature +ptx60` is used in this patch, which corresponds to the highest
PTX version that CUDA 9.2 can support.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D97198

Files:
  clang/lib/Driver/ToolChains/Cuda.cpp
  clang/test/Driver/Inputs/libomptarget/libomptarget-nvptx-cuda_102-sm_35.bc
  clang/test/Driver/Inputs/libomptarget/libomptarget-nvptx-sm_35.bc
  clang/test/Driver/openmp-offload-gpu.c
  openmp/libomptarget/deviceRTLs/nvptx/CMakeLists.txt
  openmp/libomptarget/deviceRTLs/nvptx/src/target_impl.cu

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D97198.325457.patch
Type: text/x-patch
Size: 10678 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/openmp-commits/attachments/20210222/793e8ef3/attachment-0001.bin>