[PATCH] D140264: [OpenMP] Improve AMDGPU Plugin

[Johannes Doerfert via Phabricator via llvm-commits]

jdoerfert created this revision.
jdoerfert added reviewers: ye-luo, kevinsala, jhuber6, tianshilei1992, JonChesterfield, carlo.bertolli.
Herald added subscribers: kosarev, kerbowa, guansong, bollu, tpr, dstuttard, yaxunl, jvesely, kzhuravl.
Herald added a project: All.
jdoerfert requested review of this revision.
Herald added subscribers: llvm-commits, sstefan1, wdng.
Herald added a project: LLVM.

With this patch we:

- pick more sensible defaults for the number of teams, inspired by the old plugin, and configured via LIBOMPTARGET_AMDGPU_TEAMS_PER_CU.
- check the input signal of a kernel launch late, after the queue lock was taken, to avoid a barrier packet more often.
- copy the kernel arguments in one swoop into the appropriate memory.
- manually specialize the callbacks to avoid potential indirect calls.


https://reviews.llvm.org/D140264

Files:
  llvm/include/llvm/Frontend/OpenMP/OMPGridValues.h
  openmp/libomptarget/plugins-nextgen/amdgpu/src/rtl.cpp
  openmp/libomptarget/plugins-nextgen/common/PluginInterface/PluginInterface.cpp
  openmp/libomptarget/plugins-nextgen/common/PluginInterface/PluginInterface.h

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D140264.483774.patch
Type: text/x-patch
Size: 9888 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20221217/8477b491/attachment.bin>