This is an archive of the discontinued LLVM Phabricator instance.

Changed wrong ROCDL instructions in GPU lowering.
ClosedPublic

Authored by dfki-jugr on Jan 28 2020, 2:14 AM.

Details

Summary

In the scope of the lowering phase from GPU to ROCDL, the intructions for the conversion patterns seems to be wrong.
According to https://github.com/ROCm-Developer-Tools/HIP/blob/master/include/hip/hcc_detail/math_fwd.h the instructions need two underscores in the beginning instead of one.

Diff Detail

Event Timeline

dfki-jugr created this revision.Jan 28 2020, 2:14 AM
herhut accepted this revision.Jan 28 2020, 2:20 AM

Nice catch!

We do not currently have end-to-end tests for these but as this shows we should.

This revision is now accepted and ready to land.Jan 28 2020, 2:20 AM

Unit tests: fail. 62254 tests passed, 1 failed and 827 were skipped.

failed: MLIR.Conversion/GPUToROCDL/gpu-to-rocdl.mlir

clang-tidy: pass.

clang-format: fail. Please format your changes with clang-format by running git-clang-format HEAD^ or applying this patch.

Build artifacts: diff.json, clang-tidy.txt, clang-format.patch, CMakeCache.txt, console-log.txt, test-results.xml

Pre-merge checks is in beta. Report issue. Please join beta or enable it for your project.

rriddle requested changes to this revision.Jan 28 2020, 9:16 AM

Please fix the clang format issue and add/fix the test.

This revision now requires changes to proceed.Jan 28 2020, 9:16 AM

Fixed test cases and format issues.

Unit tests: pass. 62255 tests passed, 0 failed and 827 were skipped.

clang-tidy: pass.

clang-format: pass.

Build artifacts: diff.json, clang-tidy.txt, clang-format.patch, CMakeCache.txt, console-log.txt, test-results.xml

Pre-merge checks is in beta. Report issue. Please join beta or enable it for your project.

rriddle accepted this revision.Jan 29 2020, 11:34 PM
This revision is now accepted and ready to land.Jan 29 2020, 11:34 PM
This revision was automatically updated to reflect the committed changes.