This is an archive of the discontinued LLVM Phabricator instance.

[mlir][VectorToGPU] Fix support for i4, col-major operand support
ClosedPublic

Authored by christopherbate on Jun 17 2022, 10:51 AM.

Details

Summary

For the conversion to nvgpu mma.sync and ldmatrix pathways, the code
was missing support for the i4 data type. While fixing this, another
bug was discoverd that caused the number of ldmatrix tiles calculated for
certain operand types and configurations to be incorrect. This change
fixes both issues and adds additional tests.

Diff Detail

Event Timeline

christopherbate requested review of this revision.Jun 17 2022, 10:51 AM
ThomasRaoux accepted this revision.Jun 17 2022, 10:59 AM
This revision is now accepted and ready to land.Jun 17 2022, 10:59 AM

Add fix/test for tf32 col/col/row