This is an archive of the discontinued LLVM Phabricator instance.

[mlir] Preserve lexicographic order after loop collapsing.
ClosedPublic

Authored by pifon2a on Jan 29 2021, 11:52 AM.

Details

Summary

Currently, for a scf.parallel (i,j,k) after the loop collapsing to 1D is done, the
IVs would be traversed as for an scf.parallel(k,j,i).

Diff Detail

Event Timeline

pifon2a created this revision.Jan 29 2021, 11:52 AM
pifon2a requested review of this revision.Jan 29 2021, 11:52 AM
bkramer accepted this revision.Jan 29 2021, 11:55 AM
bkramer added a subscriber: bkramer.

looks good

This revision is now accepted and ready to land.Jan 29 2021, 11:55 AM
pifon2a updated this revision to Diff 320189.Jan 29 2021, 12:27 PM

Fix single-parallel-loop-collapsing.mlir.test

This revision was landed with ongoing or failed builds.Jan 29 2021, 12:33 PM
This revision was automatically updated to reflect the committed changes.