This is an archive of the discontinued LLVM Phabricator instance.

[mlir][sparse] Moving <P,I,V>-invariant parts of SparseTensorStorage to base
ClosedPublic

Authored by wrengr on Apr 1 2022, 11:55 AM.

Details

Summary

This reorganization helps to clean up the changes needed for D122060.

Work towards fixing: https://github.com/llvm/llvm-project/issues/51652

Depends On D122625

Diff Detail

Event Timeline

wrengr created this revision.Apr 1 2022, 11:55 AM
wrengr requested review of this revision.Apr 1 2022, 11:55 AM
wrengr updated this revision to Diff 419826.Apr 1 2022, 12:02 PM

re-adding todo lost when splitting the diff

wrengr added inline comments.Apr 6 2022, 4:13 PM
mlir/lib/ExecutionEngine/SparseTensorUtils.cpp
318–321

Note to self: After everything else in this CL gets reviewed, be sure to adjust this to match D123166 and any subsequent changes between now and then.

wrengr marked an inline comment as not done.Apr 6 2022, 4:14 PM
wrengr updated this revision to Diff 421047.Apr 6 2022, 5:01 PM

Rebasing for D123166. Also removing a bunch of inline keywords, per MLIR style-guide.

wrengr updated this revision to Diff 421058.Apr 6 2022, 6:09 PM

Using checkedMul in finalizeSegment, now that D123166 renders it potentially unsafe.

aartbik accepted this revision.Apr 8 2022, 10:05 AM

Thanks for breaking this out in a separate CL.

mlir/lib/ExecutionEngine/SparseTensorUtils.cpp
173–178

This is clear doc!

This revision is now accepted and ready to land.Apr 8 2022, 10:05 AM