This is an archive of the discontinued LLVM Phabricator instance.

[mlir] [VectorOps] [integration_test] Sparse matrix times vector (DOT version)
ClosedPublic

Authored by aartbik on Jul 24 2020, 5:21 PM.

Details

Summary

Integration test that illustrates the gather operation with a
real-world operation expressed in mostly the Vector dialect.
Uses jagged diagonal storage.

Diff Detail

Event Timeline

aartbik created this revision.Jul 24 2020, 5:21 PM
aartbik retitled this revision from [mlir] [VectorOps] Sparse matrix times vector (DOT version) to [mlir] [VectorOps] [integration_test] Sparse matrix times vector (DOT version).Jul 24 2020, 5:31 PM
bondhugula accepted this revision.Jul 26 2020, 12:46 AM
bondhugula added a subscriber: bondhugula.
bondhugula added inline comments.
mlir/integration_test/Dialect/Vector/CPU/test-sparse-dot-matvec.mlir
1–4

Nit: for my information, why aren't we having all on one line? Looks more readable that way since we are anyway breaking at arbitrary places.

28

Rephrase the last part and terminate sentence.

233–257

Should these all be CHECK-NEXT's?

This revision is now accepted and ready to land.Jul 26 2020, 12:46 AM
aartbik updated this revision to Diff 280987.Jul 27 2020, 11:18 AM
aartbik marked 4 inline comments as done.

addressed comments

mlir/integration_test/Dialect/Vector/CPU/test-sparse-dot-matvec.mlir
1–4

I guess I have been doing this out of 80 col habit. In my vi editor, this version reads a bit easier than the wrap-around version

This revision was automatically updated to reflect the committed changes.