This is an archive of the discontinued LLVM Phabricator instance.

[mlir] Use dynamic_tensor_from_elements in shape.broadcast conversion
ClosedPublic

Authored by silvas on Oct 18 2020, 9:43 PM.

Details

Summary

Now, convert-shape-to-std doesn't internally create memrefs, which was
previously a bit of a layering violation. The conversion to memrefs
should logically happen as part of bufferization.

Diff Detail

Event Timeline

silvas created this revision.Oct 18 2020, 9:43 PM
silvas requested review of this revision.Oct 18 2020, 9:43 PM
herhut accepted this revision.Oct 19 2020, 4:45 AM

Thanks for cleaning this up!

mlir/test/Conversion/ShapeToStandard/shape-to-standard.mlir
309

Nit: indentation

This revision is now accepted and ready to land.Oct 19 2020, 4:45 AM
silvas updated this revision to Diff 299179.Oct 19 2020, 2:37 PM
silvas marked an inline comment as done.

update