[Mosaic TPU] Improve vector layout inference for vector.shape_cast #23792
+67
−55
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
[Mosaic TPU] Improve vector layout inference for vector.shape_cast
This both simplifies and strengthens the pass. The loop over candidate result implicit
dims was not useful beyond the initial no-op version. I don't think that there's any
implicit dim that would in fact satisfy the checks in a way that the current implementation
doesn't.
What's more important, however, is that the previous implementation never tried to alter
the implicit dim of the input layout, which caused unnecessary failures when inputs came
in with an implicit_dim. Now the rules are more explicit and will proactively adjust the
input layout to make the shape_cast realizable.