Kernel IR refactoring: part 4.1#284
Merged
tlemo merged 59 commits into20_7_6_develfrom Aug 12, 2020
Merged
Conversation
This reverts commit fc09c1b5a7240701da093406753908eba6f41e1d.
csarofeen
requested changes
Aug 11, 2020
Owner
csarofeen
left a comment
There was a problem hiding this comment.
Blocking for now as Indexing is being rebased.
csarofeen
approved these changes
Aug 12, 2020
|
|
||
| if (a->buffer()->getValType().value() == ValType::KirTensorView) { | ||
| const auto tv = a->buffer()->as<kir::TensorView>(); | ||
| TORCH_INTERNAL_ASSERT(tv->domain()->nDims() > 0); |
Owner
There was a problem hiding this comment.
I'm not 100% certain we won't actually send a 0-dim tensor view to this. That would indicate it's just a scalar.
Owner
There was a problem hiding this comment.
Actually it looks like we were supporting it.
Collaborator
Author
There was a problem hiding this comment.
Good question. Yes, the code was in there but never used as far as I can tell.
More importantly, this is a good example where we can consider different trade-offs between Fusion IR and Kernel IR: in the former the benefit of treating 0-dimensional Tensors as scalars makes sense. We can, and should, eliminate this duality of representing scalars during lowering.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
A small followup to the previous iteration, completing a few unfinished parts around kir::Allocate.