Improve deabstraction SSA promotion logic, fixing SR-8395 #18400
Conversation
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
- Handle the simple case of storing into a struct, to handle the common pattern of a store through a struct_element_addr into a Tensor. - Fix handling of begin_access to be a separate loop. Before, we were adding new entries to the use list that we're iterating over. This was almost fine (no invalid iterators or anything) but the entries are added to the start of the list so we wouldn't see them. - Fix a bug this exposed in graph lowering where we wouldn't handle aggregate wrappers around elements in a Tensor initializer.
the code used by TFPartition.
|
@swift-ci please test tensorflow |
1 similar comment
|
@swift-ci please test tensorflow |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
Improve deabstraction SSA promotion logic, fixing SR-8395
pattern of a store through a struct_element_addr into a Tensor.
adding new entries to the use list that we're iterating over. This
was almost fine (no invalid iterators or anything) but the entries
are added to the start of the list so we wouldn't see them.
aggregate wrappers around elements in a Tensor initializer.