Draft: [LV] Outer-loop vectorization in the default vectorizer codepath #128202
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
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Outdated suggestions cannot be applied.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.
This is a draft MR to get feedback if something like this would be considered
a good-enough approach by current maintainers to merge into LLVM. I would split
it into smaller pieces if the general direction is not conflicting with current plans.
It implements outer-loop vectorization outside the VPlan-native path. Minimal
LoopAccessAnalysis support for non-innermost loops was added relying on the
!llvm.loop.parallel_accessesmetadata.Unlike for the VPlan-native path, inner loops with non-invariant trip-counts or
non-uniform inductions are supported, and the quality of the emitted code is better
than that of the current VPlan-native path.
A implementation very close to this one (#124432 required some changes but
also simplified this MR a lot) was successfully tested in combination with
basic LAA MemoryDepChecker support (not part of this MR) for outer loops on the
llvm-test-suite and SPEC (~3000 loops, outer-loop vect. was forced).
As a real-world motivational example, this loop
can be looked at. Performance is more than doubled when outer-loop vectorizing it.
Some code for the VPWidenPHIRecipes is duplicated from #128187.