JIT: Don't run fgComputeBlockWeights in MinOpts
#110451
Merged
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
ghost
added
the
area-CodeGen-coreclr
Contributor
|
Tagging subscribers to this area: @JulieLeeMSFT, @jakobbotsch |
Contributor
Author
|
cc @dotnet/jit-contrib, @AndyAyersMS PTAL. Small MinOpts diffs from what looks like different block placement: This seems to stem from some blocks (not) being marked as rarely-run now, leading |
This was referenced Dec 5, 2024
AndyAyersMS
approved these changes
Dec 6, 2024
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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Outdated suggestions cannot be applied.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.
Block weights don't seem particularly useful in MinOpts. Skipping
fgComputeBlockWeightswhen we aren't optimizing had few diffs locally, so this looks like an opportunity to save some TP.(As an aside, I suspect we can soon replace
fgComputeBlockWeightsandoptSetBlockWeightswith profile synthesis. After the existing flow opt and unreachable block removal passes, I think we ought to compute the loop data structures, run profile synthesis, and then run loop opts. Since earlier phases don't seem too dependent on profile fidelity, but loop inversion/cloning/etc. stand to benefit from it, this is a natural checkpoint for running synthesis.)