JIT: fix count reconstruction problem #100385
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
In large methods with lots of irreducible loops we may find reconstructed counts reaching very large values. Since profile counts in practice won't ever be much larger than say 10^12, detect when reconstructed counts exceed this value, and stop the algorithm. We may eventually decide to rerun in "hard blend" mode where we intentionally limit the edge likelihood ranges. But this should do for now. Closes dotnet#100350.
ghost
added
the
area-CodeGen-coreclr
|
@jakobbotsch PTAL No diffs expected |
|
/azp run runtime-coreclr libraries-jitstress, runtime-coreclr jitstress |
|
Azure Pipelines successfully started running 2 pipeline(s). |
|
/azp run runtime-coreclr libraries-jitstress, runtime-coreclr jitstress |
|
Azure Pipelines successfully started running 2 pipeline(s). |
|
Jitstress failure is timeout. @BruceForstall looks like Jakob is away -- can you take a look? |
BruceForstall
approved these changes
Mar 28, 2024
51 tasks
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Outdated suggestions cannot be applied.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.
In large methods with lots of irreducible loops we may find reconstructed counts reaching very large values.
Since profile counts in practice won't ever be much larger than say 10^12, detect when reconstructed counts exceed this value, and stop the algorithm.
We may eventually decide to rerun in "hard blend" mode where we intentionally limit the edge likelihood ranges. But this should do for now.
Closes #100350.