Fix performance issue in ResourceLeakAnalysis.qll #862
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
There was a problem hiding this comment.
PR Overview
This pull request updates the change note for the ResourceLeakAnalysis.qll performance fix, documenting the removal of the recursive alias lookup and the switch from GlobalValueNumbering to HashCons-based alias comparison to improve performance.
- Updated change note to reflect aliasing changes for improved performance
- Clarified behavior differences and performance impact for RULE-22-16, ERR57-CPP, and A15-1-4
Reviewed Changes
| File | Description |
|---|---|
| change_notes/2025-02-20-rule-22-16-update-aliasing-for-performance.md | Updated change note documenting aliasing updates and performance improvements |
Copilot reviewed 3 out of 3 changed files in this pull request and generated no comments.
Tip: Copilot code review supports C#, Go, Java, JavaScript, Markdown, Python, Ruby and TypeScript, with more languages coming soon. Learn more
|
/test-performance |
|
🏁 Beep Boop! Performance testing for this PR has been initiated. Please check back later for results. Note that the query package generation step must complete before testing will start so it might be a minute. |
|
🏁 Beep Boop! Performance testing complete! See below for performance of the last 3 runs vs your PR. Times are based on predicate performance. You can find full graphs and stats in the PR that was created for this test in the release engineering repo. 🏁 Below are the slowest predicates for the last 2 releases vs this PR. |
knewbury01
reviewed
Mar 5, 2025
There was a problem hiding this comment.
@MichaelRFairhurst maybe not totally necessary - but can we rm this commented out line in this PR too? https://github.com/github/codeql-coding-standards/blob/main/c/misra/src/rules/RULE-22-16/MutexObjectsNotAlwaysUnlocked.ql#L48
we can also check if that actually should be commented out or not at the same time as this perf improvement? is probably just something leftover but good to check!
…-qll-performance-issue
knewbury01
approved these changes
Mar 10, 2025
knewbury01
deleted the
michaelrfairhurst/fix-resource-leak-analysis-qll-performance-issue
branch
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.
Description
After trying a lot of things, this is where I ended up, but open to additional ideas.
Regarding
getAnAliasRecursivewhich this PR removes: The intention was to ensure that if code locksxwhich is an alias ofy, and there's an unlock call onzwhich is alias ofy(but notx), then we don't want to report that. This wasn't previously tested. The newly added test fails even withgetAnAliasRecursive()as its incorrectly defined. Further, since the identity of an allocation (TJustResource) is based on aDataFlow::Node, and the out arg dataflow node is different from the expr dataflow node, it's not trivial to handle this edge case.Constraining
getAnAlias(node)isn't simply constrained toisAllocate(node)for two minor reasons:Self::, it would require additional refactoring and increase complexitygetAnAliasRecursiveor something like it. This could require something likeConfig::getAnAlias+(node), so that can't require its argument to be an allocate in that case.I then experimented with
inline_late,pragma[only_bind_into](...), and other various options, but none were providing a benefit. Eventually I copiedgetAnAlias()intoResourceLeakAnalysis.qlland noticed that no matter what changes I made, I couldn't get good performance from it on opencv as long as it used GVN, and I didn't have the same issue withhashCons. I did also measure a slight performance improvement by addingbindingset[node]togetAnAlias()and so included that as well.Change request type
.ql,.qll,.qlsor unit tests)Rules with added or modified queries
RULE-22-16ERR57-CPPA15-1-4Release change checklist
A change note (development_handbook.md#change-notes) is required for any pull request which modifies:
If you are only adding new rule queries, a change note is not required.
Author: Is a change note required?
🚨🚨🚨
Reviewer: Confirm that format of shared queries (not the .qll file, the
.ql file that imports it) is valid by running them within VS Code.
Reviewer: Confirm that either a change note is not required or the change note is required and has been added.
Query development review checklist
For PRs that add new queries or modify existing queries, the following checklist should be completed by both the author and reviewer:
Author
As a rule of thumb, predicates specific to the query should take no more than 1 minute, and for simple queries be under 10 seconds. If this is not the case, this should be highlighted and agreed in the code review process.
Reviewer
As a rule of thumb, predicates specific to the query should take no more than 1 minute, and for simple queries be under 10 seconds. If this is not the case, this should be highlighted and agreed in the code review process.