-
Notifications
You must be signed in to change notification settings - Fork 1.7k
Rust: Fix bad join #19714
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Rust: Fix bad join #19714
Conversation
``` [2025-06-10 15:21:37] (83s) Tuple counts for TypeMention::PathTypeReprMention.aliasResolveTypeAt/1#dispred#ae16f47b#fbf#cpe#13/2@i5#5a5451wi after 57.7s: 205999 ~1028% {3} r1 = JOIN `TypeMention::TypeMention.resolveTypeAt/1#dispred#a125c821#bff#reorder_1_0_2#prev_delta` WITH `_TypeAlias::Generated::TypeAlias.getTypeRepr/0#dispred#5fd7e521_TypeMention::PathTypeReprMention#a35__#join_rhs#1` ON FIRST 2 OUTPUT Rhs.2, Lhs.2 'result', Rhs.3 'this' {3} | AND NOT `project#TypeMention::pathGetTypeParameter/2#fd301eed#2`(FIRST 2) 204000 ~982% {2} | SCAN OUTPUT In.2 'this', In.1 'result' 5499 ~448% {7} r2 = JOIN `TypeMention::TypeMention.resolveTypeAt/1#dispred#a125c821#bff#reorder_0_2_1#prev_delta` WITH `_GenericArgList::GenericArgList.getTypeArg/1#dispred#1c780027_PathSegment::Generated::PathSegment.ge__#join_rhs#1` ON FIRST 1 OUTPUT Rhs.1, Rhs.2 'this', Rhs.3, Lhs.1 'result', _, Lhs.2, _ {5} | REWRITE WITH Tmp.4 := "", Tmp.6 := "", Out.4 := InverseAppend(Tmp.4,In.5,Tmp.6) KEEPING 5 2500 ~398% {5} | SCAN OUTPUT In.0, In.2, In.4, In.1 'this', In.3 'result' 0 ~0% {2} | JOIN WITH `TypeMention::TypeMention.resolveTypeAt/1#dispred#a125c821#bff#reorder_0_2_1#prev` ON FIRST 3 OUTPUT Lhs.3 'this', Lhs.4 'result' 9000 ~489% {7} r3 = JOIN `TypeMention::TypeMention.resolveTypeAt/1#dispred#a125c821#bff#reorder_0_2_1#prev_delta` WITH `_PathSegment::Generated::PathSegment.getGenericArgList/0#dispred#e072653b__Path::Generated::Path.get__#join_rhs#1` ON FIRST 1 OUTPUT Rhs.1, Rhs.2 'this', Rhs.3, Lhs.1, _, Lhs.2, _ {5} | REWRITE WITH Tmp.4 := "", Tmp.6 := "", Out.4 := InverseAppend(In.5,Tmp.4,Tmp.6) KEEPING 5 5000 ~411% {5} | SCAN OUTPUT In.4, In.0, In.1 'this', In.2, In.3 10461919500 ~610% {6} | JOIN WITH `TypeMention::TypeMention.resolveTypeAt/1#dispred#a125c821#bff#reorder_1_0_2#prev` ON FIRST 1 OUTPUT Lhs.3, Rhs.1, Lhs.1, Lhs.2 'this', Lhs.4, Rhs.2 'result' 500 ~0% {5} | JOIN WITH `GenericArgList::GenericArgList.getTypeArg/1#dispred#1c780027_021#join_rhs` ON FIRST 2 OUTPUT Lhs.2, Rhs.2, Lhs.4, Lhs.3 'this', Lhs.5 'result' 0 ~0% {2} | JOIN WITH `TypeMention::pathGetTypeParameter/2#fd301eed` ON FIRST 3 OUTPUT Lhs.3 'this', Lhs.4 'result' 204000 ~982% {2} r4 = r1 UNION r2 UNION r3 203964 ~982% {2} | AND NOT `TypeMention::PathTypeReprMention.aliasResolveTypeAt/1#dispred#ae16f47b#fbf#cpe#13#prev`(FIRST 2) return r4 ```
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Pull Request Overview
This PR adds pragma[nomagic]
annotations to two methods in the Rust CodeQL QLL files to improve join behavior when running the SummaryStats.ql
query.
- Add
pragma[nomagic]
toPathTypeReprMention.aliasResolveTypeAt
- Add
pragma[nomagic]
toGenericArgListImpl.getTypeArg
Reviewed Changes
Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.
File | Description |
---|---|
rust/ql/lib/codeql/rust/internal/TypeMention.qll | Insert pragma[nomagic] before aliasResolveTypeAt |
rust/ql/lib/codeql/rust/elements/internal/GenericArgListImpl.qll | Insert pragma[nomagic] before getTypeArg |
Comments suppressed due to low confidence (2)
rust/ql/lib/codeql/rust/internal/TypeMention.qll:133
- The Javadoc comment above
aliasResolveTypeAt
is now separated from the method by thepragma
. This may prevent the comment from being associated with the method in generated docs—consider moving the comment block above the pragma or placing the pragma before the comment.
pragma[nomagic]
rust/ql/lib/codeql/rust/elements/internal/GenericArgListImpl.qll:29
- The documentation comment for
getTypeArg
is now separated by thepragma
, which may break its association with the method. Consider relocating the pragma or the comment so that the doc block immediately precedes the method signature.
pragma[nomagic]
#19584 follow-up. The bad join appears to only happen when running the
SummaryStats.ql
query.DCA reports an average 75 % analysis time speedup.