-
Notifications
You must be signed in to change notification settings - Fork 1.4k
Update ROOT and Cling to LLVM20 #19242
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
Conversation
Test Results 22 files 22 suites 4d 0h 9m 15s ⏱️ Results for commit 0204fdd. ♻️ This comment has been updated with latest results. |
9ee153a
to
253faee
Compare
b01f14c
to
d2749b8
Compare
4e6fac4
to
1731a44
Compare
67fd319
to
c090439
Compare
c090439
to
818bfc3
Compare
@smuzaffar, the CI looks good and everything seems to be working reasonably well with the LLVM 20 upgrade. Could you check this PR against CMSSW? |
@devajithvs , cms-sw#227 is running cmssw CI tests for this change |
cmssw CI tests cms-sw#227 (comment) look good |
TemplateArgument can return a QualType that may contain unqualified parts. Before (incorrect): __gnu_cxx::__common_pool_policy<__pool, true> After (correct): __gnu_cxx::__common_pool_policy<__gnu_cxx::__pool, true>
LLVM 20 produce `RecoveryExpr` for dependent errors where LLVM 18 returned `ExprError`. In our type/name lookup path, this can leave a <recovery-expr> node and later codeGen asks the AST to constant-evaluate that initializer and we hit this assertion: `VarDecl::evaluateValueImpl: !Init->isValueDependent()`.
Avoid cases like: avoid `class__attribute__` with LLVM20 Fixes the failure on alma9: `gtest-tree-dataframe-dataframe-snapshot-emptyoutput`
Skip dumping TopLevelStmt if they are invalid. The change llvm/llvm-project@4b70d17bcf caused an invalid `TopLevelStatement` to be added to the AST, which can trigger assertions when dumping.
Also temporarily failing disable `testHashRecursiveRemove` test on Windows
template_arguments() has the types, but getTemplateArgs() has the pack argument, so their counts may be different. Add test for failing CMS build with LLVM 20. Co-authored-by: Devajith Valaparambil Sreeramaswamy <[email protected]>
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.
Great work! LGTM!
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.
Looks all good!
This Pull request:
Update ROOT and Cling to LLVM20:
TODO:
Changes or fixes:
Checklist:
This PR fixes #