[llvm-branch-commits] [clang] Thread Safety Analysis: Track try-acquired capabilities as a ternary try-held state (PR #220635)

Aaron Puchert via llvm-branch-commits llvm-branch-commits at lists.llvm.org
Sun Nov 15 16:19:16 PST 2026


aaronpuchert wrote:

> What I did locally was create a new `FactEntry` kind for representing capabilities returned from `try_lock` calls, and the 3-state semantics only applies to this new kind of `FactEntry`. WDYT?

Scoped capabilities can also have try-acquire functions. I'm not sure how we handle this at the moment and what the best way is to handle it, but it seems having a capability conditionally acquired is orthogonal to whether it is scoped or not.

https://github.com/llvm/llvm-project/pull/220635


More information about the llvm-branch-commits mailing list