diff options
| author | Baranov Victor <bar.victor.2002@gmail.com> | 2025-03-28 16:33:58 +0300 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2025-03-28 21:33:58 +0800 |
| commit | d6dcd985c03bec4b77872be00a81d92454fedc32 (patch) | |
| tree | 9e514a00cf4d36ed47411fb68eb0ffe78401479b /clang/lib/Frontend/CompilerInvocation.cpp | |
| parent | ecdbd26ba4b6017200b7773d41a270aa1fa2bd71 (diff) | |
| download | llvm-d6dcd985c03bec4b77872be00a81d92454fedc32.zip llvm-d6dcd985c03bec4b77872be00a81d92454fedc32.tar.gz llvm-d6dcd985c03bec4b77872be00a81d92454fedc32.tar.bz2 | |
[clang-tidy] Fix `thread_local` false positives in `misc-use-internal-linkage` check (#132573)
Based on C++ standard (see issue
https://github.com/llvm/llvm-project/issues/131679) and
[StackOverflow](https://stackoverflow.com/questions/22794382/are-c11-thread-local-variables-automatically-static)
`thread_local` variables are implicitly `static` so we should not
suggest adding `static` on a `thread_local` variables. I'd appreciate if
someone else will confirm this too because reading standard is tricky.
However, many people still use `static` and `thread_local` together:
[github
code-search](https://github.com/search?type=code&q=%22static+thread_local%22+language%3AC%2B%2B).
Maybe disabling warnings on `thread_local` should be made as a flag?
WDYT?
Closes https://github.com/llvm/llvm-project/issues/131679.
Diffstat (limited to 'clang/lib/Frontend/CompilerInvocation.cpp')
0 files changed, 0 insertions, 0 deletions