rniwa wrote: > Looks harmless. I wonder what happens if you add this attribute but the > function definition violates this. I suspect that such a mismatch wouldn't be > detected, thus would remain in silence. Would that be possible/worth it to > check this attribute if we have a definition? So that we could say, that this > attribute needs to be in the header, but the implementation would also > include this and check if it's in-sync with the definition. WDYT?
Yeah, I was thinking that it ought to be a new checker. Should I include it in this PR? https://github.com/llvm/llvm-project/pull/177839 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
