[clang] [analyzer] Fix FN for UCRT fopen ignoring the lack of `restrict` qualifier (PR #236747)

Balázs Benics via cfe-commits cfe-commits at lists.llvm.org
Tue Nov 3 01:32:48 PST 2026


steakhal wrote:

> I would also support never distinguishing functions that only differ in the presence of `restrict`. I think it is the more natural approach given that e.g. overloading also doesn't distinguish functions based on `restrict`.

This seems like a strong argument for dropping `restrict` matching. Why did we have that in the first place, we must have had some solid arguments for doing all of this work.

https://github.com/llvm/llvm-project/pull/236747


More information about the cfe-commits mailing list