rsmith added a comment. In D88518#2308085 <https://reviews.llvm.org/D88518#2308085>, @mstorsjo wrote:
> This broke use of setjmp for mingw on x86_64. Thanks, should be fixed in rG8fb2a235b0f22dedba72b8b559ba33171a8dcd09 <https://reviews.llvm.org/rG8fb2a235b0f22dedba72b8b559ba33171a8dcd09>. Now that we properly recognize `_setjmp` even if it has a weird signature, the MSVC-specific check for exactly one argument was breaking the MinGW case. I've removed that check and instead we only apply the custom MSVC-specific codegen rule if the declaration happens to be of the right form (but we apply the `returns_twice` attribute regardless). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D88518/new/ https://reviews.llvm.org/D88518 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
