Re: [PATCH] D14653: [libcxx] Introduce the mechanism for fixing -fno-exceptions test failures.

scott-0 added a subscriber: scott-0. scott-0 added a comment. It would be nice if the no-exceptions library called an intermediate helper, for example `__libcxx_noexceptions_abort()`, instead of calling `abort()` directly. Then the user and the tests could provide a replacement for `__libcxx_n

Re: [PATCH] D12366: Avoid unnecessarily storing vtable pointers in more destructor cases

scott-0 added a comment. Thanks for the review and advice; I'll give `undef` a try. It's a much simpler approach. http://reviews.llvm.org/D12366 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo

[PATCH] D12366: Avoid unnecessarily storing vtable pointers in more destructor cases

scott-0 created this revision. scott-0 added a subscriber: cfe-commits. This makes an existing optimization apply in more cases. http://reviews.llvm.org/D12366 Files: lib/CodeGen/CGClass.cpp test/CodeGenCXX/destructors.cpp Index: test/CodeGenCXX/destructors.cpp =

[PATCH] D12365: Make SideEffectFinder publicly available (nfc)

scott-0 created this revision. scott-0 added a subscriber: cfe-commits. I'm going to use SideEffectFinder in my subsequent patch. http://reviews.llvm.org/D12365 Files: include/clang/AST/SideEffectFinder.h lib/AST/Expr.cpp Index: lib/AST/Expr.cpp =

RE: [PATCH] RE: [cfe-dev] missing return statement for non-void functions in C++

> Please see updated patch file attached, which now includes a fixed/added > regression test. I think it's a good idea; two minor remarks: +if (!CGM.getCodeGenOpts().OptimizeSize) { Should this be 'else if'? I imagine there's no use emitting a trap after the sanitizer has emitted a missing_