AaronBallman wrote: > > We could maybe presume glibc on ELF targets > > Isn't that part of the triple? `aarch64-linux-gnu` gnu -> glibc (as opposed > to `aarch64-linux-musl` -> musl, or `aarch64-linux-android` -> bionic)
Oh -- I wasn't aware we tracked that in the triple, that's awesome! I'll give that a shot. > Hopefully the number of constructors that expect getting arguments is > extremely rare. But I do worry that there is code out there that assumes > glibc behaviour, so it seems to me that err_ctor_dtor_attr_on_non_void_func > needs to be a warning-as-error that can be downgraded? Hopefully we can avoid emitting the diagnostic at all for glibc, then we can leave it as a hard error for the other cases. My concern about turning that into a warning which defaults to an error is that a less thorough programmer may shut the warning up rather than realize their CRT doesn't support that. https://github.com/llvm/llvm-project/pull/67360 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
