rjmccall added a comment. In https://reviews.llvm.org/D44883#1048485, @rjmccall wrote:
> In https://reviews.llvm.org/D44883#1048439, @lebedev.ri wrote: > > > In https://reviews.llvm.org/D44883#1048400, @rjmccall wrote: > > > > > Yeah, like I said, I'm just worried about the usability of having > > > multiple dimensions of warning group here. Like, do we need both > > > -Wself-assign-field-builtin and -Wself-assign-field-overloaded? > > > > > > I'm not saying that is not a valid concern. I'm simply following the > > pre-existing practice, which is, as far i'm aware, to split the diag groups > > if it makes sense. > > > I agree. In general, I think this would be fine; my only concern is because > we do already have some splitting along a different dimension, so we do need > to stop and think about it. Maybe the answer is that Sorry, I didn't finish this thought. Maybe we should just treat -Wself-assign-field as an accident of history; we can roll it generally into -Wself-assign and just allow that specific warning to be separately enabled if they want. Or if it isn't a very old flag, maybe we can just remove it entirely. Repository: rC Clang https://reviews.llvm.org/D44883 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
