erichkeane wrote: > > I'm not sure I buy the justification > > FWIW, this came from a post-commit review comment: [#137166 > (comment)](https://github.com/llvm/llvm-project/pull/137166#discussion_r2070615023) > > > , and the flag-diag there is perhaps 'cute' (and frankly, if we find > > ourself doing that again, I would vastly prefer we do something like this > > in tablegen/Diagnostics directly to auto generate/check these things in a > > way that cannot be called. > > Yeah, at some point we may want to update the diagnostics engine so it can > track groups directly instead of by warning, or at the very least, hide this > kind of hack behind an interface on `DiagnosticsEngine`.
I'd seen that, I still am not sure I buy the justification anyway, but if others want it and you're sympathetic, I am not against it. The Diagnostics hack is wacky/hurts me, but this is the one 'pass' :D https://github.com/llvm/llvm-project/pull/138266 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
