On Jan 27, 2025, Richard Biener <richard.guent...@gmail.com> wrote: > What I was saying that the conservative tree_could_trap_p could say > 'yes' to a certain encoding of a ref but 'no' to another if in reality > the ref can never trap. We of course cannot (apart from bugs in > tree_could_trap_p) turn a for-sure trap into a not-trap by simply > rewriting the ref.
I see. Yeah, that makes sense. > So I think we want this bit in (and it's dependences), but 'k, I'll split that bit out, after some clarification: > (I see the assert is no longer in the patch). That's because the assert went in as part of an earlier patch. I take it it should be backed out along with the to-be-split-out bits above, right? -- Alexandre Oliva, happy hacker https://FSFLA.org/blogs/lxo/ Free Software Activist GNU Toolchain Engineer More tolerance and less prejudice are key for inclusion and diversity Excluding neuro-others for not behaving ""normal"" is *not* inclusive
