nikic added a comment. In D99121#2650146 <https://reviews.llvm.org/D99121#2650146>, @nlopes wrote:
> Given that it gives a decent speedup for some important workload, and > provided it doesn't regress others [...] I think that second point is going to need some evidence (at least in the form of transform stats I guess). D96881 <https://reviews.llvm.org/D96881> is another recent attempt to work around pointer types, and it looks like that one did regress others. This patch already shows regressions inside InstCombine itself, so that's not a great sign. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D99121/new/ https://reviews.llvm.org/D99121 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
