kovdan01 wrote: > It's also possible that merging `PointerAuthQualifier` into `Mask` inside > `Qualifiers` would make some of the low-level handling more efficient.
I'm now applying this approach and it looks like that the overhead becomes significantly lower. A new invariant has to be introduced: with `PointerAuthenticationMode::None`, the whole `Data` field of `PointerAuthQualifier` must be 0 (which looks reasonable as for me). I'll publish the changes as soon as I double-check that the observable behavior is not changed and obtain final compiler-time-tracker results - so, stay tuned https://github.com/llvm/llvm-project/pull/84384 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
