ilya-biryukov wrote: Some benchmarks show an even bigger regression of up to 40% (although we're not sure if they're clear from noise). We would want to look a little more into what's causing this and reland this in a way that does not cause significant regression or prove that it's impossible (hopefully that's not the case).
https://github.com/llvm/llvm-project/pull/123962 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
