efriedma-quic wrote: > We usually use Tuningxxx due to hardware defects, which implies some hardware > preform good while some bad. But here the problem is we haven't fully address > regression to make it a net profit for performance.
I don't really see a difference between the two: if there's a compiler deficiency, that's still a code generation problem, not a issue with the underlying hardware feature. Messing with the hardware features has other effects, like turning off assembler support. https://github.com/llvm/llvm-project/pull/153751 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
