[llvm] [CodeGen][AArch64] Set min jump table entries to 13 for AArch64 targets (PR #71166)
David Sherwood via llvm-commits
llvm-commits at lists.llvm.org
Wed Nov 15 02:11:30 PST 2023
david-arm wrote:
> I think it's actually noise, don't think I see improvements on the V2. Perhaps that's still okay because it is shows improvements on the V1. But now I am actually wondering if the V1 improvement is an actual improvement, or do we get lucky with secondary effects such as different function/loop/block alignments?
I ran this quite a few times on neoverse-v1 hardware and consistently saw an improvement, particularly with https://github.com/llvm/llvm-project/pull/72132 applied as well. Although I appreciate there is a lot of noise in perlbench. From observation it seems that GCC also takes a similar approach when deciding to create jump tables for perlbench.
https://github.com/llvm/llvm-project/pull/71166
More information about the llvm-commits
mailing list