Thanks for letting me know. I could use some help to repro because I'm not
very familiar with that benchmark or ARM32.
1. Can you provide the unoptimized IR for "BZ2_decompress"?
2. What is the particular flavor/CPU of ARM32 to target?
3. Was there a speed regression in addition to the size regress
I have a hunch about what went wrong. Please see if this commit changes
anything for you:
https://reviews.llvm.org/rGca6e117d8634
On Wed, Jul 14, 2021 at 11:12 AM Sanjay Patel
wrote:
> Thanks for letting me know. I could use some help to repro because I'm not
> very familiar with that benchmark
Hi Maxim,
> We use Nvidia TK1s (Cortex-A15) for benchmarking on 32-bit ARM.
That's a bit old, I used Cortex-A57 as the closest to that.
> LTO tends to increase functions due to additional inlining, which increases
> scheduling regions,
> which increases opportunities for the 1st scheduler for i
Successfully identified regression in *llvm* in CI configuration
tcwg_bmk_llvm_tx1/llvm-release-aarch64-spec2k6-Oz. So far, this commit has
regressed CI configurations:
- tcwg_bmk_llvm_tx1/llvm-release-aarch64-spec2k6-Oz
Culprit:
commit f645cea8f63e76f4d1ed291da3f61768cbd6abf4
Author: Chen Zh
Successfully identified regression in *gcc* in CI configuration
tcwg_bmk_gnu_tk1/gnu-release-arm-spec2k6-O2_LTO. So far, this commit has
regressed CI configurations:
- tcwg_bmk_gnu_tk1/gnu-release-arm-spec2k6-O2_LTO
Culprit:
commit 268d509d67efac45f01b356602036e1dc7c6935e
Author: Andrew Stubb
