[PATCH] D101414: [AMDGPU] Disable the scalar IR, SDWA and load store vectorizer passes at -O1

[Nico Weber via Phabricator via llvm-commits]

thakis added a comment.

In D101414#2737424 <https://reviews.llvm.org/D101414#2737424>, @bsaleil wrote:

> In D101414#2737354 <https://reviews.llvm.org/D101414#2737354>, @thakis wrote:
>
>> This breaks incremental builders: http://45.33.8.238/linux/45830/step_12.txt
>>
>> Ptal. You'll need a -o /dev/null and an `rm -f` for the temp file to clean up bots (can remove the latter after a bit)
>
> @thakis I already committed a fix to avoid generating the .s file. Should I add the `rm -f` directly into the RUN lines ? Is something like `RUN: rm %S/llc-pipeline.s -f` ok ?

Yes, that sounds perfect. See e.g. clang/test/CoverageMapping/coroutine.cpp for an example :)


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D101414/new/

https://reviews.llvm.org/D101414