[llvm] [DXIL] Implement pow lowering (PR #86733)

Farzon Lotfi via llvm-commits llvm-commits at lists.llvm.org
Wed Apr 24 13:39:25 PDT 2024


farzonl wrote:

> Hi, I'm seeing an assertion failure that seems to be due to this patch. I bisected it to this patch. I'm including a reproducer (not reduced) and my bisect script.
> 
> Failing bot: https://ci.chromium.org/ui/p/fuchsia/builders/ci/clang_toolchain.ci.core.arm64-debug-subbuild/b8749763985778255185/overview Logs: https://logs.chromium.org/logs/fuchsia/buildbucket/cr-buildbucket/8749760927129551041/+/u/clang/build/stdout
> 
> I bisected with
> 
> ```
> git bisect good  0f61051f541a5b8cfce25c84262dfdbadb9ca688
> git bisect bad d3f6a88a1fb36f94c71940514e576821c6cc3ade
> git bisect run bisect.sh
> ```
> 
> I checked the good and bad SHA's multiple times, and the good commit is one commit prior to yours.
> 
> Note if you want to use the bisect script you may need to update paths. The `.cpp` and other `.sh` are from the generated reproducer, so should work if you update the path to clang in the script.
> 
> I have creduce running now, so I can probably post a smaller reproducer in a while.
> 
> In the meantime, can you take a look or possibly revert this until a fix is ready?
> 
> [repro.zip](https://github.com/llvm/llvm-project/files/15100080/repro.zip)

This seems very unlikely this change is in an experimental backend that isn’t enabled on any build bots.

https://github.com/llvm/llvm-project/pull/86733


More information about the llvm-commits mailing list