[llvm] [AMDGPU] Always lower s/udiv64 by constant to MUL (PR #100723)

Steven Johnson via llvm-commits llvm-commits at lists.llvm.org
Fri Aug 2 14:16:07 PDT 2024

Previous message: [llvm] [AMDGPU] Always lower s/udiv64 by constant to MUL (PR #100723)
Next message: [llvm] [AMDGPU] Always lower s/udiv64 by constant to MUL (PR #100723)
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

steven-johnson wrote:

Enclosed is a Halide-generated .ll that will assert-fail when run through `llc` with this PR in place. We should probably revert this PR for now.
[bad.ll.zip](https://github.com/user-attachments/files/16475365/bad.ll.zip)


https://github.com/llvm/llvm-project/pull/100723

Previous message: [llvm] [AMDGPU] Always lower s/udiv64 by constant to MUL (PR #100723)
Next message: [llvm] [AMDGPU] Always lower s/udiv64 by constant to MUL (PR #100723)
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

More information about the llvm-commits mailing list