[llvm] [AMDGPU] In promote-alloca, if index is dynamic, sandwich load with bitcasts to reduce number of extractelements as they have large expansion in the backend. (PR #171253)

Kevin Choi via llvm-commits llvm-commits at lists.llvm.org
Mon Dec 15 21:36:51 PST 2025

Previous message: [llvm] [AMDGPU] In promote-alloca, if index is dynamic, sandwich load with bitcasts to reduce number of extractelements as they have large expansion in the backend. (PR #171253)
Next message: [llvm] [AMDGPU] In promote-alloca, if index is dynamic, sandwich load with bitcasts to reduce number of extractelements as they have large expansion in the backend. (PR #171253)
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

https://github.com/choikwa edited https://github.com/llvm/llvm-project/pull/171253

More information about the llvm-commits mailing list