[llvm] [NVPTX] Make i16x2 a native type and add supported vec instructions (PR #65432)

George Karpenkov via llvm-commits llvm-commits at lists.llvm.org
Thu Sep 14 06:40:46 PDT 2023

Previous message: [compiler-rt] [CMake] [compiler-rt] fix architecture checks in runtime builds (PR #66374)
Next message: [llvm] [IRSymtab] Replace linear time lookup with DenseSet (PR #66376)
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

cheshire wrote:

This causes a nearly ~2x regression in many Triton kernels on Ampere, we'll post a reproducer.

https://github.com/llvm/llvm-project/pull/65432

Previous message: [compiler-rt] [CMake] [compiler-rt] fix architecture checks in runtime builds (PR #66374)
Next message: [llvm] [IRSymtab] Replace linear time lookup with DenseSet (PR #66376)
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

More information about the llvm-commits mailing list