[llvm] [SelectionDAG] Avoid double rounding by using a libcall for int (>i32) to bf16 conversions (PR #71658)

Eli Friedman via llvm-commits llvm-commits at lists.llvm.org
Thu Nov 16 10:07:22 PST 2023

Previous message: [llvm] [SelectionDAG] Avoid double rounding by using a libcall for int (>i32) to bf16 conversions (PR #71658)
Next message: [llvm] 1e8c17e - [AMDGPU] Allow folding to FMAMK with SGPR and immediate operand on GFX10+ (#72258)
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

efriedma-quic wrote:

I'm specifically concerned that this isn't changing the behavior for i32->bf16.

https://github.com/llvm/llvm-project/pull/71658

Previous message: [llvm] [SelectionDAG] Avoid double rounding by using a libcall for int (>i32) to bf16 conversions (PR #71658)
Next message: [llvm] 1e8c17e - [AMDGPU] Allow folding to FMAMK with SGPR and immediate operand on GFX10+ (#72258)
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

More information about the llvm-commits mailing list