[Mlir-commits] [mlir] [mlir][Math][LLVMSPV] Add Math to LLVM-SPV conversion pass (PR #198370)

Akhil Goel llvmlistbot at llvm.org
Fri May 29 14:51:34 PDT 2026


akhilgoe wrote:

Yes, this is relevant for XeVM/SPIR-V targets. XeVM already lowers fastmath ops to native OpenCL intrinsics, so feels natural to move this to the existing pipeline.

https://github.com/llvm/llvm-project/pull/198370


More information about the Mlir-commits mailing list