mirror of
https://github.com/llvm/llvm-project.git
synced 2025-04-24 23:56:06 +00:00

This patch invokes a pass when compiling for an AMDGPU target to lower math operations to AMD GPU library calls library calls instead of libm calls.