[all-commits] [llvm/llvm-project] 375b70: [libclc] Improve performance and precision of reci...

[Joseph Huber via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 375b70bafedafa71fabfb0e8b1e759daeef0f703
      https://github.com/llvm/llvm-project/commit/375b70bafedafa71fabfb0e8b1e759daeef0f703
  Author: Joseph Huber <huberjn at outlook.com>
  Date:   2026-06-14 (Sun, 14 Jun 2026)

  Changed paths:
    M libclc/clc/lib/amdgpu/CMakeLists.txt
    A libclc/clc/lib/amdgpu/math/clc_recip_fast.cl
    A libclc/clc/lib/amdgpu/math/clc_recip_fast.inc
    M libclc/clc/lib/generic/CMakeLists.txt

  Log Message:
  -----------
  [libclc] Improve performance and precision of reciprocal functions (#203805)

Summary:
Small change to improve the performance and output of functions using
the reciprocol. This makes these functions *byte-for-byte* identical
with their OCML counterparts in ROCm.

Additionally ensure tanpi is correctly rounded to match ROCm.



To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications