[all-commits] [llvm/llvm-project] 742008: libclc: Define work_group_barrier (#184780)

[Matt Arsenault via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 742008d4e003b5c2bae6364c137829ef0dfecdf6
      https://github.com/llvm/llvm-project/commit/742008d4e003b5c2bae6364c137829ef0dfecdf6
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2026-03-05 (Thu, 05 Mar 2026)

  Changed paths:
    M libclc/opencl/lib/amdgcn/SOURCES
    R libclc/opencl/lib/amdgcn/synchronization/barrier.cl
    M libclc/opencl/lib/generic/SOURCES
    M libclc/opencl/lib/generic/async/wait_group_events.cl
    A libclc/opencl/lib/generic/synchronization/work_group_barrier.cl
    M libclc/opencl/lib/ptx-nvidiacl/SOURCES
    R libclc/opencl/lib/ptx-nvidiacl/synchronization/barrier.cl

  Log Message:
  -----------
  libclc: Define work_group_barrier (#184780)

Previously only the old barrier name was implemented. Define this
as an indirection around the new name, and move it to common code.
The target implementations are already provided by
__clc_work_group_barrier,
so targets were unnecessarily duplicating these.

This also fixes the default scope, which should be
memory_work_group_scope. Previously this was guessing that
if the flags included global memory, it makes the scope
device which is not the case.



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