[all-commits] [llvm/llvm-project] 7a1017: Revert "[AMDGPU] Mark mbcnt as convergent"

[Sameer Sahasrabuddhe via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 7a101798b7539743de08adc9ea7d7f748cab37eb
      https://github.com/llvm/llvm-project/commit/7a101798b7539743de08adc9ea7d7f748cab37eb
  Author: Sameer Sahasrabuddhe <sameer.sahasrabuddhe at amd.com>
  Date:   2023-06-30 (Fri, 30 Jun 2023)

  Changed paths:
    M clang/test/CodeGenOpenCL/builtins-amdgcn.cl
    M llvm/include/llvm/IR/IntrinsicsAMDGPU.td
    M llvm/test/CodeGen/AMDGPU/atomic_optimizations_buffer.ll
    M llvm/test/CodeGen/AMDGPU/atomic_optimizations_global_pointer.ll
    M llvm/test/CodeGen/AMDGPU/atomic_optimizations_local_pointer.ll
    M llvm/test/CodeGen/AMDGPU/atomic_optimizations_raw_buffer.ll
    M llvm/test/CodeGen/AMDGPU/atomic_optimizations_struct_buffer.ll

  Log Message:
  -----------
  Revert "[AMDGPU] Mark mbcnt as convergent"

This reverts commit 37114036aa57e53217a57afacd7f47b36114edfb.

The output of mbcnt does not depend on other active lanes, and hence it is not
convergent. The original change was made as a possible fix for

https://github.com/ROCm-Developer-Tools/HIP/issues/3172

But changing mbcnt does not fix that issue.

Reviewed By: ruiling, foad, yaxunl

Differential Revision: https://reviews.llvm.org/D153953