[all-commits] [llvm/llvm-project] a349ac: [flang][OpenMP] Allocate `allocatable` init temps ...

[Kareem Ergawy via All-commits]

  Branch: refs/heads/users/ergawy/alloc_on_stack
  Home:   https://github.com/llvm/llvm-project
  Commit: a349ac2be7e54c65193e9c70cdcbd3ca8a7a4248
      https://github.com/llvm/llvm-project/commit/a349ac2be7e54c65193e9c70cdcbd3ca8a7a4248
  Author: ergawy <kareem.ergawy at amd.com>
  Date:   2025-10-17 (Fri, 17 Oct 2025)

  Changed paths:
    M flang/lib/Lower/Support/PrivateReductionUtils.cpp
    M flang/test/Lower/OpenMP/DelayedPrivatization/target-private-allocatable.f90

  Log Message:
  -----------
  [flang][OpenMP] Allocate `allocatable` init temps on the stack for GPUs

Temps needed for the allocatable reduction/privatization init regions
are now allocated on the heap all the time. However, this is performance
killer for GPUs since malloc calls are prohibitively expensive.
Therefore, we should do these allocations on the stack for GPU reductions.



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