[all-commits] [llvm/llvm-project] 100602: [AMDGPU] Multi dword spilling for unaligned tuples

[Aaditya via All-commits]

  Branch: refs/heads/users/easyonaadit/amdgpu/split-unaligned-spills-into-chunks
  Home:   https://github.com/llvm/llvm-project
  Commit: 100602e76b5825464dd97d063bd2c1ea97a8d798
      https://github.com/llvm/llvm-project/commit/100602e76b5825464dd97d063bd2c1ea97a8d798
  Author: Aaditya <Aaditya.AlokDeshpande at amd.com>
  Date:   2026-03-09 (Mon, 09 Mar 2026)

  Changed paths:
    M llvm/lib/Target/AMDGPU/SIRegisterInfo.cpp
    M llvm/test/CodeGen/AMDGPU/vgpr-spill.mir

  Log Message:
  -----------
  [AMDGPU] Multi dword spilling for unaligned tuples

While spilling unaligned tuples, rather than breaking the
spill into 32-bit accesses, spill the first register as a
single 32-bit spill, and spill the remainder of the tuple
as an aligned tuple.
Some additional bookkeeping is required in the spilling
loop to manage the state.



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