[PATCH] D114643: [AMDGPU] Aggressively fold immediates in SIFoldOperands

[Jay Foad via Phabricator via llvm-commits]

foad added inline comments.


================
Comment at: llvm/test/CodeGen/AMDGPU/GlobalISel/flat-scratch.ll:82-86
+; GFX10-NEXT:    v_mov_b32_e32 v2, 15
 ; GFX10-NEXT:    v_lshlrev_b32_e32 v1, 2, v1
-; GFX10-NEXT:    v_add_nc_u32_e32 v0, v2, v0
-; GFX10-NEXT:    v_add_nc_u32_e32 v1, v2, v1
-; GFX10-NEXT:    scratch_store_dword v0, v3, off
+; GFX10-NEXT:    v_add_nc_u32_e32 v0, 4, v0
+; GFX10-NEXT:    v_add_nc_u32_e32 v1, 4, v1
+; GFX10-NEXT:    scratch_store_dword v0, v2, off
----------------
sebastian-ne wrote:
> Not really related to this patch, but shouldn’t we be able to inline v2 (15) into the scratch_store?
No, v2 is the value being stored, and it has to be in a vgpr for that instruction (not even an inline constant is allowed).


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D114643/new/

https://reviews.llvm.org/D114643