[llvm] [AMDGCN][SIWholeQuadMode] Handle case when SI_KILL_I1_TERMINATOR -1,0 is not the only terminator (PR #122922)
Juan Manuel Martinez CaamaƱo via llvm-commits
llvm-commits at lists.llvm.org
Mon Jan 27 08:24:04 PST 2025
================
@@ -3361,7 +3361,7 @@ define amdgpu_ps void @test_for_deactivating_lanes_in_wave32(ptr addrspace(6) in
; GFX9-W64-NEXT: s_buffer_load_dword s0, s[0:3], 0x0
; GFX9-W64-NEXT: s_waitcnt lgkmcnt(0)
; GFX9-W64-NEXT: v_cmp_le_f32_e64 vcc, s0, 0
-; GFX9-W64-NEXT: s_andn2_b64 s[4:5], exec, vcc
----------------
jmmartinez wrote:
SI optimize exec mask pre-RA considers this as a unused exec copy and removes the copy (s[4:5] is not used).
https://github.com/llvm/llvm-project/pull/122922
More information about the llvm-commits
mailing list