[PATCH] D80754: AMDGPU/GlobalISel: cmp/select method for insert element

[Stanislav Mekhanoshin via Phabricator via llvm-commits]

rampitec added inline comments.


================
Comment at: llvm/test/CodeGen/AMDGPU/GlobalISel/insertelement.ll:58
 ; GPRIDX:       ; %bb.0: ; %entry
-; GPRIDX-NEXT:    s_mov_b32 s0, s2
-; GPRIDX-NEXT:    s_mov_b32 s1, s3
-; GPRIDX-NEXT:    s_mov_b32 s2, s4
-; GPRIDX-NEXT:    s_mov_b32 s3, s5
-; GPRIDX-NEXT:    s_mov_b32 s4, s6
-; GPRIDX-NEXT:    s_mov_b32 s5, s7
-; GPRIDX-NEXT:    s_mov_b32 s6, s8
-; GPRIDX-NEXT:    s_mov_b32 s7, s9
-; GPRIDX-NEXT:    s_mov_b32 m0, s11
-; GPRIDX-NEXT:    s_nop 0
-; GPRIDX-NEXT:    s_movreld_b32 s0, s10
+; GPRIDX-NEXT:    s_cmp_eq_u32 s11, 0
+; GPRIDX-NEXT:    s_cselect_b32 s0, s10, s2
----------------
arsenm wrote:
> I think the register indexing looks better if the index is uniform
It may look better, but it is not faster.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D80754/new/

https://reviews.llvm.org/D80754