[llvm] [AMDGPU] Inplace FI elimination during PEI for scalar copy instruction (PR #99556)

[Matt Arsenault via llvm-commits]

================
@@ -368,9 +368,9 @@ body:   |
     ; GFX8-NEXT: S_NOP 0, implicit-def $sgpr56_sgpr57_sgpr58_sgpr59_sgpr60_sgpr61_sgpr62_sgpr63
     ; GFX8-NEXT: V_CMP_EQ_U32_e32 0, killed $vgpr0, implicit-def $vcc, implicit $exec
     ; GFX8-NEXT: S_CMP_EQ_I32 $sgpr4, $sgpr5, implicit-def $scc
+    ; GFX8-NEXT: $vgpr0 = V_MOV_B32_e32 64, implicit $exec
----------------
arsenm wrote:

This test also should get gfx10, gfx11, and gfx12 runlines (maybe leave for a separate commit though) 

https://github.com/llvm/llvm-project/pull/99556