[llvm] [AMDGPU] Optimize the register uses if offset inlinable (PR #101676)

[Pankaj Dwivedi via llvm-commits]

================
@@ -706,8 +706,7 @@ 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
-    ; GFX8-NEXT: $vgpr0 = V_MAD_U32_U24_e64 killed $vgpr0, 64, $sgpr32, 0, implicit $exec
+    ; GFX8-NEXT: $vgpr0 = V_MAD_U32_U24_e64 64, 64, $sgpr32, 0, implicit $exec
----------------
PankajDwivedi-25 wrote:

that's really great point, adding another test.

https://github.com/llvm/llvm-project/pull/101676