[llvm] [MachineLICM] Use `RegisterClassInfo::getRegPressureSetLimit` (PR #119826)

[Pengcheng Wang via llvm-commits]

================
@@ -31,13 +31,13 @@ define float @flat_agent_atomic_fmin_ret_f32__amdgpu_no_fine_grained_memory(ptr
 ; GFX940-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
 ; GFX940-NEXT:    flat_load_dword v3, v[0:1]
 ; GFX940-NEXT:    s_mov_b64 s[0:1], 0
-; GFX940-NEXT:    v_max_f32_e32 v2, v2, v2
 ; GFX940-NEXT:  .LBB0_1: ; %atomicrmw.start
 ; GFX940-NEXT:    ; =>This Inner Loop Header: Depth=1
 ; GFX940-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
 ; GFX940-NEXT:    v_mov_b32_e32 v5, v3
+; GFX940-NEXT:    v_max_f32_e32 v4, v2, v2
----------------
wangpc-pp wrote:

So, should I fix these changes in this patch? Or should I just go ahead?

https://github.com/llvm/llvm-project/pull/119826