[llvm] [AMDGPU][MC] Fix disassemble of image_gather4 with d16 (PR #114609)

[Mirko Brkušanin via llvm-commits]

================
@@ -235,5 +235,5 @@
 # VI: image_gather4 v[252:255], v1, s[8:15], s[12:15] dmask:0x3 ; encoding: [0x00,0x03,0x00,0xf1,0x01,0xfc,0x62,0x00]
 0x00,0x03,0x00,0xf1,0x01,0xfc,0x62,0x00
 
-# VI: image_gather4 v[252:255], v1, s[8:15], s[12:15] dmask:0x1 unorm glc slc tfe lwe da ; encoding: [0x00,0x71,0x03,0xf3,0x01,0xfc,0x62,0x00]
+# VI: image_gather4 v[252:253], v1, s[8:15], s[12:15] dmask:0x1 unorm glc slc tfe lwe da ; encoding: [0x00,0x71,0x03,0xf3,0x01,0xfc,0x62,0x00]
----------------
mbrkusanin wrote:

Previous default was V4 variant so it would start with VReg_128: v[252:255]
Now default is V2 so it starts with VReg_64: v[252:253]
And I think both fail when trying to convert to V5 and VReg_160: v[252:**256**] which would be invalid.
I'm not sure what the original test was checking since this is one of "incorrect" encodings

https://github.com/llvm/llvm-project/pull/114609