[llvm] [AArch64] Remove copy in SVE/SME predicate spill and fill (PR #81716)

[Sam Tebbs via llvm-commits]

================
@@ -57,11 +57,11 @@ tracksRegLiveness: true
 body:             |
   bb.1:
     ; CHECK-LABEL: name: inlineasm_virt_reg_output
-    ; CHECK: INLINEASM &"mov ${0:w}, 7", 0 /* attdialect */, 1310730 /* regdef:PPR2_with_psub_in_PNR_3b_and_PPR2_with_psub1_in_PPR_p8to15 */, def %0
+    ; CHECK: INLINEASM &"mov ${0:w}, 7", 0 /* attdialect */, {{[0-9]+}} /* regdef:GPR32common */, def %0
     ; CHECK-NEXT: [[COPY:%[0-9]+]]:gpr(s32) = COPY %0
     ; CHECK-NEXT: $w0 = COPY [[COPY]](s32)
     ; CHECK-NEXT: RET_ReallyLR implicit $w0
-    INLINEASM &"mov ${0:w}, 7", 0 /* attdialect */, 1310730 /* regdef:GPR32common */, def %0:gpr32common
+    INLINEASM &"mov ${0:w}, 7", 0 /* attdialect */, 1769482 /* regdef:GPR32common */, def %0:gpr32common
----------------
SamTebbs33 wrote:

Good point, I've changed the .mir tests to explicitly match the regclass numbers.

https://github.com/llvm/llvm-project/pull/81716