[AArch64][SME] Fix bug on SMELd1St1 (#118109)

Patch[1] has update intrinsic interface for ld1/st1, while based on
ARM's document, "If the intrinsic also has a vnum argument, the ZA slice
number is calculated by adding vnum to slice.". But the "vnum" did not
work for our realization now, this patch fix this point.


[1]ee31ba0

Author: wwwatermiao

clang/lib/CodeGen/CGBuiltin.cpp

@@ -10646,7 +10646,7 @@ Value *CodeGenFunction::EmitSMELd1St1(const SVETypeFlags &TypeFlags,
   NewOps.push_back(Ops[2]);
 
   llvm::Value *BasePtr = Ops[3];
-
+  llvm::Value *RealSlice = Ops[1];
   // If the intrinsic contains the vnum parameter, multiply it with the vector
   // size in bytes.
   if (Ops.size() == 5) {
@@ -10658,10 +10658,13 @@ Value *CodeGenFunction::EmitSMELd1St1(const SVETypeFlags &TypeFlags,
         Builder.CreateMul(StreamingVectorLengthCall, Ops[4], "mulvl");
     // The type of the ptr parameter is void *, so use Int8Ty here.
     BasePtr = Builder.CreateGEP(Int8Ty, Ops[3], Mulvl);
+    RealSlice = Builder.CreateZExt(RealSlice, Int64Ty);
+    RealSlice = Builder.CreateAdd(RealSlice, Ops[4]);
+    RealSlice = Builder.CreateTrunc(RealSlice, Int32Ty);
   }
   NewOps.push_back(BasePtr);
   NewOps.push_back(Ops[0]);
-  NewOps.push_back(Ops[1]);
+  NewOps.push_back(RealSlice);
   Function *F = CGM.getIntrinsic(IntID);
   return Builder.CreateCall(F, NewOps);
 }