[clang] [Clang][SME2] Add multi-vector add/sub builtins (PR #69725)

[Kerry McLaughlin via cfe-commits]

================
@@ -9571,22 +9571,17 @@ Value *CodeGenFunction::EmitSVEStructStore(const SVETypeFlags &TypeFlags,
   Value *BasePtr = Ops[1];
 
   // Does the store have an offset?
-  if (Ops.size() > 3)
+  if (Ops.size() > (2 + N))
----------------
kmclaughlin-arm wrote:

This change was intended to be part of this patch. Now that EmitSVEStructLoad is being called after GetAArch64SVEProcessedOperands, the first operand will have been split into a series of vector extracts and the number of operands expected here will change depending on the value of N.

https://github.com/llvm/llvm-project/pull/69725