[llvm] [AArch64][SVE] Remove isSVECC() in favour of changing the calling convention (PR #152742)

[Benjamin Maxwell via llvm-commits]

================
@@ -1546,40 +1482,33 @@ define tailcc void @f15(double %d, <vscale x 4 x i32> %vs, [9 x i64], i32 %i) {
 ; CHECK-LABEL: f15:
 ; CHECK:       .seh_proc f15
 ; CHECK-NEXT:  // %bb.0:
-; CHECK-NEXT:    addvl sp, sp, #-1
----------------
MacDue wrote:

This change is due to the explicit `tailcc` on the function. I'm not 100% sure on the correct behaviour here, `LowerCall` only uses the SVE CC if the original CC is `C` or `Fast`. 

https://github.com/llvm/llvm-project/pull/152742