[AArch64][SME] Fix 'addvl' addressing to scavenged stackslot.

In https://reviews.llvm.org/D159196 we avoided stackslot scavenging
when there was no FP available. But in the case where FP is available
we need to actually prefer using the FP over the BP.

This change affects more than just SME, but in general it should be
an improvement, since any slot above the (address pointed to by) FP
is always closer to FP than BP, so it makes sense to favour using
the FP to address it.

This also fixes the issue for SME, where this is not just preferred
but required.

Author: sdesmalen-arm

llvm/lib/Target/AArch64/AArch64FrameLowering.cpp

@@ -2757,7 +2757,11 @@ StackOffset AArch64FrameLowering::resolveFrameOffsetReference(
       bool FPOffsetFits = !ForSimm || FPOffset >= -256;
       PreferFP |= Offset > -FPOffset && !SVEStackSize;
 
-      if (MFI.hasVarSizedObjects()) {
+      if (FPOffset >= 0) {
+        // If the FPOffset is positive, that'll always be best, as the SP/BP
+        // will be even further away.
+        UseFP = true;
+      } else if (MFI.hasVarSizedObjects()) {
         // If we have variable sized objects, we can use either FP or BP, as the
         // SP offset is unknown. We can use the base pointer if we have one and
         // FP is not preferred. If not, we're stuck with using FP.
@@ -2769,11 +2773,6 @@ StackOffset AArch64FrameLowering::resolveFrameOffsetReference(
         // else we can use BP and FP, but the offset from FP won't fit.
         // That will make us scavenge registers which we can probably avoid by
         // using BP. If it won't fit for BP either, we'll scavenge anyway.
-      } else if (FPOffset >= 0) {
-        // Use SP or FP, whichever gives us the best chance of the offset
-        // being in range for direct access. If the FPOffset is positive,
-        // that'll always be best, as the SP will be even further away.
-        UseFP = true;
       } else if (MF.hasEHFunclets() && !RegInfo->hasBasePointer(MF)) {
         // Funclets access the locals contained in the parent's stack frame
         // via the frame pointer, so we have to use the FP in the parent

llvm/test/CodeGen/AArch64/sme-streaming-mode-changing-call-disable-stackslot-scavenging.ll

@@ -60,19 +60,17 @@ define void @test_no_stackslot_scavenging_with_fp(float %f, i64 %n) #0 "frame-po
 ; CHECK-NEXT:    stp x24, x19, [sp, #112] // 16-byte Folded Spill
 ; CHECK-NEXT:    addvl sp, sp, #-1
 ; CHECK-NEXT:    lsl x9, x0, #3
+; CHECK-NEXT:    mov x8, sp
 ; CHECK-NEXT:    mov x19, sp
-; CHECK-NEXT:    addvl x8, x19, #1
+; CHECK-NEXT:    str s0, [x29, #28] // 4-byte Folded Spill
 ; CHECK-NEXT:    add x9, x9, #15
-; CHECK-NEXT:    str s0, [x8, #92] // 4-byte Folded Spill
-; CHECK-NEXT:    mov x8, sp
 ; CHECK-NEXT:    and x9, x9, #0xfffffffffffffff0
 ; CHECK-NEXT:    sub x8, x8, x9
 ; CHECK-NEXT:    mov sp, x8
 ; CHECK-NEXT:    //APP
 ; CHECK-NEXT:    //NO_APP
 ; CHECK-NEXT:    smstop sm
-; CHECK-NEXT:    addvl x8, x19, #1
-; CHECK-NEXT:    ldr s0, [x8, #92] // 4-byte Folded Reload
+; CHECK-NEXT:    ldr s0, [x29, #28] // 4-byte Folded Reload
 ; CHECK-NEXT:    bl use_f
 ; CHECK-NEXT:    smstart sm
 ; CHECK-NEXT:    sub sp, x29, #64