[AMDGPU] When allocating VGPRs, VGPR spills are not part of the prologue

PRs llvm#69924 and llvm#72140 modified SIInstrInfo::isBasicBlockPrologue to skip
over EXEC modifications and spills when allocating VGPRs. But treating
VGPR spills as part of the prologue can confuse the register allocator
as in llvm#109294, so restrict it to SGPR spills, which were inserted during
SGPR allocation which is done in an earlier pass.

Fixes: llvm#109294
Fixes: SWDEV-485841
Change-Id: I328fb2edfca8110ea36c94812e60bc1d7663c266

Author: jayfoad

llvm/lib/Target/AMDGPU/SIInstrInfo.cpp

@@ -8902,8 +8902,9 @@ bool SIInstrInfo::isBasicBlockPrologue(const MachineInstr &MI,
   // FIXME: Copies inserted in the block prolog for live-range split should also
   // be included.
   return IsNullOrVectorRegister &&
-         (isSpill(Opcode) || (!MI.isTerminator() && Opcode != AMDGPU::COPY &&
-                              MI.modifiesRegister(AMDGPU::EXEC, &RI)));
+         (isSGPRSpill(Opcode) ||
+          (!MI.isTerminator() && Opcode != AMDGPU::COPY &&
+           MI.modifiesRegister(AMDGPU::EXEC, &RI)));
 }
 
 MachineInstrBuilder