[Turboprop] Avoid calling RepresentationFor in mid-tier allocator.

RepresentationFor is quite hot in MidTierAllocator profiles. To
optimize this, instead stash the representation in the
VirtualRegisterData and pass that about consistently instead of
passing the virtual_register int and having to retrieve both
representation and VirtualRegisterData for the vreg multiple times.

This improves mid-tier allocation time by ~8% on Octane benchmarks.

BUG=v8:9684

Change-Id: Ied01fbdab013c278da022d1df321b08fbfc68a4c
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2768618Reviewed-by: Santiago Aboy Solanes <solanes@chromium.org>
Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73508}

src/compiler/backend/mid-tier-register-allocator.h

@@ -44,7 +44,6 @@ class MidTierRegisterAllocationData final : public RegisterAllocationData {
   }
 
   VirtualRegisterData& VirtualRegisterDataFor(int virtual_register);
-  MachineRepresentation RepresentationFor(int virtual_register);
 
   // Add a gap move between the given operands |from| and |to|.
   MoveOperands* AddGapMove(int instr_index, Instruction::GapPosition position,