[X86] AMD Piledriver (BdVer2): fine-tune some latencies

[X86] AMD Piledriver (BdVer2): fine-tune some latencies

Based on llvm-exegesis measurements.

Now that llvm-exegesis is ~2 magnitudes faster, and is a bit smarter,
it is now possible to continue cleanup of the scheduler model.

With this, there are no more latency inconsistencies for the
opcodes that produce stable measurements, and only a few inconsistencies
for unstable measurements (MMX_* opcodes, opcodes that llvm-exegesis
measures by chaining - CMP, TEST, BT, SETcc, CVT, MOV, etc.)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@357169 91177308-0d34-0410-b5e6-96231b3b80d8