Timeouts are now specified in milliseconds instead of seconds.

TimeoutSolvers that hit a timeout no longer makes the wrapped solver
useless for all subsequent invocations.

This commit introduces a new XlangAnalysisPhase that run
all the xlang phase as well as the AnalysisPhase. It updates
the Main accordingly.

The reason for this change is to be able to correctly control
the --functions option as well as transforming each VerificationCondition
about function postcondition into loop invariant.

The previous solution was to use some mutable states inside the FunDef object.
Those are cleaned by this commit. To do so, it was necessary to change the
transformation phases signature in order to return along with the modified
program a Set or Map (depending on which phase) of freshly introduced
functions and their correspondance in the original program.

One small change that was necessary was to not print the verification report
in the analysis phase but only in the Main. This allows the XlangAnalysisPhase
to update correctly the verification conditions in the report before it gets
printed. This is also arguably a better design decision to have it printed
in the Main since it was returned by the AnalysisPhase.

FairZ3Solver no longer displays counter-examples. As a result, they were
not shown when doing verification. The fix is to have AnalysisPhase
display the counter-examples when it wants to.

Improve performance of FairZ3 by lowering unrolling to the z3 level.  Improve synthesis profiling script.

Improve performance of FairZ3:

  Make function templates and unlocking/unrolling work directly at the z3 level for performance reasons.

  Implement push-pop at the unrolling-bank level. Works around a z3 bug.

  Z3 apparently side-effects during check-assumptions, causing a following
  check without assumptions to produce unreliable results. We work around this
  by backtracking to the state before the check-assumptions using
  push/pop.

Improve synthesis profiling/benchmarking:

  Move benchrmarking suite away from tests. Synthesis Rule application
  benchmark can now use a --rule option specifying what rule to apply.

  Add Testcase to evaluate CEGIS unrolling of different depth.
  Benchmarking script is now generated via the 'bench' sbt task.