This adds a few features:
* Parallel execution (and all other golang test runner features)
* Colored diffs of output
* Make and kati run in subdirs, all output is preserved
* Full logs saved of raw and normalized output per-testcase
* TODO(c/<testcase>) support to mark individual testcases as expected
  failures, instead of the full test.

Removed support for the [go]kati serialization tests

Change-Id: Id6be80a3168ccff3fa3a970fe089986f1aa066dd

The find man page says that "expression expression" is equivalent to
"expression -a expression".

This reduces the number of shell commands we need to run during regen
check of one of the internal android trees by ~10%, which reduced the
time spent in regen checking by ~60%.

before: *kati*: shell time (regen): 5.842119 / 516
after:  *kati*: shell time (regen): 2.377083 / 462

Change-Id: I177f37cd7e12625fb2dbfa371f3d79cb625c84fe

Now FindEmulator supports symlinks which refer

- a file in FindEmulator or
- a file which isn't a directory nor a symlink

Also findleaves.py follows symbolic links.

When FindEmulator is sure the target directory does not exist,
we do not need to run the shell command.

[C++] Don't sort find/ls results