Introduce `--profile` support to the benchmark helper script, which will
invoke `perf` on Linux. Additionally, add a `--flamegraph` output
option based on that.
It can be useful to report the commit ID during benchmarks to track down
regressions; leverage the addition of that in `git version` during
benchmark runs.
Teach the benchmark script how to clone the git or kernel repositories,
which is useful to have a larger corpus of test data. If a benchmark
script wants to `clone git` or `clone linux`, then this will be done.
Callers probably want to specify `BENCHMARK_GIT_REPOSITORY` to a
previously cloned local repository so that the script does not download
the repository repeatedly.
