Upgrade to clap 3.

[nnethercote]

This requires moving away from the deprecated clap_app! macro. Clap's
derive API (formerly StructOpt) is nicer anyway, because it's not
stringly-typed.

There is quite a bit of furniture moving in execute.rs to better
separate the perf tools used for bench_* subcommands from those used
for the profile subcommands.

[nnethercote]

@epage: this may be of interest after all my questions today, and I'd be happy to hear any suggestions you have. Thanks!

[Mark-Simulacrum]

Would you be up for bumping clap in database as well? That should be significantly easier (the options are much more limited, I suspect) -- it would be nice to avoid depending on 2.x and 3.x at the same time.

[epage]

I didn't dig into verifying clap_app -> derive but instead focused on how the derive is being used.

[nnethercote]

Would you be up for bumping clap in database as well? That should be significantly easier (the options are much more limited, I suspect) -- it would be nice to avoid depending on 2.x and 3.x at the same time.

No problem. I didn't even know that usage was there. I'll do it in a follow-up PR to avoid scope creep here.

[nnethercote]

Thanks @epage for the helpful suggestions :)

[Mark-Simulacrum]

@nnethercote I think this broadly seems good to me, but it looks like there are some ongoing(?) conversations on potential cleanups/simplifications that could be done -- can you comment when you want to move ahead?

[nnethercote]

@Mark-Simulacrum I'm happy with the current code as it is.

[nnethercote]

Having said that, I just added a commit to update database to clap 3 as well. It's much simpler.

[nnethercote]

Note: The same "Test profiling" failure is happening on some other PRs, so it isn't caused by this PR's changes.