WIP: Injects counters before every block (experimental only)

[richkadel]

Uses the span from the start of the first statement to the end of the Terminator.

This simple approach does introduce some poor coverage results. For
example, a function that only calls macros (like println!, for
example) may only have counters injected at the macro site, and these
may never get counted, for some reason. (I'm seeing some functions with
coverage counts of zero, even though they clearly should be called.)

I'm developing a follow-up change to this one that injects counters more
precisely, by inspecting the Statement and Terminator types, and making
different decisions regarding what blocks should be counted, and what
spans to use to generate the code regions.

[rust-highfive]

r? @petrochenkov

(rust_highfive has picked a reviewer for you, use r? to override)

[richkadel]

r? @tmandry

[richkadel]

@tmandry - I tried to push a specific commit on the branch, but got other commits, including future changes.

Ignore this PR for now (it's draft anyway), and I'll add a note when it's ready to take a look at.

Thanks!

[richkadel]

@tmandry - Fixed. You can see the changes specific to this PR in the last commit. (The first 2 commits are from the PR currently in review.)

ec050f2

This is just what you looked at yesterday during our meeting, simply adding counters to each BasicBlock.

As discussed, I'll do some more investigating of SourceScope and Span values in the contributing Statement and Terminator components each BasicBlock.

And as noted, SourceScope may or may not need to be set on the injected calls based on its surrounding context, depending on how it's used. (I have a feeling it's not, but need to confirm. If it's really only relevant to variables and debugging, as we speculated yesterday, the injected BB's do not introduce user-relevant variables; the temporary variable to capture the () return result of the call is eventually thrown away, in codegen.) If not, I can probably just use a default value, like OUTERMOST_SOURCE_SCOPE.

I may also push a parallel PR with the work I did to be more selective about which BasicBlocks to use, based on CFG branching, because I thought that had promise. But even with that approach, extracting reasonable Spans (and maybe SourceScope, if required) also needs work. So the investigation will help in both approaches.

[bors]

☔ The latest upstream changes (presumably #74837) made this pull request unmergeable. Please resolve the merge conflicts.

[Dylan-DPC-zz]

@richkadel if you rebase and resolve the latest conflicts, we can get this reviewed if possible

[richkadel]

The gist of this PR is incorporated in #75828 instead.