Replace non-atomic load with an atomic one

[james7132]

Fixes #70 by replacing the non-atomic load in get_inner with an atomic one.

[notgull]

You should add a test that makes sure the issue isn't still active.

[james7132]

You should add a test that makes sure the issue isn't still active.

Tried earlier, and it's surprisingly hard to get miri to fire with a minimal repro. I'll take another stab at this soon.

[james7132]

Added a test, made sure it fails with miri on master, and then added miri to the CI to ensure that the test (and others) are tested against miri for all future PRs.

[james7132]

As a final sanity check, I reran the benchmarks in the repo, and it seems like this shouldn't negatively affect performance, at least not on x86 platforms:

get                     time:   [2.6138 ns 2.6172 ns 2.6216 ns]
                        change: [-2.8757% -2.6826% -2.4831%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
  6 (6.00%) high mild
  4 (4.00%) high severe

insert                  time:   [9.9271 ns 10.080 ns 10.221 ns]
                        change: [-6.3852% -3.7294% -1.0880%] (p = 0.01 < 0.05)
                        Performance has improved.
Found 4 outliers among 100 measurements (4.00%)
  4 (4.00%) high m

[Amanieu]

We are waiting on #69 to fix CI. This will bump the Rust version used in CI, so you will be able to use black_box for benchmarks.