ci: Add criterion performance regressions to PR workflows

[scottgerring]

This PR follows on from #2705, introducing a workflow that publishes a criterion regression comparison between main and the current branch.

It cannot be merged and the build will not succeed until #2705 is in place.

Merge requirement checklist

• CONTRIBUTING guidelines followed
• Unit tests added/updated (if applicable)
• Appropriate CHANGELOG.md files updated for non-trivial, user-facing changes
• Changes in public API reviewed (if applicable)

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 79.3%. Comparing base (7954252) to head (e9b77ca).
Report is 2 commits behind head on main.

Additional details and impacted files

@@          Coverage Diff          @@
##            main   #2706   +/-   ##
=====================================
  Coverage   79.3%   79.3%           
=====================================
  Files        123     123           
  Lines      22670   22670           
=====================================
  Hits       17986   17986           
  Misses      4684    4684           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[cijothomas]

Closes #251

[scottgerring]

@cijothomas this is good to go.

I'm not sure if the permissions are going to work for this to submit to the PR feedback from forks. I am hopeful that once the workflow is in main it will, but we'll see. If it doesn't I'll spend some more time looking at if we can set the perms up differently to achieve this (perhaps with repository settings) after this goes in.

Regardless, you always see the performance comparison in the job run:

https://github.com/open-telemetry/opentelemetry-rust/actions/runs/13538282948/job/37833792773?pr=2706

[cijothomas]

@cijothomas this is good to go.

I'm not sure if the permissions are going to work for this to submit to the PR feedback from forks. I am hopeful that once the workflow is in main it will, but we'll see. If it doesn't I'll spend some more time looking at if we can set the perms up differently to achieve this (perhaps with repository settings) after this goes in.

Regardless, you always see the performance comparison in the job run:

https://github.com/open-telemetry/opentelemetry-rust/actions/runs/13538282948/job/37833792773?pr=2706

merged. let us see !

[scottgerring]

@cijothomas benchmark_collect_histogram intermittently fails. I don't want to introduce a CI check that breaks the build. I'd rather comment out this assertion for now with a TODO so we can get this in. What do you think?

thread 'main' panicked at opentelemetry-sdk/benches/metric.rs:348:36:
index out of bounds: the len is 0 but the index is 0
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

fn benchmark_collect_histogram(b: &mut Bencher, n: usize) {
    let r = SharedReader(Arc::new(ManualReader::default()));
    let mtr = SdkMeterProvider::builder()
        .with_reader(r.clone())
        .build()
        .meter("sdk/metric/bench/histogram");

    for i in 0..n {
        let h = mtr.u64_histogram(format!("fake_data_{i}")).build();
        h.record(1, &[]);
    }

    let mut rm = ResourceMetrics {
        resource: Resource::builder_empty().build(),
        scope_metrics: Vec::new(),
    };

    b.iter(|| {
        let _ = r.collect(&mut rm);
        assert_eq!(rm.scope_metrics[0].metrics.len(), n); // Fails here
    })

[cijothomas]

@cijothomas benchmark_collect_histogram intermittently fails. I don't want to introduce a CI check that breaks the build. I'd rather comment out this assertion for now with a TODO so we can get this in. What do you think?

thread 'main' panicked at opentelemetry-sdk/benches/metric.rs:348:36:
index out of bounds: the len is 0 but the index is 0
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

fn benchmark_collect_histogram(b: &mut Bencher, n: usize) {
   let r = SharedReader(Arc::new(ManualReader::default()));
   let mtr = SdkMeterProvider::builder()
       .with_reader(r.clone())
       .build()
       .meter("sdk/metric/bench/histogram");

   for i in 0..n {
       let h = mtr.u64_histogram(format!("fake_data_{i}")).build();
       h.record(1, &[]);
   }

   let mut rm = ResourceMetrics {
       resource: Resource::builder_empty().build(),
       scope_metrics: Vec::new(),
   };

   b.iter(|| {
       let _ = r.collect(&mut rm);
       assert_eq!(rm.scope_metrics[0].metrics.len(), n); // Fails here
   })

yes, we can comment it out with TODO, and fix that separately.