fix(otlp): avoid panic in grpc retry without tokio runtime

[Wassbdr]

Summary

Fix panic when experimental-grpc-retry is enabled and exports run from a thread without an active Tokio runtime (e.g. thread-based BatchSpanProcessor).

Fixes #3432.

Root cause

The tonic retry wrapper used Tokio runtime delays unconditionally in retry-enabled builds. The thread-based processors run exports on plain threads without a Tokio reactor, so retry backoff triggered panic (there is no reactor running).

Changes

• Make tonic retry runtime-aware:
  • use Tokio when a Tokio runtime is active
  • fallback to opentelemetry_sdk::runtime::NoAsync otherwise
• Update tonic traces/logs/metrics exporters to use the updated wrapper signature
• Add regression test for no-runtime path in tonic retry wrapper
• Document fix in opentelemetry-otlp/CHANGELOG.md

Validation

Ran locally:

cargo fmt --all -- --check
cargo test -p opentelemetry-otlp --lib test_retry_wrapper_falls_back_without_tokio_runtime --features experimental-grpc-retry
cargo test -p opentelemetry-otlp --lib exporter::tonic::tests --features experimental-grpc-retry

All passed (1/1 + 25/25).

[linux-foundation-easycla]

The committers listed above are authorized under a signed CLA.

• ✅ login: Wassbdr / name: Wassbdr (68b8903, ac1ce14, bed7371, dd409ff)

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 83.4%. Comparing base (ef8ff2f) to head (ac1ce14).

Additional details and impacted files

@@           Coverage Diff           @@
##            main   #3472     +/-   ##
=======================================
+ Coverage   83.3%   83.4%   +0.1%     
=======================================
  Files        128     128             
  Lines      25225   25280     +55     
=======================================
+ Hits       21026   21103     +77     
+ Misses      4199    4177     -22     

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[scottgerring]

Per comment on #3432 , let's hold off on this one until we know what we are doing with the Runtime trait, as it may help