[ehds]

What problem does this PR solve?

Issue Number: resolve

Problem Summary:

What is changed and the side effects?

LatencyRecorder defines latency_us = latency * latency_scale_factor, with latency_scale_factor defaulting to 1. Therefore, callback durations passed to the recorder should be in microseconds under the default configuration.

A callback lasting over approximately 2.15 seconds could produce an IntRecorder overflow warning.

Changed:

Event dispatcher read and write callback durations should measured in microsecond values.

Side effects:

• Performance effects:

• Breaking backward compatibility:

---

Check List:

• Please make sure your changes are compilable.
• When providing us with a new feature, it is best to add related tests.
• Please follow Contributor Covenant Code of Conduct.

[Copilot]

🟢 Approval recommended

No active review issues are reported for the callback timing changes from nanoseconds to microseconds.

No open issues 🎉

What changed in this PR

This PR corrects event dispatcher callback timing to use microseconds, matching LatencyRecorder’s default unit.

Changes:

• Switches read and write callback measurements from nanoseconds to microseconds in both dispatcher backends.
• Renames timing variables to start_us, keeping names consistent with the recorded values.

File	Description
src/​brpc/​event_dispatcher_kqueue.cpp	Records kqueue read/write callback durations in microseconds.
src/​brpc/​event_dispatcher_epoll.cpp	Records epoll read/write callback durations in microseconds.

_{Review effort: Balanced}

[chenBright]

LGTM