add cpu sampling feature

[adriaardila]

Note

Medium Risk
Touches the event-loop hot path and starts a process-wide daemon plus at-fork reinit. Overhead is designed to be tiny, but sampling is on by default and uses sys._current_frames().

Overview
CPU-bound slow tasks now get stack attribution, matching what IO events already have from audit hooks.

A daemon watchdog samples the loop thread after an arming delay (default: half the slow-task threshold) and attaches aggregated stacks with counts to SlowTaskEvent.cpu_stack_samples. detect_slow_tasks() starts this automatically (cpu_sampling=False to opt out); call start_cpu_sampling() first to customize interval, arm delay, idle wake-up, and sample cap.

Sampling follows the thread that published the slice (not only main), restarts after fork() for gunicorn --preload, filters aiocop frames, and floors sleep at 1ms so config cannot busy-loop. GIL-holding C calls still starve the sampler — few samples on a long slice is treated as a signal.

^{Reviewed by Cursor Bugbot for commit f31c3d6. Bugbot is set up for automated code reviews on this repo. Configure here.}

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, have a team admin enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit bb12754. Configure here.}