[roll] Roll fuchsia [diagnostics] Fix hard spin in detect-stall When `stop_on_idle_timeout_millis` is configured to 0 or a very small value, the debounce interval is 0, causing fasync::Timer to expire immediately on its first poll. If a stream stall is detected, but there are outstanding responders or control handles, the unbind attempt `Arc::try_unwrap` fails. The code then synchronously loops inside `poll_next` to retry after the next debounce interval. With a 0 debounce interval, the timer immediately fires again, resulting in a tight synchronous infinite loop that never yields to the async executor (consuming 100% CPU and making zero syscalls). This change resolves the issue in two ways: 1. Enforces a minimum 1ms debounce interval if less than 1ms or negative is given. 2. Yields Poll::Pending rather than looping when unbinding fails. This ensures control is returned to the event loop so that other tasks can run and drop their outstanding responders. Original-Original-Bug: 515836103 Test: Maunal testing on Sorrel and existing unit tests. Original-Original-Reviewed-on: https://fuchsia-review.googlesource.com/c/fuchsia/+/1621787 Original-Original-Revision: 453021096cb210d0873eea8ddf6a145220fcc698 GitOrigin-RevId: 18cecc57352fac7b4c8192e2687d658d811021f4 Change-Id: I1e9e66d9e5debe25f0bc2b74089092b222220a4e
This repository contains Fuchsia's Global Integration manifest files.
All changes should be made to the internal version of this repository. Our infrastructure automatically updates this version when the internal one changes.
Currently all changes must be made by a Google employee. Non-Google employees wishing to make a change can ask for assistance in one of the communication channels documented at get involved.
First install Jiri.
Next run:
$ jiri init $ jiri import minimal https://fuchsia.googlesource.com/integration $ jiri update
Third party projects should have their own subdirectory in ./third_party.