[roll] Roll fuchsia [fdomain] Try to handle datagram sockets correctly `zx_socket_read` specifies that when a socket is in datagram mode, a read returns as many bytes as the buffer allows and then discards the rest of the datagram. This became a problem when we started allowing the client to drop a read future, as there's no longer a 1:1 correlation between read calls on the host and read syscalls on the target. This means the length specified when you do a read may not be the length given to the read call. We mitigate this by simply endeavoring to always return a full datagram from the FDomain, ignoring the read argument, and then implementing the truncation on the client side when we're past the IO phase and know the users intentions. The one gap in this strategy is that the FDomain can't know the socket is a datagram socket if we don't have the `INSPECT` right. In such a case things will probably still work if the client doesn't drop read futures or uses a consistent datagram size. Dropping read futures is pretty rare behavior anyway, so the bug should be hard to trigger. Original-Fixed: b/388039830 Original-Reviewed-on: https://fuchsia-review.googlesource.com/c/fuchsia/+/1217686 Original-Revision: fd84b1b2469637855288a7015b09122dcbe60068 GitOrigin-RevId: 57750f5f35458989101e712fb8f2818e4489d7c4 Change-Id: Ia40dfa0ea2d39b4c953ef87edfbf7eca6cef9cd4
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.