[webkit-changes] [WebKit/WebKit] 1f6c23: REGRESSION(283414 at main): [WPE][GTK] Network proces...

[Michael Catanzaro]

  Branch: refs/heads/main
  Home:   https://github.com/WebKit/WebKit
  Commit: 1f6c2306d3ed8a591d2e94880b7271703d1906ed
      https://github.com/WebKit/WebKit/commit/1f6c2306d3ed8a591d2e94880b7271703d1906ed
  Author: Michael Catanzaro <mcatanzaro at redhat.com>
  Date:   2024-09-20 (Fri, 20 Sep 2024)

  Changed paths:
    M Source/WebKit/Platform/IPC/unix/ConnectionUnix.cpp

  Log Message:
  -----------
  REGRESSION(283414 at main): [WPE][GTK] Network process crash when writing to pid socket
https://bugs.webkit.org/show_bug.cgi?id=280003

Reviewed by Carlos Garcia Campos.

If the UI process closes its pid socket connection to the auxiliary
process before the auxiliary process is able to successfully write the
pid, the auxiliary process should not crash. That's sort of weird, but
since 283414 at main it's now expected that this might happen if the
ProcessLauncher gets destroyed before the process launches.

In practice, I *suspect* this is actually happening for the
NetworkProcessProxy's ProcessLauncher, because I don't know how else to
explain all the network process crashes I am seeing. That's probably a
bug, and I hesitate to land a change that hides a bug, but it really is
sometimes expected and shouldn't cause a crash. E.g. If the user presses
Ctrl+T and then Ctrl+Q in quick succession, the UI process could very
plausibly quit before a web process launches. Process launching takes a
couple hundred milliseconds on my computer.

* Source/WebKit/Platform/IPC/unix/ConnectionUnix.cpp:
(IPC::sendPIDToPeer):

Canonical link: https://commits.webkit.org/283981@main



To unsubscribe from these emails, change your notification settings at https://github.com/WebKit/WebKit/settings/notifications