diff options
| author | bemasc@chromium.org <bemasc@chromium.org@0039d316-1c4b-4281-b951-d872f2087c98> | 2014-01-08 00:29:20 +0000 |
|---|---|---|
| committer | bemasc@chromium.org <bemasc@chromium.org@0039d316-1c4b-4281-b951-d872f2087c98> | 2014-01-08 00:29:20 +0000 |
| commit | a30b21cd28c48c7bbfe09dcb40b991fff9b3150e (patch) | |
| tree | 466b949dfcf5a2e8170bce9484c20e2b8f4f991b /ipc/DEPS | |
| parent | 5865860ff717c87754f93d784ee9aa3e505770a2 (diff) | |
| download | chromium_src-a30b21cd28c48c7bbfe09dcb40b991fff9b3150e.zip chromium_src-a30b21cd28c48c7bbfe09dcb40b991fff9b3150e.tar.gz chromium_src-a30b21cd28c48c7bbfe09dcb40b991fff9b3150e.tar.bz2 | |
Revert 242936 "When stopping QTKit video capture, call stopRunni..."
This CL removed a deadlock on camera removal, but created a new one
on shutdown while a camera is running.
> When stopping QTKit video capture, call stopRunning before removeInput:.
>
> Empirically, this prevents a browser hang when the active camera is
> unplugged. The unplug event triggers both stopping the capturer and
> enumerating the available input devices at the same time. It seems
> that calling removeInput: on a running capturer after an error
> produces a deadlock, but stopping the capturer first resolves the
> problem.
>
> BUG=285053
>
> Review URL: https://codereview.chromium.org/120963003
TBR=bemasc@chromium.org
Review URL: https://codereview.chromium.org/126883005
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@243445 0039d316-1c4b-4281-b951-d872f2087c98
Diffstat (limited to 'ipc/DEPS')
0 files changed, 0 insertions, 0 deletions