Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(185)

Issue 255783007: Leak instead of deadlock if executeHardwareAction fails (Closed)

Created:
6 years, 8 months ago by boliu
Modified:
6 years, 7 months ago
CC:
chromium-reviews, android-webview-reviews_chromium.org
Visibility:
Public.

Description

Leak instead of deadlock if executeHardwareAction fails If executeHardwareAction fails in onDetachedFromWindow, then HardwareRenderer destructor will deadlock later when being destroyed on UI thread. In this case, run the destructor on UI without deadlock but leak GL resources instead. Achieve this by running always running release gl even on UI thread, and making the allow_gl global to be a true thread local, which can be set to true on UI thread. BUG= TBR=benm Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=266429

Patch Set 1 #

Unified diffs Side-by-side diffs Delta from patch set Stats (+17 lines, -35 lines) Patch
M android_webview/browser/gl_view_renderer_manager.h View 2 chunks +0 lines, -6 lines 0 comments Download
M android_webview/browser/gl_view_renderer_manager.cc View 1 chunk +0 lines, -13 lines 0 comments Download
M android_webview/browser/hardware_renderer.h View 2 chunks +3 lines, -1 line 0 comments Download
M android_webview/browser/hardware_renderer.cc View 1 chunk +5 lines, -6 lines 0 comments Download
M android_webview/java/src/org/chromium/android_webview/AwContents.java View 1 chunk +9 lines, -9 lines 0 comments Download

Messages

Total messages: 12 (0 generated)
boliu
6 years, 8 months ago (2014-04-26 00:54:30 UTC) #1
hush (inactive)
lgtm
6 years, 8 months ago (2014-04-26 01:17:29 UTC) #2
hush (inactive)
The CQ bit was checked by hush@chromium.org
6 years, 8 months ago (2014-04-26 01:17:36 UTC) #3
commit-bot: I haz the power
CQ is trying da patch. Follow status at https://chromium-status.appspot.com/cq/boliu@chromium.org/255783007/1
6 years, 8 months ago (2014-04-26 02:05:12 UTC) #4
commit-bot: I haz the power
The CQ bit was unchecked by commit-bot@chromium.org
6 years, 8 months ago (2014-04-26 02:05:14 UTC) #5
commit-bot: I haz the power
No LGTM from a valid reviewer yet. Only full committers are accepted. Even if an ...
6 years, 8 months ago (2014-04-26 02:05:15 UTC) #6
boliu
tbr benm
6 years, 8 months ago (2014-04-26 03:14:14 UTC) #7
boliu
The CQ bit was checked by boliu@chromium.org
6 years, 8 months ago (2014-04-26 03:14:18 UTC) #8
commit-bot: I haz the power
CQ is trying da patch. Follow status at https://chromium-status.appspot.com/cq/boliu@chromium.org/255783007/1
6 years, 8 months ago (2014-04-26 03:17:09 UTC) #9
commit-bot: I haz the power
Change committed as 266429
6 years, 7 months ago (2014-04-28 05:45:42 UTC) #10
benm (inactive)
Could we have a clearer log message? Also should it be log.w ? Maybe something ...
6 years, 7 months ago (2014-04-28 12:02:05 UTC) #11
boliu
6 years, 7 months ago (2014-04-28 19:56:43 UTC) #12
Message was sent while issue was closed.
On 2014/04/28 12:02:05, benm wrote:
> Could we have a clearer log message? Also should it be log.w ?
> 
> 
> Maybe something like "Unable to free GL resources. Has the Window leaked?"
> 
> Sent from my Android
> On 26 Apr 2014 04:17, <mailto:commit-bot@chromium.org> wrote:
> 
> > CQ is trying da patch. Follow status at
> > https://chromium-status.appspot.com/cq/boliu@chromium.org/255783007/1
> >
> >
> > https://codereview.chromium.org/255783007/
> >
> 
> To unsubscribe from this group and stop receiving emails from it, send an
email
> to mailto:chromium-reviews+unsubscribe@chromium.org.

Yes. I'll fix it in a follow up. Also occured to me it's not exactly right,
since executeHardwareAction is expected to fail in a software view tree.

Powered by Google App Engine
This is Rietveld 408576698