diff options
| author | thakis@chromium.org <thakis@chromium.org@0039d316-1c4b-4281-b951-d872f2087c98> | 2010-10-02 03:24:13 +0000 |
|---|---|---|
| committer | thakis@chromium.org <thakis@chromium.org@0039d316-1c4b-4281-b951-d872f2087c98> | 2010-10-02 03:24:13 +0000 |
| commit | 3cafc369b174594f0f7e873890aff267233d2209 (patch) | |
| tree | 3def7c63b322ae96439fc13c0103ab7407ccd4f0 /base/thread.h | |
| parent | 6de328fddf450413595d75c2fe3cf2c34f7cee4f (diff) | |
| download | chromium_src-3cafc369b174594f0f7e873890aff267233d2209.zip chromium_src-3cafc369b174594f0f7e873890aff267233d2209.tar.gz chromium_src-3cafc369b174594f0f7e873890aff267233d2209.tar.bz2 | |
Revert 61269 (broke lots of browser and ui tests, like on your try run) - [GTK] delay reference count check in OwnedWidgetGtk
It's a common problem that a GtkWidget will be referenced by an event that's currently on the callstack. These ref holders ignore the "destroy" signal, so when we go to destroy the owned widget, the refcount == 1 DCHECK fails. This change delays that DCHECK until the current stack has unwound, so we don't have to keep running into this.
BUG=none
TEST=in debug builds, Escape in a constrained window doesn't DCHECK, dragging a tab doesn't cause a DCHECK, pressing the custom frame [x] in lucid doesn't DCHECK; in general, no additional DCHECKs
Review URL: http://codereview.chromium.org/3416032
TBR=estade@chromium.org
Review URL: http://codereview.chromium.org/3564007
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@61283 0039d316-1c4b-4281-b951-d872f2087c98
Diffstat (limited to 'base/thread.h')
0 files changed, 0 insertions, 0 deletions