diff options
| author | jrg@chromium.org <jrg@chromium.org@0039d316-1c4b-4281-b951-d872f2087c98> | 2010-04-13 19:46:14 +0000 |
|---|---|---|
| committer | jrg@chromium.org <jrg@chromium.org@0039d316-1c4b-4281-b951-d872f2087c98> | 2010-04-13 19:46:14 +0000 |
| commit | 8d69af8f41601bf9b4e4e440982477d01a46396c (patch) | |
| tree | f15560727b82bc9ef9e4839c9472cbb9f7228b11 /chrome_frame/chrome_frame_unittest_main.cc | |
| parent | 47aee181fc8b2593f13c1bc6f8ae31a5e546bccf (diff) | |
| download | chromium_src-8d69af8f41601bf9b4e4e440982477d01a46396c.zip chromium_src-8d69af8f41601bf9b4e4e440982477d01a46396c.tar.gz chromium_src-8d69af8f41601bf9b4e4e440982477d01a46396c.tar.bz2 | |
A common pattern for Chrome is for dialogs to "own themselves"
(e.g. fire and forget). As implemented on the Mac, the termination
chain (e.g. when OK is pressed) is to call [NSApp endSheet:].
didEndSheet then closes the window, and the controller (set as the
delegate of the window) gets a windowWillClose: callback (so the
controller knows to delete itself).
The problem here was that the window was ordered out (offscreen) but
not actually closed, so we didn't clean up properly.
BUG=40429
TEST=no valgrind suppression needed and unit_tests stay happy.
nib change: window controller (file's owner) is the delegate of the window
(for both nibs)
Review URL: http://codereview.chromium.org/1654003
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@44383 0039d316-1c4b-4281-b951-d872f2087c98
Diffstat (limited to 'chrome_frame/chrome_frame_unittest_main.cc')
0 files changed, 0 insertions, 0 deletions