Merge lp:~raof/gnome-session/fix-bug-623700 into lp:~ubuntu-desktop/gnome-session/ubuntu

Proposed by Chris Halse Rogers
Status: Merged
Merge reported by: Robert Ancell
Merged at revision: not available
Proposed branch: lp:~raof/gnome-session/fix-bug-623700
Merge into: lp:~ubuntu-desktop/gnome-session/ubuntu
Diff against target: 51 lines (+31/-0)
3 files modified
debian/changelog (+8/-0)
debian/patches/102_fix_negative_transition_in_idletime.patch (+22/-0)
debian/patches/series (+1/-0)
To merge this branch: bzr merge lp:~raof/gnome-session/fix-bug-623700
Reviewer Review Type Date Requested Status
Ubuntu Desktop Pending
Review via email: mp+33605@code.launchpad.net

Description of the change

Fix a corner-case in the session idle code: bug #623700

To post a comment you must log in.
Revision history for this message
Robert Ancell (robert-ancell) wrote :

There doesn't seem to agreement upstream as to if this is a correct fix (and I suspect not enough people, including me, know enough about this to say either way).
Is there a test case to trigger it?
Also, is this intended for maverick or natty or both?

Preview Diff

[H/L] Next/Prev Comment, [J/K] Next/Prev File, [N/P] Next/Prev Hunk
1=== modified file 'debian/changelog'
2--- debian/changelog 2010-08-06 09:46:40 +0000
3+++ debian/changelog 2010-08-25 03:05:57 +0000
4@@ -1,3 +1,11 @@
5+gnome-session (2.31.6-0ubuntu2) maverick; urgency=low
6+
7+ * debian/patches/102_fix_negative_transition_in_idletime.patch:
8+ - Fix race in idle handling which could result in a session being marked
9+ as idle not being marked as active on user activity. (LP: #623700)
10+
11+ -- Christopher James Halse Rogers <raof@ubuntu.com> Wed, 25 Aug 2010 11:51:55 +1000
12+
13 gnome-session (2.31.6-0ubuntu1) maverick; urgency=low
14
15 * New upstream version:
16
17=== added file 'debian/patches/102_fix_negative_transition_in_idletime.patch'
18--- debian/patches/102_fix_negative_transition_in_idletime.patch 1970-01-01 00:00:00 +0000
19+++ debian/patches/102_fix_negative_transition_in_idletime.patch 2010-08-25 03:05:57 +0000
20@@ -0,0 +1,22 @@
21+Description: Fix race condition in session idle->active detection
22+ .
23+ There's a chance that a user could generate some X input on the same tick that
24+ the active->idle threshold is reached. In this case the idle alarm will get
25+ fired, but because IDLETIME never *passed* the threshold the idle->active
26+ alarm will never get fired.
27+Author: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com>
28+Bug: https://bugzilla.gnome.org/show_bug.cgi?id=627903
29+Bug-Ubuntu: https://bugs.launchpad.net/bugs/623700
30+
31+=== modified file 'gnome-session/gs-idle-monitor.c'
32+--- a/gnome-session/gs-idle-monitor.c 2009-07-14 20:36:58 +0000
33++++ b/gnome-session/gs-idle-monitor.c 2010-08-24 22:43:28 +0000
34+@@ -459,6 +459,7 @@
35+ watch->xalarm_positive = XSyncCreateAlarm (GDK_DISPLAY (), flags, &attr);
36+ }
37+
38++ attr.trigger.wait_value = _int64_to_xsyncvalue (_xsyncvalue_to_int64 (watch->interval) - 1);
39+ attr.trigger.test_type = XSyncNegativeTransition;
40+ if (watch->xalarm_negative != None) {
41+ g_debug ("GSIdleMonitor: updating alarm for negative transition wait=%lld",
42+
43
44=== modified file 'debian/patches/series'
45--- debian/patches/series 2010-08-06 08:43:11 +0000
46+++ debian/patches/series 2010-08-25 03:05:57 +0000
47@@ -15,3 +15,4 @@
48 97_fix_query_end_session_crash.patch
49 100_fix_xsmp_stop_crash.patch
50 101_screen_lock_on_suspend.patch
51+102_fix_negative_transition_in_idletime.patch

Subscribers

People subscribed via source and target branches