Closed
Bug 826195
Opened 12 years ago
Closed 12 years ago
When leaving print preview: "ASSERTION: print object has mismatching pres shell and pres context: 'false', file /mozilla/layout/printing/nsPrintData.cpp, line 108" followed by "ASSERTION: mPresShell->GetPresShell() is null"
Categories
(Core :: Print Preview, defect)
Tracking
()
RESOLVED
FIXED
mozilla20
People
(Reporter: dholbert, Assigned: heycam)
References
Details
(Keywords: assertion)
Attachments
(1 file)
782 bytes,
patch
|
dholbert
:
review+
|
Details | Diff | Splinter Review |
STR:
0. Be using a debug build
1. Visit about:blank (or perhaps any site...?)
2. File | Print Preview (on Linux / Windows)
3. Click "close" button
ACTUAL RESULTS:
###!!! ASSERTION: print object has mismatching pres shell and pres context: 'false', file /mozilla/layout/printing/nsPrintData.cpp, line 108
###!!! ASSERTION: mPresShell->GetPresShell() is null: 'false', file /mozilla/layout/printing/nsPrintData.cpp, line 111
I hit this using a debug build from up-to-date mozilla-central, from this cset:
https://hg.mozilla.org/mozilla-central/rev/6955309291ee
Reporter | ||
Comment 1•12 years ago
|
||
Actually, looks like these assertions were just added, in bug 818626's cset:
https://hg.mozilla.org/mozilla-central/rev/28f9ce570aa1
I can reproduce this in the earliest m-c-debug nightly to include that patch...
https://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/2013/01/2013-01-02-mozilla-central-debug/
...so I think this has just been broken since bug 818626 landed.
Blocks: 818626
Keywords: regression,
regressionwindow-wanted
Assignee | ||
Comment 2•12 years ago
|
||
Tweak assertion. Turns out it is valid for the nsPresContext to no longer have a reference to the PresShell, as long as Destroy() has already been called on the PresShell (as far as I can tell). I'll assert that over in bug 818626.
Reporter | ||
Comment 3•12 years ago
|
||
Comment on attachment 697823 [details] [diff] [review]
patch
Looks like this addresses the first assertion-failure, but not the second one ("mPresShell->GetPresShell() is null")
It sounds like we might just want to drop that second assertion, based on comment 2...?
r=me with that, or something like it. (The assertion message also needs s/mPresShell/mPresContext/, if we happen to keep it. :))
Attachment #697823 -
Flags: review?(dholbert) → review+
Assignee | ||
Comment 4•12 years ago
|
||
Thanks, removed that other assertion entirely.
https://hg.mozilla.org/integration/mozilla-inbound/rev/9e624b73a293
Comment 5•12 years ago
|
||
Status: ASSIGNED → RESOLVED
Closed: 12 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla20
You need to log in
before you can comment on or make changes to this bug.
Description
•