Closed
Bug 796115
Opened 12 years ago
Closed 12 years ago
Crash in mozilla::FrameLayerBuilder::ClearManagerData with preserve-3d
Categories
(Core :: Graphics, defect)
Tracking
()
RESOLVED
FIXED
mozilla19
Tracking | Status | |
---|---|---|
firefox18 | --- | verified |
People
(Reporter: jruderman, Unassigned)
References
Details
(Keywords: crash, regression, testcase)
Crash Data
Attachments
(4 files)
Crashes in mozilla::FrameLayerBuilder::ClearManagerData.
Probably a regression from bug 539356. Same crash signature as bug 795591 and bug 795646, which are fixed.
Reporter | ||
Comment 1•12 years ago
|
||
Reporter | ||
Comment 2•12 years ago
|
||
Only crashes debug builds??
Comment 3•12 years ago
|
||
On Windows: bp-22bcb072-b9ea-4309-aed7-440972121001.
Crash Signature: [@ mozilla::FrameLayerBuilder::ClearManagerData] → [@ mozilla::FrameLayerBuilder::ClearManagerData]
[@ mozilla::FrameLayerBuilder::ClearManagerData(nsIFrame*, mozilla::LayerManagerData*)]
OS: Mac OS X → All
Hardware: x86_64 → All
Version: Trunk → 18 Branch
Comment 4•12 years ago
|
||
Attachment #666797 -
Flags: review?(roc)
Comment 5•12 years ago
|
||
This dump shows the problem, because of WrapPreseve3DList, we end up with frames referenced in discrete subtrees.
We only store the LayerManagerData* for one of these on the frame, and the other is left dangling when the frame is removed.
I think we're going to need to store a list of LayerManagerData* per frame. We still want to only store the outermost LMD when we have nested ones though, so I'm going to need to work out a way to detect the difference.
Attachment #666797 -
Flags: review?(roc) → review+
Comment 6•12 years ago
|
||
Fixed by bug 795674.
Status: NEW → RESOLVED
Closed: 12 years ago
Resolution: --- → FIXED
Updated•12 years ago
|
Comment 7•12 years ago
|
||
Bug 795674 was uplifted to Aurora.
Comment 8•12 years ago
|
||
Verified the crash for the testcase in comment 1 for 2012-10-01 Nightly
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:18.0) Gecko/18.0 Firefox/18.0
Build ID: 20121001030603
My crash signature is different from the ones mentioned in bug.
Signature: xul.dll@0x40fd2 | xul.dll@0x40fbe | xul.dll@0x132953 | xul.dll@0x1329d4 | xul.dll@0x2785a | xul.dll@0x1f0e7a | xul.dll@0x730a2 | xul.dll@0x4c01f | xul.dll@0x43d7b2 | xul.dll@0x43d840 | xul.dll@0x115060 | xul.dll@0x1124fd | xul.dll@0x11256c | xul.dll@0x...
http://crash-stats.mozilla.com/report/index/bp-969a36d8-a194-401d-b870-d31362121128
http://crash-stats.mozilla.com/report/index/bp-a2af3e9f-da2d-4c3b-bdd1-69f332121128
Verified testcase from comment 1 also for
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:18.0) Gecko/18.0 Firefox/18.0 Beta 1
Build ID: 20121121075611
and
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:20.0) Gecko/20.0 Firefox/20.0 latest Nightly
Build ID: 20121127030907
No crashes found for these versions.
Can this be moved to "verified" considering that my crash signature is different from the ones mentioned in this bug?
Updated•12 years ago
|
You need to log in
before you can comment on or make changes to this bug.
Description
•