[Hardy] exa limits pixmap width to 8188

Bug #205599 reported by DocHoliday52090
46
Affects Status Importance Assigned to Milestone
xorg-server (Ubuntu)
Fix Released
Medium
Timo Aaltonen

Bug Description

Binary package hint: firefox-3.0

I logged on to xbox.com and clicked on the friend's list link. The page seems to have alot going on, and I barely see it load before firefox 3 beta 4 closes out completely.

I ran it from terminal and received this message:

The program 'firefox' received an X Window System error.
This probably reflects a bug in the program.
The error was 'BadAlloc (insufficient resources for operation)'.
  (Details: serial 38321 error_code 11 request_code 53 minor_code 0)
  (Note to programmers: normally, X errors are reported asynchronously;
   that is, you will receive the error a while after causing it.
   To debug your program, run it with the --sync command line
   option to change this behavior. You can then get a meaningful
   backtrace from your debugger if you break on the gdk_x_error() function.)

The error does not happen on Firefox 3 Beta 4 on Windows.

Using Ubuntu 8.04 LTS Beta but it has occurred on 7.10 as well.

ProblemType: Bug
Architecture: i386
Date: Sun Mar 23 12:16:55 2008
DistroRelease: Ubuntu 8.04
Package: firefox 3.0~b4+nobinonly-0ubuntu1
PackageArchitecture: all
ProcEnviron:
 PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games
 LANG=en_US.UTF-8
 SHELL=/bin/bash
SourcePackage: firefox-3.0
Uname: Linux 2.6.24-12-generic i686

Tags: apport-bug

Related branches

Revision history for this message
DocHoliday52090 (docholiday52090) wrote :
Revision history for this message
DocHoliday52090 (docholiday52090) wrote :

I just want to add that this is reoccurring. It is impossible for me to navigate to that page without Firefox crashing.

Revision history for this message
BrandonTomlinson (druke) wrote :

possible dupe of bug #205568

Revision history for this message
Michael Rooney (mrooney) wrote :

Hi Doc,

Can you give a little more detail in what you had to do to produce this bug? Did you just go to xboxlive.com and select "View Friends" from the "My Xbox" drop down? Or did you have to log in? Just clicking the My Xbox link had no bad effects for me other than displaying the alert: "live.xbox.com:443 uses an invalid security certificate".

Revision history for this message
DocHoliday52090 (docholiday52090) wrote :

I have to log in first and use the drop down box to get "My Friends". Crashes every time.

Revision history for this message
DocHoliday52090 (docholiday52090) wrote :

Oops, "View Friends".

I'm mincing my words today. I'll just include the picture:

Revision history for this message
Matthew Tighe (tighem) wrote :

This does not crash for me. Can you try running from a terminal and see if you get any error messages? Also, perhaps an extension is causing this. You may try disabling your extensions and seeing if the crashing stops.

Changed in firefox-3.0:
status: New → Incomplete
Revision history for this message
Tawsi (n-launchpad-tom-walker-co-uk) wrote :

I have this problem too with the 64-bit version of Hardy.

It doesn't seem to be specifically related to Firefox 3. It also happens in Epiphany but not Konqueror.

It still happens if I run Firefox 2 on a remote machine, running gutsy, and use X11 forwarding over ssh.

It doesn't happen if I run firefox3 on hardy 64-bit with X11 forwarding to display it on OS X.

Revision history for this message
DocHoliday52090 (docholiday52090) wrote :

"This does not crash for me. Can you try running from a terminal and see if you get any error messages? Also, perhaps an extension is causing this. You may try disabling your extensions and seeing if the crashing stops."

Extensions are all disabled, and the message in the terminal is provided in the original post.

Revision history for this message
Tawsi (n-launchpad-tom-walker-co-uk) wrote :

I went through a backtrace and the problem seems to be to do with decoding gifs, which I think was the same problem as people had with howtoforge

The offending gif is

http://live.xbox.com/xweb/lib/images/MyXboxSkinImageSheet.gif

It's got a size of 10000x30

Revision history for this message
DocHoliday52090 (docholiday52090) wrote :

Confirmed. I crash on that. It just closes out completely without warning.

What can be done?

Revision history for this message
Matthew Tighe (tighem) wrote :

It displays properly for me, but since it does not for both you guys I'm confirming this (in xulrunner-1.9, which is the rendering engine for Firefox and Epiphany).

Changed in firefox-3.0:
assignee: nobody → mozilla-bugs
status: Incomplete → Confirmed
Revision history for this message
Matthew Tighe (tighem) wrote :

Also, I don't think it's a memory issue. I put the live CD in a virtual machine running 256M of RAM and it worked fine. Something you guys have in common, though for sure.

Revision history for this message
Tawsi (n-launchpad-tom-walker-co-uk) wrote :

It seems it crashes with any 10000x30 image I make as long as its not a pure colour, although flipping single pixels seems to mean it wont crash until you zoom in, I guess it's smart enough not to store them as a 10000x30 image if it's all one colour.

Revision history for this message
Tawsi (n-launchpad-tom-walker-co-uk) wrote :

Seems Evince will crash in a similar manner if you zoom the image to 100%

Revision history for this message
Tawsi (n-launchpad-tom-walker-co-uk) wrote :

It also happens with stuff like xloadimage, and xpaint

xpaint vanishes on my desktop but is quite happy to load the image when forwarded to my mac
xloadimage seems to avoid crashing but doesn't display the image, it does have an _XError if I set it as a breakpoint. Again it's quite happy to work if I forward it to X11 on OS X.

Revision history for this message
Tawsi (n-launchpad-tom-walker-co-uk) wrote :

I can recreate the problem using a simple C file that uses Xcreatepixmap. It works fine up to 8188x1 and then dies at 8189x1 for 24bits.

Revision history for this message
Matthew Tighe (tighem) wrote :

Sounds like this is not a firefox issue per se but xwindows problem under hardy.

Changed in xulrunner-1.9:
assignee: mozilla-bugs → nobody
Revision history for this message
Tawsi (n-launchpad-tom-walker-co-uk) wrote :

Yeah it seems to be something to do with EXA as the problem goes away when I switch to using XAA with
option "AccelMethod" "XAA"
in the device section of xorg.conf

I'm using the intel graphics driver
xserver-xorg-video-intel

Revision history for this message
Tawsi (n-launchpad-tom-walker-co-uk) wrote :

I ran the whole xserver through gdb and its same problem as mentioned in this post on the xorg list

http://lists.freedesktop.org/archives/xorg/2007-June/025579.html

I've attached a patch that implements the same change as suggested in the post, which fixes all the issues I've mentioned. Don't know if this is the best way to do it since I've never looked at the code before and it does seem to be a superfluous check if the aim is to limit the maximum size of the pixmap.

Revision history for this message
Matthew Tighe (tighem) wrote :

Please see post from Tawsi who found patch for this particular problem.

Changed in xorg-server:
assignee: nobody → ubuntu-core-dev
assignee: ubuntu-core-dev → nobody
Revision history for this message
Scott Kitterman (kitterman) wrote :

Please do not assign bugs to teams unless they have a team specific policy that they be assigned.

Matthew Tighe (tighem)
Changed in xorg-server:
assignee: nobody → desktop-bugs
Changed in xorg-server:
assignee: desktop-bugs → nobody
Revision history for this message
Tawsi (n-launchpad-tom-walker-co-uk) wrote :
Revision history for this message
Timo Aaltonen (tjaalton) wrote :

Fixed in git, thanks!

Changed in xorg-server:
assignee: nobody → tjaalton
importance: Undecided → Medium
status: Confirmed → Fix Committed
Revision history for this message
Launchpad Janitor (janitor) wrote :

This bug was fixed in the package xorg-server - 2:1.4.1~git20080131-1ubuntu6

---------------
xorg-server (2:1.4.1~git20080131-1ubuntu6) hardy; urgency=low

  * 157_fix_exa_pixmap_width.diff:
    A patch from upstream to allow wider pixmaps with EXA (LP: #205599)
  * 158_xkb_wrapping.diff:
    A patch reviewed by upstream that prevents keys from getting stuck under
    certain situations. (LP: #194214)
  * 51_xkb-and-loathing.diff:
    Re-enable. Ignore SIGALRM around calls to Popen()/Pclose() to fix a hang
    when opening menus in OpenOffice.org. (LP: #204137)

 -- Timo Aaltonen <email address hidden> Sun, 30 Mar 2008 19:18:32 +0300

Changed in xorg-server:
status: Fix Committed → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.