could be a themes crash or a sync crash I guess

Most likely not themes; GTK is very chatty with theme errors (and will continue anyway 
with "fatal" errors) so we have to whitelist them. They look like they come from 
chrome because of our error handler.

My money is on sync. Do you have a stack trace?

Adding some folks who may know more about this AsyncDNS thingamajig.  
It looks to me just like the two URLs we set up for talk connectivity fail to get 
resolved for reasons unknown, though I'm still searching for what could cause the seg 
fault..

This'll be pretty tough to debug without a backtrace or a repro case.

Or maybe the revision number of the last known working build, so as to narrow down 
the list of possible breaking changes.

got the same issue on ubuntu karmic 64bit:

[5555:5667:2077666045:ERROR:chrome/browser/sync/notifier/base/async_dns_lookup.cc(94)] (talk.google.com)
error: 1
[5555:5678:2077902438:ERROR:chrome/browser/sync/notifier/base/async_dns_lookup.cc(94)] (talkx.l.google.com)
error: 1
Segmentation fault

Stack trace attached.

chromium-browser_Issue_33091_strace.txt 160 KB   View   Download

The stack trace is of r37082.
Issue appears for me at revision 36982.
Last functional revision: 36724

That's not a stack trace but an strace (system trace) run, which is less helpful.  However, 
the last known good revision does help.  Thanks!

Another thing that might be helpful would be the last page or so of logs when running 
with --log-level=0 on a failing case.  That would provide more clues.

I will see if I can repro this.

Sorry.
here is the full gdb output.

gdb-chromium-issue-33091.txt 46.1 KB   View   Download

Thanks, that gdb run is very helpful!

I think http://src.chromium.org/viewvc/chrome?view=rev&revision=36787 is what 
exposed this.  Before that CL it would just swallow the connection errors.

Tracing through the code in login.cc, auto_reconnect_ isn't being initialized because 
network_status is NULL (because there's no Linux implementation yet).  Zork has a CL 
for the network_status implementation out for review.  That should fix it.  I don't 
think we need to gracefully handle the null network_status case gracefully.

Assigning to zork, since he's doing the CL that'll fix this.  The fix should be merged into 
the beta branch, too (306 now).

(No comment was entered for this change.)

The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=37342 

------------------------------------------------------------------------
r37342 | zork@chromium.org | 2010-01-27 15:41:50 -0800 (Wed, 27 Jan 2010) | 6 lines
Changed paths:
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/sync/engine/syncapi.cc?r1=37342&r2=37341
   A http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/sync/notifier/base/linux/async_network_alive_linux.cc
   D /trunk/src/chrome/browser/sync/notifier/base/linux/network_status_detector_task_linux.cc
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/chrome.gyp?r1=37342&r2=37341

Add an implementation for async_network_alive to Linux

BUG=33091
TEST=none

Review URL: http://codereview.chromium.org/552162
------------------------------------------------------------------------

Bug fixed for me as of r37376
Thanks!

The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=37599 

------------------------------------------------------------------------
r37599 | zork@chromium.org | 2010-01-29 17:46:45 -0800 (Fri, 29 Jan 2010) | 8 lines
Changed paths:
   M http://src.chromium.org/viewvc/chrome/branches/307/src/chrome/browser/sync/engine/syncapi.cc?r1=37599&r2=37598
   A http://src.chromium.org/viewvc/chrome/branches/307/src/chrome/browser/sync/notifier/base/linux/async_network_alive_linux.cc
   D /branches/307/src/chrome/browser/sync/notifier/base/linux/network_status_detector_task_linux.cc
   M http://src.chromium.org/viewvc/chrome/branches/307/src/chrome/chrome.gyp?r1=37599&r2=37598

Merge 37342 - Add an implementation for async_network_alive to Linux

BUG=33091
TEST=none

Review URL: http://codereview.chromium.org/552162

TBR=zork@chromium.org
------------------------------------------------------------------------

This appears to be fixed and merged to 307. Updating status.

 Issue 34435  has been merged into this issue.

Merged: removing formerge label.

Milestone 5 fixed bugs, assigning for verification. 

Verified, build 7.0.517.44.
Closing.

Chrome Version       : 4.0.306.0~svn20100124r36982-0ubuntu2~ucd1~karmic
OS + version : Ubuntu 9.10
CPU architecture (32-bit / 64-bit): 32-bit
window manager : Gnome
<b>URLs (if applicable) :</b>
Behavior in Firefox 3.x (if applicable): Browser stays open.
<b>Behavior in Chrome for Windows (optional):</b>

<b>What steps will reproduce the problem?</b>
<b>1.</b>
<b>2.</b>
<b>3.</b>

<b>What is the expected result?</b>

Browser opens and stays open for basic web browsing.

<b>What happens instead?</b>

After upgrade to the latest daily build program opens and after a few
seconds closes. Text of error:

[22606:22606:18749507376:ERROR:chrome/app/chrome_dll_main.cc(223)] GTK
theme error: Theme directory scalable/apps/24 of theme OSX_Iconset has no
size field

[22606:22606:18749510777:ERROR:chrome/app/chrome_dll_main.cc(223)] GTK
theme error: Theme directory  scalable/stock/128 of theme OSX_Iconset has
no size field

[22606:22816:18757248541:ERROR:chrome/browser/sync/notifier/base/async_dns_lookup.cc(94)]
(talk.google.com) error: 11001
[22606:22972:18782435244:ERROR:chrome/browser/sync/notifier/base/async_dns_lookup.cc(94)]
(talkx.l.google.com) error: 11001
Segmentation fault


<b>Please provide any additional information below. Attach a screenshot</b>
<b>and backtrace if possible.</b>