Bug 697 - web2 segfaults - can't start up on SugarBay
Summary: web2 segfaults - can't start up on SugarBay
Status: RESOLVED WORKSFORME
Alias: None
Product: OE-Core
Classification: Build System, Metadata & Runtime
Component: core (show other bugs)
Version: unspecified
Hardware: Other x86_64
: Medium normal
Target Milestone: 1.0 M4
Assignee: Edwin Zhai
QA Contact:
URL:
Whiteboard:
Depends on:
Blocks:
 
Reported: 2011-02-02 21:37 UTC by Tom Zanussi
Modified: 2011-02-23 17:01 UTC (History)
5 users (show)

See Also:
OS type for building Yocto: ---
Type of Regression: ---
Verified:
Documentation change: ---


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Tom Zanussi 2011-02-02 21:37:54 UTC
Invoking the Web Browser in Sato causes it to start up and then immediately disappear.  The following message was found in dmesg:

web2[650]: segfault at bbadbeef ip 00007f259db89a1d sp 00007fff63afaa60 error 6 in libwebkitgtk-1.0.so.0.4.0[7f259ca1a000+1473000]

Looking at the code, this appears to come from a JavaScript CRASH() assert macro, but I haven't dug further.

This seems to have appeared recently, since it worked fine until my most recent pull from poky/master.

I also haven't been able to test it yet on another machine to verify whether it's just the SugarBay, or just x86_64, etc.
Comment 1 Dexuan Cui 2011-02-17 22:36:58 UTC
Hi Tom,
Just now Edwin and I tested the bug On SugerBay using
http://viggo.jf.intel.com/~tzanussi/yocto/misc/poky-image-sato-live-sugarbay-20110211082154.hddimg.bz2 (from the file name, I think the file was built In Feb 11?) and we didn't find any segfault.

Maybe somebody has fixed the bug??
Comment 2 Edwin Zhai 2011-02-17 23:51:46 UTC
mark it WORKSFORME, as latest image seems to work.

Tom,
could you pls. confirm it?
Thanks,
Comment 3 Tom Zanussi 2011-02-18 14:39:05 UTC
I just built the latest sugarbay and can't verify this, because now the web app itself is gone.
Comment 4 Edwin Zhai 2011-02-20 18:54:26 UTC
(In reply to comment #3)
> I just built the latest sugarbay and can't verify this, because now the web app
> itself is gone.

RP disabled webkit due to the long time of building @ 9d7aa109.

So maybe you need revert this commit to verify it.

Thanks,
edwin
Comment 5 Edwin Zhai 2011-02-21 17:08:00 UTC
CC RP,
Should we disable web-webkit and webkit-gtk for a long time? If so, we can close this low priority bug without bothering TOM to try.

Thanks,
edwin
Comment 6 Tom Zanussi 2011-02-21 22:15:31 UTC
(In reply to comment #4)
> (In reply to comment #3)
> > I just built the latest sugarbay and can't verify this, because now the web app
> > itself is gone.
> 
> RP disabled webkit due to the long time of building @ 9d7aa109.
> 
> So maybe you need revert this commit to verify it.
> 
> Thanks,
> edwin

I reverted this and tried again with code from 2/18 (commit 8048b714cd57c89aecfac438af3c7094a770fb9b) and am still seeing the segfault after starting the browser.
Comment 7 Edwin Zhai 2011-02-21 22:35:05 UTC
(In reply to comment #6) 
> I reverted this and tried again with code from 2/18 (commit
> 8048b714cd57c89aecfac438af3c7094a770fb9b) and am still seeing the segfault
> after starting the browser.

Tom,
I'll try to build on my side, pls. tell me your changeset.

Thanks,
edwin
Comment 8 Kevin Tian 2011-02-21 22:36:24 UTC
also Tom could you share your image some places? that way Edwin can also take a look...
Comment 9 Tom Zanussi 2011-02-22 11:49:17 UTC
(In reply to comment #8)
> also Tom could you share your image some places? that way Edwin can also take a
> look...

Here's the image with 9d7aa109 reverted and that segfaults for me:

http://viggo.jf.intel.com/~tzanussi/yocto/misc/poky-image-sato-live-sugarbay-20110222183817.hddimg.bz2
Comment 10 Dexuan Cui 2011-02-22 17:56:18 UTC
(In reply to comment #9)
> (In reply to comment #8)
> > also Tom could you share your image some places? that way Edwin can also take a
> > look...
> Here's the image with 9d7aa109 reverted and that segfaults for me:
> http://viggo.jf.intel.com/~tzanussi/yocto/misc/poky-image-sato-live-sugarbay-20110222183817.hddimg.bz2

Let's reopen the bug. We'll try the new img.
Comment 11 Dexuan Cui 2011-02-22 19:02:31 UTC
(In reply to comment #10)
> (In reply to comment #9)
> > (In reply to comment #8)
> > > also Tom could you share your image some places? that way Edwin can also take a
> > > look...
> > Here's the image with 9d7aa109 reverted and that segfaults for me:
> > http://viggo.jf.intel.com/~tzanussi/yocto/misc/poky-image-sato-live-sugarbay-20110222183817.hddimg.bz2
> Let's reopen the bug. We'll try the new img.

Hi Tom,
Unfortunately, we still can't reproduce the issue with this new img... :-(
Any special steps we missed? -- we just used dd to write your .hddimg file into a U-disk and booted our SugarBay from it.

Or, we don't have the exactly same type of host as yours???

The info of our SugarBay host is:

Sugar Bay Internal Eng Beta
Sandy Bridge C0 / Cougar Point B0
Los Lunas SCPT CRB Fab A
Comment 12 Tom Zanussi 2011-02-23 07:27:46 UTC
(In reply to comment #11)
> (In reply to comment #10)
> > (In reply to comment #9)
> > > (In reply to comment #8)
> > > > also Tom could you share your image some places? that way Edwin can also take a
> > > > look...
> > > Here's the image with 9d7aa109 reverted and that segfaults for me:
> > > http://viggo.jf.intel.com/~tzanussi/yocto/misc/poky-image-sato-live-sugarbay-20110222183817.hddimg.bz2
> > Let's reopen the bug. We'll try the new img.
> 
> Hi Tom,
> Unfortunately, we still can't reproduce the issue with this new img... :-(
> Any special steps we missed? -- we just used dd to write your .hddimg file into
> a U-disk and booted our SugarBay from it.
> 
> Or, we don't have the exactly same type of host as yours???
> 
> The info of our SugarBay host is:
> 
> Sugar Bay Internal Eng Beta
> Sandy Bridge C0 / Cougar Point B0
> Los Lunas SCPT CRB Fab A

Hi,

There are no special steps required other than what you did.  My hardware is more recent than yours apparently.  Here's the information from my hw:

Intel desktop board 01
scpt_crb
Los Lunas CRB Fab Id 1
Sandy Bridge i7 3400Mhz Stepping D2 / Cougar Point Stepping 03 (B1 Stepping)

It's strange that hw would matter for something like this.

I can see if it happens on another 64-bit machine e.g. the Jasper Forest...
Comment 13 Edwin Zhai 2011-02-23 17:01:57 UTC
(In reply to comment #12)

> Hi,
> 
> There are no special steps required other than what you did.  My hardware is
> more recent than yours apparently.  Here's the information from my hw:
> 
> Intel desktop board 01
> scpt_crb
> Los Lunas CRB Fab Id 1
> Sandy Bridge i7 3400Mhz Stepping D2 / Cougar Point Stepping 03 (B1 Stepping)
> 
> It's strange that hw would matter for something like this.
> 
> I can see if it happens on another 64-bit machine e.g. the Jasper Forest...

Tom,
before I dd the img, I used dd if=/dev/zero of=/dev/sdg bs=1M count=512 to flush the header of the usb disk.

Sometimes, probably cause troubles without it.