Bug 9231 - Jive: Crashed while rebooting. Takes more than 10 minutes to reboot.
: Jive: Crashed while rebooting. Takes more than 10 minutes to reboot.
Status: RESOLVED WORKSFORME
Product: SB Controller
Classification: Unclassified
Component: Power Management
: unspecified
: Other Other
: -- normal (vote)
: 7.3
Assigned To: Richard Titmuss
:
Depends on:
Blocks:
  Show dependency treegraph
 
Reported: 2008-08-20 15:21 UTC by Wallace Lai
Modified: 2008-10-30 05:06 UTC (History)
3 users (show)

See Also:
Category: ---


Attachments
Putty log for this bug. (133.33 KB, text/plain)
2008-08-20 15:21 UTC, Wallace Lai
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Wallace Lai 2008-08-20 15:21:41 UTC
Created attachment 3845 [details]
Putty log for this bug.

HW:  MP Jive

Note:  Similar issue had happened in the past without the dongle and putty.  Dean and David had come and checked it.  This time, the AutoJiveSmokeTestsSys reproduced it with the dongle connected and putty running.

Steps to Reproduce:
1.  Downgrade Jive to the stable 7.1 r2447 FW.
2.  Upgrade to jive_7.3_r2860.bin.
3.  After the update, the Jive automatically reboots.
4.  Notice the Jive crashed.  The screen is black.  The back light is lit though.
5.  Wait about 15 minutes.  The Jive is on again.  However, none of the keys works.
Comment 1 Richard Titmuss 2008-08-21 04:15:38 UTC
The logs indicate that this jive is sick. For example:

  mapped channel 10 to 2
  libpng error: IDAT: CRC error
  Unable to handle kernel NULL pointer dereference at virtual address 00000000
  pgd = c3500000
  [00000000] *pgd=33a29031, *pte=00000000, *ppte=00000000
  Internal error: Oops: 17 [#1]
  Modules linked in: gspi8xxx(P) gspi(P)
  CPU: 0    Tainted: P        (2.6.22-P7 #1)
  PC is at generic_file_llseek+0x14/0xc0
  LR is at mini_fo_llseek+0xd0/0x108

This is a very unusually error in libpng, followed by a kernel crash in the file filesystem. On reboot we see lots of:

  bad magic number for tty struct (4:2) in tty_poll

That indicates a filesystem problem.

In case this is a 'soft' error in the overlay filesystem please try to perform a factory reset on the jive. Does that fix the problem?

If not this looks like a hardware failure. We need to analyze the ram and flash to determine the cause of the failure.

Wallace, I am correct in thinking this is the same jive that you had problems with last week? If so this is probably a duplicate of bug 9115 and bug 9105, where a similar crashes are seen in the log. On closer inspection of the logs this is the same unit:

  Setting ethaddr: 00:04:20:1A:01:67

If this error is repeatable on many jives then it is possible the toolchain changes when converting to poky have caused an error - but this is very unlikely.

Comment 2 Richard Titmuss 2008-08-21 04:15:54 UTC
*** Bug 9115 has been marked as a duplicate of this bug. ***
Comment 3 Richard Titmuss 2008-08-21 04:16:15 UTC
*** Bug 9105 has been marked as a duplicate of this bug. ***
Comment 4 Wallace Lai 2008-08-21 08:25:49 UTC
There had been at least two Jive units that exibited problems while booting.

Yes, this and all other Jives had been factory reseted many times in the past several weeks.  Sometimes more than once a day.

Who should this Jive be given to for ram and flash analysis?
Comment 5 Richard Titmuss 2008-08-21 08:33:28 UTC
Wallace you should ask David to take a look at it.
Comment 6 Chris Owens 2008-08-21 09:44:33 UTC
Please follow up with David, Wallace.
Comment 7 Wallace Lai 2008-08-21 14:51:15 UTC
David has that Jive now.
Comment 8 Wallace Lai 2008-10-14 09:17:21 UTC
Have not seen this bug for a while.  Some Jives with 7.3 FW continue to refuse boot up all the way.  However, they do not resume booting after 15 minutes.  This bug as written has not been observed on other units.
Comment 9 Richard Titmuss 2008-10-30 05:06:02 UTC
Closing as works for me. Wallace if this happens again please reopen the bug.