Odd behaviour

ld999 · Post by **ld999** » Wed Nov 02, 2005 2:51 pm

Just stopped ZM and had this issue, also my machine crashed with lots of kernel debug messages on the screen overnight and had to power off:

Stopping ZoneMinder:
Message from syslogd@Host at Wed Nov 2 14:06:55 2005 ...
Host kernel: Bad page state at free_hot_cold_page (in process 'zmc', page c12e9d40)

Message from syslogd@Host at Wed Nov 2 14:06:55 2005 ...
Host kernel: flags:0x20000014 mapping:00000000 mapcount:1048576 count:0 (Not tainted)

Message from syslogd@Host at Wed Nov 2 14:06:55 2005 ...
Host kernel: Backtrace:

Message from syslogd@Host at Wed Nov 2 14:06:55 2005 ...
Host kernel: Trying to fix it up, but a reboot is needed

Nov 2 14:06:55 Host zma_m1[3686]: INF [Got signal (Terminated), exiting]
Nov 2 14:06:55 Host zmc_d0[3681]: INF [Got TERM signal, exiting]
Nov 2 14:06:55 Host zmc_d0[3681]: ERR [Sync failure for frame 8 buffer 9(1): Interrupted system call]
Nov 2 14:06:55 Host kernel: Bad page state at free_hot_cold_page (in process 'zmc', page c12e9d40)
Nov 2 14:06:55 Host kernel: flags:0x20000014 mapping:00000000 mapcount:1048576 count:0 (Not tainted)
Nov 2 14:06:55 Host kernel: Backtrace:
Nov 2 14:06:55 Host kernel: [<c015416f>] bad_page+0x8a/0xbd
Nov 2 14:06:55 Host kernel: [<c0154c53>] free_hot_cold_page+0x52/0xd1
Nov 2 14:06:55 Host kernel: [<c0160df4>] zap_pte_range+0x100/0x1eb
Nov 2 14:06:55 Host kernel: [<c0160f4e>] unmap_page_range+0x6f/0x9d
Nov 2 14:06:55 Host kernel: [<c0161075>] unmap_vmas+0xf9/0x369
Nov 2 14:06:55 Host kernel: [<c0163d85>] handle_mm_fault+0x191/0x2d9
Nov 2 14:06:55 Host kernel: [<c0167412>] exit_mmap+0xa3/0x268
Nov 2 14:06:55 Host kernel: [<c011d9fa>] mmput+0x23/0x1be
Nov 2 14:06:55 Host kernel: [<c012441c>] do_exit+0xc3/0x506
Nov 2 14:06:55 Host kernel: [<c0124941>] do_group_exit+0xb6/0x1dc
Nov 2 14:06:55 Host kernel: [<c0105912>] do_IRQ+0x53/0x85
Nov 2 14:06:55 Host kernel: [<c010392d>] syscall_call+0x7/0xb
Nov 2 14:06:55 Host kernel: Trying to fix it up, but a reboot is needed
Nov 2 14:07:05 Host su(pam_unix)[5050]: session closed for user apache

Anyone seen this?

Am running 1.21.3 CTU version of Fedora, kernel 2.6.12-1.1372_FC3.

KR
LD

jameswilson · Post by **jameswilson** » Wed Nov 02, 2005 3:55 pm

i havnt seen this but my inital thoughts are you have a bad stick of ram. Do you have a linux boot disk with memtest on it? id try that first

zherebet · Post by **zherebet** » Thu Nov 03, 2005 7:19 am

I've seen this before - *shudder* - I had a box that just did those messages, and then it would hang. Ports would be open, but they wouldn't respond, needed a cold reboot.

Can you post what your set up is? Motherboard/vid card/#inputs/etc that you're using?

I had a problem with an asus mobo (K8S-MX) and two happauge bt878 cards doing this. I disabled one of the cards - and the system stopped throwing the errors. I ended up ordering new hardware in the end - however I have not had the chance to troubleshoot.

My guess is two fold:

1) The motherboard cannot handle the amount of traffic on the PCI bus and barfs
2) Or the second capture card was not working properly and needed to be scrapped.

I did not get a chance to test these theories out.

Phil - what do you think? Can the overloading of the PCI bus (coupled with a cheap motherboard) break the system? If so - I saw a post for multiple PCI bus channel mobos, would they help?

Cheers,

Ilya

zherebet · Post by **zherebet** » Thu Nov 03, 2005 7:21 am

Oh regarding RAM - could be a problem, however I don't believe so - I've had very small captures running with a 512 MB stick, and one card ran fine even when i turned up the resolution on one feed.

zherebet · Post by **zherebet** » Thu Nov 03, 2005 7:23 am

Sorry for the excess posts - but I keep thinking back to what I ran into - I also saw not only ZMC throwing this memory swap error, but I saw other processes, like sshd, apache, mysql, and others throw similar message errors.

ld999 · Post by **ld999** » Thu Nov 03, 2005 5:43 pm

Hi,

I have a custom PC with an ASUS A7M266 MB, 2 x PC210 256MB DIMMS, AMD Athlon 1.4, Nvidia TNT2 Ultra, Intel 10/100/1000.

Capture card is a 4 port Provideo http://www.provideo.com.tw/PV-149P.htm

I have ran Memtest86, twice and no errors have been found.

I have disabled all unnecessary services, hopefully a s/w issue not h/w.

KR
LD

Post by **cordel** » Thu Nov 03, 2005 8:08 pm

If you run into any more issues please let me know as I would be interested in taking a peak to work it out.
Just email me the info.
Regards,
Cordel

ld999 · Post by **ld999** » Fri Nov 04, 2005 1:37 pm

Switched back to my original machine so will soon see if it is h/w or s/w.

jameswilson · Post by **jameswilson** » Fri Nov 04, 2005 1:59 pm

i had a board that reported mem issues no matter what sticks i put it in it turned out to be a mobo issue but it showed up in memtest, plus it locked up totally it didnt recover. it would work for abot 2 weeks before it failed

ld999 · Post by **ld999** » Tue Nov 08, 2005 4:47 pm

I think you are right, running off another machine now with no issues.

jameswilson · Post by **jameswilson** » Tue Nov 08, 2005 5:18 pm

glad you have a working zm now