[mythtv-users] Re: axel's patched kernel kernel panics with i vtv -- stock fedora one doesn't

Bill Oberman Oberman at BodyMedia.Com
Sun Feb 1 17:14:15 EST 2004 

I had this problem last week, but only posted to the ivtv mailing list.  I
don't know the how's and why's, but I came to the same solution of renaming
msp3400.o and tuner.o.

Next time I should cross post to myth's list?

william

-----Original Message-----
From: Jarod C. Wilson [mailto:jcw at wilsonet.com] 
Sent: Sunday, February 01, 2004 1:49 PM
To: Discussion about mythtv
Subject: Re: [mythtv-users] Re: axel's patched kernel kernel panics with
ivtv -- stock fedora one doesn't

On Feb 1, 2004, at 11:18 AM, Scott White wrote:

> Loads fine if I remove the cx88 rpm.  The cx88 driver loads fine with  
> or
> without ivtv
>
> Looks like the culprit is
>
> /lib/modules/2.4.22-1.2163.nptl_34.rhfc1.at/updates/drivers/video/ 
> media/tuner.o
>
> If I remove this and depmod -a, all is fine in ivtv land.  The cx8800
> module even loads without it.  I get tuner.o from
>
> /lib/modules/2.4.22-1.2163.nptl_34.rhfc1.at/kernel/drivers/media/ 
> video/tuner.o
>
> I am guessing this is the same for you bttv users.

Yeah, that's right in line with my guess below, (that it was probably  
either the msp3400 or tuner module actually causing the problem) and my  
system was bombing out just after loading up the tuner module. I  
believe tuner is now part of the video4linux-kmdl package now, which  
gets pulled in by both bttv and cx88. So it would appear something has  
changed in the tuner module recently that is incompatible with ivtv...  
I'm guessing it isn't specific to Axel's rpms, but a more general  
thing, for anyone using the very latest v4l components from Gerd along  
with ivtv.


> On Sun, 2004-02-01 at 18:16, Scott White wrote:
>> Hmm, looks like I may be hitting a related problem.  This rig is
>> unchanged and was working 12 hours ago with RH9 and Alex's RPMs and  
>> with
>> my own kernel which includes dvb.
>>
>> [root at linserv etc]# modprobe ivtv
>> Segmentation fault
>>
>> [root at linserv root]# lsmod | grep ivtv
>> ivtv                   89324   1  (initializing)
>> videodev                8928   0  [ivtv]
>> i2c-algo-bit            8552   0  [ivtv]
>> i2c-core               20868   0  [ivtv msp3400 saa7115 tuner
>> i2c-algo-bit]
>>
>> [root at linserv root]# rpm -q -a | egrep \
>> '(kmdl|kernel|ivtv|bttv|v4l|cx88)' | sort
>> alsa-kmdl-2.4.22-1.2163.nptl_34.rhfc1.at-1.0.2-20.rhfc1.at
>> cx88-kmdl-2.4.22-1.2163.nptl_34.rhfc1.at-0.0.1-5.rhfc1.at
>> ivtv-0.1.9-24.rhfc1.at
>> ivtv-firmware-17_21288-1.at
>> ivtv-kmdl-2.4.22-1.2163.nptl_34.rhfc1.at-0.1.9-24.rhfc1.at
>> kernel-2.4.22-1.2163.nptl_34.rhfc1.at
>> kernel-utils-2.4-9.1.101.fedora
>> perl-Video-ivtv-0.12-3
>>
>>
>> Feb  1 17:58:10 linserv kernel: i2c-core.o: i2c core module version
>> 2.8.2 (20031211)
>> Feb  1 17:58:10 linserv kernel: Linux video capture interface: v1.00
>> Feb  1 17:58:10 linserv kernel: ivtv: version 0.1.9 (release) loading
>> Feb  1 17:58:10 linserv kernel: ivtv: SGarray_size = 320,  
>> DSGarray_size
>> = 16
>> Feb  1 17:58:10 linserv kernel: ivtv: Found an iTVC16 based chip
>> Feb  1 17:58:10 linserv kernel: PCI: Found IRQ 5 for device 00:0a.0
>> Feb  1 17:58:10 linserv kernel: ivtv: Unreasonably low latency timer,
>> setting to 64 (was 32)
>> Feb  1 17:58:10 linserv kernel: tuner: chip found @ 0xc2
>> Feb  1 17:58:10 linserv kernel: Unable to handle kernel NULL pointer
>> dereference at virtual address 000003be
>> Feb  1 17:58:10 linserv kernel:  printing eip:
>> Feb  1 17:58:10 linserv kernel: c026a29c
>> Feb  1 17:58:10 linserv kernel: *pde = 0810a067
>> Feb  1 17:58:10 linserv kernel: *pte = 00000000
>> Feb  1 17:58:10 linserv kernel: Oops: 0000
>> Feb  1 17:58:10 linserv kernel: ivtv msp3400 saa7115 tuner videodev
>> i2c-algo-bit i2c-core parport_pc lp parport nfsd lockd sunrpc autofs4
>> 3c59x floppy sg scsi_mod serial microcode keybdev mo
>> Feb  1 17:58:10 linserv kernel: CPU:    0
>> Feb  1 17:58:10 linserv kernel: EIP:    0060:[<c026a29c>]    Not  
>> tainted
>> Feb  1 17:58:10 linserv kernel: EFLAGS: 00010097
>> Feb  1 17:58:10 linserv kernel:
>> Feb  1 17:58:10 linserv kernel: EIP is at vsnprintf [kernel] 0x2ec
>> (2.4.22-1.2163.nptl_34.rhfc1.at)
>> Feb  1 17:58:10 linserv kernel: eax: 000003be   ebx: 0000000a   ecx:
>> 000003be   edx: fffffffe
>> Feb  1 17:58:10 linserv kernel: esi: c03ad87a   edi: 00000000   ebp:
>> c03adc5f   esp: c80edd18
>> Feb  1 17:58:10 linserv kernel: ds: 0068   es: 0068   ss: 0068
>> Feb  1 17:58:10 linserv kernel: Process modprobe (pid: 2416,
>> stackpage=c80ed000)
>> Feb  1 17:58:10 linserv kernel: Stack: c03ad876 c03adc5f ffffffff
>> ffffffff 0000000a ffffffff 00000000 00000002
>> Feb  1 17:58:10 linserv kernel:        ffffffff ffffffff c03ad860
>> 00000246 c839b7a8 00000000 c011cfe4 c03ad860
>> Feb  1 17:58:10 linserv kernel:        00000400 d093b71b c80edd90
>> 00000061 00000061 00000000 c839b700 c839b700
>> Feb  1 17:58:10 linserv kernel: Call Trace:   [<c011cfe4>] printk
>> [kernel] 0x64 (0xc80edd50)
>> Feb  1 17:58:10 linserv kernel: [<d093b71b>] .rodata.str1.32 [tuner]
>> 0x81b (0xc80edd5c)
>>
>>
>> On Sun, 2004-02-01 at 11:28, Jarod C. Wilson wrote:
>>> On Feb 1, 2004, at 1:23 AM, Axel Thimm wrote:
>>>
>>>> On Sun, Feb 01, 2004 at 12:57:36AM -0800, Reza Naima wrote:
>>>>> I removed Axel's kernel, and all the various ivtv/bttv drivers.  I
>>>>> reinstalled the kernel RPMs, and built the ivtv driver by hand.  It
>>>>> seemed to work this time.
>>>>>
>>>>> I'm guessing something must have gotten out of whack.  I'm just  
>>>>> glad I
>>>>> didn't have to do a full rebuild like I had planned.
>>>>
>>>> O.K., so it's the ivtv driver. Which version are you using now? The
>>>> 0.1.9 released version or the CVS snapshot?
>>>> -- 
>>>> Axel.Thimm at physik.fu-berlin.de
>>>
>>> This may be a more general problem, actually. There's an almost
>>> identical posting on the v4l mailing lists for someone using bttv
>>> 0.9.12 with a pcHDTV card. My oopses looked nearly identical, but  
>>> with
>>> ivtv inserted instead of bttv in most places. That leads me to  
>>> believe
>>> it may not actually be the bttv module itself, but rather something
>>> like the msp3400 or tuner modules causing the problem. See below...
>>> (FYI, the oops below is just after msp3400 loads, mine was just after
>>> tuner loaded, so who knows what the real culprit is). I can provide a
>>> full listing like below for my system if it helps.

-- 
Jarod C. Wilson, RHCE

Got a question? Read this first...
     http://catb.org/~esr/faqs/smart-questions.html
MythTV, Fedora Core & ATrpms documentation:
     http://wilsonet.com/mythtv/
MythTV Searchable Mailing List Archive
     http://www.gossamer-threads.com/archive/MythTV_C2/

More information about the mythtv-users mailing list