Skip to contentSkip to conversation

mantis crashes

DISCUSSION Vidar Tyldum Hansen 3 participants 11 messages Apr 13, 2010 – Nov 20, 2010 Media (V4L/DVB)

lore
Share & tools
lore
#0 lore raw

Hello list,

I am having issues with my TerraTec Cinergy C DVB-C, described in detail
here: https://bugzilla.kernel.org/show_bug.cgi?id=15750

The essence is that machine would reboot or hang at random intervals if the card is actively used (mythtv or tvheadend running for instance). Just loading the mantis module and let the card idle does not trigger anything.

I am trying to figure out if this is a driver or hardware issue, and by enabling more verbose logging on the mantis module I ended up with the syslog output attached to the bug report. It shows binary garbage and function calls for parts of the kernel not in use by the machine in question right before the crash occurred.

I have not been able to find a different card to test with yet.

So, based on this, is it possible to conclude whom to blame for the crash?

Signature · 2 lines
       Vidar Tyldum Hansen
                                 [email protected]               PGP: 0x3110AA98
#1 lore raw
Quoted 9 lines · Vidar Tyldum Hansen

Hello list,

I am having issues with my TerraTec Cinergy C DVB-C, described in detail
here: https://bugzilla.kernel.org/show_bug.cgi?id=15750

The essence is that machine would reboot or hang at random intervals if the card is actively used (mythtv or tvheadend running for instance). Just loading the mantis module and let the card idle does not trigger anything.

Does the 2.6.31-20-generic Ubuntu kernel come bundled with the mantis driver, or did you build this yourself? If so, from where and which version?

AFAIK, mantis wasn't included in mainline until 2.6.33. I've just filed a wishlist bug for Debian's 2.6.32 so I'd really be interested in any Ubuntu work on this...

Quoted 10 lines

I am trying to figure out if this is a driver or hardware issue, and by enabling more verbose logging on the mantis module I ended up with the syslog output attached to the bug report. It shows binary garbage and function calls for parts of the kernel not in use by the machine in question right before the crash occurred.

I have not been able to find a different card to test with yet.

So, based on this, is it possible to conclude whom to blame for the crash?

FWIW I have exactly the same card with no such problems observed. I am using a standard Debian 2.6.32-4-amd64 kernel with mantis driver taken from vanilla 2.6.34-rc4 (after patching mantis-input.c to compensate for the recent ir layer changes). This is also what I submitted to Debian in the wishlist bug report.

Bjørn

#2 lore raw
Quoted 7 lines · Bjørn Mork
Quoted 5 lines · Vidar Tyldum Hansen

Hello list,

I am having issues with my TerraTec Cinergy C DVB-C, described in detail
here: https://bugzilla.kernel.org/show_bug.cgi?id=15750

Quoted 8 lines
Quoted 4 lines

The essence is that machine would reboot or hang at random intervals if the card is actively used (mythtv or tvheadend running for instance). Just loading the mantis module and let the card idle does not trigger anything.

Does the 2.6.31-20-generic Ubuntu kernel come bundled with the mantis driver, or did you build this yourself? If so, from where and which version?

I should have clarified this, my apologies.
I am running the latest V4L from the official V4L HG repository. I have bluntly built and installed it on top of the Ubuntu kernel, replacing V4L entirely.

Other users with same card and kernel has reported no issues with the same configuration. And the card works for a while...

I realize this is not optimal, though.

Quoted 3 lines

AFAIK, mantis wasn't included in mainline until 2.6.33. I've just filed a wishlist bug for Debian's 2.6.32 so I'd really be interested in any Ubuntu work on this...

Don't know why I didn't think of that. I'll see if I can do similar for Ubuntu to signal the need.

Quoted 8 lines
Quoted 8 lines

I am trying to figure out if this is a driver or hardware issue, and by enabling more verbose logging on the mantis module I ended up with the syslog output attached to the bug report. It shows binary garbage and function calls for parts of the kernel not in use by the machine in question right before the crash occurred.

I have not been able to find a different card to test with yet.

Quoted 8 lines
Quoted 2 lines

So, based on this, is it possible to conclude whom to blame for the crash?

FWIW I have exactly the same card with no such problems observed. I am using a standard Debian 2.6.32-4-amd64 kernel with mantis driver taken from vanilla 2.6.34-rc4 (after patching mantis-input.c to compensate for the recent ir layer changes). This is also what I submitted to Debian in the wishlist bug report.

Interesting. I will see if I can get it working similarly on the Ubuntu .31 kernel.

Quoted 7 lines

Bjørn

--
To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

Signature · 2 lines
       Vidar Tyldum Hansen
                                 [email protected]               PGP: 0x3110AA98
#3 lore raw

I have the exact same symptoms with a technisat cablestar HD2,
also a twinhan clone like the terratec.
Glad to know this isn't a broken hardware issue (that is it seems highly unlikely we have the exact same problem for it to be a hardware malfunction ).

Have used the .31 and .32 and .33 kernel in ubuntu under karmic and lucid. With a own compiled mantis driver from intuxication and linuxtv mercurial repos.

So it kinda looks like this is a ubuntu thing? Or are there alternative conclusions?

Hans

#4 lore raw
Quoted 8 lines · hans van den Bogert

I have the exact same symptoms with a technisat cablestar HD2,
also a twinhan clone like the terratec.
Glad to know this isn't a broken hardware issue (that is it seems highly unlikely we have the exact same problem for it to be a hardware malfunction ).

Have used the .31 and .32 and .33 kernel in ubuntu under karmic and lucid. With a own compiled mantis driver from intuxication and linuxtv mercurial repos.

I have been trying since Hardy :) Always the same.
Also tried the jusst.de and s2-liplianin trees back and forth ever since.

But I have met users with the same clone card and same Ubuntu who had no issues.

Looks like I will be able to try an identical card very soon. I'll let you know how that works out.

Quoted 9 lines

So it kinda looks like this is a ubuntu thing? Or are there alternative conclusions?

Hans

--
To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

Signature · 2 lines
       Vidar Tyldum Hansen
                                 [email protected]               PGP: 0x3110AA98
#5 lore raw
In reply to hans van den Bogert
Quoted 2 lines · hans van den Bogert

So it kinda looks like this is a ubuntu thing? Or are there alternative conclusions?

Yesterday I was able to lend a card from someone who has the same setup as me (Ubuntu Karmic with latest kernel and using s2-lipnianin) and same Terratec card. He too uses the blunt aproach and just build the v4l-tree and installs it on top of the Ubuntu kernel.

After 5 hours of runtime the system crashed again. This time I had a keyboard hooked up, which flashed caps-lock and scroll-lock. Kernel panic, in other words. Different binay garbage in my syslog this time.

So the card hardware is ruled out, and since I know can establish that it is a kernel panic I am almost rcertainly ruling out hardware in general. The system is stable if the card is not used.

Things left to try hardware-wise is to switch to a different PCI slot, but I guess I'll go down the new kernel route... I'll have to do some research regarding v4l versions in .31, .32 and .33 to figure out to which kernel I can 'backport' mantis without replacing v4l completely.

I am by no means a developer so this will hurt.

Signature · 2 lines
       Vidar Tyldum Hansen
                                 [email protected]               PGP: 0x3110AA98
#6 lore raw
Quoted 4 lines · Vidar Tyldum Hansen

Things left to try hardware-wise is to switch to a different PCI slot, but I guess I'll go down the new kernel route... I'll have to do some research regarding v4l versions in .31, .32 and .33 to figure out to which kernel I can 'backport' mantis without replacing v4l completely.

I attached a patch for 2.6.32 to http://bugs.debian.org/577264

This keeps most of the existing DVB subsystem intact, only adding new files with the exception of the necessary one-liner in tda10021.c to avoid unwanted binding to tda10023.

I based the patch on the driver in 2.6.34-rc4, but updating it with newer versions should be trivial.

Bjørn

#7 lore raw
Quoted 7 lines · Bjørn Mork
Quoted 4 lines · Vidar Tyldum Hansen

Things left to try hardware-wise is to switch to a different PCI slot, but I guess I'll go down the new kernel route... I'll have to do some research regarding v4l versions in .31, .32 and .33 to figure out to which kernel I can 'backport' mantis without replacing v4l completely.

Quoted 8 lines

I attached a patch for 2.6.32 to http://bugs.debian.org/577264

This keeps most of the existing DVB subsystem intact, only adding new files with the exception of the necessary one-liner in tda10021.c to avoid unwanted binding to tda10023.

I based the patch on the driver in 2.6.34-rc4, but updating it with newer versions should be trivial.

Great. I will give that a go.
Latest development:
* Swapped cards with known working one: same instability.
* PSU found to be faulty, replaced. Same instability.
* Upgraded from 2.6.31 to 2.6.32 (Ubuntu Lucid). Same instability.

So now I will revert to stock Lucid kernel and give your patch a go. But I have a feeling this has something might be some incompatibility or bug between the card and the motherboard (Asus P5N7A-VM). This, though is beyond me...

Signature · 2 lines
       Vidar Tyldum Hansen
                                 [email protected]               PGP: 0x3110AA98
#8 lore raw
Quoted 5 lines · Vidar Tyldum Hansen <vidar <at> tyldum.com> writes:

So now I will revert to stock Lucid kernel and give your patch a go. But I have a feeling this has something might be some incompatibility or bug between the card and the motherboard (Asus P5N7A-VM). This, though is beyond me...

Late reply, but
Probably is nvidia/bios/asus incompatibility, I used a M3N78-pro

No further luck getting this to work?

#9 lore raw

Den 21.08.2010 21:11, skrev Hans van den Bogert:

Quoted 8 lines
Quoted 5 lines · Vidar Tyldum Hansen <vidar <at> tyldum.com> writes:

So now I will revert to stock Lucid kernel and give your patch a go. But I have a feeling this has something might be some incompatibility or bug between the card and the motherboard (Asus P5N7A-VM). This, though is beyond me...

Quoted 4 lines

Late reply, but
Probably is nvidia/bios/asus incompatibility, I used a M3N78-pro

No further luck getting this to work?

No, I gave up in the end. Mucking around with this also messes with my V4L so I decided to wait for Ubuntu to ship the module.

Signature · 2 lines
Vidar Tyldum
                              [email protected]               PGP: 0x3110AA98
#10 lore raw

Den 31.08.2010 08:02, Vidar Tyldum:

Quoted 11 lines

Den 21.08.2010 21:11, skrev Hans van den Bogert:

Quoted 10 lines
Quoted 5 lines · Vidar Tyldum Hansen <vidar <at> tyldum.com> writes:

So now I will revert to stock Lucid kernel and give your patch a go. But I have a feeling this has something might be some incompatibility or bug between the card and the motherboard (Asus P5N7A-VM). This, though is beyond me...

Late reply, but
Probably is nvidia/bios/asus incompatibility, I used a M3N78-pro

And for the record, ditching the Asus P5N7A-VM fixed the problem in the end.

Signature · 2 lines
Vidar Tyldum
                              [email protected]               PGP: 0x3110AA98