Nebula DigiTV DVB-T (PCI) no longer works

Bug #368215 reported by touc
98
This bug affects 16 people
Affects Status Importance Assigned to Milestone
linux (Ubuntu)
Won't Fix
Undecided
Unassigned

Bug Description

Binary package hint: kaffeine

My Nebula DigiTV DVB-T (PCI) card does not work under Jaunty, either the i686 or x86_64 version. This is a multi-boot system and when I boot into Intrepid (v2.6.27-11-generic#1 SMP i686) it works fine.

I did a clean install and when I try to scan for channels using Kaffeine it gets nothing and displays the following:

fred@cartman:/media$ kaffeine
kbuildsycoca running...
Reusing existing ksycoca
/dev/dvb/adapter0/frontend0 : opened ( Zarlink MT352 DVB-T )
0 EPG plugins loaded for device 0:0.
Loaded epg data : 0 events (0 msecs)
fred@cartman:/media$ DvbCam::probe(): /dev/dvb/adapter0/ca0: : No such file or directory
Using DVB device 0:0 "Zarlink MT352 DVB-T"
tuning DVB-T to 177500000 Hz
inv:2 bw:1 fecH:9 fecL:9 mod:6 tm:2 gi:4 hier:4
.FE_READ_STATUS: Remote I/O error

Transponders: 1/57
scanMode=0
it's dvb 2!

Invalid section length or timeout: pid=17

Invalid section length or timeout: pid=0

Frontend closed
Using DVB device 0:0 "Zarlink MT352 DVB-T"
tuning DVB-T to 184500000 Hz
inv:2 bw:1 fecH:9 fecL:9 mod:6 tm:2 gi:4 hier:4
.FE_READ_STATUS: Remote I/O error

/var/log/messages looks like:

Apr 27 21:52:36 cartman kernel: [ 10.928721] bttv: driver version 0.9.17 loaded
Apr 27 21:52:36 cartman kernel: [ 10.928726] bttv: using 8 buffers with 2080k (520 pages) each for capture
Apr 27 21:52:36 cartman kernel: [ 10.928782] bttv: Bt8xx card found (0).
Apr 27 21:52:36 cartman kernel: [ 10.928802] bttv 0000:05:07.0: PCI INT A -> Link[APC2] -> GSI 17 (level, low) -> IRQ 17
Apr 27 21:52:36 cartman kernel: [ 10.928812] bttv0: Bt878 (rev 17) at 0000:05:07.0, irq: 17, latency: 32, mmio: 0xd2000000
Apr 27 21:52:36 cartman kernel: [ 10.929515] bttv0: detected: Nebula Electronics DigiTV [card=104], PCI subsystem ID is 0071:0101
Apr 27 21:52:36 cartman kernel: [ 10.929520] bttv0: using: Nebula Electronics DigiTV [card=104,autodetected]
Apr 27 21:52:36 cartman kernel: [ 10.929654] bttv0: tuner absent
Apr 27 21:52:36 cartman kernel: [ 10.929719] bttv0: registered device video1
Apr 27 21:52:36 cartman kernel: [ 10.929742] bttv0: registered device vbi0
Apr 27 21:52:36 cartman kernel: [ 10.929760] bttv0: PLL: 28636363 => 35468950 .. ok
Apr 27 21:52:36 cartman kernel: [ 10.961077] bttv0: add subdevice "dvb0"
Apr 27 21:52:36 cartman kernel: [ 10.961206] input: bttv IR (card=104) as /devices/pci0000:00/0000:00:09.0/0000:05:07.0/input/input6
Apr 27 21:52:36 cartman kernel: [ 10.994975] synaptics was reset on resume, see synaptics_resume_reset if you have trouble on resume
Apr 27 21:52:36 cartman kernel: [ 11.013425] bt878: AUDIO driver version 0.0.0 loaded
Apr 27 21:52:36 cartman kernel: [ 11.013628] bt878: Bt878 AUDIO function found (0).
Apr 27 21:52:36 cartman kernel: [ 11.013645] bt878 0000:05:07.1: PCI INT A -> Link[APC2] -> GSI 17 (level, low) -> IRQ 17
Apr 27 21:52:36 cartman kernel: [ 11.013648] bt878_probe: card id=[0x1010071],[ Nebula Electronics DigiTV ] has DVB functions.
Apr 27 21:52:36 cartman kernel: [ 11.013655] bt878(0): Bt878 (rev 17) at 05:07.1, irq: 17, latency: 32, memory: 0xd2001000
Apr 27 21:52:36 cartman kernel: [ 11.018482] scsi 6:0:0:0: Direct-Access VIA-P VT6205-DevB 2.82 PQ: 0 ANSI: 2
Apr 27 21:52:36 cartman kernel: [ 11.019101] scsi 6:0:0:1: Direct-Access VIA-P VT6205-DevM 2.82 PQ: 0 ANSI: 2
Apr 27 21:52:36 cartman kernel: [ 11.021611] sd 6:0:0:0: [sdc] Attached SCSI removable disk
Apr 27 21:52:36 cartman kernel: [ 11.021740] sd 6:0:0:0: Attached scsi generic sg4 type 0
Apr 27 21:52:36 cartman kernel: [ 11.022689] sd 6:0:0:1: [sdd] Attached SCSI removable disk
Apr 27 21:52:36 cartman kernel: [ 11.022772] sd 6:0:0:1: Attached scsi generic sg5 type 0
Apr 27 21:52:36 cartman kernel: [ 11.050652] DVB: registering new adapter (bttv0)
Apr 27 21:52:36 cartman kernel: [ 11.058791] usbcore: registered new interface driver snd-usb-audio
Apr 27 21:52:36 cartman kernel: [ 11.142234] ACPI: PCI Interrupt Link [APCJ] enabled at IRQ 22
Apr 27 21:52:36 cartman kernel: [ 11.142241] Intel ICH 0000:00:04.0: PCI INT A -> Link[APCJ] -> GSI 22 (level, low) -> IRQ 22
Apr 27 21:52:36 cartman kernel: [ 11.308995] DVB: registering adapter 0 frontend 0 (Zarlink MT352 DVB-T)...
A
.
.
.
Apr 27 21:54:50 cartman kernel: [ 149.693129] mt352_read_register: readreg error (reg=137, ret==-6)
Apr 27 21:54:50 cartman kernel: [ 149.693715] mt352_read_register: readreg error (reg=138, ret==-6)
Apr 27 21:54:53 cartman kernel: [ 152.692542] mt352_write() to reg 89 failed (err = -6)!
Apr 27 21:55:01 cartman kernel: [ 160.599305] mt352_read_register: readreg error (reg=137, ret==-6)
Apr 27 21:55:01 cartman kernel: [ 160.599806] mt352_read_register: readreg error (reg=138, ret==-6)
Apr 27 21:55:01 cartman kernel: [ 160.600787] mt352_write() to reg 51 failed (err = -6)!
Apr 27 21:55:01 cartman kernel: [ 160.601285] mt352_write() to reg 5d failed (err = -6)!
Apr 27 21:55:01 cartman kernel: [ 160.700793] mt352_read_register: readreg error (reg=0, ret==-6)
Apr 27 21:55:02 cartman kernel: [ 161.085569] mt352_read_register: readreg error (reg=20, ret==-6)
Apr 27 21:55:02 cartman kernel: [ 161.086069] mt352_read_register: readreg error (reg=21, ret==-6)
Apr 27 21:55:02 cartman kernel: [ 161.086867] mt352_read_register: readreg error (reg=9, ret==-6)
Apr 27 21:55:02 cartman kernel: [ 161.090498] mt352_read_register: readreg error (reg=0, ret==-6)
Apr 27 21:55:02 cartman kernel: [ 161.404535] mt352_read_register: readreg error (reg=0, ret==-6)
Apr 27 21:55:02 cartman kernel: [ 161.405039] mt352_write() to reg 51 failed (err = -6)!
Apr 27 21:55:02 cartman kernel: [ 161.405536] mt352_write() to reg 5d failed (err = -6)!

ProblemType: Bug
Architecture: i386
DistroRelease: Ubuntu 9.04
ExecutablePath: /usr/bin/kaffeine
Package: kaffeine 0.8.7-1ubuntu5
ProcEnviron:
 SHELL=/bin/bash
 LANG=en_GB.UTF-8
 LANGUAGE=
SourcePackage: kaffeine
Uname: Linux 2.6.28-11-generic i686

Revision history for this message
touc (stephen-bainbridge) wrote :
Revision history for this message
Roberto Tyley (roberto-tyley) wrote :

I can confirm this issue on my Jaunty box (which was an upgrade from Intrepid rather than a full install). My dmesg is attached, which contains all the mt352 messages - I don't think this bug is particular to kaffeine tho, as I don't use it- I'm using MythTV.

Revision history for this message
Antti Seppälä (antseppa) wrote :

Exactly same thing happened to me. I upgraded from Intrepid to Jaunty and mt352 stopped working. Shows similar error messages in my dmesg already during bootup (probably when mythbackend starts).

Revision history for this message
Chris Gresty (chris-gresty) wrote :

Similar issue here - running a fresh installation of Jaunty on x86_64.

I'm trying to set up MythTV - this can identify the Nebula DigiTV card correctly, but cannot scan for channels. I get the same messages in syslog as the original poster.

Revision history for this message
chipsugar (ubuntu-chipsugar) wrote :

This is actually probably a kernel problem. I only use Gentoo and Mandriva myself at the moment but in kernel >2.6.28.* the tuning and channel switching fail whereas in <2.6.27.* the card works fine.

Revision history for this message
Jonathan Thomas (echidnaman) wrote :

Reassigning to the kernel package, thanks.

affects: kaffeine (Ubuntu) → linux (Ubuntu)
Revision history for this message
Peter Springett (peterspringett) wrote :
Revision history for this message
James Harris (james-harris-1) wrote :

Same issue. Why did I upgrade...?

Any time the machine tries to record it gets ret of -6 as follows.

Sep 9 01:21:06 hfx1 kernel: [ 277.606234] mt352_read_register: readreg error (reg=137, ret==-6)
Sep 9 01:21:06 hfx1 kernel: [ 277.606869] mt352_read_register: readreg error (reg=138, ret==-6)
Sep 9 01:21:06 hfx1 kernel: [ 277.647684] mt352_write() to reg 51 failed (err = -6)!
Sep 9 01:21:06 hfx1 kernel: [ 277.648629] mt352_write() to reg 5d failed (err = -6)!
Sep 9 01:21:06 hfx1 kernel: [ 277.699342] mt352_read_register: readreg error (reg=0, ret==-6)
Sep 9 01:21:06 hfx1 kernel: [ 277.739990] mt352_read_register: readreg error (reg=0, ret==-6)
Sep 9 01:21:06 hfx1 kernel: [ 277.742616] mt352_read_register: readreg error (reg=20, ret==-6)
Sep 9 01:21:06 hfx1 kernel: [ 277.743249] mt352_read_register: readreg error (reg=21, ret==-6)
etc etc

Revision history for this message
MarcRandolph (mrand) wrote :

Usually regressions like this are fixed in the following kernel releases, but it doesn't appear to be the case here since there are now reports of this still not working in 9.10.

Changed in linux (Ubuntu):
status: New → Confirmed
Revision history for this message
MarcRandolph (mrand) wrote :

Bug #366606 (marked as dup of this) points out that it is trying to access a Nebula DigiTV PCI card with a Zarlink mt352 chip and driver.

Revision history for this message
Ardo Beltza (ardo-beltza) wrote :

Hello,

I apologise for posting the following text to an earlier closed bug report, so thought I'd better repeat it here where it seems relevant.

I am experiencing the same issue as the OP.

I am using the Nebula Electronics DigiTV card. In 8.04 and 8.10 this worked perfectly with MythTV. After installing 9.04 the card seems to be detected but channel scanning indicates a Signal/Noise: of 98% and 0% Signal Strength. No channels are detected. Today I installed Ubuntu 9.10 beta and allowed all the updates to be installed. The issue remains with all channels indicating "Timed out, no signal".

It appears to me that this is a kernel issue which arrose during the change from 8.10 to 9.04. I read somewhere that somebody with the Nebula problem on 9.04 loaded an older kernel and the card started working again.

I have attached my dmesg from 9.10 beta and would be pleased to help with any tesing.

Thanks.

Revision history for this message
my4keys82 (michael-handford) wrote :

Same problem here, clean install of 9.04 and updated to 2.6.28-15-generic.
8.10 worked fine.

Revision history for this message
Tom Allerton (tom-other) wrote :

Exactly the same problem here too. I've tested it with every major kernel release using Ubuntu 9.04. It works fine in 2.6.27-11, but the problem starts with 2.6.28-1 and still happens in 2.6.32-rc3.

Revision history for this message
James Lambert (jms-lambert) wrote :

I also have this issue on Mythbuntu 9.10, with the Nebula card and mt352 errors. The only difference is my Nebula is one of the newer models that does have the mt352 chip rather than the nxt6000. As soon as I boot the machine I'll get the mt352 errors in dmesg and the card won't work, however running "modprobe -vr dvb_bt8xx" then "modprobe -v dvb_bt8xx" will restore the functionality of the card until the next reboot.

Revision history for this message
rob_rupe (rpo-deactivatedaccount) wrote :

I have the the same issue with my install of Ubuntu x86_64 Karmic Koala 9.10 and the older design of the PCI card(NXT6000). I can confirm the above steps using modprobe gets things working. But it still puts hundreds of the following error messages in dmesg

[ 618.264917] DVB: adapter 0 frontend 0 frequency limits undefined - fix the driver

And it only lasts till the computer gets turned off.

Revision history for this message
redu (reino.urala.gmail) wrote :

Accidentally Nebula DVB-T now works.
No modprobing tricks required.

Compiled latest Liplianian drivers, to use my DVB-S2 card (Azurewave):
http://mercurial.intuxication.org/hg/s2-liplianin
Now DVB-S2 card works, and accidentally also this Nebula PCI DVB-T works.
Possibly Liplianin also fixed this Nebula bug? Well, checked only today. These computer things may sometimes show peculiar behaviours?

Karmic 9.10
redu

Revision history for this message
alistair (alistair-tyeurgain) wrote :

This kernel bug seems to have been fixed in 2.6.32. I noticed that there were a lot of changes in the dvb system in 2.6.32, so I compiled a kernel from the rc8 sources and, hay presto, the nebula card works fine!

A

Revision history for this message
Seb James (sebjames) wrote :

I've just put a 2.6.32rc8 vanilla kernel on my i386 Ubuntu 9.04 system and it does not solve the problem for me. I have the search for channels problem and the same errors in the system as other posters ( mt352_write() to reg 5d failed (err = -6)! etc).

Revision history for this message
Seb James (sebjames) wrote :

I can confirm that the process of removing, then re-inserting the dvb_bt8xx kernel module works for me with i386 Ubuntu 9.04 and a Nebula DigiTV DVB-T (PCI) card. I shall see if I can stick that in my startup scripts as a workaround.

Revision history for this message
redu (reino.urala.gmail) wrote :

alistair wrote:
>
This kernel bug seems to have been fixed in 2.6.32. I noticed that there were a lot of changes in the dvb system in 2.6.32, so I compiled a kernel from the rc8 sources and, hay presto, the nebula card works fine!
A
>
Confirmed: Using Ubuntu Karmic + brand new kernel 2.26.32, Nebula DVB-T works.

But after installing latest Liplianian drivers, to use my DVB-S2 card (Azurewave):
http://mercurial.intuxication.org/hg/s2-liplianin,
Nebula DVB-T again requires that modprobe remove, modprobe trick.

redu

Revision history for this message
Tom Allerton (tom-other) wrote :

I've also tried using Ubuntu Karmic with the 2.26.32 kernel from http://kernel.ubuntu.com/~kernel-ppa/mainline/, but I have the same errors in dmesg when I try to tune. I don't think it has been fixed in this kernel release. The modprobe trick does not work for me either.

I tried with both the 64 and 32 bit versions of Karmic, just in case, but get the same problem.

Revision history for this message
redu (reino.urala.gmail) wrote :

>
Confirmed: Using Ubuntu Karmic + brand new kernel 2.26.32, Nebula DVB-T works.

But after installing latest Liplianian drivers, to use my DVB-S2 card (Azurewave):
http://mercurial.intuxication.org/hg/s2-liplianin,
Nebula DVB-T again requires that modprobe remove, modprobe trick.
redu
>
Edit report:
Now new Liplianin drivers: s2-liplianin-7212833be10d.zip from
http://mercurial.intuxication.org/hg/s2-liplianin,
also includes a working driver for nebula dvb-t.
No modprobing tricks needed.

Karmic, kernel 2.26.32, core2duo 2.5G
redu

Revision history for this message
Mark Ellison (tmarkellison+ubuntu) wrote :

I had the problem complained of in the original bug notice after an update of my Karmic Mythbuntu (AMD 64 3000+, NVidia 5200, Nebula DVB-T with NXT6000, Nebula DVB-T with MT352). I solved my problems using the above suggestions (installing a 2.6.32 kernel, and setting up a script to do the module 'refresh'). Details can be found at <a href="http://stealthlizard.blogspot.com/2010/01/returning-mythbuntu-box-to.html">my blog</a>.

Revision history for this message
Seb James (sebjames) wrote :

I can confirm that the process of removing, then re-inserting the dvb_bt8xx kernel module is required with i386 Ubuntu 9.10 and i386 Ubuntu 10.04 when using the Nebula DigiTV (PCI) card.

I'm still using the workaround of removing, then re-inserting the module (with a script) every time I want to watch TV on the machine, but it's pretty hacky :(

Revision history for this message
redu (reino.urala.gmail) wrote : Re: [Bug 368215] Re: Nebula DigiTV DVB-T (PCI) no longer works

Seb James wrote:
> I can confirm that the process of removing, then re-inserting the
> dvb_bt8xx kernel module is required with i386 Ubuntu 9.10 and i386
> Ubuntu 10.04 when using the Nebula DigiTV (PCI) card.
>
> I'm still using the workaround of removing, then re-inserting the module
> (with a script) every time I want to watch TV on the machine, but it's
> pretty hacky :(
>
>
Confirmed,
but this old trick seems to work for my Ubuntu Lynx 10.04. Also works
for Karmic 9.10, No need to re-insert the module manually every time.
This may not work in all machines?

Add tree lines to /etc/rc.local

sudo gedit /etc/rc.local
modprobe -vr dvb_bt8xx
sleep 10
modprobe -v dvb_bt8xx

redu

Revision history for this message
Paul Barron (paul-barron) wrote :
Download full text (4.8 KiB)

Please help,
I had Mythtv setup with separate backend and frontend, running under Mythbuntu Hardy 8.10 had 2 nebula DigiTV cards with TDED4 tuners in the backend
During play back of the recording the video would freeze, I therefore decided to
install the backend and frontend on what was the frontend. and upgrade to Mythbuntu 10.04, but found when scanning for channels in Mythgbackend setup
the scanner gives "scanning channel xx No Lock for channels 21 to 68 and lists the following information for each channel
channel xx offset 2 -- Timed out, no signal
After channel 68 has been tried the message No channels found appears.
thoulght this may have been resolved in 10.10 so upgraded, deleted the partitions on the system disk an installed 10.10, am getting the same results, I came across this post and on looking at dmesg the module is bt878
verified by
sudo modprobe -vr bt878
and getting message module busy

tried
sudo gedit /etc/rc.local
modprobe -vr bt878
sleep 10
modprobe -v bt878
no change,
then increasing Signal Timeout to 3000ms
Tuning Timeout to 5000ms
and DVB Tuning Delay to 2000
and ran the scan again with the same result.

the only difference I can see is that the module number is different
bttv: driver version 0.9.17 in Hardy
bttv: driver version 0.9.18

Working Mythbuntu Hardy Backend only syslog

Nov 28 18:15:31 hub kernel: [ 196.562569] bttv: driver version 0.9.17 loaded
Nov 28 18:15:31 hub kernel: [ 196.562582] bttv: using 8 buffers with 2080k (520 pages) each for capture
Nov 28 18:15:31 hub kernel: [ 196.562733] bttv: Bt8xx card found (0).
Nov 28 18:15:31 hub kernel: [ 196.562783] ACPI: PCI Interrupt 0000:01:0c.0[A] -> GSI 25 (level, low) -> IRQ 21
Nov 28 18:15:31 hub kernel: [ 196.562801] bttv0: Bt878 (rev 17) at 0000:01:0c.0, irq: 21, latency: 32, mmio: 0xfc003000
Nov 28 18:15:31 hub kernel: [ 196.574645] bttv0: detected: Nebula Electronics DigiTV [card=104], PCI subsystem ID is 0071:0101
Nov 28 18:15:31 hub kernel: [ 196.574655] bttv0: using: Nebula Electronics DigiTV [card=104,autodetected]
Nov 28 18:15:31 hub kernel: [ 196.574700] bttv0: gpio: en=00000000, out=00000000 in=00ff00cf [init]
Nov 28 18:15:31 hub kernel: [ 196.575095] bttv0: tuner absent
Nov 28 18:15:31 hub kernel: [ 196.575142] bttv0: registered device video0
Nov 28 18:15:31 hub kernel: [ 196.575172] bttv0: registered device vbi0
Nov 28 18:15:31 hub kernel: [ 196.575195] bttv0: PLL: 28636363 => 35468950 .. ok
Nov 28 18:15:31 hub kernel: [ 196.620049] bttv0: add subdevice "dvb0"
Nov 28 18:15:31 hub kernel: [ 196.620167] input: bttv IR (card=104) as /devices/pci0000:01/0000:01:0c.0/input/input3

Not working Mythbuntu Maverick syslog
Dec 18 16:56:16 paul-K7VT4APro kernel: [ 19.639131] bttv: driver version 0.9.18 loaded
Dec 18 16:56:16 paul-K7VT4APro kernel: [ 19.639141] bttv: using 8 buffers with 2080k (520 pages) each for capture
Dec 18 16:56:16 paul-K7VT4APro kernel: [ 19.639249] bttv: Bt8xx card found (0).
Dec 18 16:56:16 paul-K7VT4APro kernel: [ 19.639285] alloc irq_desc for 18 on node -1
Dec 18 16:56:16 paul-K7VT4APro kernel: [ 19.639291] alloc kstat_irqs on node -1
Dec 18 16:56:16 paul-K7VT4APro kernel: [ 19.639309] bttv 00...

Read more...

Revision history for this message
CoolDreamZ (chris-thompson-stormfolio) wrote :

I found this problem when upgrading to 10.04.

I tried manually removing/re-inserting the bt878 kernel module whilst running mythtv-backend-setup and I was able to scan channels on the DigiTV card OK. As soon as I re-started mythtv-backend the problem returned. I also tried the rc.local workaround without any luck.

Revision history for this message
CoolDreamZ (chris-thompson-stormfolio) wrote :

I have made a little more progress on this. The mythtv-backend daemon is started using "upstart" in 10.04 and using rc.local means that the "fix" is applied afterhe backend process has already started (and therefore does not work). A partial solution is to add the modprobe workaround to the mythtv-backend upstart job configuration file: /etc/init/mythtv-backend.conf.

This works 100% for me when manually starting the mythtv-backend - using "sudo stop mythtv-backend" followed by "sudo start mythtv-backend".

Unfortunately I can't get this to work automatically on a re-boot of the machine. I suspect either a subtle timing issue during the boot process or a problem with Upstart or the mythtv-upstart job. Unfortunately Upstart seems very opaque to me and I am unable to debug this issue further despite several hours of research.

I am also wondering if the original bug is also caused by a boot sequence/timing/upstart issue?

I have attached the modified Upstart job configuration file in case this works for anyone else.

Revision history for this message
CoolDreamZ (chris-thompson-stormfolio) wrote :
Revision history for this message
Brad Figg (brad-figg) wrote : Unsupported series, setting status to "Won't Fix".

This bug was filed against a series that is no longer supported and so is being marked as Won't Fix. If this issue still exists in a supported series, please file a new bug.

This change has been made by an automated script, maintained by the Ubuntu Kernel Team.

Changed in linux (Ubuntu):
status: Confirmed → Won't Fix
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Duplicates of this bug

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.