Continuity Counter Errors

[josmi]

Yes, it's the first 8 bytes that appears to be left from the last packet that used the same buffer in memory. I suppose it could be driver related but since it's 8 bytes in the "middle" of a 64KB buffer my first guess would be HW. I'd be happy to test in Windows too if I could get some advice on how to do it.

[josmi]

I now have Windows 7 running from a USB drive but I could still use some advice on suitable test SW.

[jeroen]

Any update on this issue?
Did you give up or found a workaround/solution? Did you try windows and different hardware?

[zinonino]

Any soution for this errors, I have TBS 6981 and have the same erros ?

[kvaster]

I have same problems - TBS6985. There are no problems with SD channels at all, but i'm having errors with HD channels when more then one adapter is in use. There are no problems in case only one adapter is used. Any chances to fix that ? (i'm using latest drivers with msi interrupts on).

I have ASUS Z87 Deluxe motherboard, running gentoo with latest driver on it.

[CrazyFin]

This so interesting to read!

I was searching on the web for solutions on "continuity counter errors" . I use TVHeadend 3.9 on Ubuntu 12.04 LTS with Digital Devices Cine S2 dual tuner card with DuoFlex add on cards to get quad tuners per tuner card.

I am experiencing EXACTLY the same issues! I also get CC-errors and artifacts/pixelation if I have one of the tuners tuned into an HD channel and then tune into a 2nd HD channel on the same card. It is ALWAYS the 2nd and 4th tuners on each card that creates these problem... So if I have 2 cards (Cine S2) with 4 tuners per card and then disable every 2nd tuner it all works fine but this means that I am using only 50% tuner capacity... As soon as I enable every 2nd tuner on each subcard the CC errors starts to appear as soon as those tuners are being used by HD channels. Problem does not appear when tuning in to SD channels.

I see in the forum thread viewtopic.php?f=142&t=8868 that another person using the TBS6985 has IDENTICAL issues as me using a TOTALLY different brand of DVB S2 cards BUT we all use V4L-linux-drivers that seems to be crappy.

If I remove my 2nd card AND remove the DuoFlex addon card from the remaining Cine S2 card so that I will only have 2 tuners left, I STILL get the same problem, i.e. tune into HD channel on tuner 1 works fine but as soon as I tune into a HD channel on tuner 2 I start to get artifacts/pixelation and CC errors.

Looks like the Linux drivers do not protect key/shared PCI resources properly causing these CC errors?
Seems to be Linux-drivers only since if I test the same card with windows based solutions I do not get these problems at all.

[slaine]

I have the same problem with cc errors (about 50 errors/hour). But if I run stress tool on Ubuntu (enough for 1 core of 4: “stress – c 1”) the cc errors disappeared (0 errors for 5 hours). I’m not Linux/HW expert so I don’t know what does it means. Can someone help me to interpret/solve this problem?
My system:
MB: Supermicro X10SBA (J1900 chipset), Bios v.: R 1.0c
TV card: TBS6985
OS: Ubuntu server 14.0.4.1 (3.13.0-35-generic), TBS Driver – 140819, MSI enabled,
TVheadend: 3.4.28

Thanx and sorry for my English.

[Mythological]

For those having these problems with TVHeadEnd but only on certain channels or only at random times, try this. Under Configuration | Stream click on the pass steam profile, then in the sidebar you should see an option for "Rewrite PMT" which is checked by default (at least in the version I have). Uncheck it and then save it.

This setting was in a different place in earlier versions (in the default Recording profile) but it was unchecked by default there. After installing the new version I started experiencing all kinds of errors, but only on certain channels, until I found the new location of that setting and unchecked it. I am not saying this will fix it in all cases but it is certainly worth a try!

[LSD]

For me disabling C6/C7 support in the BIOS on an Intel Celeron J1900 MB did the trick of getting rid of discontinuity errors. The other option was to "stress" the system with e.g. stress or burnP6 on all cores so that the CPU did not enter C6/C7 states. But simply disabling it in BIOS saves resources and power (allthough you loose a few watts) and now the system idles on C1 most of the time.

LSD

[tzimmo]

I'm experiencing the same issue. I have two systems and using DVB-T/DVB-T2:

1) Asus M4A88T-M with TBS6284 quad tuner. Current setup: adapter0=idle, adapter1=hd, adapter2=sd, adapter3=hd. There are no problems with it, no continuity counter errors.

2) Asus M4A89GTD-PRO with 2x TBS284 quad tuner and because of the current issues also Hauppauge WinTV Nova-T(D) 500 PCI dual tuner. Current setup: adapter0=hd, adapter1=idle, adapter2=idle, adapter3=idle, adapter4=hd, adpater5=sd, adapter6=sd, adapter7=sd, adapter8(nova)=sd, adapter9(nova)=sd.

On the second system, you maybe noticed that I am using only one out of 4 tuners on the card. Previously that channel was DVB-T (SD) and there were no issues whatsoever. But about a month ago that multiplex was changed to DVB-T2 (HD) and the problems started. I'm only able to use one tuner out of 4 to get uncorrupted TS stream. If I use any of the other 3 tuners, I'm starting to get continuity errors and they look exactly like described here. There are corrupted packets that are out of sequence, probably partial copies of some previous packets as when I compare the recordings between system 1 and system 2, system 2 produces slightly larger output files per program, like 188, 376 or 564 bytes larger per hour. Removing those extraneous duplicate/corrupted packets seem to generate at least sometimes a fully working, identical stream to what system 1 gets (where there is no issue for some reason). I haven't yet tried switching to MSI interrupts or trying to see if there are any power saving settings in BIOS that I could try to disable. As I have two adapters in system 2, I have tried to move that multiplex (UHF, 618 MHz) to the other card, i.e., adapter numbers 4-7, but exactly the same issue happens there: that frequency has to be the only one tuned there or the stream gets corrupted.

For some reason the other DVB-T2 multiplex (VHF, 184.5 MHz) does not suffer from this issue but it can be stored while the other 3 tuners are also storing full stream, as long as I don't tune that problematic 618 MHz frequency to the other tuners on the same adapter. However, even if I have both HD streams on the same adapter, it only corrupts this 618 MHz frequency, 184.5 MHz has no continuity issues.

Any ideas how I could further debug the issue to maybe help resolving an issue in the driver (or in the HW)? I guess the manufacturer has been running stress tests on their adapters, receiving 4x high bandwidth HD channels at the same time and checking the continuity counters for each PID?