Re: S2-3200 switching-timeouts on 2.6.38

Re: S2-3200 switching-timeouts on 2.6.38

[H. Ellenberger]

Hi list,

Follow up to: [1]

@Manu: Your argumentation is inconsistent and lacks any proof.

Running a full scan of Astra 19.2 E with Kaffeine together with cards model 
Skystar HD and Twinhan/Azurewave VP-1041 results in a channel list of approx 
400 stations only. When I apply my patch then almost all stations are found:

patch in tuner: 400 stations found, not usable, 
patch in tuner + demod: 1127 stations found, better but less than without 
tuner patch.
patch in demod only: 1145 stations found, slightly more than with tuner patch.

This proves that the patch in demod stb0899 is still a mandatory must.

Discussions in vdr-portal.de [2] have shown that this patch solved the year 
old problems with these cards for everybody.

My conjecture is that maybe your patch in the code of stb6100 tuner might 
improve the situation for weak signals. With high signal levels as seen from 
Astra 19.2 the modification of the tuner code does not show any positive 
effect, while the cards without my patch in the demodulator code are not 
usable at all!!

So I kindly ask to include this patch.

Thanks and best regards

H .Ellenberger

[1] http://www.spinics.net/lists/linux-media/msg30490.html
[2] http://www.vdr-portal.de/board/thread.php?threadid=99603

Re: S2-3200 switching-timeouts on 2.6.38

[Julian Scheel]

Hi,

Am 23.03.2011 19:19, schrieb H. Ellenberger:
> @Manu: Your argumentation is inconsistent and lacks any proof.
>
> Running a full scan of Astra 19.2 E with Kaffeine together with cards model
> Skystar HD and Twinhan/Azurewave VP-1041 results in a channel list of approx
> 400 stations only. When I apply my patch then almost all stations are found:
>
> patch in tuner: 400 stations found, not usable,

With KNC TV-Station DVB-S2 (which is stb0899 as well) I can't confirm 
this issue. Channel search finds > 1000 channels at any time.

> patch in tuner + demod: 1127 stations found, better but less than without
> tuner patch.
> patch in demod only: 1145 stations found, slightly more than with tuner patch.

Have you done both searches at (almost) same time? On astra network 
there are quite some channels, which are only available and announced at 
certain time, which can make a noticable difference in the amount of 
channels found.

-Julian

Re: S2-3200 switching-timeouts on 2.6.38

[Steffen Barszus]

2011/3/24 Julian Scheel <julian@jusst.de>:
> Hi,
>
> Am 23.03.2011 19:19, schrieb H. Ellenberger:
>>
>> @Manu: Your argumentation is inconsistent and lacks any proof.
>>
>> Running a full scan of Astra 19.2 E with Kaffeine together with cards
>> model
>> Skystar HD and Twinhan/Azurewave VP-1041 results in a channel list of
>> approx
>> 400 stations only. When I apply my patch then almost all stations are
>> found:
>>
>> patch in tuner: 400 stations found, not usable,
>
> With KNC TV-Station DVB-S2 (which is stb0899 as well) I can't confirm this
> issue. Channel search finds > 1000 channels at any time.
>
>> patch in tuner + demod: 1127 stations found, better but less than without
>> tuner patch.
>> patch in demod only: 1145 stations found, slightly more than with tuner
>> patch.
>
> Have you done both searches at (almost) same time? On astra network there
> are quite some channels, which are only available and announced at certain
> time, which can make a noticable difference in the amount of channels found.

It has been done some work (and effort), to prove its having positive
effect. you should allways
have more then 1000 stations at any time on that SAT, if you see that maximum
number of stations is somewhere at 1600. everything around 400 is
broken hardware or driver.

Please dont consider his post as single experience of a single user. The general
perception was more like "Wow i can finally use that card and don't
throw it away."
 from 10+ Users which i know of. So no matter if its considered done
right or not.
The tuning patch is out of question of improving the usability of the
affected cards.

The request has been, to prove that the changes at the stb6100 is not
enough to actually
fix the problem. This prove has been done now . The changes on the
stb6100 is not enough,
the changes at the stb0899 still has quite positive effect.

So taking into account that the patch improves situation and doesn't
have negative side
effects for others, it might be wise to consider inclusion of the
patch, even if it might not
be the perfect solution from technical/theoretical perspective.

I dont want to complain until the patch is in, its just: We have a
problem, we have a solution.
Lets use it and make this hardware useful for the people and lets do
it right when time permits.

There is still a handful of other patches for the same hardware
floating around to improve other parts of that driver.

I mean this hardware is known to work "less then perfect" since years, really.

So please do something about it. Accept the helping hands.

Thanks

Re: S2-3200 switching-timeouts on 2.6.38

[Julian Scheel]

Am 24.03.2011 19:32, schrieb Steffen Barszus:
> It has been done some work (and effort), to prove its having positive
> effect. you should allways
> have more then 1000 stations at any time on that SAT, if you see that maximum
> number of stations is somewhere at 1600. everything around 400 is
> broken hardware or driver.

Agreed, but this is why I asked about the difference between the channel 
search with stb6100 patch only and both patches applied.
Nobody says that 400 channels are expected behaviour.
But the difference of less than 20 channels between both patches could 
easily be caused by temporal availability of services. This is all I 
said and all I suggested to proof.

Even a simple diff of both searches would be helpful. Just to check 
which channels are missing.

I am not in the position to accept or reject this driver, just wanted to 
be helpful in finding out if it's really needed.

> Please dont consider his post as single experience of a single user. The general
> perception was more like "Wow i can finally use that card and don't
> throw it away."
>   from 10+ Users which i know of. So no matter if its considered done
> right or not.
> The tuning patch is out of question of improving the usability of the
> affected cards.
>
> The request has been, to prove that the changes at the stb6100 is not
> enough to actually
> fix the problem. This prove has been done now . The changes on the
> stb6100 is not enough,
> the changes at the stb0899 still has quite positive effect.
>
> So taking into account that the patch improves situation and doesn't
> have negative side
> effects for others, it might be wise to consider inclusion of the
> patch, even if it might not
> be the perfect solution from technical/theoretical perspective.
>
> I dont want to complain until the patch is in, its just: We have a
> problem, we have a solution.
> Lets use it and make this hardware useful for the people and lets do
> it right when time permits.
>
> There is still a handful of other patches for the same hardware
> floating around to improve other parts of that driver.
>
> I mean this hardware is known to work "less then perfect" since years, really.
>
> So please do something about it. Accept the helping hands.
>
> Thanks
> --
> To unsubscribe from this list: send the line "unsubscribe linux-media" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: S2-3200 switching-timeouts on 2.6.38

[Paul Franke]

Hi list,

Follow up to: [1]

I did testing with a TechniSat SkyStar HD 2 (1ae4:0003).

- without a patch: card unuseable, gets locks only rarely
- with Manu's patch: card useable, gets locks, but sometimes this still 
needs some seconds
- with the mentioned stb0899 patch: card useable, gets locks, most time 
or nearly always fast (and faster as with Manu's patch)

In practice the stb0899 patch does a better job for zap frenzies, so I 
kindly ask to include this patch.

Thanks and best regards

Paul

[1] http://www.spinics.net/lists/linux-media/msg30537.html

Re: S2-3200 switching-timeouts on 2.6.38

[Manu Abraham]

On Fri, Mar 25, 2011 at 8:31 PM, Paul Franke <paul.franke@uni-rostock.de> wrote:
> Hi list,
>
> Follow up to: [1]
>
> I did testing with a TechniSat SkyStar HD 2 (1ae4:0003).
>
> - without a patch: card unuseable, gets locks only rarely
> - with Manu's patch: card useable, gets locks, but sometimes this still
> needs some seconds
> - with the mentioned stb0899 patch: card useable, gets locks, most time or
> nearly always fast (and faster as with Manu's patch)
>
> In practice the stb0899 patch does a better job for zap frenzies, so I
> kindly ask to include this patch.
>

Can you guys, (who seem to really need the said patch) please provide
me the STB0899 chipset version on your card ?

For eg: what I have here 3 different silicon versions
C6L STB0899 VQ940KDA12 22 9BS VQ MLT 22 012
C2L STB0899 VQ636DSK    22 09C  VQ MLT 22 656
C1L STB0899 VQ608FSD    22 OPD VQ MLT 22 628

Best Regards,
Manu

Re: S2-3200 switching-timeouts on 2.6.38

[Paul Franke]

 > Can you guys, (who seem to really need the said patch) please provide
 > me the STB0899 chipset version on your card ?

Sure. For a TechniSat SkyStar HD 2 (1ae4:0003)...

 > C6L STB0899 VQ940KDA12 22 9BS VQ MLT 22 012
 > C2L STB0899 VQ636DSK    22 09C  VQ MLT 22 656
 > C1L STB0899 VQ608FSD    22 OPD VQ MLT 22 628

STB0899 VQ719RZM 22 OPO VQ MLT 22 815     001

Thanks and best regards

Paul

S2-3200 switching-timeouts on 2.6.38

[Rico Tzschichholz]

[-- Attachment #1: Type: text/plain, Size: 334 bytes --]

Hello,

I would like to know if there is any intention to include this patch
soon? https://patchwork.kernel.org/patch/244201/

Currently using 2.6.38 results in switching-timeouts on my S2-3200 and
this patch fixes this for good.

So it would be nice to have it in 2.6.39.

Thank you and best regards,
Rico Tzschichholz


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 900 bytes --]

Re: S2-3200 switching-timeouts on 2.6.38

[Randy Dunlap]

On Mon, 21 Mar 2011 20:46:23 +0100 Rico Tzschichholz wrote:

> Hello,
> 
> I would like to know if there is any intention to include this patch
> soon? https://patchwork.kernel.org/patch/244201/

There are MANY posted but unmerged patches in patchwork from the linux-media
mailing list.  What is going on (or not going on) with patch merging?


> Currently using 2.6.38 results in switching-timeouts on my S2-3200 and
> this patch fixes this for good.
> 
> So it would be nice to have it in 2.6.39.


thanks,
---
~Randy
*** Remember to use Documentation/SubmitChecklist when testing your code ***

Re: S2-3200 switching-timeouts on 2.6.38

[Manu Abraham]

On Tue, Mar 22, 2011 at 1:46 AM, Randy Dunlap <rdunlap@xenotime.net> wrote:
> On Mon, 21 Mar 2011 20:46:23 +0100 Rico Tzschichholz wrote:
>
>> Hello,
>>
>> I would like to know if there is any intention to include this patch
>> soon? https://patchwork.kernel.org/patch/244201/
>
> There are MANY posted but unmerged patches in patchwork from the linux-media
> mailing list.  What is going on (or not going on) with patch merging?

Actually, quite a lot of effort was put in to get that part right. It
does the reverse thing that's to be done.
The revamped version is here [1] If the issue persists still, then it
needs to be investigated further.

[1] http://www.mail-archive.com/linuxtv-commits@linuxtv.org/msg09214.html


Best Regards,
Manu

Re: S2-3200 switching-timeouts on 2.6.38

[Bjørn Mork]

Manu Abraham <abraham.manu@gmail.com> writes:
> On Tue, Mar 22, 2011 at 1:46 AM, Randy Dunlap <rdunlap@xenotime.net> wrote:
>> On Mon, 21 Mar 2011 20:46:23 +0100 Rico Tzschichholz wrote:
>>
>>> Hello,
>>>
>>> I would like to know if there is any intention to include this patch
>>> soon? https://patchwork.kernel.org/patch/244201/
>>
>> There are MANY posted but unmerged patches in patchwork from the linux-media
>> mailing list.  What is going on (or not going on) with patch merging?
>
> Actually, quite a lot of effort was put in to get that part right. It
> does the reverse thing that's to be done.
> The revamped version is here [1] If the issue persists still, then it
> needs to be investigated further.
>
> [1] http://www.mail-archive.com/linuxtv-commits@linuxtv.org/msg09214.html

So the patch state should be "Rejected" and not "Under Review".

Would certainly help us all if the patchwork state was updated whenever
a patch actually was processed...


Bjørn

Re: S2-3200 switching-timeouts on 2.6.38

[Mauro Carvalho Chehab]

Em 22-03-2011 06:53, Bjørn Mork escreveu:
> Manu Abraham <abraham.manu@gmail.com> writes:
>> On Tue, Mar 22, 2011 at 1:46 AM, Randy Dunlap <rdunlap@xenotime.net> wrote:
>>> On Mon, 21 Mar 2011 20:46:23 +0100 Rico Tzschichholz wrote:
>>>
>>>> Hello,
>>>>
>>>> I would like to know if there is any intention to include this patch
>>>> soon? https://patchwork.kernel.org/patch/244201/
>>>
>>> There are MANY posted but unmerged patches in patchwork from the linux-media
>>> mailing list.  What is going on (or not going on) with patch merging?
>>
>> Actually, quite a lot of effort was put in to get that part right. It
>> does the reverse thing that's to be done.
>> The revamped version is here [1] If the issue persists still, then it
>> needs to be investigated further.
>>
>> [1] http://www.mail-archive.com/linuxtv-commits@linuxtv.org/msg09214.html
> 
> So the patch state should be "Rejected" and not "Under Review".
> 
> Would certainly help us all if the patchwork state was updated whenever
> a patch actually was processed...

Yes, but the thing is that somebody needs to manually update me about status change,
as patchwork doesn't provide any way for it. Patchwork's permission for a project
is all or nothing: or you completely own a project, or you have just read-only
access. There's no way, currently, for me to delegate a patch to someone without
allowing that person to touch at the non-delegated patches. Also, patchwork
also doesn't allow me to change contributor status (there's just one global admin
that creates and/or modifies account access on patchwork).

Anyway, I've updated the status for #244201.

Thanks,
Mauro.

Re: S2-3200 switching-timeouts on 2.6.38

[Manu Abraham]

On Tue, Mar 22, 2011 at 3:23 PM, Bjørn Mork <bjorn@mork.no> wrote:
> Manu Abraham <abraham.manu@gmail.com> writes:
>> On Tue, Mar 22, 2011 at 1:46 AM, Randy Dunlap <rdunlap@xenotime.net> wrote:
>>> On Mon, 21 Mar 2011 20:46:23 +0100 Rico Tzschichholz wrote:
>>>
>>>> Hello,
>>>>
>>>> I would like to know if there is any intention to include this patch
>>>> soon? https://patchwork.kernel.org/patch/244201/
>>>
>>> There are MANY posted but unmerged patches in patchwork from the linux-media
>>> mailing list.  What is going on (or not going on) with patch merging?
>>
>> Actually, quite a lot of effort was put in to get that part right. It
>> does the reverse thing that's to be done.
>> The revamped version is here [1] If the issue persists still, then it
>> needs to be investigated further.
>>
>> [1] http://www.mail-archive.com/linuxtv-commits@linuxtv.org/msg09214.html
>
> So the patch state should be "Rejected" and not "Under Review".
>
> Would certainly help us all if the patchwork state was updated whenever
> a patch actually was processed...

Sure, I could've pointed it out to Mauro, but handling multiple
things, that slipped off my mind.

Best Regards,
Manu

Re: S2-3200 switching-timeouts on 2.6.38

[Rico Tzschichholz]

[-- Attachment #1: Type: text/plain, Size: 721 bytes --]

Hello Manu,

> Actually, quite a lot of effort was put in to get that part right. It
> does the reverse thing that's to be done.
> The revamped version is here [1] If the issue persists still, then it
> needs to be investigated further.
>
> [1] http://www.mail-archive.com/linuxtv-commits@linuxtv.org/msg09214.html

I am not sure how this is related to stb6100?

Does that mean the current stb0899 patch [2] isnt ready to be proposed
for 2.6.39 yet? Or does the stb6100 patch has a better design to solve
this issue which should be adapted for stb0899 then?
I was hoping to see it included before the merge window is closed again.

[2] https://patchwork.kernel.org/patch/244201/

Best regards,
Rico


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 900 bytes --]

Re: S2-3200 switching-timeouts on 2.6.38

[Bjørn Mork]

Rico Tzschichholz <ricotz@t-online.de> writes:

>> Actually, quite a lot of effort was put in to get that part right. It
>> does the reverse thing that's to be done.
>> The revamped version is here [1] If the issue persists still, then it
>> needs to be investigated further.
>>
>> [1] http://www.mail-archive.com/linuxtv-commits@linuxtv.org/msg09214.html
>
> I am not sure how this is related to stb6100?
>
> Does that mean the current stb0899 patch [2] isnt ready to be proposed
> for 2.6.39 yet? Or does the stb6100 patch has a better design to solve
> this issue which should be adapted for stb0899 then?

I believe the point was that the real issue was a noise problem in the
stb6100 tuner driver, and that the stb0899 frontend patch just papered
over this.

> I was hoping to see it included before the merge window is closed again.
>
> [2] https://patchwork.kernel.org/patch/244201/

Please test the driver with the patch Manu refers to and report if there
still are issues.  The patch is now upstream in the 2.6.38 kernel, so
testing it should be fairly easy.  

Do you still have this issue with a plain 2.6.38 driver?


Bjørn

Re: S2-3200 switching-timeouts on 2.6.38

[Manu Abraham]

Hello Rico,

On Tue, Mar 22, 2011 at 5:03 PM, Rico Tzschichholz <ricotz@t-online.de> wrote:
> Hello Manu,
>
>> Actually, quite a lot of effort was put in to get that part right. It
>> does the reverse thing that's to be done.
>> The revamped version is here [1] If the issue persists still, then it
>> needs to be investigated further.
>>
>> [1] http://www.mail-archive.com/linuxtv-commits@linuxtv.org/msg09214.html
>
> I am not sure how this is related to stb6100?
>
> Does that mean the current stb0899 patch [2] isnt ready to be proposed
> for 2.6.39 yet? Or does the stb6100 patch has a better design to solve
> this issue which should be adapted for stb0899 then?
> I was hoping to see it included before the merge window is closed again.
>
> [2] https://patchwork.kernel.org/patch/244201/


stb0899 is a channel decoder (or demodulator in other words) while the
stb6100 is a
tuner which provides I-Q components to the demod.

When a tuner locks to a transponder, in the spectrum in the absence of
a signal, it will be contain White (Gaussian) noise. In such a case
the demod has a hard time to lock to the signal. In this particular
case, we had a bit of luck additionally, ie we  found a case where the
stb0899 demodulator functioned perfectly well with another tuner, but
with the same hardware configuration. This helped in narrowing the bug
to the tuner and hence the fix.

The one in patchwork, does modify the step size but that doesn't
reduce the white noise, which is something like a lucky dip. (similar
to what Bjorn pointed out in another post.) I am not really sure
whether modifying the step size of any benefit/disadvantage, but need
to do some research on that aspect.

Best Regards,
Manu
]

Re: S2-3200 switching-timeouts on 2.6.38

[Rico Tzschichholz]

[-- Attachment #1: Type: text/plain, Size: 146 bytes --]

Hello Manu,

you are absolutely right, thanks for pointing this out.
Looking forward to the stb0899-patch inclusion.

Best Regards,
Rico


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 900 bytes --]

Re: S2-3200 switching-timeouts on 2.6.38

[P. van Gaans]

On 03/21/2011 08:46 PM, Rico Tzschichholz wrote:
> Hello,
>
> I would like to know if there is any intention to include this patch
> soon? https://patchwork.kernel.org/patch/244201/
>
> Currently using 2.6.38 results in switching-timeouts on my S2-3200 and
> this patch fixes this for good.
>
> So it would be nice to have it in 2.6.39.
>
> Thank you and best regards,
> Rico Tzschichholz
>

Hello,

I also have a Technotrend S2-3200. Bought it 4 years ago. Support for 
Linux seemed to be close at the time. I was a bit of a fool for thinking 
that. It's been on the shelf for most of the time.

Anyway, I've applied that patch, and it turns the S2-3200 from being 
worth hardly more than a paperweight to a functional DVB-S card. I don't 
understand why this patch does not get included.

Only DVB-S2 transponders don't consistently lock properly in kaffeine. 
I'll test what "scan" does tomorrow, see if it's a kaffeine-specific 
problem or not. But even just DVB-S working properly is a massive 
difference for this card.

If it helps, I'll check the silicon version of my card tomorrow as well.

Best regards,

P. van Gaans

Re: S2-3200 switching-timeouts on 2.6.38

[P. van Gaans]

On 08/11/2011 01:49 AM, P. van Gaans wrote:
> On 03/21/2011 08:46 PM, Rico Tzschichholz wrote:
>> Hello,
>>
>> I would like to know if there is any intention to include this patch
>> soon? https://patchwork.kernel.org/patch/244201/
>>
>> Currently using 2.6.38 results in switching-timeouts on my S2-3200 and
>> this patch fixes this for good.
>>
>> So it would be nice to have it in 2.6.39.
>>
>> Thank you and best regards,
>> Rico Tzschichholz
>>
>
> Hello,
>
> I also have a Technotrend S2-3200. Bought it 4 years ago. Support for
> Linux seemed to be close at the time. I was a bit of a fool for thinking
> that. It's been on the shelf for most of the time.
>
> Anyway, I've applied that patch, and it turns the S2-3200 from being
> worth hardly more than a paperweight to a functional DVB-S card. I don't
> understand why this patch does not get included.
>
> Only DVB-S2 transponders don't consistently lock properly in kaffeine.
> I'll test what "scan" does tomorrow, see if it's a kaffeine-specific
> problem or not. But even just DVB-S working properly is a massive
> difference for this card.
>
> If it helps, I'll check the silicon version of my card tomorrow as well.
>
> Best regards,
>
> P. van Gaans
> --
> To unsubscribe from this list: send the line "unsubscribe linux-media" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>

Hello,

Here I am again. My card can be seen at http://tinypic.com/r/2hewzyu/7

Chip:

C2L STB0899 VQ628NDY 220QQ VQ MLT 22 645

So far, scan-s2 won't compile: "scan.c:51:2: error: #error scan-s2 
requires Linux DVB driver API version 5.0!" (I am doing something wrong 
undoubtedly) and I cannot find a way to make w_scan scan specific 
frequencies. It must be possible, but right now I can't figure it out. 
Scan doesn't support DVB-S2.

Perhaps someone else who has an S2-3200 can test this, if not, I will 
eventually get to it I guess, but I'm not entirely sure when.

Best regards,

Pim