Seca card stops responding

[RempageR1]

Revision

6642

Issue Description

Log messages appear:
2012/04/28 21:57:23 599E80 r ERROR in IO_Serial_Read (errno=0 Success)
2012/04/28 21:57:47 599E80 r ERROR in IO_Serial_Read (errno=0 Success)
2012/04/28 21:57:50 599E80 r ERROR in IO_Serial_Read (errno=0 Success)
2012/04/28 21:58:02 599E80 r ERROR in IO_Serial_Read (errno=0 Success)
2012/04/28 21:58:05 599E80 r ERROR in IO_Serial_Read (errno=0 Success)
2012/04/28 21:58:13 599E80 r ERROR in IO_Serial_Read (errno=0 Success)
2012/04/28 21:58:16 599E80 r ERROR in IO_Serial_Read (errno=0 Success)

Card is no longer used

When the issue occurs

After a few hours, mostly under heavy use

How the issue is reproducable

Happens after few hours. More info can be provided when asked.

<Don't forget to ATTACH (NOT post here as text) a log file of oscam in debug mode (start oscam with -d255) and to remove the brackets around the fields! Otherwise, your ticket will be closed as you seem to be unable to read.>

[Deas]

the problem happens earlier - post the last lines before this error.

[RempageR1]

Replying to Deas:

the problem happens earlier - post the last lines before this error.

Got this so far:
oscam.log:2012/04/28 22:47:52 5FEEB8 r ERROR in IO_Serial_Read (errno=0 Success)
oscam.log:2012/04/28 22:47:52 5FEEB8 r ERROR, function call IO_Serial_Read (reader, timeout + IFD_TOWITOKO_TIMEOUT, size, buffer) returns error.
oscam.log:2012/04/28 22:47:52 5FEEB8 r ERROR, function call Phoenix_Receive (reader, data, size, reader->read_timeout) returns error.
oscam.log:2012/04/28 22:47:52 5FEEB8 r ERROR, function call ICC_Async_Receive (reader, 1, buffer + recv) returns error.

Log turned over before I could get more. Have it running debug 255 and 1M logfile now. Update you when I have more.

[Deas]

and what hardware do you use? you just throw some lines from the log and some one should fix it...

[jedpod]

I noticed the same messages with N3 card and internal reader.
I put my comments here: ​http://oscam.to/wbb3/index.php?page=Thread&postID=113283#post113283
(see bottom)

The point is I don't use Phoenix neither another serial reader but apparently it was involved somehow and it drives to permanent decoding failure.
Small change in source code sort the things-out. Details presented into mentioned thread.
It would be interesting to confirm if this change can help with SECA card too.

[RempageR1]

Replying to Deas:

and what hardware do you use? you just throw some lines from the log and some one should fix it...

Like I said I will add more info
These are the lines:

---

2012/04/29 09:21:23 5FEEB8 r IO: Receiving:
2012/04/29 09:21:23 5FEEB8 r ERROR in IO_Serial_Read (errno=0 Success)
2012/04/29 09:21:23 5FEEB8 r ERROR, function call IO_Serial_Read (reader, timeout + IFD_TOWITOKO_TIMEOUT, size, buffer) returns error.
2012/04/29 09:21:23 5FEEB8 r ERROR, function call Phoenix_Receive (reader, data, size, reader->read_timeout) returns error.
2012/04/29 09:21:23 5FEEB8 r ERROR, function call ICC_Async_Receive (reader, 1, buffer + recv) returns error.
2012/04/29 09:21:23 5FEEB8 r IFD Transmit:
2012/04/29 09:21:23 5FEEB8 C1 3C 01 5C 5C
2012/04/29 09:21:23 5FEEB8 r IO: Sending:
2012/04/29 09:21:23 5FEEB8 C1 3C 01 5C 5C
2012/04/29 09:21:23 5FEEB8 r IFD Transmit succesful

---

Running Vuplus Duo with VTI image 4.1 using internal reader. If you need anything let me know. This problem isn't in 6611 (but the latest version does use the local card more, that`s why I would like to run it)

[RempageR1]

Complete logfile

[Deas]

follow the link posted by jedpod and try his code

[RempageR1]

Replying to Deas:

follow the link posted by jedpod and try his code

I then need to get myself a compiler for my Vuplus box. Do not know how do to that so need to look it up. Can be a while.

[theparasol]

Tried jedpod code on dm800hd pvr with mediaguard3 / seca card again and again not working with seca (0100:00006a), please post your code as patch / diff

[RempageR1]

Tried compiling using ​http://www.streamboard.tv:8001/wiki/crosscompiling but cannot install cmake as it wants to remove important packages. Sorry, that`s where my linux skills end....

[jedpod]

Replying to RempageR1:

Tried compiling using ​http://www.streamboard.tv:8001/wiki/crosscompiling but cannot install cmake as it wants to remove important packages. Sorry, that`s where my linux skills end....

If you want to test my compilation see here: http://www.streamboard.tv/wbb2/thread.php?postid=410709#post410709

Config I use for compilation I attached here: http://www.streamboard.tv/wbb2/thread.php?threadid=33934

I have no idea why it could operate correctly with my Vu Duo internal readers? One smart guy comments this solution as T0 protocol break-out. Maybe it is not a general solution but I'm pretty happy with this. No more oscam restarts required.
I'm too stupid to make deep analyse. But last compilation comes from yesterday and till today no problem was detected. Previous compilation I made week ago (6620) and at least 24 hours no problem was detected. After 24 hours cron did restart of oscam.

[RempageR1]

Replying to jedpod:

Replying to RempageR1:

Tried compiling using ​http://www.streamboard.tv:8001/wiki/crosscompiling but cannot install cmake as it wants to remove important packages. Sorry, that`s where my linux skills end....

If you want to test my compilation see here: http://www.streamboard.tv/wbb2/thread.php?postid=410709#post410709

Config I use for compilation I attached here: http://www.streamboard.tv/wbb2/thread.php?threadid=33934

I have no idea why it could operate correctly with my Vu Duo internal readers? One smart guy comments this solution as T0 protocol break-out. Maybe it is not a general solution but I'm pretty happy with this. No more oscam restarts required.
I'm too stupid to make deep analyse. But last compilation comes from yesterday and till today no problem was detected. Previous compilation I made week ago (6620) and at least 24 hours no problem was detected. After 24 hours cron did restart of oscam.

Wel you are smarter then I am when it comes to these things. I`m running your version now and will let you know how it works out!

[RempageR1]

Hi jedpod,

I used the build for about an hour without any issues. However, I need the loadbalancer feature to be available so I cannot let it running. Could you supply a build with loadbalancer enabled?

Thank you.

Edit: Tested with latest svn, there the problem starts after only five minutes:

---

2012/05/01 19:55:32 59AE68 r ERROR in IO_Serial_Read (errno=0 Success)
2012/05/01 19:55:32 59AE68 r ERROR, function call IO_Serial_Read (reader, timeout + IFD_TOWITOKO_TIMEOUT, size, buffer) returns error.
2012/05/01 19:55:32 59AE68 r ERROR, function call Phoenix_Receive (reader, data, size, reader->read_timeout) returns error.
2012/05/01 19:55:32 59AE68 r ERROR, function call ICC_Async_Receive (reader, 1, buffer + recv) returns error.
2012/05/01 19:55:32 59AE68 r ERROR in IO_Serial_Write u=0 to_send=92 (errno=0 Success)
2012/05/01 19:55:32 59AE68 r ERROR, function call IO_Serial_Write (reader, char_delay, to_send, buffer+sent) returns error.
2012/05/01 19:55:32 59AE68 r ERROR, function call Phoenix_Transmit (reader, sent, size, 0, 0) returns error.
2012/05/01 19:55:32 59AE68 r ERROR, function call ICC_Async_Transmit(reader, MAX (Lc - sent, 0), data + sent) returns error.
2012/05/01 19:55:32 59AE68 r ERROR, function call Protocol_T0_Command returns error.

---

I am hopefull your fix works....

[RempageR1]

Hi Jedpod,

Finally got it compiled myself and testing ATM. Tested with your config change but still same problem. Now edited this (line 524 @ icc_async.c). I think you wanted that in but did not put it in the forum:

---

#elif defined(SCI_DEV)

call (Sci_Transmit (reader, sent, size)); the internal reader will provide the delay

---

Edit: Resulted in freezing pictures. Do not know what is wrong but cannot fix this myself.

[jedpod]

Sorry for late answer - I lost this ticket number.

Good to hear that my compilation can work more stable than official release - it is in-line with my observation.

I've got few more days of testing - freezing wasn't observed neither once. But need to mention that after 1-2 days times increase to 920 ms. Due to new svn releases I wasn't tested any version more than 2 days.
Will try to compile my simplified version with loadbalancer set to on this afternoon and will attach it here.

Update (2012-05-04):
New compilation done - see here: http://www.streamboard.tv/wbb2/thread.php?postid=410709#post410709

Update (2012-05-06)
4 days and still no problem, I can watch all channels, no OSCAM's restart is required, although time is 920 ms.

[RempageR1]

I went back to the svnp build as there was no benefit from using the self compiled ones.

As a workaround I now run this script every minute:

---

test=tail -n 500 /tmp/oscam.log | grep 'ERROR in IO_Serial_Read' -c
if [ $test -gt 8 ]
then

error=tail -n 500 /tmp/oscam.log | grep 'ERROR in IO_Serial_Read'
wget ​http://1.1.1.1:80/readers.html?label=cardreader\&action=disable -O /dev/null
sleep 10
wget ​http://1.1.1.1:80/readers.html?label=cardreader\&action=enable -O /dev/null
sleep 10
wget ​http://1.1.1.1:80/readerstats.html?label=cardreader\&action=resetstat\&rc=5 -O /dev/null
sleep 10
wget ​http://1.1.1.1:80/readerstats.html?label=cardreader\&action=resetstat\&rc=4 -O /dev/null
timeofReset=date
echo "Reader reset at" $timeofReset "with count of" $test >> /var/log/oscamRestart
echo $error >> /var/log/oscamRestart
echo "clear" > /tmp/oscam.log

fi

---

So the card gets reset when errors are seen and the not-found and time-out entries are removed. Also, the retry time of loadbalancer is set to 60 seconds so if a station failes to decode, it will be tried again the next minute.

Further I have added all allowed stations to the oscam.services file so no pointless reads are made for stations I do not have.

[Deas]

according to #2549 fixed in r6833

[RempageR1]

Confirmed working in r6833, thank you!