cx8800 and V4L driver issue

[bernhart2002@…]

The cx8800 driver does not accept ioctl command VIDIOC_STREAMOFF and this causes the card to fail with eno: Invalid argument (22) . I guess the same problem exist with the bttv driver in file mythtv/mythtv/libs/libmythtv/v4lchannel.cpp

line 102

if (driver_name == "bttv")

should be

if (driver_name == "bttv"

driver_name == "cx8800")

sorry I do not know how to make a patch

[bernhart2002@…]

sorry the wiki formating changed the or

 if (driver_name == "bttv" || driver_name == "cx8800")

[danielk]

Fixes #10519. Adds workaround for cx8800 driver. This card reports V4L2_CAP_STREAMING but doesn't support streaming. The bug has been reported on the #linuxtv IRC channel.

[danielk]

Was fixed by [8bf3888ab0ac5d45af3065d5c9a04c0fbf7aab3e].

[danielk]

Any chance you can test 10519-v1.patch ?

in the mythtv directory just do a

   cat 10519-v1.patch | patch -p2
   make && sudo make install

I got some clarification from mchehab in #linuxtv. Streaming is supported with the cx8800 driver, but it only works if you close the data streaming file descriptor first. This patch attempts to do that but I don't have any way to test it.

[danielk]

Updated for latest master

[bernhart2002@…]

mjpeg backend log

[bernhart2002@…]

mpeg 2 card setup

[danielk]

Thanks Bernhart, can you rerun the mpeg2 card test with "-v record,channel"

The patch doesn't address the same problem with NuppelVideoRecorder?, it only does the close/reopen in the MPEGRecorder so there is no need to test that right now.

[Simon Kenyon <simon@…>]

10519-v2.patch fixes #9830 for me!

Happy days are here again

The skies above are clear again

So let's sing a song of cheer again

Happy days are here again

[bernhart2002@…]

MPG2 card test with -v record,channel

[bernhart2002@…]

Hey I ran the test with the options you requested and have attached the log file (mpg2-vrc.txt). Please let me know if you would like to try something else

Thanks for all the help

Eric

[danielk]

Eric, from the log it looks like cx8800 doesn't advertise that it supports V4L2_CAP_STREAMING so the original patch would be unnecessary.

From what I can tell this card may not actually support MPEG-2 encoding in hardware. If you can attach the output of "grep cx88 /var/log/syslog" I can confirm this, basically when cx88-blackbird is loaded it should print a message saying that mpeg is supported.

Note: I've lowered the priority and cleared the milestone because this looks like it could simply be an unsupported device or miss-configuration based on the latest info. This doesn't mean I won't try to help if I can.

[bernhart2002@…]

grep cx88 /var/log/syslog

[bernhart2002@…]

Hey, I've uploaded the output of grep cx88 /var/log/syslog

Also I'm good if you want to close this bug since you applied the changes to the v4l analog code that I requested which makes the card work with mythtv and the WAF is good again. However if you want to keep trying different things I don't mind helping out.

regards Eric

[bernhart2002@…]

grep cx88 /var/log/dmesg

[bernhart2002@…]

hey, also included the dmesg log as well.

regards,

eric

[Jonathan <henbit28441@…>]

I am getting the same error as the original reporter with the cx23885 driver which does support V4L2_CAP_STREAMING.

Mythbackend log:​http://pastebin.com/K2hLu36k

[Jonathan <henbit28441@…>]

I just found out the cx23885 driver falsely reports V4L2_CAP_STREAMING capability. However, the workaround at the top of this ticket does not resolve the issue.

Replying to Jonathan <henbit28441@…>:

I am getting the same error as the original reporter with the cx23885 driver which does support V4L2_CAP_STREAMING.

Mythbackend log:​http://pastebin.com/K2hLu36k

[danielk]

It looks like the original issue was resolved by the V4L usage changes so I'm closing this as fixed.

Eric, the IRQ loop reported in the first syslog would seem to indicate a driver issue.

Jonathan, there is little we can do if the driver falsely reports capabilities. Please report that to the V4L devs so they can address the problem.