[Linux-kernel-mentees] [RFC PATCH resend] vidtv: adapter->mdev was set too late
hverkuil at xs4all.nl
Mon Feb 1 10:30:42 UTC 2021
On 30/01/2021 13:44, Daniel Almeida wrote:
> Hi Hans (sending this again since it also did not make it to the ML..)
> > (resend again, looks like the previous two attempts never reached the ML)
> > Daniel, how did you test the media controller feature? I don't really
> > understand how your merged patch for MC support could ever work. It's
> > the reason I've marked my patch as an RFC since I wonder if something
> > accidentally got messed up along the way.
> > Regards,
> > Hans
> I still have much to learn on the subsystem.. what I did to test it was
> trying out this command:
> media-ctl -d platform:vidtv -p
> Media controller API version 5.11.0
> Media device information
> driver vidtv
> model vidtv
> bus info platform:vidtv
> hw revision 0x0
> driver version 5.11.0
> I was under the impression that this was enough to show everything was
> in order with the patch.
It should also show the actual topology, and it doesn't.
With this patch it is now be possible to use 'v4l2-compliance -M platform:vidtv'
to check the media device for compliance.
Note that the vidtv.rst documentation needs some more work:
1) There is one more reference to 'modprobe dvb_vidtv_bridge', which should be
'modprobe vidtv' (220.127.116.11)
2) 18.104.22.168.1: the dvb-fe-tool output after modprobe vidtv looks slightly different
since [DVBT] is the default delivery system, not [DVBC/ANNEX_A]. The frequency
range output also looks different (likely also because of the different delivery
system). This is confusing, so make sure that the output is in line with what
you see after loading vidtv for the first time.
3) 22.214.171.124.2: we need proper scan files for vidtv that allows us to test the
various delivery systems. They need to be stored somewhere: either as part of
v4l-utils (contrib/test) or in https://git.linuxtv.org/dtv-scan-tables.git/.
4) 126.96.36.199.2: this example is very confusing since it refers to two different
channel files, without showing the contents of either:
$ dvbv5-scan channel.conf
5) 188.8.131.52.3: this refers to yet another dvb_channel.conf file.
More information about the Linux-kernel-mentees