On 03/16/2015 10:04 PM, Antti Palosaari wrote:
On 03/16/2015 10:44 PM, Benjamin Larsson wrote:
On 03/15/2015 11:57 PM, Benjamin Larsson wrote:
Some devices have 2 demodulators, when this is the case
make the slave demod be listed first. Enumerating the slave
first will help legacy applications to use the hardware.
Ignore this patch for now. Stuff gets broken if applied.
I will not apply it even you fix it. I don't like idea adding such hack
to kernel in order to workaround buggy applications. There is many older
devices having similar situation, having 2 demods - one for DVB-T and
one for DVB-C. So there has been surely enough time for app developers
to add support for multiple frontends... laziness.
Quite same happened for DVB-T2 support. I added initially hack for
CXD2820R driver in order to support DVB-T2 even applications are not
supporting it. That was implemented using trick driver did DVB-T2 tune
when DVB-T tune fails. After many years I did another DVB-T2 driver
Si2168 and didn't add such hack anymore. And what was result; almost all
applications were still lacking proper DVB-T2 support, after many many
years...
So I will never ever add any hacks to driver to workaround application
support as application developers are so lazy to add support or new
things if there is some workaround available.
regards
Antti
While I agree with your reasoning, the point with this patch is just to
register one demod before another. If that is a hack or not I don't
know. I have used it locally for test purposes and it is not needed for
actual use so consider the patch withdrawn.
MvH
Benjamin Larsson
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
