Re: [PATCH 10/75] V4L/DVB: declare MODULE_FIRMWARE for modules using XC2028 and XC3028L tuners

[hermann pitton <hermann-pitton@arcor.de>]

Hi,

Am Sonntag, den 08.11.2009, 01:20 -0200 schrieb Mauro Carvalho Chehab:
> Hi Ben,
> 
> > > It's not clear to me what this MODULE_FIRMWARE is going to be used
> > > for, but if it's for some sort of module dependency system, then it
> > > definitely should *not* be a dependency for em28xx.  There are lots of
> > > em28xx based devices that do not use the xc3028, and those users
> > > should not be expected to go out and find/extract the firmware for
> > > some tuner they don't have.
> > 
> > This information is currently used by initramfs builders to find
> > required firmware files.  I also use this information in the Debian
> > kernel upgrade script to warn if a currently loaded driver may require
> > firmware in the new kernel version and the firmware is not installed.
> > This is important during the transition of various drivers from built-in
> > to separate firmware.
> > 
> > Neither of these uses applies to TV tuners, but the information may
> > still be useful in installers.
> 
> > > Also, how does this approach handle the situation where there are two
> > > different possible firmwares depending on the card using the firmware.
> > >  As in the example above, you the xc3028 can require either the xc3028
> > > or xc3028L firmware depending on the board they have.  Does this
> > > change now result in both firmware images being required?
> > 
> > It really depends on how the information is used.  Given that there are
> > many drivers that load different firmware files for different devices
> > (or even support multiple different versions with different names), it
> > would be rather stupid to treat these declaration as requirements.
> 
> I agree. An interesting case happens with devices that uses tda10046 DVB demods.
> They have the firmware stored internally on their eeprom. Those firmwares can be
> replaced by a different version loaded in ram, but, in general, they work
> properly with the eeprom one. So, even having the firmware load code there,
> the firmware at /lib/firmware is optional.

Mauro, that could lead to some misunderstanding of the current use
conditions, at least on saa7134.

Minor annotations, the tda10046 does not store firmware internally, but
there are cards which have an extra eeprom to store such firmware.

If such a firmware eeprom is found and correctly configured, the driver
in all cases will load the firmware from that eeprom and all other
firmware in /lib/firmware is ignored.

If not, the firmware will be loaded from /lib/firmware.

For all what I know, firmware revisions 26 and 29 are both valid
"enough", correspondents to what we can load either from TechnoTrend or
LifeView with the getfirmware script, and might be either stored in an
extra eeprom or loaded from host/file.

Prior revisions had issues with missing freqs, in what combination ever.

We also can't totally exclude, given the whole mass of such, that in
some cases firmware eeproms might exist on some boards, but are not
correctly configured and load from host only because of that.

The tendency seen overall is that competitors save the few cents for an
extra firmware eeprom these days ...

Cheers,
Hermann

> -
> 
> I don't see any reason why we should add MODULE_FIRMWARE for v4l/dvb devices.
> As you said, its primary usage is focused on booting a machine, and none
> of those devices would affect booting. 
> 
> As you pointed, the secondary usage doesn't seem to apply to those devices as
> well, and seems to be distro-specific, since different distros use different
> methods to check for firmware dependencies, generally relying at the package
> metadata. To make things worse, several of those firmwares still don't have any
> redistribution rights license that would be required for its inclusion on a distro
> package.
> 
> Also, as this macro have no current usage that would make sense for those
> drivers, I'm afraid that, as time goes by, people will simply forget to
> keep it updated, since they'll need to add the same firmware name on two
> different places.
> 
> That's said, for now, the better is to not add those macros for the devices
> under /drivers/media. They'll just waste some space at the object file, and
> require an additional maintenance care for no good reason.
> 
> Cheers,
> Mauro