Re: non-free firmware in kernel modules, aggregation and unclear copyright notice.

[Jeff Garzik <jgarzik@pobox.com>]

Brian Gerst wrote:
> Jeff Garzik wrote:
>
> > Josselin Mouette wrote:
> >
> > > Finally, you shouldn't forget that, technically speaking, using hotplug
> > > for uploading the firmware is much more flexible and elegant than
> > > including it in the kernel. Upgrading the firmware and the module should
> > > be two independent operations. People who are advocating the current
> > > situation are refusing technical improvements just because they are
> > > brought by people they find convenient to call "zealots".
> >
> >
> >
> > This is highly amusing, coming from someone who does not maintain a 
> > driver with a firmware.
> >
> > The current firmware infrastructure is too primitive.  Compiling the 
> > firmware into the driver is much easier on the driver maintainers and 
> > users, presently.
> >
> > Repeating myself,
> >
> > * Most firmwares are a -collection- of images and data.  The firmware 
> > infrastructure should load an -archive- of firmwares and associated 
> > data values.
>
>
> The firmware interface should only be concerned with getting the blob of 
> data into kernel space.  Once it is in kernel space the driver can parse 
> out whatever archive format it is in.  Take for example the ihex code 
> that was posted recently.  Similar code could be written to parse out a 
> tarball, cpio archive, etc.

The archive format for firmware data collections must be standardized, 
and generic code for loading such collections needs to be written, not 
duplicated into each driver.

Obviously the driver-specific data inside the archive is, as the phrase 
implies, driver-specific.

	Jeff



-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/