Re: treewide: License cleanup - RedHat originated

[Thomas Gleixner <tglx@linutronix.de>]

Richard!

On Sun, May 22 2022 at 13:33, Richard Fontana wrote:
> On Sun, May 22, 2022 at 10:55 AM Thomas Gleixner <tglx@linutronix.de> wrote:
> I assume you're selecting files that have some sort of Red Hat or
> Sistina copyright notice. Just as a disclaimer, I can't speak to
> copyrights in these files that may be held by other organizations or
> individuals (and I can't say definitively whether any file bearing a
> Red Hat or Sistina copyright notice has or retains any copyright owned
> by Red Hat). With that said:
>
>> 1) this file is released under the gpl
>
> I am fine with saying that any Red Hat copyrights (including any
> Sistina copyrights acquired by Red Hat) in these can be represented
> with SPDX-License-Identifier: GPL-2.0.
>
> I would note that the following files did not seem to have any Red Hat
> or Sistina copyright notices:
>
>>    drivers/md/dm-log-writes.c
>>    drivers/md/dm-ps-queue-length.c
>>    drivers/md/dm-ps-service-time.c
>>    drivers/md/dm-unstripe.c
>>    drivers/md/dm-zero.c

Oops, yes. They are clearly flagged as not Red Hat.

>> 2) this file is released under the lgpl
>>
>>    drivers/md/dm-core.h
>>    drivers/md/dm-log-userspace-base.c
>>    drivers/md/dm-log-userspace-transfer.c
>>    drivers/md/dm-log-userspace-transfer.h
>>    drivers/md/dm-log.c
>>    drivers/md/dm-rq.h
>>    drivers/md/dm.h
>>    include/linux/device-mapper.h
>>    include/linux/dm-dirty-log.h
>
> For these, if the question is just about what version of the LGPL we
> should treat these as, I'd be fine with representing them as
> SDPX-License-Identifier: LGPL-2.1.
>
> However, and I realize this may go beyond my 'remit' as someone
> consulted for linux-spdx stuff or open some additional bag of worms,
> I'm wondering if these would be better off just relicensed under
> GPLv2, and thus represented as SPDX-License-Identifier: GPL-2.0. Even
> treating this code as having been LGPLv2.x-licensed, this would be
> authorized or at any rate contemplated by the largely-overlooked
> LGPLv2.0/LGPLv2.1 section 3. Suggesting this because it would seem to
> help marginally with some of the goals of this initiative and also
> because it's not obvious to me why LGPL would have been preferred over
> GPLv2 for these files to begin with, assuming they weren't copied from
> some unrelated LGPL-licensed project. I've cc'd Jonathan Brassow and
> Mike Snitzer in case they have any thoughts on this.

Let's see :)

Thanks,

        Thomas