[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <bpnl6godljzxtzatg7juobvitx6dhnpzayaljqlyq7tnbpzou7@odpzrosrpvq7>
Date: Fri, 15 Nov 2024 11:44:14 +0100
From: Uwe Kleine-König <ukleinek@...nel.org>
To: Werner Sembach <wse@...edocomputers.com>
Cc: Christoph Hellwig <hch@...radead.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>, Christoffer Sandberg <cs@...edo.de>
Subject: Re: EXPORT_SYMBOL_GPL with GPLv3
Hello Werner,
On Tue, Nov 12, 2024 at 09:34:29AM +0100, Werner Sembach wrote:
> Am 12.11.24 um 06:00 schrieb Christoph Hellwig:
> > On Mon, Nov 11, 2024 at 07:40:25PM +0100, Werner Sembach wrote:
> > > Hi,
> > >
> > > asking at the source: Can EXPORT_SYMBOL_GPL-exports be used with out of tree
> > > GPLv3 licensed modules?
> > No. The kernel as whole is licencsed as GPLv2, not GPLv2+.
>
> Yes, but the documentation for EXPORT_SYMBOL_GPL only talks about
> "GPL-compatible" and external linking for EXPORT_SYMBOL-exports is even
> legal with proprietary code.
AFAIK there is no agreement about the legality of proprietary code using
symbols published with EXPORT_SYMBOL. EXPORT_SYMBOL_GPL was created with
the intend to at least have a way to express a symbol to be used by
GPL(v2) code only.
So the safe assumption about EXPORT_SYMBOL functions and variables is
only: "The copyright holder did not expressively prohibit proprietary
callers. This might be because they are ok with proprietary callers,
but maybe they just didn't consider it necessary to share their
opinion.".
> So to my understanding, it comes down to what EXPORT_SYMBOL_GPL means
> and not what's the license of the code behind it is.
Here my approach to clearify the meaning:
https://lore.kernel.org/linux-kernel/20241115103842.585207-2-ukleinek@kernel.org
Best regards
Uwe
Download attachment "signature.asc" of type "application/pgp-signature" (489 bytes)
Powered by blists - more mailing lists