Received-SPF: pass (google.com: best guess record for domain of linux-crypto-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
From:   Toke =?utf-8?Q?H=C3=B8iland-J=C3=B8rgensen?= <thoiland@redhat.com>
To:     "Jason A. Donenfeld" <Jason@zx2c4.com>, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org
Cc:     "Jason A. Donenfeld" <Jason@zx2c4.com>,
        David Miller <davem@davemloft.net>,
        Greg KH <gregkh@linuxfoundation.org>,
        Linus Torvalds <torvalds@linux-foundation.org>,
        Herbert Xu <herbert@gondor.apana.org.au>,
        linux-crypto@vger.kernel.org
Subject: Re: [PATCH RFC net-next] net: WireGuard secure network tunnel
In-Reply-To: <20191120203538.199367-1-Jason@zx2c4.com>
References: <20191120203538.199367-1-Jason@zx2c4.com>
Date:   Thu, 21 Nov 2019 11:29:16 +0100
Message-ID: <877e3t8qv7.fsf@toke.dk>
MIME-Version: 1.0
Content-Type: text/plain; charset=WINDOWS-1252
Content-Transfer-Encoding: quoted-printable
Sender: linux-crypto-owner@vger.kernel.org
Precedence: bulk

"Jason A. Donenfeld" <Jason@zx2c4.com> writes:

> RFC Note:
>   This is a RFC for folks who want to play with this early, because
>   Herbert's cryptodev-2.6 tree hasn't yet made it into net-next. I'll
>   repost this as a v1 (possibly with feedback incorporated) once the
>   various trees are in the right place. This compiles on top of the
>   Frankenzinc patchset from Ard, though it hasn't yet received suitable
>   testing there for me to call it v1 just yet. Preliminary testing with
>   the usual netns.sh test suite on x86 indicates it's at least mostly
>   functional, but I'll be giving things further scrutiny in the days to
>   come.

Hi Jason

Great to see this! Just a few small comments for now:

> +/*
> + * Copyright (C) 2015-2019 Jason A. Donenfeld <Jason@zx2c4.com>. All Rig=
hts Reserved.
> + */

Could you please get rid of the "All Rights Reserved" (here, and
everywhere else)? All rights are *not* reserved: this is licensed under
the GPL. Besides, that phrase is in general dubious at best:
https://en.wikipedia.org/wiki/All_rights_reserved

> +=09MAX_QUEUED_INCOMING_HANDSHAKES =3D 4096, /* TODO: replace this with D=
QL */
> +=09MAX_STAGED_PACKETS =3D 128,
> +=09MAX_QUEUED_PACKETS =3D 1024 /* TODO: replace this with DQL */

Yes, please (on the TODO) :)

FWIW, since you're using pointer rings I think the way to do this is
probably to just keep the limits in place as a maximum size, and then
use DQL (or CoDel) to throttle enqueue to those pointer rings instead of
just letting them fill.

Happy to work with you on this (as I believe I've already promised), but
we might as well do that after the initial version is merged...

-Toke