Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752471AbcD3T07 (ORCPT ); Sat, 30 Apr 2016 15:26:59 -0400 Received: from shadbolt.e.decadent.org.uk ([88.96.1.126]:35506 "EHLO shadbolt.e.decadent.org.uk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752022AbcD3T05 (ORCPT ); Sat, 30 Apr 2016 15:26:57 -0400 Message-ID: <1462043884.17662.7.camel@decadent.org.uk> Subject: Re: [PATCH 3.2 085/115] veth: =?UTF-8?Q?don=E2=80=99t?= modify ip_summed; doing so treats packets with bad checksums as good. From: Ben Hutchings To: Ben Greear , Sabrina Dubroca Cc: Hannes Frederic Sowa , linux-kernel@vger.kernel.org, stable@vger.kernel.org, akpm@linux-foundation.org, "David S. Miller" , Vijay Pandurangan , Cong Wang , netdev@vger.kernel.org, Evan Jones , Nicolas Dichtel , Phil Sutter , Toshiaki Makita , xiyou.wangcong@gmail.com Date: Sat, 30 Apr 2016 21:18:04 +0200 In-Reply-To: <5722140E.80702@candelatech.com> References: <5720E1F0.9010203@candelatech.com> <1461780469.5102.0.camel@decadent.org.uk> <1461801603.3971874.591751457.2DB91B98@webmail.messagingengine.com> <572155F4.10405@candelatech.com> <20160428102953.GA7656@bistromath.localdomain> <5722140E.80702@candelatech.com> Content-Type: multipart/signed; micalg="pgp-sha512"; protocol="application/pgp-signature"; boundary="=-llYl7b5xpvvmwis5NeY8" X-Mailer: Evolution 3.18.5.1-1 Mime-Version: 1.0 X-SA-Exim-Connect-IP: 80.110.71.10 X-SA-Exim-Mail-From: ben@decadent.org.uk X-SA-Exim-Scanned: No (on shadbolt.decadent.org.uk); SAEximRunCond expanded to false Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2194 Lines: 56 --=-llYl7b5xpvvmwis5NeY8 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable On Thu, 2016-04-28 at 06:45 -0700, Ben Greear wrote: >=20 > On 04/28/2016 03:29 AM, Sabrina Dubroca wrote: [...] > > Your use case is invalid, it just happened to work because of a > > bug.=C2=A0=C2=A0If you want the stack to fill checksums so that you wan= t capture > > and reinject packets, you have to disable checksum offloading (or > > compute the checksum yourself in userspace). > Disabling checksum offloading or computing in user-space (and then > recomputing in veth to verify the checksum?) is a huge performance loss. >=20 > Maybe we could add a socket option to enable Cong's patch on a per-socket > basis?=C2=A0=C2=A0That way my use-case can still work and you can have th= is new > behaviour by default? It does sound like a useful option to have. =C2=A0If there are other applications that depend on veth's checksum-fixing behaviour and are being distributed in binary form, then a per-device option might be necessary so users can keep those applications working before they're updated. Ben. --=20 Ben Hutchings Tomorrow will be cancelled due to lack of interest. --=-llYl7b5xpvvmwis5NeY8 Content-Type: application/pgp-signature; name="signature.asc" Content-Description: This is a digitally signed message part -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAABCgAGBQJXJQTsAAoJEOe/yOyVhhEJdsIP/3erTuDrsRZy0r5zIaSReJGT hCXyg+G+YIpGfHcPA8/US8P3cAespJyplKzIsxAYIQq84rMVVefWqYuF05Ctr1Yx 7vhLUvc5qWPLksVkNHMAfIUSl7nj2a1rZbFMfRg4+2El61Icvd8EHGfdqMH5PldA nHWV2mlM6M95reO7ilgbNu7OwgCKGmbcxRgn6tOABFNR7rnUGbk2Yq45oy7H3Ny5 cX5NtFt+CvmTDpva5ENAg9T8d7qzBtaCtMy+wd7lmD2herw+smIZbyoPFcTe+mmU e8CEzYQMWAsbo/OpoOOadAt5ojOHjV0/WQ2onuYzqzHsFhL/x/tplnECvjrMsPrK UofsCKow9s+B6mUW08QIgD/gD9wBUZuMSALip5qisfbkWK26Mzyk62rnqbvxx8N2 Pxq0EFVJ+kMMlXbUzMDajXPDRgo1Xsy3+V7tE8T/qgOQ5iOzA31xXOr9h9ViSit2 v6p4aCgvH5pMoTdysW2r5cFJdW8P2dV5VmCgf9yF/WNSnJiQ5coHDpqGLsqiwFhL REr7CQN3YWRs9JOopLnrljg/5RG157GJb4mVXbsJlNs7eNFktm9KQVATuO+XsAeY uQQFL2SYLNJCUZUkqMawme4tME6Gq5M7chsPKJFxLk2iXmUwKcO46t572ZjDQKrR DKlCNNg11e9XpZ1DsGWs =msz9 -----END PGP SIGNATURE----- --=-llYl7b5xpvvmwis5NeY8--