From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752331AbaKZG0F (ORCPT ); Wed, 26 Nov 2014 01:26:05 -0500 Received: from mail-ie0-f182.google.com ([209.85.223.182]:50348 "EHLO mail-ie0-f182.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752041AbaKZG0C (ORCPT ); Wed, 26 Nov 2014 01:26:02 -0500 Message-ID: <1416983160.29427.52.camel@edumazet-glaptop2.roam.corp.google.com> Subject: Re: [PATCH net-next] tun/macvtap: use consume_skb() instead of kfree_skb() in tun/macvtap_do_read() From: Eric Dumazet To: Jason Wang Cc: davem@davemloft.net, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, mst@redhat.com Date: Tue, 25 Nov 2014 22:26:00 -0800 In-Reply-To: <1416982084-12212-1-git-send-email-jasowang@redhat.com> References: <1416982084-12212-1-git-send-email-jasowang@redhat.com> Content-Type: text/plain; charset="UTF-8" X-Mailer: Evolution 3.10.4-0ubuntu2 Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, 2014-11-26 at 14:08 +0800, Jason Wang wrote: > To be more friendly with drop monitor, we should use comsume_skb() instead > of kfree_skb() in tun/macvtap_do_read(). Otherwise, the packets will be > marked as dropped. > > Signed-off-by: Jason Wang > --- > drivers/net/macvtap.c | 2 +- > drivers/net/tun.c | 2 +- > 2 files changed, 2 insertions(+), 2 deletions(-) > > diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c > index 42a80d3..74d1b23 100644 > --- a/drivers/net/macvtap.c > +++ b/drivers/net/macvtap.c > @@ -862,7 +862,7 @@ static ssize_t macvtap_do_read(struct macvtap_queue *q, > } > iov_iter_init(&iter, READ, iv, segs, len); > ret = macvtap_put_user(q, skb, &iter); > - kfree_skb(skb); > + consume_skb(skb); > break; > } > > diff --git a/drivers/net/tun.c b/drivers/net/tun.c > index ac53a73..d7edeaf 100644 > --- a/drivers/net/tun.c > +++ b/drivers/net/tun.c > @@ -1363,7 +1363,7 @@ static ssize_t tun_do_read(struct tun_struct *tun, struct tun_file *tfile, > > iov_iter_init(&iter, READ, iv, segs, len); > ret = tun_put_user(tun, tfile, skb, &iter); > - kfree_skb(skb); > + consume_skb(skb); > > return ret; > } These patches keep going on, but they are wrong. If you care about drop monitor, then you want to not hide drops but precisely report them. Surely tun_put_user() can return an error, and then packet _is_ dropped.