From mboxrd@z Thu Jan  1 00:00:00 1970
From: David Barmann <david.barmann@stackpath.com>
Subject: Re: [PATCH net-next] sock: Reset dst when changing sk_mark via
 setsockopt
Date: Wed, 7 Nov 2018 17:07:32 -0600
Message-ID: <20181107230732.GA21388@konacove.com>
References: <20181107223327.GA21018@konacove.com>
 <4a6b111a-c0d2-e12c-bc42-3c723a0297bc@gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Cc: Eric Dumazet <eric.dumazet@gmail.com>
To: netdev@vger.kernel.org
Return-path: <netdev-owner@vger.kernel.org>
Received: from konacove.com ([69.16.156.219]:47562 "EHLO konacove.com"
        rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
        id S1726896AbeKHIkH (ORCPT <rfc822;netdev@vger.kernel.org>);
        Thu, 8 Nov 2018 03:40:07 -0500
Content-Disposition: inline
In-Reply-To: <4a6b111a-c0d2-e12c-bc42-3c723a0297bc@gmail.com>
Sender: netdev-owner@vger.kernel.org
List-ID: <netdev.vger.kernel.org>

> On 11/07/2018 02:33 PM, David Barmann wrote:

> > When setting the SO_MARK socket option, the dst needs to be reset so
> > that a new route lookup is performed.
> > 
> > This fixes the case where an application wants to change routing by
> > setting a new sk_mark.  If this is done after some packets have already
> > been sent, the dst is cached and has no effect.
> > 

> > +		} else {
> > +			struct dst_entry *dst = sk_dst_get(sk);
> >  			sk->sk_mark = val;
> > +			sk_dst_reset(sk);
> 
> 
> sk_dst_get() and dst_release() seems extra overhead ?
> 
> net/ipv4/ip_sockglue.c do_ip_setsockopt() has a similar handling for IP_TOS,
> and it only calls sk_dst_reset() (If the new TOS is different than the current one)
> 
> 
> 
> So I would suggest :
> 
> 	if (!ns_capable(...)) {
>         	ret = -EPERM;
> 	} else if (val != sk->sk_mark) {
> 		sk->sk_mark = val;
> 		sk_dst_reset(sk);
> 	}


Thanks Eric, I'll redo that and resubmit.

-David