netdev - Re: [PATCH net-next] net: better IFF_XMIT_DST

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Thu, 02 Oct 2014 13:06:53 -0700
From:	Eric Dumazet <eric.dumazet@...il.com>
To:	Julian Anastasov <ja@....bg>
Cc:	David Miller <davem@...emloft.net>, netdev <netdev@...r.kernel.org>
Subject: Re: [PATCH net-next] net: better IFF_XMIT_DST_RELEASE support

On Thu, 2014-10-02 at 22:34 +0300, Julian Anastasov wrote:
> 	Hello,
> 
> On Thu, 2 Oct 2014, Eric Dumazet wrote:
> 
> > From: Eric Dumazet <edumazet@...gle.com>
> > 
> > Testing xmit_more support with netperf and connected UDP sockets,
> > I found strange dst refcount false sharing.
> > 
> > Current handling of IFF_XMIT_DST_RELEASE is not optimal.
> > 
> > dropping dst in validate_xmit_skb() is certainly too late in case
> > packet was queued by cpu X but dequeued by cpu Y
> > 
> > The logical point to take care of drop/force is in __dev_queue_xmit()
> > before even taking qdisc lock.
> 
> 	Does it hurt skb_dst usage in net/sched/, for example,
> dst->tclassid in net/sched/cls_route.c ?

Hmm good point...

Then, if a qdisc like that is used, I guess we can remove
IFF_XMIT_DST_RELEASE from dev->priv_flags

Maybe I should add a dev->dst_are_needed counter, instead using a flag.

Thanks


--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html