netdev - Re: [RFC PATCH] udp4: Don't take socket reference in receive path

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-ID: <1391728138.10160.23.camel@edumazet-glaptop2.roam.corp.google.com>
Date:	Thu, 06 Feb 2014 15:08:58 -0800
From:	Eric Dumazet <eric.dumazet@...il.com>
To:	Tom Herbert <therbert@...gle.com>
Cc:	David Miller <davem@...emloft.net>,
	Linux Netdev List <netdev@...r.kernel.org>,
	Eric Dumazet <edumazet@...gle.com>
Subject: Re: [RFC PATCH] udp4: Don't take socket reference in receive path

On Thu, 2014-02-06 at 14:40 -0800, Tom Herbert wrote:

> >
> The rationale for SLAB_DESTROY_BY_RCU might be different for UDP than
> TCP. For instance, in the DNS example small connected UDP flows are
> more an issue on the client, the server (which is likely to have much
> greater load) should be using unconnected sockets.
> 

I know some servers have a non connected UDP socket acting as a
'listener' and instancing a new connected socket for ever incoming flow.

You cannot really know what model is used.

> In any case, I am still looking for a way to address this. Like I said
> in the commit log, this per packet cost for UDP processing is far too
> high at least in encapsulation path. I thought about extending
> SO__REUSEPORT to provide CPU affinity but that seems like overkill
> with its own performance implications. Alternatively, we could have
> fast path for the encapsulation using UDP offload model which bypass
> sockets completely which seems unpleasant.

Note that you can solve this before UDP layer, in GRO for example.

If layers before UDP already provide skb->sk (early demux), this could
be a socket that is plainly using call_rcu() for its destruction, and
you do not need to touch socket refcount.

Alternative would be to use a percpu refcnt for these special 'sockets'
that potentially are receiving XX millions frames per second using 48
cpus...

Kent Overstreet designed Percpu refcounts, maybe you should take a look
at this.

--
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