Date: Sat, 12 Sep 2009 13:47:27 +0400
From: Evgeniy Polyakov <zbr@ioremap.net>
To: Jamie Lokier <jamie@shareable.org>
Cc: jamal <hadi@cyberus.ca>, Eric Paris <eparis@redhat.com>,
       David Miller <davem@davemloft.net>, linux-kernel@vger.kernel.org,
       linux-fsdevel@vger.kernel.org, netdev@vger.kernel.org,
       viro@zeniv.linux.org.uk, alan@linux.intel.com, hch@infradead.org,
       balbir@in.ibm.com
Subject: Re: [PATCH 1/8] networking/fanotify: declare fanotify socket numbers
Message-ID: <20090912094727.GC24709@ioremap.net>
References: <20090911052558.32359.18075.stgit@paris.rdu.redhat.com> <20090911.114620.260824240.davem@davemloft.net> <1252697613.2305.38.camel@dhcp231-106.rdu.redhat.com> <1252704102.25158.36.camel@dogo.mojatatu.com> <20090911214254.GB19901@shareable.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20090911214254.GB19901@shareable.org>
User-Agent: Mutt/1.5.13 (2006-08-11)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 875
Lines: 19

On Fri, Sep 11, 2009 at 10:42:54PM +0100, Jamie Lokier (jamie@shareable.org) wrote:
> While skbs and netlink aren't that slow, I suspect they're an order of
> magnitude or two slower than, say, epoll or inotify at passing events
> around.

Skbs are those tiny bits which allow 100 Gbit/s forwarding between multiple
interfaces. But of course it can not be as fast as plain memory copy :)

Having one skb allocation per IO syscall will be challenging but after
all we have this for send/recv calls and achieve high performance
numbers. Idea with merging events in the same skb will be also very non-trivial.

-- 
	Evgeniy Polyakov
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/