Date: Thu, 21 Sep 2006 18:08:03 +0100 (BST) From: Robert Watson <rwatson@FreeBSD.org> To: Peter Jeremy <peterjeremy@optushome.com.au> Cc: freebsd-net@freebsd.org, freebsd-current@freebsd.org, Andre Oppermann <andre@freebsd.org> Subject: Re: Much improved sendfile(2) kernel implementation Message-ID: <20060921180711.W56349@fledge.watson.org> In-Reply-To: <20060921075903.GD960@turion.vk2pj.dyndns.org> References: <4511B9B1.2000903@freebsd.org> <20060921075903.GD960@turion.vk2pj.dyndns.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, 21 Sep 2006, Peter Jeremy wrote: > On Wed, 2006-Sep-20 23:59:13 +0200, Andre Oppermann wrote: >> I have rewritten kern_sendfile() to work in two loops, the inner which >> turns as many pages into mbufs as it can up to the free send socket buffer >> space. > > The 64K blocks sounds good but how does this interact with TCP slow start? > Is there the possibility that a couple (for some reasonably large value of > 'couple') of TCP connections slowly accepting a file could eat all the mbuf > space? In principle, existing socket buffer resource limits are the way to prevent this from happening -- sendfile() should continue to obey them as before, we are just better able to use the available buffer space if it exists. Robert N M Watson Computer Laboratory University of Cambridge
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060921180711.W56349>