Date: Fri, 31 Oct 2003 08:30:19 -0800 From: Sam Leffler <sam@errno.com> To: Hajimu UMEMOTO <ume@FreeBSD.org>, src-committers@FreeBSD.org, cvs-src@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: src/sys/netinet6 icmp6.c Message-ID: <200310310830.19903.sam@errno.com> In-Reply-To: <200310311621.h9VGLQcF082946@repoman.freebsd.org> References: <200310311621.h9VGLQcF082946@repoman.freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Friday 31 October 2003 08:21 am, Hajimu UMEMOTO wrote: > ume 2003/10/31 08:21:26 PST > > FreeBSD src repository > > Modified files: > sys/netinet6 icmp6.c > Log: > (icmp6_rip6_input) if the received data is small enough but in an > mbuf cluster, copy the data to a separate mbuf that do not use a > cluster. this change will reduce the possiblity of packet loss > in the socket layer. This may be the first of several similar commits to do this throughout IPv6, but I think optimizations like this really belong in the network drivers; or at least at that level. OTOH I'm not sure doing this is worthwhile at all. I'd be interested in real experiences to back this up. FWIW I noticed that some Linux network drivers do this. Sam
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200310310830.19903.sam>