Date: Mon, 17 Nov 2003 09:32:12 -0800 From: Sam Leffler <sam@errno.com> To: Brian Feldman <green@FreeBSD.org>, src-committers@FreeBSD.org, cvs-src@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: src/sys/netinet ip_divert.c ip_input.c Message-ID: <200311170932.12308.sam@errno.com> In-Reply-To: <200311170317.hAH3HnIW057958@repoman.freebsd.org> References: <200311170317.hAH3HnIW057958@repoman.freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sunday 16 November 2003 07:17 pm, Brian Feldman wrote: > green 2003/11/16 19:17:49 PST > > FreeBSD src repository > > Modified files: > sys/netinet ip_divert.c ip_input.c > Log: > Fix a few cases where MT_TAG-type "fake mbufs" are created on the stack, > but do not have mh_nextpkt initialized. Somtimes what's there is "1", and > the ip_input() code pukes trying to m_free() it, rendering divert sockets > and such broken. > This really underscores the need to get rid of MT_TAG. I've got most of the changes for that done along with mods to cleanup the "backdoor linkage" between parts of the network code; but it's not ready to commit yet. Switching over dummynet cleans up that code (IMO) and adds no cost as there was already a malloc'd data structure for each packet, but ipfw is a bit more tricky. The goal of all this is to eliminate all MT_TAG usage and also the global function pointer hooks (the latter are replaced by pfil_hooks which then become necessary for many packages and so will be needed in a GENERIC kernel). Sam
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200311170932.12308.sam>