Date: Fri, 28 Mar 2003 23:52:08 -0600 (CST) From: Mike Silbersack <silby@silby.com> To: Mike Silbersack <silby@FreeBSD.org> Cc: cvs-all@FreeBSD.org Subject: Re: cvs commit: src/sys/kern uipc_mbuf.c src/sys/sys mbuf.h Message-ID: <20030328234910.N3026@odysseus.silby.com> In-Reply-To: <200303290548.h2T5ma5Q007596@repoman.freebsd.org> References: <200303290548.h2T5ma5Q007596@repoman.freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
I took the "goal" parameter back out. The more I thought about it, I realized that it would only lead to confusion wrt the semantics of what would be considered a failure. Additionally, the host chain is unharmed if memory allocation fails; this allows m_defrag to be called on critical chains during a time of low memory without the concern of the chain being lost. Also, thanks to Bosko's patch to m_dup_pkthdr, I use m_getcl when needed. I'm going to beef up MBUF_FRAG_TEST so it can create chains busdma_load_mbuf cannot coalesce, then I'm going to go hunting for deviant network drivers. On Fri, 28 Mar 2003, Mike Silbersack wrote: > silby 2003/03/28 21:48:36 PST > > FreeBSD src repository > > Modified files: > sys/kern uipc_mbuf.c > sys/sys mbuf.h > Log: > Add the m_defrag routine, as discussed on committers@. This > incarnation should address the concerns of all in the discussion, > and keeps statistics which show how much it is used. > > MFC after: 2 weeks > > Revision Changes Path > 1.112 +83 -0 src/sys/kern/uipc_mbuf.c > 1.117 +1 -0 src/sys/sys/mbuf.h >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030328234910.N3026>