Date: Tue, 17 Nov 1998 21:36:47 +1000 From: Stephen McKay <syssgm@dtir.qld.gov.au> To: Michael Robinson <robinson@public.bta.net.cn> Cc: freebsd-stable@FreeBSD.ORG, syssgm@dtir.qld.gov.au Subject: Re: writev() to tcp Message-ID: <199811171136.VAA04910@nymph.dtir.qld.gov.au> In-Reply-To: <199811170204.KAA26059@public.bta.net.cn> from Michael Robinson at "Tue, 17 Nov 1998 10:04:47 %2B0800" References: <199811170204.KAA26059@public.bta.net.cn>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tuesday, 17th November 1998, Michael Robinson wrote: >Marc Slemko <marcs@znep.com> writes: >>This doesn't really have anything to do with writev() in particular. There >>is a bug in the TCP code where a packet bigger than a single mbuf (MLEN == >>108 bytes) but not big enough for a mbuf cluster (MINCLSIZE == 204 bytes) >>ends up being put into two mbufs that end up being put on the wire in two >>parts. There should be some messages about it in the archives. There are >>various fixes, but none has been made yet. > >Thanks for the information. I checked the GNATS bug database and didn't >find any PR matching this bug. > >If someone sent a PR, would this get fixed? (The Squid FAQ has a one-line >patch to uipc_socket.c.) That patch went into -current on 1998/07/06 but was removed on 1998/07/18 because various and sundry things broke. See revisions 1.41 and 1.42 of sys/kern/uipc_socket.c. There was some inconclusive discussion of the problem at the time, probably on freebsd-current. My own personal experience of using this patch was also inconclusive. Some things got better, but some things became worse and I didn't manage to prove that the patch was to blame, or that it was innocent. Feel free to experiment! Stephen. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199811171136.VAA04910>