From owner-freebsd-net Sun Jun 3 14:55:58 2001 Delivered-To: freebsd-net@freebsd.org Received: from rhymer.cogsci.ed.ac.uk (rhymer.cogsci.ed.ac.uk [129.215.144.8]) by hub.freebsd.org (Postfix) with ESMTP id 4686737B403 for ; Sun, 3 Jun 2001 14:55:55 -0700 (PDT) (envelope-from richard@cogsci.ed.ac.uk) Received: from banks.cogsci.ed.ac.uk (banks [129.215.144.55]) by rhymer.cogsci.ed.ac.uk (8.9.3/8.9.3) with ESMTP id WAA19174 for ; Sun, 3 Jun 2001 22:55:54 +0100 (BST) Received: (from richard@localhost) by banks.cogsci.ed.ac.uk (8.9.3+Sun/8.9.3) id WAA26338 for freebsd-net@freebsd.org; Sun, 3 Jun 2001 22:55:53 +0100 (BST) Date: Sun, 3 Jun 2001 22:55:53 +0100 (BST) Message-Id: <200106032155.WAA26338@banks.cogsci.ed.ac.uk> From: Richard Tobin Subject: help with mbufs To: freebsd-net@freebsd.org Organization: just say no Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org I'm working on a (user-mode) driver for the Alcatel Speedtouch USB ADSL modem. It works, but the process that relays PPP packets to the ppp program sometimes gets ENOBUFS when writing to the it (via a SOCK_DGRAM socketpair). I tried increasing kern.ipc.nmbclusters (from 1024 to 4096) and kern.ipc.nmbufs (from 4096 to 16384) but this doesn't help. Then I looked at the output of netstat -m: $ netstat -m 69/128/16384 mbufs in use (current/peak/max): 66 mbufs allocated to data 3 mbufs allocated to packet headers 64/94/4096 mbuf clusters in use (current/peak/max) 220 Kbytes allocated to network (1% of mb_map in use) 0 requests for memory denied 0 requests for memory delayed 0 calls to protocol drain routines I expected that after getting ENOBUFS, the peak value would be equal to the maximum. Does something else limit the number of mbufs available? -- Richard To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message