From owner-freebsd-hackers Mon May 12 18:06:46 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id SAA17779 for hackers-outgoing; Mon, 12 May 1997 18:06:46 -0700 (PDT) Received: from caipfs.rutgers.edu (root@caipfs.rutgers.edu [128.6.37.100]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id SAA17755; Mon, 12 May 1997 18:06:39 -0700 (PDT) Received: from darkwing.rutgers.edu (darkwing.rutgers.edu [128.6.111.4]) by caipfs.rutgers.edu (8.8.5/8.8.5) with ESMTP id VAA09875; Mon, 12 May 1997 21:06:23 -0400 (EDT) Received: (davem@localhost) by darkwing.rutgers.edu (8.8.4/8.6.9) id VAA21008; Mon, 12 May 1997 21:06:27 -0400 Date: Mon, 12 May 1997 21:06:27 -0400 Message-Id: <199705130106.VAA21008@darkwing.rutgers.edu> From: "David S. Miller" To: terry@lambert.org CC: Andrew.Gordon@net-tel.co.uk, terry@lambert.org, hackers@FreeBSD.ORG, current@FreeBSD.ORG In-reply-to: <199705121908.MAA08042@phaeton.artisoft.com> (message from Terry Lambert on Mon, 12 May 1997 12:08:31 -0700 (MST)) Subject: Re: PATCHES: NFS server locking support Sender: owner-hackers@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk From: Terry Lambert Date: Mon, 12 May 1997 12:08:31 -0700 (MST) If you can guaranteee that you can hash the handle values in user space because the handle values are not unique in the conversion part between client systems, then you're all set... and it's not a problem. F_CNVT will only be called once per hash miss in that case. Why not put lockd into the kernel as a kernel thread and avoid all of this overhead? That's what we do and it works extremely well... ---------------------------------------------//// Yow! 11.26 MB/s remote host TCP bandwidth & //// 199 usec remote TCP latency over 100Mb/s //// ethernet. Beat that! //// -----------------------------------------////__________ o David S. Miller, davem@caip.rutgers.edu /_____________/ / // /_/ ><