From owner-cvs-sys  Mon Nov  6 09:21:42 1995
Return-Path: owner-cvs-sys
Received: (from root@localhost)
          by freefall.freebsd.org (8.6.12/8.6.6) id JAA03358
          for cvs-sys-outgoing; Mon, 6 Nov 1995 09:21:42 -0800
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19])
          by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id JAA03345
          ; Mon, 6 Nov 1995 09:21:26 -0800
Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.9/8.6.9) id EAA25085; Tue, 7 Nov 1995 04:19:25 +1100
Date: Tue, 7 Nov 1995 04:19:25 +1100
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199511061719.EAA25085@godzilla.zeta.org.au>
To: bde@zeta.org.au, dyson@freefall.freebsd.org
Subject: Re: cvs commit: src/sys/ufs/ufs ufs_lookup.c
Cc: CVS-commiters@freefall.freebsd.org, cvs-sys@freefall.freebsd.org,
        j@uriah.heep.sax.de
Sender: owner-cvs-sys@FreeBSD.org
Precedence: bulk

>The sync system call (and the update daemone) works correctly
>because the MNT_ASYNC flag is
>temporarily cleared.  You argument for fsync is more valid though, and

Not good.  The flag is per file system so all operations on the file system
currently being synced are slowed down.

>Actually, I am considering supporting a couple of levels of asynchness,
>a "safe" one that just helps speed things up with little risk, and
>a really fast "no holds barred" one like Linux.  Perhaps the "safe" one
>will do meta-data updates every second or so, and the unsafe one will work
>like it does now??

I'd prefer [meta-]data updates based on synchronization points - try not
to update for half completed operations.

Bruce