From owner-freebsd-current Fri Oct 6 13:00:40 1995 Return-Path: owner-current Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id NAA11900 for current-outgoing; Fri, 6 Oct 1995 13:00:40 -0700 Received: from Sysiphos (Sysiphos.MI.Uni-Koeln.DE [134.95.212.10]) by freefall.freebsd.org (8.6.12/8.6.6) with SMTP id NAA11883 for ; Fri, 6 Oct 1995 13:00:30 -0700 Received: by Sysiphos id AA04407 (5.67b/IDA-1.5 for current@freebsd.org); Fri, 6 Oct 1995 20:59:03 +0100 Message-Id: <199510061959.AA04407@Sysiphos> From: se@zpr.uni-koeln.de (Stefan Esser) Date: Fri, 6 Oct 1995 20:59:02 +0100 X-Mailer: Mail User's Shell (7.2.6 alpha(2) 7/9/95) To: current@freebsd.org Subject: biodone: buffer already done (and system hangs) with -current Cc: Doug Rabson Sender: owner-current@freebsd.org Precedence: bulk There seems to be some problem in FreeBSD-current regarding mmap() and NFS: Just found, that "tail -300" can hang forever in disk wait, if used to access some file via NFS: $ ps -lw UID PID PPID CPU PRI NI VSZ RSS WCHAN STAT TT TIME COMMAND 200 27492 27459 0 -22 0 152 444 vmpfw D+ p0 0:00.03 tail -300 /homes/se/Mail/fvwm I've traced the point of failure to be the first access to the mmap()ed file (tail/forward.c:218). Interestingly I can access the file by other means without problems (i.e. a cat of the same file succeeds). On the console appeared 3 "biodone: buffer already done" messages per cat of the file (some 540KB). The system didn't shutdown (i.e. hung before the "syncing disks" message), when I tried to reboot. When the system was up again, I retried the tail command, and it hung again! (I.e. I can reproduce this failure, in case you got something I might test.) Any ideas ? STefan -- Stefan Esser, Zentrum fuer Paralleles Rechnen Tel: +49 221 4706021 Universitaet zu Koeln, Weyertal 80, 50931 Koeln FAX: +49 221 4705160 ============================================================================== http://www.zpr.uni-koeln.de/staff/esser/esser.html