Date: Sat, 6 Jul 1996 14:00:50 +0200 (MET DST) From: blank@sliphost37.uni-trier.de To: FreeBSD-gnats-submit@freebsd.org Subject: kern/1371: kernel doesn't flush all its buffers when told to halt the system Message-ID: <199607061200.OAA04192@sliphost37.uni-trier.de> Resent-Message-ID: <199607061400.HAA26058@freefall.freebsd.org>
next in thread | raw e-mail | index | archive | help
>Number: 1371 >Category: kern >Synopsis: kernel doesn't flush all its buffers when told to halt the system >Confidential: no >Severity: serious >Priority: medium >Responsible: freebsd-bugs >State: open >Class: sw-bug >Submitter-Id: current-users >Arrival-Date: Sat Jul 6 07:00:02 PDT 1996 >Last-Modified: >Originator: Sascha Blank >Organization: University of Trier, Germany >Release: FreeBSD 2.1-STABLE i386 >Environment: System: FreeBSD 2.1-STABLE with CTM-Patches up to #136 Architecture: Intel Pentium 133, EIDE harddisc and CD-ROM, no SCSI devices, 48 MByte RAM, Triton Chipset >Description: About one out of two attempts to shutdown and halt my system using the command "shutdown -h now" ends up like this: syncing disks... 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 giving up If this happens, then it is always buffer 4 that can't be flushed. As far as I have noticed the amount of process and/or memory load does not influence the occurence of this error. I encountered it sometimes right after shutting down and halting a system that has been rebooted a minute ago, while at other times halting my machine after running for some days under some heavier load works correctly. I am quite sure the error was introduced in one of the CTM patches >= #120 (I guess it was somewhere around #128, but I am not sure). Before that crucial patch I have never had any problems halting my machine. >How-To-Repeat: Enter "shutdown -h now" and watch, if all buffers are flushed correctly. >Fix: No idea yet. >Audit-Trail: >Unformatted:
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199607061200.OAA04192>