From owner-freebsd-hackers Fri Jan 19 14:07:30 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id OAA18156 for hackers-outgoing; Fri, 19 Jan 1996 14:07:30 -0800 (PST) Received: from mpp.minn.net (root@mpp.Minn.Net [204.157.201.242]) by freefall.freebsd.org (8.7.3/8.7.3) with ESMTP id OAA18147 for ; Fri, 19 Jan 1996 14:07:26 -0800 (PST) Received: (from mpp@localhost) by mpp.minn.net (8.7.3/8.6.9) id QAA00597; Fri, 19 Jan 1996 16:06:28 -0600 (CST) Message-Id: <199601192206.QAA00597@mpp.minn.net> Subject: Re: kern/subr_diskslice debug messages To: phk@critter.tfs.com (Poul-Henning Kamp) Date: Fri, 19 Jan 1996 16:06:28 -0600 (CST) From: "Mike Pritchard" Cc: bde@zeta.org.au, freebsd-hackers@FreeBSD.ORG In-Reply-To: <1115.822048288@critter.tfs.com> from "Poul-Henning Kamp" at Jan 19, 96 11:44:48 am X-Mailer: ELM [version 2.4 PL25 ME8b] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-hackers@FreeBSD.ORG Precedence: bulk Poul-Henning Kamp wrote: > > > >I've noticed that if I shutdown into single user mode and > > >unmount all my disks, and then do something to access my second > > >SCSI hard disk (e.g. fsck it), I will get debug output from > > >sys/kern/subr_diskslice.c. These are messages generated by the TRACE > > >macro defined in that module. These messages are only printed if ds_debug > > >is set, which it should not be, and I can't find anything that does > > >set it. One odd this is that ds_debug is declared as volatile. > > >Anyone else seen this, or have any clue why these messages are coming out? > > > > It's volatile just to stop gcc deleting it. Nothing in the kernel should > > set it. Look for array overruns etc. > > Likely to come form the sprintf or thereabout, I'm looking at it... Yep, the latest kern/subr_prf.c fixes the problem. Hopefully that also fixes some of the strange file system corruption problems I've been seeing the past few days. It also seems to have fixed the problem with core files getting junk appended to the end of the core file name as someone else reported yesterday. -- Mike Pritchard mpp@minn.net "Go that way. Really fast. If something gets in your way, turn"