Subject: Re: ffs race panic run away!
To: Patrick Welche <prlw1@cam.ac.uk>
From: Jason Thorpe <thorpej@nas.nasa.gov>
List: current-users
Date: 06/26/1997 10:26:06
On Thu, 26 Jun 1997 17:14:25 +0100 (BST) 
 prlw1@cam.ac.uk (Patrick Welche) wrote:

 > _ffs_blkfree+0x3e3	_ffs_blkfree(f8710700,a4,1000,fa484e24,f8710700)
 > _ffs_truncate+0x950	_ffs_truncate(fa484dcc)
 > _ufs_setattr+0x19a	_ufs_setattr(fa484e10)
 > _vn_open+0x2f7		_vn_open(fa484ee0,403,1a4,f819bd88,f8713f00)
 > _sys_open+0x97		_sys_open(...,...,...,0,2b1a80)
 > _syscall+0x154		(number 5)
 > 
 > Beforehand, as usual, aha0: host_stat 12
 > Afterwards, again, unknown filetype I=7840 and many DUP I = ...
 > This was during a sup, took 3 fsck's to fix, and for some reason a
 > number of files were left as directories, eg /usr/lib/libfl_p.a !?

Wow... sounds... painful.  Almost like your SCSI controller is just
acting up.  The host_stat message might be an indicator of that.  That's
not really normal at all.  If you SCSI controller is getting flaky,
that could _definitely_ cause file system problems... I recently had
that problem with an EISA 1742 ... right before it ate a bunch of file
systems, and then refused to let the machine boot (BIOS wouldn't even
come up... ouch...)

 > I think I am giving up with all this, so what's a good PCI SCSI card
 > that works with a P133 (thus hopefully avoiding the race condition)?

I use a BusLogic 958 quite happily.

Jason R. Thorpe                                       thorpej@nas.nasa.gov
NASA Ames Research Center                               Home: 408.866.1912
NAS: M/S 258-6                                          Work: 415.604.0935
Moffett Field, CA 94035                                Pager: 415.428.6939