Subject: kern/12449: K7/VIA IDE filesystem corruption persists
To: None <gnats-bugs@gnats.netbsd.org>
From: None <dogcow@redback.com>
List: netbsd-bugs
Date: 03/21/2001 15:57:33
>Number:         12449
>Category:       kern
>Synopsis:       K7/VIA IDE filesystem corruption persists
>Confidential:   no
>Severity:       critical
>Priority:       high
>Responsible:    kern-bug-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Wed Mar 21 15:58:00 PST 2001
>Closed-Date:
>Last-Modified:
>Originator:     Tom Spindler
>Release:        <NetBSD-current source date>1.5 + some 1.5.1_ALPHA pullups
>Organization:
	 Redback Networks
>Environment:
	
System: NetBSD fartmangler.redback.com 1.5.1_ALPHA NetBSD 1.5.1_ALPHA (NETZUUL) #32: Mon Feb 5 17:55:39 PST 2001 notroot@zuul.redback.com:/amd/netapp6/netbsd15/nbsrc/sys/arch/i386/compile/NETZUUL i386


>Description:
Even after the patch that manuel sent to developers to try and work around
the VIA chipset bugs with K7s was applied (namely: add
+                           APO_UDMA_PIO_MODE(chp->channel, drive) |
to sys/dev/pci/pciide.c @ line 1987, @@ apollo_setup_channel(chp) )
we're still seeing local filesystem corruption under heavy load.
	
>How-To-Repeat:
Do a disk-intense write-to-local-filesystem activity, such as
"build redback's source tree". Files will start to get corrupted
in weird and strange ways.

AFAIK, we have not seen this with SCSI-based disks.
	
>Fix:
	
>Release-Note:
>Audit-Trail:
>Unformatted: