Subject: kern/25966: IDE drives no longer detected
To: None <gnats-bugs@gnats.NetBSD.org>
From: None <schmonz-netbsd@schmonz.com>
List: netbsd-bugs
Date: 06/19/2004 02:29:16
>Number: 25966
>Category: kern
>Synopsis: IDE drives no longer detected
>Confidential: no
>Severity: critical
>Priority: high
>Responsible: kern-bug-people
>State: open
>Class: sw-bug
>Submitter-Id: net
>Arrival-Date: Sat Jun 19 02:30:00 UTC 2004
>Closed-Date:
>Last-Modified:
>Originator: Amitai Schlair
>Release: NetBSD 2.0_BETA
>Organization:
Schmonz Enterprises
>Environment:
NetBSD 2.0_BETA NetBSD 2.0_BETA (GENERIC) #1: Sat Jun 12 23:42:39 EDT 2004 schmonz@fastasspc.home.schmonz.com:/home/schmonz/build/objdir/sys/arch/macppc/compile/GENERIC macppc
>Description:
On a beige Power Macintosh G3, 2.0_BETA kernels don't detect wd0
or cd0, and hang after printing "scsibus0: waiting 2 seconds for
devices to settle...".
Additional data points:
* 1.6.2 detected the drives and ran fine
* 1.6O detected the drives and ran fine
* 2.0_BETA detects the drives and runs fine if I set wdcdebug_mask to 0x10
* 2.0_BETA with -r1.172.2.1 of sys/dev/ic/wdc.c: same as 2.0_BETA
* 2.0F: same as 2.0_BETA
* Going back to 1.6.2 (when this system started running NetBSD), I have
often seen this message once during boot, just before the root
fs type is identified:
wdc0:0:0: lost interrupt
type: ata tc_bcount: 512 tc_skip: 0
wdc0:0:0: intr with DRQ (st=0x58)
wd0: transfer error, downgrading to PIO mode 4
wd0(wdc0:0:0): using PIO mode 4
wd0c: device timeout reading fsbn 0 (wd0 bn 0; cn 0 tn 0 sn 0), retrying
wd0: soft error (corrected)
With 1.6.2 and 1.6O I ignored this message since the system ran
without further errors (and disk throughput wasn't an issue).
Some dmesg output:
http://netbsd.schmonz.com/dmesg/macppc/beige-g3-wdcdebug.txt
http://netbsd.schmonz.com/dmesg/macppc/beige-g3-wdcdebug-2.0F.txt
http://netbsd.schmonz.com/dmesg/macppc/beige-g3-1.6.2.txt
>How-To-Repeat:
Attempt to boot a 2.0_BETA or -current kernel on a beige Power Macintosh G3.
>Fix:
A workaround (perhaps the problem is timing-related?):
boot -d
write wdcdebug_mask 0x10
continue
>Release-Note:
>Audit-Trail:
>Unformatted: