Subject: Problems with wd0 on a laptop with current 19991003
To: None <>
From: Brett Lymn <>
List: current-users
Date: 10/06/1999 12:19:38
        Over the weekend I took the brave step and updated my laptop
to -current as of about 3rd Oct.  After I did this I have experienced
some lockups on the machine which look like they are related to the wd
driver.  On the console I get a message something like (I am relying
on memory here):

wd(0:0:0): missing untimeout

Once I get this message any access to the disk blocks, presumably in
the D state.  One thing I have noticed is that the message seems to be
related to the disk spinning up after it has been put into sleep mode.
I have followed the recommendations on the NetBSD web pages about
being aggressive in putting the disk to sleep whilst on battery to
save power - this does work and I was running fine on a pre-1.4
current for a long time but now with a more current current the wd
driver seems not to be handling the disk going to sleep at all well.
I have turned off the sleep timeout when on battery and the system
seems to be working fine again but at the cost of battery life.  Has
anyone any ideas?  Here is a cut of the dmesg for the wd stuff if it

Oct  3 19:07:45 rover /netbsd: wdc0 at isa0 port 0x1f0-0x1f7 irq 14
Oct  3 19:07:45 rover /netbsd: wd0 at wdc0 channel 0 drive 0: <TOSHIBA MK1926FCV
Oct  3 19:07:45 rover /netbsd: wd0: drive supports 16-sector pio transfers, lba addressing
Oct  3 19:07:45 rover /netbsd: wd0: 777MB, 1579 cyl, 16 head, 63 sec, 512 bytes/sect x 1592568 sectors
Oct  3 19:07:45 rover /netbsd: wd0: drive supports PIO mode 4, DMA mode 2
Oct  3 19:07:46 rover /netbsd: wdc1 at isa0 port 0x170-0x177 irq 15
Oct  3 19:07:46 rover /netbsd: atapibus0 at wdc1 channel 0
Oct  3 19:07:46 rover /netbsd: cd0 at atapibus0 drive 0: <TOSHIBA CD-ROM XM-1202B, 0600256534, 1635> type 5 cdrom removable
Oct  3 19:07:46 rover /netbsd: cd0: drive supports PIO mode 3

When the drive sleeps and when the machine resumes I get these:

Oct  4 15:58:54 rover /netbsd: wdc0:0:0: timeout waiting for DRQ, st=0x51, err=0x00
Oct  4 15:58:54 rover /netbsd: wd0a: device timeout writing fsbn 32448 of 32448-32463 (wd0 bn 32511; cn 32 tn 4 sn 3), retrying
Oct  4 15:58:54 rover /netbsd: wd0: soft error (corrected)

But I got those before and I believe these are normal in this case,
the untimeout message does follow one of the above messages.

Brett Lymn, Computer Systems Administrator, British Aerospace Australia