I didn't see any pointers other than pilot error raised in the
recent
thread with subject line:
Subject: Re: IDE quirk in 3.2-STABLE kernel ?
Perhaps those of you who're in support of the pilot error notion
could
have a look at PR 13174 and comment? The originator claims that his
kernel only finds wdc0 when he "auto-detects" his hard drives with
his
Award BIOS.
He doesn't specify whether he has drives connected to that
controller,
which is why I've cc'd him on this mail.
I apologize for the missing information. This problem has been posted
several times (by me) to the -stable mailing list, and I simply forgot to
include all the details with the bug report.
The only reason why this would be a critical bug is if there were drives
connected to the primary controller. In this case, I only have IDE drives,
which means that the kernel cannot mount the root filesystem without wcd0
being detected. I have two drives on wcd0: a 1.27 GB Samsung as master (with
root filesystem) and a 6.54 GB Seagate (with /usr, and others). I have an
ATAPI CD-ROM drive as master on wcd1 (which is detected just fine). I've
noticed this behaviour with every 3.- branch installation I've had on my
machine (3.2-RELEASE boot disks, 3.2-STABLE). I have 4 other OSs on the same
machine, so I would prefer not to mess with the physical configuration (i.e.
moving disks from one controller to another, etc), which would mess up their
drive configurations.
Please note that this is not a problem with IRQ's, etc. I've had
2.2.8-RELEASE and -STABLE running on the same machine with absolutely no
problems detecting wcd0 (and the default IRQ's and port addresses for both
2.2.8 and 3.2 are the same).
Any help is appreciated.
Andrew J. Lofthouse
To Unsubscribe: send mail to majord...@freebsd.org
with "unsubscribe freebsd-hackers" in the body of the message
