Bug 36313

Summary:	ATA disk not bootable anymore after cvsuping to 4.5-STABLE
Product:	Base System	Reporter:	Bjoern Engels <bjoern.engels>
Component:	kern	Assignee:	Søren Schmidt <sos>
Status:	Closed FIXED
Severity:	Affects Only Me
Priority:	Normal
Version:	4.5-STABLE
Hardware:	Any
OS:	Any

Description Bjoern Engels 2002-03-25 21:40:00 UTC

After cvsupping 4.5-STABLE today my system didn't boot anymore.
The ATA-mirror was defective and couldn't be rebuilt. I reinstalled
the machine, cvsupped again and the same thing happened again.
Booting the old kernel doesn't help either, the new ATA drivers seem
to have crashed the disk.
Here are the error messages (I had to type them down, so it's only the
relevant things, I hope it's enough information.)

[...]
atapci0: <VIA 82C686 ATA66 controller> port ... on pci0
ata0: at [...] on atapci0
ata1: at [...] on atapci0
[...]
atapci1: <Promise ATA100 controller> [...] on pci0
ata2: at [...] on atapci1
ata3: at [...] on atapci1
[...]
ata1-slave: timeout waiting for interrupt
ata1-slave: ATAPI identify failed
stray irq 7
ar0: 58610MB <ATA RAID0 array> [7471/255/63] status READY subdisks:
 0 READY ad4: 58610MB <Maxtor 4W060H4> [119081/16/63] at ata2-master UDMA100
acd0: CDROM <SR243T> at ata1-master PIO4
Mounting root from ufs:/dev/ar0s1a
ad4: READ command timeout tag=0 serv=0 - resetting
ata2: resetting devices ..
ad4: removed from configuration
ar0: ERROR - array broken
ad4: deleted from ar0 disk0
ar0: ERROR - array broken
done
ar0: reading primary partition table: error reading fsbn 0
Root mount failed: 5

How-To-Repeat: See above.

Comment 1 iedowse freebsd_committer

2002-12-08 18:39:16 UTC

State Changed
From-To: open->feedback


Was this resolved with more recent -STABLE sources?

Comment 2 iedowse 2002-12-08 20:26:45 UTC

Adding to the audit trail:

In message <200212082123.46647.bjoern.engels@mail.isis.de>, Bjoern Engels write
s:
>> Was this resolved with more recent -STABLE sources?
>
>Hi Ian,
>
>thanks for the fedback.
>No, this problem hasn't resolved yet. I tried several times to upgrade th=
>e box=20
>(4.6-RELEASE, 4.7-RELEASE, {any}-STABLE), but that helped just a little:
>after trying to boot the box with the new kernel, I saw the same errors=20
>("array broken" and so on), but I was able to reboot the machine with the=
> old=20
>4.5-RELEASE kernel normally.
>If you need any more information, please tell me, I'll attach a serial co=
>nsole=20
>to get exact boot messages.
>
>
>Thank you for any help
>
>Bjoern
>
>
>
>

Comment 3 iedowse freebsd_committer

2002-12-08 20:31:19 UTC

State Changed
From-To: feedback->open


Problem is still present in recent -STABLE.

Comment 4 iedowse freebsd_committer

2002-12-08 20:31:19 UTC

Responsible Changed
From-To: freebsd-bugs->sos


ATA-related PR.

Comment 5 KAREN THODE 2002-12-24 20:01:48 UTC

Apparently, the bug is in the ATA RAID volume mount code.

Lucas

Comment 6 Søren Schmidt freebsd_committer

2003-04-28 19:31:47 UTC

State Changed
From-To: open->closed

I cannot reproduce this problem, however since we are now at 4.8 
I suggest try using that with a fresh install and see if the 
problem reappears.