Re: sata_sil bug

Re: sata_sil bug

[Thomas Maguin]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Justin Cormack wrote:

> ata1: DMA timeout, stat 0x0
> ATA: abnormal status 0xD8 on port 0xF8807087
> scsi0: ERROR on channel 0, id 0, lun 0, CDB: 0x2a 00 17 9b d9 80 00 00
> 20 00
> Current sda: sense = 70  3
> ASC= c ASCQ= 2
> Raw sense data:0x70 0x00 0x03 0x00 0x00 0x00 0x00 0x06 0x00 0x00 0x00
> 0x00 0x0c
> end_request: I/O error, dev sda, sector 396089728
> ATA: abnormal status 0xD8 on port 0xF8807087
> ATA: abnormal status 0xD8 on port 0xF8807087
> ATA: abnormal status 0xD8 on port 0xF8807087
> 

I was able to produce this error on my system too. It occured for me, when I 
compiled the ide interface of sil_3112 as a modul in to my kernel, although I 
was using the libata driver for my harddisks. I lost more then five times my 
sdb drive. After removing the ide modul, the system was rock stable again.

Tom
- -- 

- -------------------------------------------
protect your privacy - encrypt your mails
my key is: 0x2AA933B6
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFAoIZ+yat6ziqpM7YRAolbAJkBKrk8TYGzBUU2zemTuoFqWvdwOgCgyuQk
kSy0Sgn7TPGWQ1XLn00rut0=
=VWPw
-----END PGP SIGNATURE-----

Re: sata_sil bug

[Justin Cormack]

On Tue, 2004-05-11 at 08:53, Thomas Maguin wrote:
> Justin Cormack wrote:
> 
> > ata1: DMA timeout, stat 0x0
> > ATA: abnormal status 0xD8 on port 0xF8807087
> > scsi0: ERROR on channel 0, id 0, lun 0, CDB: 0x2a 00 17 9b d9 80 00 00
> > 20 00
> > Current sda: sense = 70  3
> > ASC= c ASCQ= 2
> > Raw sense data:0x70 0x00 0x03 0x00 0x00 0x00 0x00 0x06 0x00 0x00 0x00
> > 0x00 0x0c
> > end_request: I/O error, dev sda, sector 396089728
> > ATA: abnormal status 0xD8 on port 0xF8807087
> > ATA: abnormal status 0xD8 on port 0xF8807087
> > ATA: abnormal status 0xD8 on port 0xF8807087
> > 
> 
> I was able to produce this error on my system too. It occured for me, when I 
> compiled the ide interface of sil_3112 as a modul in to my kernel, although I 
> was using the libata driver for my harddisks. I lost more then five times my 
> sdb drive. After removing the ide modul, the system was rock stable again.

I definitely didnt have sii3112 as a module. I get this happening very
very rarely. Still wonder if it might be the SiI3112 bug for which there
is a watchdog workaround - any idea if this is the case Jeff? It is
quite hard to repeat as it happened on a faulty disk and they tend to
remap sectors eventually...

Justin

sata_sil bug

[Justin Cormack]

Using sata_sil on sii3112A on 2.6.5-rc2 (sorry a bit old, will update
but I dont think there are any changes to the driver) I got an error:

ata1: DMA timeout, stat 0x0
ATA: abnormal status 0xD8 on port 0xF8807087
scsi0: ERROR on channel 0, id 0, lun 0, CDB: 0x2a 00 17 9b d9 80 00 00
20 00
Current sda: sense = 70  3
ASC= c ASCQ= 2
Raw sense data:0x70 0x00 0x03 0x00 0x00 0x00 0x00 0x06 0x00 0x00 0x00
0x00 0x0c
end_request: I/O error, dev sda, sector 396089728
ATA: abnormal status 0xD8 on port 0xF8807087
ATA: abnormal status 0xD8 on port 0xF8807087
ATA: abnormal status 0xD8 on port 0xF8807087

Then all accesses to the drive got stuck in D state.
I think the disk has bad sectors or other faults - had read errors
earlier, but this is the first bad disk I have had that has had problems
recovering. Could I have hit the state that requires the watchdog?

Justin