I just installed CentOS 5 on a computer. On boot, I get the a error message, related to 2 SATA disks. These errors did NOT appear on version 4.4 with the EXACT same hardware.
Error message follows: (etc)
I connected both disks to a Promise PDC20378 Sata Controler available on the same board and the error disappears. The configuration stays EXACTLY the same, I only switch both disks' SATA cables to the other 2 connectors.
The error was basically this one, repeated several times:
ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0 ata1.00: tag 0 cmd 0xb0 Emask 0x1 stat 0x51 err 0x4 (device error) ata1: EH complete
I have this 5 times, after which come the statements:
SCSI device sda: 241254720 512-byte hdwr sectors (123522 MB) sda: Write Protect is off sda: Mode Sense: 00 3a 00 00 SCSI device sda: drive cache: write back SCSI device sda: 241254720 512-byte hdwr sectors (123522 MB) sda: Write Protect is off sda: Mode Sense: 00 3a 00 00 SCSI device sda: drive cache: write back
(Is this duplication normal?)
The whole thing is repeated for "ata2.00" and "SCSI device sdb"
So, it is related to the use of the Intel 82801EB ICH5 controler that makes the Southbridge. I report it here in the hope that someone related to kernel development will address this problem.
As I said before, these errors are being reported all over the Net and they seem to appear with a wide range of disks, controlers and chips set brands, even with different Linux distributions. The issue seems to be related to the last few kernel versions and is perhaps a bug or ill-implemented feature of libata or ata_piix.