Re: sata_promise disk error 2.6.22-rc5 with hrt1 patch

On 6/24/07, Mikael Pettersson [off-list ref] wrote:

(cc: linux-ide added)

The 300 TX4 model is causing transient errors for several people,
and we don't yet know why. In your case, port_status 0x00001000
means that "host bus is busy more than 256 clock cycles for every
ATA I/O transfer" (quoting the docs). Basically it's a timeout.

I'm running vanilla 2.6.22-rc7 here, getting a similar error.

[  741.010863] ata2.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x2
[  741.010869] ata2.00: (port_status 0x20080000)
[  741.010875] ata2.00: cmd c8/00:08:bf:01:0c/00:00:00:00:00/ef tag 0
cdb 0x0 data 4096 in
[  741.010877]          res 50/00:00:c6:01:0c/00:00:00:00:00/ef Emask
0x2 (HSM violation)
[  741.313131] ata2: soft resetting port
[  741.463831] ata2: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[  741.587208] ata2: EH complete
[  741.640472] sd 1:0:0:0: [sdb] 1465149168 512-byte hardware sectors
(750156 MB)
[  741.644141] sd 1:0:0:0: [sdb] Write Protect is off
[  741.644144] sd 1:0:0:0: [sdb] Mode Sense: 00 3a 00 00
[  741.670736] sd 1:0:0:0: [sdb] Write cache: enabled, read cache:
enabled, doesn't support DPO or FUA

Identical card to the one otto posted.    How common is this?  I
wouldn't notice if it wasn't logged - I don't notice the blip in disk
access.