On 09/17/2010 03:26 PM, Douglas Gilbert wrote:
On 10-09-17 02:31 PM, Jeff Garzik wrote:
On 09/05/2010 11:57 AM, Douglas Gilbert wrote:
While looking at an eSATA connected external disk with
a medium error, this sense data appeared:
READ cdb: 28 00 96 4a 7a d1 00 00 01 00
duration=2816 ms
READ: Descriptor format, current; Sense key: Medium Error
Additional sense: Unrecovered read error - auto reallocate failed
Descriptor type: Information
0x00000000964a7ad1
Raw sense data (in hex):
72 03 11 04 00 00 00 0c 00 0a 80 00 00 00 00 00
96 4a 7a d1
That is pretty close to what I was expecting to see. If anything
there is too much information. That "Unrecovered read error -
auto reallocate failed" [asc/asq: 11h/04h] should just be
"Unrecovered read error" [asc/asq: 11h/0h]. See sat2r09.pdf table
105 or sat3r00.pdf table 105.
If any auto-reallocate occurred on the drive's part, it definitely
failed at that point. And auto-reallocate is a possibility.
11h/04h seems to more accurately describe the situation.
Usually if you are implementing a SCSI device (lu) there
is some latitude in asc/asq codes. However for a SAT layer
they state exactly how to translate ATA errors.
There is one published SAT standard (ANSI INCITS 431-2007)
and there will soon be another one (SAT-2). The compliant
response is asc/asq: 11h/0h since in t10 documents additional
sense matches by string which is shown in upper case.
Fair enough. Wanna write a patch? :)
BTW I never understood what asc/asq 11h/04h was about,
auto-reallocate is something that should be attempted
on a _recovered_ read error. And that in turn might fail
if the disk had run out of spare blocks. So in a contorted
sort of way, perhaps that is what "Unrecovered read error -
auto reallocate failed" means!? If so, you are most
likely incorrect on that count as well.
The ATA disk hits an unrecovered read error, which by virtue of
operation implies auto reallocate failed. What's going on inside seems
to match the error text quite closely.
Jeff
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
