Re: ahci fails with SATA DVD drive, IDE mode works

[Mark Lord <liml@rtr.ca>]

Tejun Heo wrote:
> Chuck Ebbert wrote:
>> On Sun, 01 Mar 2009 21:26:33 +0900
>> Tejun Heo <tj@kernel.org> wrote:
>>
>>>> I ran 'hdparm -I' here. Why would it send BLANK to the drive??
>>>> (There was a 1.4 GB DVD-RW in the drive when I did that.)
>>> CDB 0xa1 is BLANK alright.  Is the problem reproducible?
>>>
> 
> Now that I think about it, 0xa0 is PACKET and 0xa1 is IDENTIFY_ATAPI.
> You used "hdparm -I /dev/sr0", right?  Mark, any ideas?  Somehow
> IDENTIFY is getting wrapped inside packet and sent to the device.
..

Does it do that with the latest versions, hdparm-9.11 or hdparm-9.12 ?
(probably fine with those)

In hdparm-9.7 I changed things so that it used ATA_12 rather than ATA_16
when passing short commands through SCSI.  Worked for for disks, but failed
for optical drives -- libata doesn't process ATA_12 for ATAPI (oops).

Fixed it again in hdparm-9.11, so just upgrade.