Re: SG_IO problem on tape devices

[Douglas Gilbert <dougg@torque.net>]

James Bottomley wrote:
> On Tue, 2008-06-03 at 09:59 +0200, Arne Wiebalck wrote: 
>> Hi all,
>>
>> I got a problem using the SG_IO ioctl with our tape drives. 
>> In order to examine errors on our drives in more detail, I 
>> would like to make the sense bytes available to the 
>> application using the REQUEST_SENSE command.
>>
>> However, there's a discrepancy between the kernel output in
>> /var/log/messages and what I see using SG_IO within my 
>> application: while the kernel sees
>>
>> kernel: st0: Error with sense data: scsi1 : destination target 0, lun 0
>> kernel:         command = Space 01 00 0d d5 00
>> kernel: Info fld=0x1, Current st0: sense key Medium Error
>> kernel: Additional sense: Read retries exhausted
>> kernel: Raw sense data:0xf0 0x00 0x03 0x00 0x00 0x00 0x01 0x12 0x00 0x00
>> 0x00 0x00 0x11 0x01 0x00 0x00 0x00 0x00 0x37 0xf7 0x10 0x01 0x00 0x00
>> 0xf7 0x37
>>
>> (which is realistic) the ioctl reports something like
>>
>> 70 00 00 00 00 00 00 12  00 00 00 00 00 00 00 00
>> 00 00 00 00 00 00 00 00  10 02
>>
>> (since the first bit is not set, the sense bytes are not even valid,
>> as far as I understand).

That is a poor usage of the word "VALID" in the standard.
See: http://www.t10.org/ftp/t10/drafts/spc4/spc4r14.pdf
section 4.5.3 on "Fixed Format Sense Data".
The VALID bit (byte 0, bit 7) indicates whether there is
valid data in the information field (bytes 3 to 6 inclusive).

What the SG_IO ioctl is reporting is well-formed sense data
saying there is no error, and no additional sense qualifiers.

It has been cleaned out as James explained.

Doug Gilbert

>> So, could it be that the sense bytes are already cleared when I request
>> them? They cleared/set by the next SCSI cmd, I assume? But shouldn't
>> they be valid even then?
>>
>> I also tried the sg3-utils to query the drive's sense bytes, and they
>> report (almost) the same sense bytes as SG_IO inside my application
>> does. Sending an INQUIRY cmd using sg3_utils works fine, btw.
>>
>> Maybe that's all a plain usage error, so please find the code I use
>> below.
>>
>> Anything obvious I am doing wrong here? All comments appreciated. 
> 
> Yes: SCSI automatically requests sense in response to a check condition.
> So, the sense should be attached to the SG_IO command that receives the
> error.  You can't do an additional request sense for it because the
> sense has already been cleared by the automatic request sense the
> mid-layer did.
> 
> Does the sense data really not get returned by the SG_IO command that
> actually encounters the check condition return?
> 
> James