RE: [t13] SAT revision 8a now available

Wed, 05 Jul 2006 10:34:14 -0700

Michael,
 
I agree, the connection here isn't very clear. It might help to think of how the Check Power Mode command helps fulfill the needs of the TEST UNIT READY command rather than the reverse. The REQUEST SENSE command is supposed to provide immediate status of the power state of the drive, except for indicating when the drive is in the STOPPED state (this is for historical reasons). The TEST UNIT READY command is used to determine if the drive is in the STOPPED state or if it's ready to process commands. The initial thought with the Check Power Mode command was that it would reflect whether the ATA device is in the "STANDBY" state (mapping to the SCSI "STOPPED" state). But it turns out there are a number of ways an ATA drive can be in the "STANDBY" state, but it would confuse host software if the emulated SCSI device reported it was in the STOPPED state, for any reason other than being put in the STOPPED state as a result of processing a START STOP UNIT command. So, the SATL ends up with the requirement to "remember" whether the device is supposed to be in the "STOPPED" state or not, independent of how the ATA device responds to the Check Power Mode command. So, the Check Power Mode command is used simply to determine if the device is functional or not, after the SATL has already determined the emulated device is not supposed to be in the STOPPED state. So as far as the translation of the TEST UNIT READY command is concerned, the values returned in the Sector Count are irrelevant.
 
Hope this helps. I welcome any comments that might provide a better clarification than the one I've suggested here.
 
Regards,
Bob

From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Eschmann, Michael K
Sent: Tuesday, July 04, 2006 6:30 PM
To: [email protected]
Subject: RE: [t13] SAT revision 8a now available

Hi All,
 
As I read this specification, I don't understand how Test Unit Ready is converted to a Check Power Mode, or more importantly how the Check Power Mode result value is reported back to the requestor.  Could someone please describe how the CPM "Sector Count" values of 00h (device in standby), 80h (device in idle) and FFh (device in either Active or Idle mode) are reported back in the context of a Test Unit Ready command?
 
Thanks, MKE.

From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Sheffield, Robert L
Sent: Monday, June 26, 2006 8:27 AM
To: [email protected]
Subject: [t13] SAT revision 8a now available

SCSI / ATA Translation (SAT) revision 8a (sat-r08a.pdf) is now available on http://www.t10.org/drafts.htm and http://www.t10.org/new.htm.

The direct link is http://www.t10.org/ftp/t10/drafts/sat/sat-r08a.pdf.

SAT revision 8a incorporates 963 letter ballot comment resolutions identified with status "Completed" as documented in 06-121r2, "SAT revision 8 letter ballot comment resolution as of SATr08" (http://www.t10.org/ftp/t10/document.06/06-121r2.pdf).

Two proposals listed below incorporate resolutions to most of the remanining letter ballot comments:

There are some 70 letter ballot comments marked with the term "DISCUSS" to be addressed in the July SAT WG that are neither incorporated into the latest draft nor in the two propoposals listed above. Those remaining LB comments relate to the TEST UNIT READY command translation, the READ/WRITE BUFFER commands, the REQUEST SENSE command, the SEND DIAGNOSTIC self-test function, START STOP UNIT (SLEEP, vs STANDBY, vs STANDBY IMMEDIATE), and a handful of scattered comments that are mostly editorial.
 
Bob Sheffield

Reply via email to