This message is from the T13 list server.

So by your (Hale and Jim's) response, it's best to leave the darn drive
alone.  

I do see where some Set Features subcommands may have unsupported Sub
Command Specific values on any given drive, but that is not discernable from
the driver until an error is seen in sending the command.  So, I'll assume
its safe to just proceed when the error occurs.  

Now, error like this on an ATAPI drive should be followed by a Request
Sense, or is there anything else (like a reset) that ATAPI devices need?

Thanks, MKE.


-----Original Message-----
From: Hale Landis [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, December 11, 2001 8:29 PM
To: T13 List Server
Subject: Re: [t13] Error Handling on Set Features Error (corrected
subject)


This message is from the T13 list server.


On Tue, 11 Dec 2001 09:42:05 -0600, Wolford, Jeff wrote:
>Example: Set feature subcommand 06h - Enable power-up in standby
>Two examples (Depends on ID Word 83, bit 5 (Pwr Up supported bit)
>Example 1:
>If ID 83, bit 5 = 1 (supported)
>and the Drive returns an error to a Set Feature sub cmd 06h
>is that really a bad drive.

Yep. As a host person I would say the drive is bad because it really
doesn't support the feature and therefore ID word 83 bit 5 is
incorrect. I can't think of any really valid reason to reject this
command... All it does is enable a feature that really doesn't take
effect until the next power on (if I understand the Power Up in
Standby feature set correctly).

>Q: Does that make it ok for the drive to error that command ?

Of course if it really doesn't support the feature.

>Q: What is the recovery procedure (The question, Michael, Hale
>       and I would like to know)

As a host I would assume the drive doesn't support the feature set
and go on.

>Example 2:
>If ID 83, bit 5 = 0 (not supported)
>and the Drive returns an error to a Set Feature sub cmd 06h
>a) This is NOT a bad drive

Correct.

>b) Not sure if you can say this is definitively a bad host
>       but its getting close -:)

Absolutely nothing wrong with a host trying a command to see if the
device supports it. Nothing in the ATA/ATAPI standards say that a
host SHALL use the ID data to determine what commands are supported.
And I would strongly object to any attempt to make the standards say
this. I don't know of any interface that does not allow the host
(initiator) to try commands to see which are supported.

>Why is this not necessarily a bad host ?

It is not a bad host.

>Maybe the host wants to make sure the drive does not support
>       the command even when it says it does not.

Yep.

>Why... Well features like power-up in standby and security
>       feature can cause a MAJOR service call (i.e. system
>       won't boot).

True. I am not sure what to say. A host (BIOS, OS, etc) that allows
these features to be inappropriately used could certainly find itself
with a drive it can not access at some future time. My opinion is
this: NEVER PURCHASE A DRIVE THAT SUPPORTS THE SECURITY FEATURE SET.
I don't have any experience with the Power Up in Standby feature set,
but if the enable command was given to the drive (by an authorized
part of the BIOS, OS, etc) and the system then failed to tell the
drive to spin up when the next pwoer cycle happens I would not be
happy. Of course that leaves open the possibility that some virus
could issue the enable command (just as a virus might give your drive
some random password). Yes, the Security feature set does have a
protection option... I'm not sure the Power Up in Standby feature set
has such protection option.

>       In the case of power-up in standby, a boot floppy
>       could solve the problem (with the appropriate set
>       feature command to disable power-up in standby).
>       In the case of security, the user has created another
>       brick.

You got it!

>Q: What is the recovery procedure ?

My opinion... Obsolete the Security feature set. Protect the Power Up
in Standby feature set or modify it such that the drive shall spin up
not just on the SET FEATURE command but also on any media access
(just like real Standby mode works in most drives).


***  Hale Landis  *** [EMAIL PROTECTED] ***
*** Niwot, CO USA ***   www.ata-atapi.com   ***


Subscribe/Unsubscribe instructions can be found at www.t13.org.
Subscribe/Unsubscribe instructions can be found at www.t13.org.

Reply via email to