To: "Mark Evans" <Mark.Evans@wdc.com> Cc: owner-t10@t10.org, t10@t10.org Subject: Re: The PPC bit in the LOG SENSE command From: Kevin D Butt <kdbutt@us.ibm.com> Date: Tue, 11 Nov 2008 12:20:21 -0700 X-Message-Number: 9299 Formatted message: HTML-formatted message Mark, I believe the intent is not if a new parameter code is added, but if the value returned by parameter code x changes but the value returned by parameter code y does not change, then a PPC bit set to one specifies that parameter code x should be returned but parameter code y should not be returned. I think this is the intent. The other point you make is related to which I_T nexus this applies to. Since the text does not call out a specific I_T nexus the implication is that it relates to any I_T nexus as you point out. The other possibility is that it could relate to a change in value returned by a parameter code since the last LOG SELECT or LOG SENSE received from the same I_T nexus that issued this LOG SENSE. My suspicion is that the intent would have been the latter - that it relates to this I_T nexus only and the information needs retained on an I_T nexus basis. With that said, my devices do not support a PPC=1 so this comment is purely from reading the spec and guessing what the intent is. Thanks, Kevin D. Butt SCSI & Fibre Channel Architect, Tape Firmware MS 6TYA, 9000 S. Rita Rd., Tucson, AZ 85744 Tel: 520-799-2869 / 520-799-5280 Fax: 520-799-2723 (T/L:321) Email address: kdbutt@us.ibm.com http://www-03.ibm.com/servers/storage/ From: "Mark Evans" <Mark.Evans@wdc.com> To: <t10@t10.org> Date: 11/11/2008 10:57 AM Subject: The PPC bit in the LOG SENSE command Hello all, I was looking at the definition for the PPC bit in the LOG SENSE command in SPC-4. My issue is with when the PPC bit is set to one. List item a in the definition for the bit in SPC-4 reads: A PPC bit set to one specifies that the device server shall return a log page with parameter code values that have changed since the last LOG SELECT or LOG SENSE command. The device server shall return only those parameter codes that are greater than or equal to the contents of the PARAMETER POINTER field in ascending order of parameter codes from the specified log page? These are almost the same words used to define the meaning for the PPC bit set to one in SCSI-2. This definition places no bounds on, ??since the last LOG SELECT or LOG SENSE command??. In addition, as I read it, the only way a parameter code can change is if a new parameter code is added to a log page. To me this all means that, if the following sequence occurs: 1) a LOG SENSE command was received on one I_T nexus 20 years ago; 2) many parameter codes have been added in the interim, along with the occurrence of many power cycles, resets, etc.; 3) a LOG SENSE command from a different I_T nexus was received ten minutes ago; 4) no parameter codes have changed in the last ten minutes; and 5) a LOG SENSE command is received on the first I_T nexus, then no parameter data would be returned. Another odious implication of this definition is that context shall always be maintained identifying any new parameter code that has been added until the device server receives a LOG SENSE command or a LOG SELECT command, again, this could be on ANY I_T nexus. I think that whatever meaning this bit once had was lost a long time ago in a galaxy far, far away. Unless someone can point out something I?m missing, I think the PPC bit should be made obsolete, and I will generate a proposal to do just that. Please feel free to call or send an email to me if you have any additional questions about this stuff. Regards, Mark Evans Western Digital Corporation 5863 Rue Ferrari San Jose, CA 95138 Email: mark.evans@wdc.com Office: 408.363.5257 Fax: 408.363.5139