T10/04-242r1 Voting Results on T10 Letter Ballot 04-241r0 on Forwarding SBC-2 to First Public Review Ballot closed: 2004/08/24 12:00 noon MDT Organization Name S Vote Add'l Info --------------------------------- -------------------- - ---- ---------- Adaptec, Inc. Tim Symons P Yes Agilent Technologies Pat Thaler P No Cmnts Amphenol Interconnect Michael Wingard P Yes Brocade Robert Snively P No Cmnts Cisco Systems, Inc. Claudio DeSanti P Yes CNT David Peterson P Yes Crossroads Systems, Inc. Dexter Anderson P Yes Dallas Semiconductor James A. Lott, Jr. P Yes Dell, Inc. Kevin Marks P Yes EMC Corp. Gary S. Robinson P Yes Emulex Robert H. Nixon P Yes Cmnts ENDL Ralph O. Weber P No Cmnts FCI Douglas Wagner P Yes Foxconn Electronics Elwood Parsons P Abs Cmnts Fujitsu Mike Fitzpatrick P Yes General Dynamics DNV Hewlett Packard Co. Rob Elliott P Yes Cmnts Hitachi Cable Manchester DNV Hitachi Global Storage Tech. Dan Colegrove P Yes IBM Corp. George O. Penokie P No Cmnts Intel Corp. Robert Sheffield P Yes Iomega Corp. David Hawks P Abs Cmnts LSI Logic Corp. John Lohmeyer P Yes Cmnts Maxtor Corp. Mark Evans P No Cmnts Microsoft Corp. Emily Hill P Yes Cmnts Molex Inc. Jay Neer P Yes Panasonic Technologies, Inc Terence J. Nelson P Yes Philips Electronics DNV Pivot3, Inc. Bill Galloway P Yes PMC-Sierra Rachelle Trent P Yes QLogic Corp. DNV Quantum Corp. Jim Jones A Abs Cmnts Seagate Technology Gerald Houlder P No Cmnts Storage Technology Corp. Erich Oetting P Yes Sun Microsystems, Inc. Vit Novak P Yes Cmnts Texas Instruments Paul D. Aloisi P Yes Toshiba Yutaka Arakawa P Yes Tyco Electronics Ashlie Fan P Abs Cmnts UNISYS Ron Mathews P Yes Veritas Software Roger Cummings P Yes Ballot totals: (26:6:4:4=40) 26 Yes 6 No 4 Abstain 4 Organization(s) did not vote 40 Total voting organizations 15 Ballot(s) included comments This 2/3rds majority ballot passed. 26 Yes are more than half the membership eligible to vote minus abstentions [greater than 18] AND 26 Yes are at least 22 (2/3rds of those voting, excluding abstentions [32]) AND 26 Yes are equal to or exceed a quorum [13] Key: P Voter is principal member A Voter is alternate member Abs Abstain vote DNV Organization did not vote Cmnts Comments were included with ballot NoCmnts No comments were included with a vote that requires comments ************************************************************** Comments attached to No ballot from Pat Thaler of Agilent Technologies: 4.15.3.2 top of page 28: It is technically incorrect dictate the number of bits that the CRC generator processes at one time. I can get the same CRC result with a pure serial generator or processing any number of bits at a time. It is not clear whether this text is stating a requirement or merely giving an example. I also object to the use of "word" here as equivalent to two bytes. I don't see any definition of "word" here or in SAM 2. Industry usage on the size of a "word" varies. The Figure 3 diagram indicates that an on byte size user field is to be padded out, but that is not indicated in any requirements text and the Figure is an example. This will cause incompatabilities as some will pad for the calculation and some will not follow that example. ************************************************************** Comments attached to No ballot from Robert Snively of Brocade: 1 (E) Page: xx Location: Introduction Problem Description: "which" s/b "that" Suggested Solution: Make requested correction 2 (E) Page: 6 Location: 3.1.20 Problem Description: "A target action in response to a reset event in which the target port performs the operations described in SAM-3." s/b "A target action in response to a reset event that causes the target port to perform the "hard reset" actions described in SAM-3" Suggested Solution: Make the requested correction Make a similar correction in 3.1.24 3 (E) Page: 10 Location: 4.2 Problem Description: "which" s/b "that" In addition, the text should clarify that it is the logical block address that is 4 or 8 bytes in length by adding the text "in length" at the end of the same sentence. Suggested Solution: Make requested correction 5 (E) Page: 12 Location: 4.8 Problem Description: "defects, which may be supplied by the original manufacturer of the device or medium, that" s/b "defects (supplied by the original manufacturer of the device or medium) that" Suggested Solution: Make requested correction 6 (E) Page: 21 Location: 4.13.1.4 Problem Description: "which" s/b "that" Suggested Solution: Make requested correction 7 (E) Page: 26 Location: 4.15.3.1 Problem Description: In table 7, for the R(X) row, the text "The remainder polynomial, which is of degree less than 16." s/b "The remainder polynomial. The polynomial shall be of degree less than 16." Suggested Solution: Make the requested correction Note that this restriction on the polynomial might better be stated in the text. The same correction needs to be made for P(X). 8 (E) Page: 35 Location: 5.4.1 Problem Description: Rewrite the "which" clause to avoid it. Suggested Solution: Make requested correction 9 (E) Page: 38 Location: 5.4.2.2 Problem Description: "header, which" s/b "header that" Suggested Solution: Make the requested correction A similar change needs to be made just above Table 17. 10 (E) Page: 46 Location: 5.7 Problem Description: "The GROUP NUMBER field specifies the group into which attributes associated with the command should be collected" s/b "The GROUP NUMBER field specifies the group that collects the attributes associated with the command." Suggested Solution: Make requested correction 11 (E) Page: 48 Location: 5.9 Problem Description: "which may address" s/b "that is capable of addressing". Suggested Solution: Make requested correction 12 (E) Page: 65 Location: 5.21 Problem Description: "It is not an error to specify that the logical unit transition to a power condition in which it currently is." s/b "It is not an error to specify that the logical unit transition to its current power condition". Suggested Solution: Make requested correction 13 (E) Page: 95 Location: 5.44 Problem Description: "The LOGICAL BLOCK ADDRESS field specifies the starting LBA of the data on which an XOR operation shall be performed with the data from the medium." s/b "The LOGICAL BLOCK ADDRESS field specifies the starting LBA of the data that shall be XOR'd with the data from the medium." Suggested Solution: Make the requested correction A similar change needs to be made in the next paragraph for TRANSFER LENGTH. 14 (E) Page: 109 Location: 6.3.2.2 Problem Description: "To determine the number of blocks at which the logical unit is currently formatted the application client shall use the READ CAPACITY command" s/b "The READ CAPACITY command shall be used to determine the number of blocks currently formatted on the logical unit." Suggested Solution: Make the requested correction. A similar change is required in the last sentence of the sub-clause. A similar change is required in the last sentence of the first paragraph on page 111 A similar change is required in the last sentence of the last paragraph on page 111 15 (E) Page: 109 Location: 6.3.2.2 Problem Description: "On a MODE SENSE command, the NUMBER OF BLOCKS field indicates the number of logical blocks on the medium to which the BLOCK LENGTH field applies." s/b "On a MODE SENSE command, the NUMBER OF BLOCKS field indicates the number of logical blocks on the medium having the length specified by the BLOCK LENGTH field." Suggested Solution: Make the requested correction. A similar change is required in the second paragraph on page 111. 16 (E) Page: 114 Location: 6.3.3 Problem Description: "The NUMBER OF CACHE SEGMENTS field specifies the number of segments into which the device server shall divide the cache." s/b "The NUMBER OF CACHE SEGMENTS field specifies the number of segments that the device server shall divide the cache into." Suggested Solution: Make requested correction 17 (E) Page: 5 Location: 3.1.4 Problem Description: The definition of byte is inconsistent among the different standards. I have no particular objection to the present definition except that it does not consider the boundarie alignment requirement for the eight contiguous bits. Suggested Solution: I would suggest changing this (and SAM and SPC) to read: "A sequence of eight contiguous bits considered as a unit and aligned on character boundary". 18 (E) Page: 5 Location: 3.1.x Problem Description: Several references are missing. Suggested Solution: Add "See SAM-3" to: 3.1.10, 3.1.11, 3.1.15, Add "See SPC-3" to: 3.1.14, 19 (E) Page: 5 Location: 3.1.9 Problem Description: Abbreviations are included in the definitions sections. Include DLIST, GLIST, CLIST, and PLIST in clause 3.2 only. Do not define them except in the body of the referenced text. Suggested Solution: Make requested correction 20 (T) Page: 5 Location: 3.1.16 Problem Description: XOR is incorrectly defined. I would propose that the following definition be used: "The exclusive-OR function defined for binary arithmetic and logic. The output value is 1 if one and only one of the input values is 1. In this document, the exclusive-OR function is used to describe the operation that creates an output string of bits of length "n" by setting the "i"th bit of the string equal to the logical XOR of the "i"th bit of each of two input strings of bits, also of length "n". By extension, the term is used as an adjective to describe commands that perform this exclusive-OR function. Such operations can be used to create data redundancy that allows recovery of damaged data under certain conditions." Suggested Solution: Make the requested correction. Note that the wording about "does not define the specific polynomial" is false. It must and does define the behavior of the device during execution of the XOR commands. 21 (T) Page: 6 Location: 3.1.26 Problem Description: To use this word, you must also define the word "medium". I believe the intent is allow 3.1.38 and others to implicitly define that. However, the wording used in 3.1.28 is precisely the wording used for "medium" in SPC-3. I would propose changing 3.1.28 using the definition from SPC-3 and realphabetizing the item. That wording is: "medium: A physical entity that stores data in a nonvolatile manner (retained through a power cycle) in accordance with commands processed by the device server." Suggested Solution: Make the requested correction. Note that the word "non-volatile medium" is used only one place in the whole document and its meaning is clear there. Note that the word "volatile medium" is used only one place in the whole document, and it refers to no device of which I am aware. I believe it should be deleted conceptually. 22 (T) Page: 6 Location: 3.1.29 Problem Description: Use the text from SPC-3 instead of this text. Suggested Solution: Make requested correction 23 (T) Page: 6 Location: 3.1.30 Problem Description: Use the text from SPC-3 instead of this text. Suggested Solution: Make requested correction 24 (T) Page: 6 Location: 3.1.33 Problem Description: Change the definition to: "Medium containing data that cannot be changed by a SCSI command. The data is applied to the medium by methods outside the scope of this standard" Suggested Solution: Make requested correction 25 (T) Page: 6 Location: 3.1.35 Problem Description: Replace the definition of "reset event" with the definition from SAM-3 Suggested Solution: Make requested correction 26 (T) Page: 6 Location: 3.1.36 Problem Description: Replace the definition of "sense data" with the definition from SPC-3. Suggested Solution: Make requested correction 27 (T) Page: 7 Location: 3.1.41 Problem Description: Remove this definition. It only applies to two usages, both considered only because they are rendered obsolete. All other usages of the term do not keep older generations of data, but refer to updates of less than a complete stripe of data and can be considered a normal English use of the term. Suggested Solution: Make requested correction 28 (T) Page: 7 Location: 3.1.44 Problem Description: Remove this definition. It is inconsistent with the SPC definition of medium. What is being referred to as a medium is really a storage area for parameters outside the medium. Suggested Solution: Make requested correction 29 (E) Page: 3 Location: 2.1, Table 1 Problem Description: Remove all those not referenced in this document. For all practical purposes, the only ones referenced are ISO, IEC, INCITS, and maybe ANSI, since there is only one approved reference and 5 approved references under development Suggested Solution: Make requested correction 30 (E) Page: 7 Location: 3.2 Problem Description: Remove unreferenced abbreviations kbit and Mbit (which are defined incorrectly anyway). Suggested Solution: Make requested correction 31 (E) Page: 10 Location: 4.1 Problem Description: Delete "and may be a multiple of 512 bytes", since it adds no information. Suggested Solution: Make requested correction 32 (T) Page: 10 Location: 4.1 Problem Description: Delete "and additional information". There is no other SCSI-visible source of additional information, so it should not be mentioned. Suggested Solution: Make the requested correction. The same correction must be made on page 11, clause 4.4, next to last paragraph, last sentence 33 (T) Page: 10 Location: 4.2 Problem Description: The text concerning volatility is not correct. There is no volatile media device defined. There are devices with non-volatile caches, but the medium by definition is non-volatile. Suggested Solution: Delete second sentence of last paragraph. 34 (T) Page: 10 Location: 4.2 Problem Description: The last sentence should be changed to read: "The medium on a device may contain vendor specific information that is not addressable through the LBA. Such data may include defect management data and other device management information." Suggested Solution: Make requested correction 35 (T) Page: 12 Location: 4.6 Problem Description: This is where the non-volatile problem needs to be handled correctly. Change the second and third paragraph to read as follows: "Block devices may save mode parameters and other management information on a non-removable media or in a non-volatile memory. Such devices only need to be initialized once. Those devices without access to a non-removable media or a non-volatile memory may need initialization (possibly including formatting and/or mode parameter initialization) after each logical unit reset or power cycle prior to the processing of read or write operations. Suggested Solution: Make requested correction 36 (T) Page: 12 Location: 4.8 Problem Description: "Some block devices provide the application client control through use of the mode parameters." s/b "Some block devices allow the application client some control of this additional information using mode parameters. Suggested Solution: Make requested correction 37 (E) Page: 13 Location: 4.9 Problem Description: On the 5th paragraph, "store data that is written to the medium at a later time." s/b "store data that is to be written to the medium at a later time." Suggested Solution: Make requested correction 38 (E`) Page: 14 Location: 4.10 Problem Description: "table" should be attached to "3" with a non-breaking space Suggested Solution: Make requested correction 39 (T) Page: 14 Location: 4.10 Problem Description: SAM does not define a "current task state". The task may become the "current task" by beginning to transfer data. It may enter the "enabled task state" by being allowed to begin. I believe what is desired here is "has become the current task for the first time" instead of "has entered the current task state for the first time". That is a measurable time, as opposed to the enabled state, which is outwardly invisible. Suggested Solution: Make requested correction 40 (T) Page: 15 Location: 4.10 Problem Description: Note 4 uses the word operation, not previously defined. I believe the intent here is to warn of the danger of interrupting a sequence of commands with a reservation. The note should either be deleted or rewritten to: "When more than one application client has access to a device server, agreement is required among application clients as to when media is reserved and released. Application clients may interrupt or interfere with each other if such conventions are not followed." Suggested Solution: Make requested correction 41 (E) Page: 16 Location: 4.10, Table 3 Problem Description: The format of the key is very unclear and it should be reformatted. Suggested Solution: Make requested correction 42 (T) Page: 16 Location: 4.11 Problem Description: SPC-3 makes it fairly clear that Sense Keys and Additional Sense Codes are only loosely related. The same ASC/ASCQ may be a HARDWARE ERROR, a RECOVERED ERROR, or a MEDIUM ERROR. As such, I would propose that Table 4 and the two sentences purporting to explain its purpose be deleted. The text of the first paragraph after Table 4 should probably be separated into 3 separate paragraphs describing these 3 error presentations commonly for all commands. Any others that are common to all commands should also be specified here, but most should only be specified in the explanation of the command. Suggested Solution: Make requested correction 43 (E) Page: 17 Location: 4.12.2 Problem Description: First and second paragraphs need to be rewritten as: "The typical application of a direct-access device is a magnetic disk device. The medium is a spinning disk coated with a magnetic material that allows flux changes to be induced and recorded. An actuator positions a read-write head radially across the spinning disk, allowing the device to randomly read or write the information at any radial position. Data is stored by using the write portion of the head to record flux changes and the read portion of the head to read the recorded data. The circular path followed by the read-write head at a particular radius is called a track. The track is divided into sectors each containing blocks of stored data. If there is more than one disk spinning on a single axis and the actuator has a read-write head for each disk surface, the collection of tracks at a particular radius is called a cylinder. Some disks can access information on tracks on the same cylinder more rapidly than they can access information on adjacent tracks on the same disk surface." Suggested Solution: Make the requested change. Note that a lot of this stuff is out of date the way disks are presently implemented. 44 (E) Page: 17 Location: 4.11, Table 5 Problem Description: Question: Is table 5 and exhaustive list? If so, is it short enough and simple enough so that it should be dropped and the information included only in the command descriptions? Suggested Solution: TBD 45 (E) Page: 17 Location: 4.12.2 Problem Description: "A sector may be made up of a header, data, and a trailer" s/b "A sector may contain headers, trailers, data written in logical blocks, and radial servoing information." Suggested Solution: Make requested correction 46 (E) Page: 17 Location: 4.12.2 Problem Description: "The data field begins with a synchronizing field and a data area that contains user data" s/b "The data field contains a synchronizing field and a data area that contains an encoding of user data." Suggested Solution: Make requested correction 47 (E) Page: 17 Location: 4.12.2 Problem Description: Last sentence on page 17 should read: "Disk devices are non-volatile" or alternatively be deleted. Suggested Solution: Make either correction 48 (E) Page: 18 Location: 4.12.2 Problem Description: "though some aspects can be evaluated" s/b "though some aspects may be visible to" Suggested Solution: Make requested correction 49 (T) Page: 18 Location: 4.12.2 Problem Description: The discussion here of READ LONG and WRITE LONG, together with my knowledge of modern disk drives beg that I ask whether this usage of READ LONG and WRITE LONG should be made obsolete. Suggested Solution: Make READ LONG and WRITE LONG obsolete unless drive vendors indicate that their self-test program is inadequate to verify the error detection and correction mechanisms and host computer manufacturers and programmers guarantee that they will use the commands in a standard way to supplement the inadequate self-test programs. 50 (T) Page: 18 Location: 4.12.3 Problem Description: The last sentence says "Memory block devices may store less data than disks or tapes, and may be volatile." There is no marker that describes the volatility behavior of a memory block device. I assume that means that this should be rewritten to read "Memory block devices may store less data than disks or tapes. For compliance with SBC-2, memory block devices shall be non-volatile." Suggested Solution: Make requested correction 51 (T) Page: 24 Location: 4.14.2.6.1 Problem Description: The stopped state "may consume less power" than when Active, Idle, or Standby. Then again, it may consume more power. This seems pretty random to me. I would expect it would drop from "Stopped" to Active, Idle, or Standby using the Power Condition mode page timers if Active, Idle, and/or Standby were a lower power condition. I would then expect it to revert to Stopped when an action at the drive or a Start/Stop command with the proper bit set occurred. That way you won't be stuck in a "stopped" but high power state indefinitely just because someone forgot to emit the proper START/STOP command. Alternatively, require the stopped state to consume no more power than the standby state. Suggested Solution: Make requested correction 52 (T) Page: 25 Location: 4.15.1 Problem Description: Last paragraph should read: "If the logical unit is formatted with protection information and the EMDP bit is set to one in the Disconnect-Reconnect mode page (see SPC-3), then checking of the logical block reference tag within the service delivery subsystem without reference to the modified data pointers and logical block alignments causes false errors when logical blocks are transmitted out of order." Suggested Solution: Make requested correction 53 (E) Page: 29 Location: 4.15.5 Problem Description: "fields are defined" s/b "fields is defined", since the subject is "description". Suggested Solution: Make requested correction 54 (T) Page: 29 Location: 4.16 Problem Description: The grouping function should either be deleted or the following required functionality MUST be defined: a) The support of the grouping function MUST be defined either in a mode page or in the INQUIRY command. b) The maximum number of groups supported MUST be either fixed at 32 OR a parameter in a mode page MUST indicate the maximum number of groups supported. c) The capture setup commands and capture information presentation commands SHOULD be defined. This may not be ready for normative text yet, but an informative annex is important here so that it doesn't develop into series of non-interoperable vendor-pair specific solutions for this function. Suggested Solution: Make the requested change or delete the function. 55 (E) Page: 33 Location: Table 9 Problem Description: Item e, "supported of and only if" s/b "supported if and only if", or in plain English, "supported only if". Suggested Solution: Make the requested correction in Brocade 57. 56 (T) Page: 31 Location: Table 9 Problem Description: SYNCHRONIZE CACHE does not apparently deal directly with protection information. I believe the Protection Information column should have no for both the 10 and 16 byte versions Suggested Solution: Make requested correction 57 (T) Page: 33 Location: Table 9 Problem Description: Items e and f make the referenced commands mandatory if a certain bit is set to one. In fact, the commands happen to be mandatory for SMC-2, but not for SCC-2. The terminology should be for note e: If the SCCS bit is set to one in the standard INQUIRY data, these commands shall be supported as required by SCC-2. If the SCCS bit is set to zero, these commands shall not be supported. and for note f should be: If the MCHGR bit is set to one in the standard INQUIRY DATA, these commands shall be supported as specified by SMC-2. If the MCHGR bit is set to zero, thse commands shall not be supported. Suggested Solution: Make requested correction 58 (T) Page: 31 Location: Table 9 Problem Description: REPORT LUNS should be mandatory for direct access devices. Suggested Solution: Make requested correction 59 (T) Page: 34 Location: Table 10/11/12 Problem Description: The values for reserved operation code/service action codes are "reserved for direct-access devices". In fact, they are reserved in their respective documents for the future use of any type of device. Change the text to read simply "Reserved" Suggested Solution: Make requested correction 60 (T) Page: 34 Location: 5.3 Problem Description: The SERVICE ACTION IN (16) operation code is supposedly defined in SPC-3 according to the reference. In fact, there is no normative definition of the 9E that I can find either there or in any of the other expected fundemental standards. Where is it defined? Suggested Solution: Reference the correct document for the definition of SERVICE ACTION IN (16). If there is not such document, now is the chance to fix SPC-3 so that it is defined. 61 (T) Page: 43 Location: 5.4.2.4.4 Problem Description: Propose to obsolete Bytes from index format address descriptor. Modern disks do not use it. Suggested Solution: Make requested correction 62 (T) Page: 44 Location: 5.4.2.4.5 Problem Description: Propose to obsolete Physical sectgor format address descriptor. Modern disks do not use it. Suggested Solution: Make the requested correction. As a side effect, the "Translate Address" diagnostic pages shall also be made obsolete. 63 (T) Page: 30 Location: Table 9 Problem Description: Propose to create a new type of "Type" called "D" for READ (6) and WRITE (6). The D type should be called "depricated". The depricated type should be mandatory for a target to implement in compliance with SBC-2, but should be obsolete for an application client to implement in compliance with SBC-2. That will allow READ (6) and WRITE (6) to be made obsolete in the next generation of the standard. Suggested Solution: Make requested correction 64 (E) Page: 50 Location: Table 33 Problem Description: The table needs to be restructured so that it is properly terminated on a page by page basis. That probably requires dividing it into one or two "values" per page. The title of the "value" column s/b "RDPROTECT Value". The term "Shall not" for the Extended INQUIRY Data VPD page column is unclear. I assume the proper wording is something like "GRD_CHK = X", "APP_CHK = X", or "REF_CHK = X", depending on the row. Suggested Solution: Make requested correction 65 (T) Page: 52 Location: note b Problem Description: Unsupported functions were previously reserved bits, and therefore should be ignored by the target receiving such a command. The READ should proceed normally. The text should be changed to read: "If the logical unit does not support protection informaton, the RDPROTECT field is ignored and the command executes normally." Suggested Solution: Make requested correction 66 (T) Page: 52 Location: note i Problem Description: The restrictions on the execution of the commands when RTO_EN is set to one are overly restrictive. The text: "If the RTO_EN bit is set to one, the .. CODE." s/b replaced with: "If the RTO_EN bit is set to one, the device server shall assume for checking purposes that the LOGICAL BLOCK REFERENCE TAG for each block has been set to the value that it would have been set to if the RTO_EN bit were zero and perform the desired checking, if any." Suggested Solution: Make the requested correction. Similar corrections are required for the write commands. 67 (T) Page: 55 Location: 5.13 Problem Description: The restrictions on the execution of the READ (32) command are overly restrictive. The text: "If the RTO_EN ... the command." s/b replaced with: "If the RTO_EN bit is set to zero in the long read capacity data (see 5.15), the device server shall perform checking of the data protection fields as required by the RDPROTECT field. The device server shall terminate the command with CHECK CONDITION status with a sense key of ABORTED COMMAND and indications of the appropriate protection check if the protection information read by the device server conflicts with the information specified by the parameters in the READ (32) command." Suggested Solution: Make the requested correction. Similar corrections are required for the write commands. Additional late comments may be developed by Brocade and will be numbered higher for convenience. Late Brocade Comments: 68 (E) Page: 56 Location: 5.14, Table 39 Problem Description: The concept of "short" and "long" is not reflected in the name of the command. Wherever "short read capacity data" is used, the term "READ CAPACITY (10) data" should be used. Wherever "long read capacity data" is used, the term "READ CAPACITY (16) data" should be used. Suggested Solution: Make requested corrections. 69 (T) Page: 12 Location: 4.7 Problem Description: The definition of "implicit head of queue" seems to me to be overreaching. I believe that simple queueing should allow for implicit head of queue, but it concerns me that ORDERED also allows for it. I would prefer to see ORDERED not allowed for those commands if it has no meaning. Alternatively, ORDERED should over-ride the implicit head of queue. Suggested Solution: Select proper solution and implement it. Note that this effects SPC-3 also. 70 (T) Page: 36 Location: 5.4.1 Problem Description: The FMTPINFO bit and RTO_REQ bit are dependent. When processing them in the READ CAPACITY command, they are processed independently. In order to allow the READ CAPACITY command to operate correctly, the RTO_REQ bit should not be ignored when the FMTPINFO bit is set to zero, but rather, if the INQUIRY command indicates the data protection is supported, and the FMTPINFO bit is set to zero, the RTO_REQ bit shall be set to zero. If it is set to one, it should cause an INVALID PARAMETERS IN COMMAND check condition. Suggested Solution: Make requested corrections. 71 (T) Page: 60 Location: 5.16 Problem Description: Brocade 61 and 62 suggest making obsolete two of the formats. It is claimed that those formats are the ones people need for defect management in clause 5.16. I propose making obsolete the READ DEFECT DATA commands. Suggested Solution: Make requested corrections. 72 (T) Page: 61 Location: 5.18 Problem Description: The text: "The most recent data written, or to be written, in the addressed field shall be returned." implies that cached data may be returned, which makes the command useless. The text should be changed to read: "The most recent data written on the recording media at the addressed field shall be returned. If the data for the specified logical block has not yet been stored on the recording media, it shall be flushed from the cache and stored on the media before the information is returned from the media." Suggested Solution: Make requested corrections. 73 (E) Page: 63 Location: 5.20 Problem Description: The term "default value of FFF..." really is not correct. This is a marker value that indicates an initialized value that has not been written. I would proposed that the term "initialized value of FFF..." be used instead for all cases describing the protection information. Suggested Solution: Make the requested corrections in all applicable locations. 74 (E) Page: 69 Location: Tables 58-60 Problem Description: The problems previously mentioned, including formatting and the use of "shall not" must also be corrected here. Suggested Solution: Make requested corrections. 75 (E) Page: 71 Location: Note a Problem Description: Note a, while stated correctly, gives the impression that the indicated failure will always occur, though it only occurs on cases other than case 0. I propose the sentence be pre-fixed with "For these cases, " Suggested Solution: Make the requested corrections in all applicable locations. 76 (T) Page: 71 Location: Note g Problem Description: This case is interesting and requires careful thought. The logical block application tag field of FFFFh is symptomatic of a block that has not been changed since formatting, but not much else. It may be that there was a choice to set it to FFFFh by the application client. Why should checking be disabled unless all the fields are set to FFFFh? Suggested Solution: Please verify that this was the intended action and indicate why? If it needs to be corrected here, there are a number of other tables to also be corrected. 77 (T) Page: 84 Location: 5.32 Problem Description: The restriction of the RTO_EN to one while using the 32-byte command appears to be unnecessary. See Brocade 67. Suggested Solution: Make the corresponding changes as in Brocade 67. 78 (T) Page: 85 Location: 5.33 Problem Description: The restrictions requiring the RTO_EN bit to be 0 for the following commands are overly restrictive, as described in Brocade 66. Note that for the XDWRITE commands, performing the XDWRITE with no checking is the only way to obtain the necessary XOR function for the case where RTO_EN is equal 1. WRITE AND VERIFY (10), WRITE AND VERIFY (12), WRITE AND VERIFY (16), WRITE SAME (10), WRITE SAME (16), XDWRITE (10), XDWRITE (32), XDWRITEREAD (10), XDWRITEREAD (32) Note that for the XDWRITE and XDWRITEREAD commands, performing the XDWRITE with no checking is the only way to obtain the necessary XOR function for the case where RTO_EN is equal 1 Suggested Solution: Make the corresponding changes as in Brocade 66. 79 (T) Page: 87 Location: 5.36 Problem Description: The restriction of the RTO_EN to one while using the 32-byte command appears to be unnecessary. See Brocade 67. Suggested Solution: Make the corresponding changes as in Brocade 67. 80 (T) Page: 92 Location: 5.41 Problem Description: The restriction of the RTO_EN to one while using the 32-byte command appears to be unnecessary. See Brocade 67. Suggested Solution: Make the corresponding changes as in Brocade 67. 81 (T) Page: 93 Location: 5.42 Problem Description: The statement: "If the XORPINFO bit is set to one and the device server does not support protection information..." violates the conventions for reserved fields. The device server is not required to check for this condition, since XORPINFO was previously a reserved bit. The command should, under these conditions, be performed exactly as if the XORPINFO was zero. Note that this does not apply to XDWRITEREAD commands, since they were not defined in a previous standard. Suggested Solution: Make requested corrections. 82 (T) Page: 94 Location: 5.43 Problem Description: See Brocade 81. The same is true for XDREAD (32) if it was contained in SBC. Suggested Solution: Make requested corrections. 83 (T) Page: 99 Location: 5.4.8 Problem Description: See Brocade 81. The same is true for XPWRITE (32) if it was contained in SBC. Suggested Solution: Make requested corrections. 84 (T) Page: 101 Location: 6.1.1 Problem Description: Are we ready to get rid of the "translate address output/input diagnostic pages? If so, they should be made obsolete at this time. Suggested Solution: Make requested corrections. 85 (E) Page: 105 Location: 6.2.2 Problem Description: I assume this uses the format from table 186 from SPC-3, but this is not explicitly made clear. Could you please provide a reference indicating something like "See the log page structure and page codes for all device types specified in SPC-3." The structure does not specify a field size for each counter, which implies that it is vendor specific. Suggested Solution: Make requested corrections. 86 (T) Page: 107 Location: Table 101 Problem Description: The WP bit is specified here, but it is indicated that it cannot be set with a MODE SELECT command. How then is it set through the SCSI interface? I did not find an SPC-2 mechanism for setting it and I don't believe a hardware-only mechanism is appropriate. That leaves only here to set it. I propose that the first sentence after Table 101 be corrected to indicate that WP is defined and controlled for disk drives by this parameter. Suggested Solution: Make requested corrections. ************************************************************** Comments attached to Yes ballot from Robert H. Nixon of Emulex: Emulex #1 PDF Page 1 Page 1 Figure 1: In the leftmost box, 1477 s/b/ 14776 Emulex #2 PDF Page 1 page 1 clause 1: "provide the following" breaks grammatical flow to the items of the following list. Delete it. Emulex #3 PDF Page 1 Unmarked set by bnixon Emulex #4 PDF Page 1 Page 1 Clause 1: "It indicates the applicability of a standard to the implementation of a given transport" does not seem to be relevant to a command set standard. I'm not convinced that it is even true. Delete it. Emulex #5 PDF Page 3 Page 3 subclause 2.2: For reliable location of documents at the ANSI and INCITS stores, change "ANSI NCITS.318:1998" to "ANSI INCITS 318-1998" Emulex #6 PDF Page 5 Page 5 subclause 3.1.17: The definition of extent uses the ambiguous concept of "continuous logical blocks". Change "continuous logical blocks" to "logical blocks occupying contiguous logical block addresses" Emulex #7 PDF Page 6 Page 6 subclause 3.1.39: SACL is referenced but is not in the abbreviations list. (Yes, I know it is within the next entry in the glossary, but only because I was reading serially through the spec 8-) Emulex #8 PDF Page 8 Page 8 subclause 3.3.10:The equivalence to reserved in the definition of restricted would force conflict, as this standard would require setting to zero a value that another standard might require to be nonzero. Change "A restricted bit, byte, word, or field shall be treated as a reserved bit, byte, word or field for the purposes of the requirements defined in this standard" to "A restricted bit, byte, word or field shall be set to zero, or in accordance with another SCSI standard. Recipients are not required by this standard to check restricted bits, bytes, words or fields for zero values" Emulex #9 PDF Page 10 Page 10 subclause 4.1: in the second paragraph, "In addition," doesn't add anything. Change "In addition, a logical block length..." to "A logical block length..." Emulex #10 PDF Page 11 Page 11 subclause 4.4: in the third line of the fourth regular paragraph, an s/b and Emulex #11 PDF Page 14 Page 14 subclause 4.9: The last sentence of the paragraph fragment at the top of page 14 has too many consecutive verbs and not enough meaning. Change "The FUA_NV bit specifies allows the device server" to "The FUA_NV bit specifies whether the device server is allowed" Emulex #12 PDF Page 15 Page 15 table 3: Why are the READ DEFECT DATA and READ LONG commands conflicted by Write type reservations? Emulex #13 PDF Page 17 Page 17 subclause 4.12.2: In the second paragraph from the bottom of the page, "user" should be "application client". Users, in the common sense of the term, are people, and can't issue SCSI commands. Emulex #14 PDF Page 20 Page 20 subclause 4.13.1.2.3: In the fifth line, e.g s/b i.e Emulex #15 PDF Page 20 Page 20 subclause 4.13.1.2.3: The last three sentences appear to be repeating the rest of the subclause. Delete the last three sentences. Emulex #16 PDF Page 26 Page 26 subclause 4.16.2: In the last paragraph, the wording strongly suggests the LOGICAL BLOCK REFERENCE TAG appears in the application client data buffer. In the second sentence change "contain" to "have a LOGICAL BLOCK REFERENCE TAG value equal to", in the third sentence change "contain" to "have", and in the fourth sentence change "contains" to "has". For consistency. (part of prior comment) Change "contain" to "have" Emulex #18 PDF Page 26 (part of prior comment) change "contains" to "has" Emulex #19 PDF Page 60 Page 60 subclause 5.16: I can't make sense of the first sentence of the second paragraph on page 60: how can the number of address descriptors the SCSI device has assigned contain a value that is insufficient to transfer all of the address descriptors? Isn't the "number ... assigned" the same as "all"? Also, the third sentence seems to contradict note 16. ************************************************************** Comments attached to No ballot from Ralph O. Weber of ENDL: ENDL 1 PDF pg 5, pg v SPI-4 was dedicated to Gene Milligan. But it appears that the practice ended with SPI-5. Is it appropriate to reinstate the practice now? Probably not. ENDL 2 PDF pg 5, pg v <> It seems unlikely the Habitat for Humanity has maintained a memorial account for Gene Milligan for all the years that have passed since his passing in 2001. If the dedication is kept, this sentence should be removed. ENDL 3 PDF pg 6, pg vi Remove Revision Information before submitting dpANS to Public Review. ENDL 4 PDF pg 11, pg xi Remove change bars before submitting dpANS to Public Review. ENDL 5 PDF pg 11, pg xi, Contents Per ISO style guide, the table of contents should use only two levels of indenting, Clauses and Sub-Clauses. ENDL 6 PDF pg 21, pg xxi, Introduction <> [s/b] <> Unlike SBC, there is only one device-type model in SBC-2. ENDL 7 PDF pg 22, pg 1, 1 Scope <> I cannot find any clauses in this standard that apply to devices other than SCSI block device. This needs rewording. ENDL 8 PDF pg 22, pg 1, 1 Scope <> [s/b] < ENDL 9 PDF pg 22, pg 1, 1 Scope <> What is the difference between this list entry and the previous one <>? The second, redundant entry should be deleted. ENDL 10 PDF pg 22, pg 1, 1 Scope <> Since there is only one type of SCSI block device defined by this standard, this list entry should be deleted. ENDL 11 PDF pg 24, pg 3, 2.1 Normative references overview <> [s/b] <> ENDL 12 PDF pg 26, pg 5, 3.1 Definitions Add <> as the last sentence in the definitions of data-in buffer, data-out buffer, and domain. ENDL 13 PDF pg 26, pg 5, 3.1 Definitions Add a glossary entry for error correcting code (ECC) ENDL 14 Technical PDF pg 26, pg 5, 3.1.16 exclusive-or (XOR) <> ??? How does a polynomial relate to an XOR function? Delete <> ENDL 15 PDF pg 26, pg 5, 3.1.16 exclusive-or (XOR) Several changes are proposed to use 'extent' as defined here. If the proposed changes are not made, then the glossary entry for 'extent' should be deleted because it is not referenced anywhere in this standard (as currently written). ENDL 16 PDF pg 27, pg 6, 3.1 Definitions Add glossary entry for I_T nexus loss event. ENDL 17 PDF pg 27, pg 6, 3.1 Definitions <<3.1.26 media: Plural of medium.>> There is no glossary entry for medium, meaning that medium is used in its normal English usage. It would seem that media is likewise used in its normal English usage. Remove this unnecessary glossary entry. ENDL 18 PDF pg 27, pg 6, 3.1.29 power cycle <> [s/b] <> ENDL 19 PDF pg 27, pg 6, 3.1 Definitions <<3.1.33 read-only medium: Medium that is not capable of being changed. The medium contains data prepared in a manner not defined by this standard.>> The term read-only medium is not used anywhere in this standard. This glossary entry should be removed. ENDL 20 PDF pg 27, pg 6, 3.1.35 reset event The definition of reset event is not consistent with SAM-3. Update the definition to be consistent with SAM-3, probably by emulating the definition of logical unit reset event. ENDL 21 PDF pg 27, pg 6, 3.1.36 sense data <> [s/b] <<<> ENDL 22 PDF pg 27, pg 6, 3.1 Definitions <<3.1.40 storage array conversion layer (SACL): Converts input logical unit numbers to output logical unit numbers and may convert input LBAs to output LBAs. See SCC-2.>> Neither storage array conversion layer nor SACL are used anywhere in this standard. This glossary entry should be deleted. ENDL 23 Technical PDF pg 28, pg 7, 3.1 Definitions <<3.1.41 update: To write new data to a logical block without destroying the previous data. After a logical block has been updated, a normal read returns the most recent generation of the data. Earlier generations are still available after the update.>> This definition appears to be specific to optical memory devices which are obsolete in this standard. This definition should be deleted. ENDL 24 Technical PDF pg 28, pg 7, 3.1 Definitions Add either a glossary entry or keyword definition for 'vendor specific'. ENDL 25 PDF pg 28, pg 7, 3.2 Symbols and abbreviations <> [s/b] <> ENDL 26 PDF pg 28, pg 7, 3.2 Symbols and abbreviations <> ID is not used anywhere in this standard. Delete this acronym. ENDL 27 PDF pg 28, pg 7, 3.2 Symbols and abbreviations kbit is not used anywhere in this standard. Delete this acronym. ENDL 28 PDF pg 28, pg 7, 3.2 Symbols and abbreviations Mbit is not used anywhere in this standard. Delete this acronym. ENDL 29 PDF pg 28, pg 7, 3.2 Symbols and abbreviations <> SACL is not used anywhere in this standard. This acronym should be deleted. ENDL 30 PDF pg 30, pg 9, 3.4 Conventions <> All instances of 'increased readability found in this standard use spaces, not underscores. If maintenance of the underscore convention is desired, deviations from it have been marked in other comments. Otherwise, it might be easier to change the convention. ENDL 31 PDF pg 31, pg 10, 4.1 General <> [s/b] <> ENDL 32 PDF pg 31, pg 10, 4.1 General <> [s/b] <> ENDL 33 PDF pg 31, pg 10, 4.1 General <> [s/b] <> ENDL 34 PDF pg 31, pg 10, 4.1 General <> [s/b] <> ENDL 35 PDF pg 31, pg 10, 4.1 General <> This sentence repeats what was said in the previous paragraph. It should be deleted. ENDL 36 Technical PDF pg 31, pg 10, 4.1 General <> Is it really true that different logical blocks can have different block lengths (as opposed to different physical block lengths)? If so, is this under the control of the application client via a FORMAT UNIT command? 6.3.1 says that there is only one mode descriptor for the entire logical unit, so it seems unlikely that the cited statements are true. Probably, these two sentences should be deleted. If not, they should be changed to <> ENDL 37 PDF pg 31, pg 10, 4.1 General Start a new paragraph for <> ENDL 38 PDF pg 31, pg 10, 4.2 Direct-access device type model overview <> This sentence is unnecessary because the use of e.g. in the previous sentence clearly indicates that the cited commands are just examples. This sentence should be deleted. ENDL 39 PDF pg 31, pg 10, 4.2 Direct-access device type model overview <> [s/b] <> for consistence with the description of read operations. ENDL 40 PDF pg 31, pg 10, 4.2 Direct-access device type model overview <> [s/b] <> ENDL 41 PDF pg 31, pg 10, 4.2 Direct-access device type model overview <> [s/b] <> ENDL 42 PDF pg 31, pg 10, 4.3.1 Removable medium overview <> This sentence should be moved to a paragraph by itself, preferably immediately preceding this paragraph in which it currently appears. ENDL 43 PDF pg 32, pg 11, 4.3.2 Removable medium with an attached medium changer <> [s/b] <> ENDL 44 PDF pg 32, pg 11, 4.3.2 Removable medium with an attached medium changer <> [s/b] <> ENDL 45 PDF pg 32, pg 11, 4.3.2 Removable medium with an attached medium changer <> [s/b] <> ENDL 46 PDF pg 32, pg 11, 4.4 Logical blocks <> [s/b] <> ENDL 47 PDF pg 32, pg 11, 4.4 Logical blocks <> [s/b] <> ENDL 48 PDF pg 32, pg 11, 4.4 Logical blocks <> [s/b] <> ENDL 49 Technical PDF pg 32, pg 11, 4.4 Logical blocks <> [s/b] <> ENDL 50 PDF pg 32, pg 11, 4.4 Logical blocks <> [s/b] <> ENDL 51 PDF pg 32, pg 11, 4.4 Logical blocks <> [s/b] <> ENDL 52 PDF pg 32, pg 11, 4.5 Ready state <> [s/b] <> ENDL 53 PDF pg 32, pg 11, 4.5 Ready state <> [s/b] <> ENDL 54 PDF pg 33, pg 12, 4.5 Ready state <> [s/b] <> ENDL 55 Technical PDF pg 33, pg 12, 4.6 Initialization <> [s/b] <> ENDL 56 PDF pg 33, pg 12, 4.6 Initialization <> [s/b] <> ENDL 57 PDF pg 33, pg 12, 4.6 Initialization <> [s/b] <> ENDL 58 PDF pg 33, pg 12, 4.7 Implicit HEAD OF QUEUE command processing <> [s/b] <> ENDL 59 PDF pg 33, pg 12, 4.8 Medium defects <> [s/b] <> ENDL 60 Technical PDF pg 33, pg 12, 4.8 Medium defects <> Control of what? Mode parameters in what mode pages? Add specific details and cross references. ENDL 61 PDF pg 33, pg 12, 4.8 Medium defects <> [s/b] <> ENDL 62 PDF pg 33, pg 12, 4.8 Medium defects <> This sentence is redundant with the paragraph that precedes it and should be deleted. ENDL 63 PDF pg 33, pg 12, 4.8 Medium defects <> [s/b] <> ENDL 64 PDF pg 33, pg 12, 4.8 Medium defects <> [s/b] <> ENDL 65 PDF pg 33, pg 12, 4.8 Medium defects <> [s/b] <> ENDL 66 PDF pg 34, pg 13, 4.8 Medium defects <> this statement belongs in 5.4.2.2, not in the model. ENDL 67 PDF pg 34, pg 13, 4.8 Medium defects <> [s/b] <> ENDL 68 PDF pg 34, pg 13, 4.8 Medium defects <> [s/b] <> ENDL 69 PDF pg 34, pg 13, 4.8 Medium defects <> [s/b] <> ENDL 70 PDF pg 34, pg 13, 4.9 Cache memory <> [s/b] <> ENDL 71 PDF pg 34, pg 13, 4.9 Cache memory Delete <> ENDL 72 PDF pg 34, pg 13, 4.9 Cache memory <> [s/b] <> ENDL 73 PDF pg 34, pg 13, 4.9 Cache memory The 3rd and 4th paragraphs in this subclause should be the 2nd and 3rd paragraphs. ENDL 74 PDF pg 34, pg 13, 4.9 Cache memory <> [s/b] <> ENDL 75 PDF pg 34, pg 13, 4.9 Cache memory <> [s/b] <> ENDL 76 PDF pg 34, pg 13, 4.9 Cache memory <> [s/b] <> [twice in this paragraph] ENDL 77 PDF pg 34, pg 13, 4.9 Cache memory <> [s/b] <> ENDL 78 PDF pg 35, pg 14, 4.9 Cache memory <> This topic belongs in a separate paragraph and the description should be as complete as that given for the FUA bit. ENDL 79 PDF pg 35, pg 14, 4.9 Cache memory What happens when the DPO and FUA_NV bits are both set to one? What happens when the DPO, FUA, AND FUA_NV bits are all set to one? The model is not complete. ENDL 80 PDF pg 35, pg 14, 4.9 Cache memory <> [s/b] <> ENDL 81 Technical PDF pg 35, pg 14, 4.9 Cache memory <> This sentence seems unrelated to the topic being discussed in this paragraph. Since the SYNCHRONIZE CACHE command is covered later in this subclause, this sentence should be deleted. ENDL 82 PDF pg 35, pg 14, 4.9 Cache memory <> [s/b] <> ENDL 83 PDF pg 35, pg 14, 4.9 Cache memory <> [s/b] <> to use the defined term [twice in the a,b,c list] ENDL 84 PDF pg 35, pg 14, 4.9 Cache memory <> [s/b] <> ENDL 85 PDF pg 35, pg 14, 4.9 Cache memory <> [s/b] <> ENDL 86 PDF pg 35, pg 14, 4.9 Cache memory <> [s/b] <> ENDL 87 PDF pg 35, pg 14, 4.9 Cache memory <> [s/b] <> ENDL 88 PDF pg 35, pg 14, 4.9 Cache memory <> [s/b] <> ENDL 89 PDF pg 35, pg 14, 4.9 Cache memory <> [s/b] <> ENDL 90 PDF pg 35, pg 14, 4.9 Cache memory <<(see 6.3.3) writeable by the MODE SELECT command>> [s/b] <<(see 6.3.3), writeable by the MODE SELECT command,>> [i.e., add two commas] ENDL 91 PDF pg 35, pg 14, 4.9 Cache memory <> [s/b] <> ENDL 92 PDF pg 35, pg 14, 4.10 Reservations <> [s/b] <> [2 times in this paragraph] ENDL 93 PDF pg 35, pg 14, 4.10 Reservations <> This statement belongs in clause 1, not here. ENDL 94 PDF pg 36, pg 15, 4.10 Reservations <> The third paragraph from the bottom of the previous page explicitly says that this note is false. Delete the note. ENDL 95 PDF pg 36, pg 15, 4.10 Reservations It the table 3 heading, <> [s/b] <> [4 times ENDL 96 PDF pg 36, pg 15, 4.10 Reservations Since only persistent reservations are covered by this standard, the use of <<[B]>> should be removed from the table 3 heading and the key. ENDL 97 PDF pg 36, pg 15, 4.10 Reservations The ERASE (10)/(12) commands are not defined by this standard. Remove the ERASE (10)/(12) commands from table 3. ENDL 98 PDF pg 36, pg 15, 4.10 Reservations The MEDIUM SCAN command is not defined by this standard. Remove the MEDIUM SCAN command from table 3. ENDL 99 PDF pg 36, pg 15, 4.10 Reservations The READ GENERATION command is not defined by this standard. Remove the READ GENERATION command from table 3. ENDL 100 PDF pg 36, pg 15, 4.10 Reservations The READ UPDATED BLOCK command is not defined by this standard. Remove the READ UPDATED BLOCK command from table 3. ENDL 101 PDF pg 36, pg 15, 4.10 Reservations The UPDATE BLOCK command is not defined by this standard. Remove the UPDATE BLOCK command from table 3. ENDL 102 PDF pg 36, pg 15, 4.10 Reservations There should be a double line at the bottom of the first page of table 3. ENDL 103 PDF pg 37, pg 16, 4.10 Reservations <> [s/b] <> [4 times in the table 3 footnotes] ENDL 104 PDF pg 37, pg 16, 4.10 Reservations Per the ISO style guide, the table 3 key must appear every page on which the table appears (i.e., it needs to be a table footer row) ENDL 105 PDF pg 37, pg 16, 4.11 Error reporting <> [s/b] <> ENDL 106 PDF pg 37, pg 16, 4.11 Error reporting <> [s/b] <> ENDL 107 PDF pg 37, pg 16, 4.11 Error reporting <> [s/b] <> ENDL 108 PDF pg 38, pg 17, 4.11 Error reporting <> [s/b] <> ENDL 109 PDF pg 38, pg 17 <<4.12 Examples>> [s/b] <<4.12 Direct-access device examples>> ENDL 110 PDF pg 38, pg 17 <<4.12.1 Examples overview>> [s/b] <<4.12.1 Overview>> ENDL 111 PDF pg 38, pg 17, 4.12.2 Rotating media <> [s/b] <> ENDL 112 PDF pg 38, pg 17, 4.12.2 Rotating media <> [s/b] <> ENDL 113 PDF pg 38, pg 17, 4.12.2 Rotating media <> [s/b] <> ENDL 114 PDF pg 38, pg 17, 4.12.2 Rotating media <> [s/b] <> ENDL 115 PDF pg 38, pg 17, 4.12.2 Rotating media <> This information is covered better in 4.6. Delete everything in this paragraph except the last sentence. ENDL 116 PDF pg 39, pg 18, 4.12.2 Rotating media <> [s/b] <> ENDL 117 PDF pg 39, pg 18, 4.12.2 Rotating media <> Is the discussion of sectors and tracks appropriate in SBC-2? This sentence can be deleted because its concepts are covered in 4.8. ENDL 118 PDF pg 39, pg 18, 4.12.2 Rotating media <> [s/b] <> ENDL 119 PDF pg 39, pg 18, 4.12.2 Rotating media <> [s/b] <> ENDL 120 PDF pg 39, pg 18, 4.12.2 Rotating media <<'mirror copies.'>> [s/b] <> ENDL 121 Technical PDF pg 39, pg 18, 4.12.2 Rotating media Should the relationship between rotating media and cache memory be discussed? ENDL 122 PDF pg 39, pg 18, 4.12.3 Memory media <> [s/b] <> ENDL 123 Technical PDF pg 39, pg 18, 4.12.3 Memory media When is memory media ready? This is explained for rotating media, why not for memory media? ENDL 124 Technical PDF pg 39, pg 18, 4.12.3 Memory media Is memory media removable? This is discussed for rotating media, why not for memory media? ENDL 125 Technical PDF pg 39, pg 18, 4.12.3 Memory media How does memory media manage defects? This is summarized for rotating media, why not for memory media? ENDL 126 Technical PDF pg 39, pg 18, 4.12.3 Memory media Does memory media use cache memory? ENDL 127 PDF pg 39, pg 18, 4.13.1 Model for XOR commands overview <> [s/b] <> ENDL 128 PDF pg 39, pg 18, 4.13.1 Model for XOR commands overview, Global in this subclause <> [s/b] <> & <> [s/b] <> [3 instances] ENDL 129 PDF pg 39, pg 18, 4.13.1 Model for XOR commands overview <> [s/b] <> ENDL 130 PDF pg 39, pg 18, 4.13.1 Model for XOR commands overview <> [s/b] <> ENDL 131 PDF pg 39, pg 18, 4.13.1 Model for XOR commands overview <> [s/b] <> ENDL 132 PDF pg 39, pg 18, 4.13.1 Model for XOR commands overview <> [s/b] <> Post letter ballot SAM-3 will no longer define interconnect. ENDL 133 PDF pg 39, pg 18, 4.13.1 Model for XOR commands overview <> Should begin as a new paragraph and [s/b] <> ENDL 134 PDF pg 39, pg 18, 4.13.1 Model for XOR commands overview <> [s/b] <> Alternatively, specific details may be added describing 'the following operating modes'. ENDL 135 PDF pg 39, pg 18 There is a 4.13.1 but no 4.13.2. Change from: 4.13 Model for XOR commands 4.13.1 Model for XOR commands overview 4.13.1.1 Storage array controller supervised XOR operations 4.13.1.1.1 Storage array controller supervised XOR operations overview 4.13.1.1.2 Update write operation (storage array controller supervised) 4.13.1.1.3 Regenerate operation (storage array controller supervised) 4.13.1.1.4 Rebuild operation (storage array controller supervised) 4.13.1.2 Additional array subsystem considerations 4.13.1.2.1 Additional array subsystem considerations overview 4.13.1.2.2 Buffer full status handling 4.13.1.2.3 Access to an inconsistent stripe 4.13.1.3 Error handling considerations 4.13.1.3.1 Error handling considerations overview 4.13.1.3.2 Primary errors - errors resulting directly from the primary command 4.13.1.4 XOR data retention requirements to: 4.13 Model for XOR commands 4.13.1 Model for XOR commands overview 4.13.2 Storage array controller supervised XOR operations 4.13.2.1 Storage array controller supervised XOR operations overview 4.13.2.2 Update write operation 4.13.2.3 Regenerate operation 4.13.2.4 Rebuild operation 4.13.3 Additional array subsystem considerations 4.13.3.1 Additional array subsystem considerations overview 4.13.3.2 Buffer full status handling 4.13.3.3 Access to an inconsistent stripe 4.13.4 Error handling considerations 4.13.4.1 Error handling considerations overview 4.13.4.2 Primary errors - errors resulting directly from the primary command 4.13.5 XOR data retention requirements ENDL 136 PDF pg 39, pg 18, 4.13.1.1.1 Storage array controller supervised XOR operations overview <> [s/b] <> ENDL 137 PDF pg 39, pg 18, 4.13.1.1.1 Storage array controller supervised XOR operations overview <> explains nothing and should be deleted. ENDL 138 PDF pg 39, pg 18, 4.13.1.1.1 Storage array controller supervised XOR operations overview <> [s/b] <> ENDL 139 PDF pg 39, pg 18, 4.13.1.1.1 Storage array controller supervised XOR operations overview <> [s/b] <> ENDL 140 PDF pg 40, pg 19, 4.13.1.1.2 Update write operation <<(storage array controller supervised)>> adds no value to this (and two other subclause headings and be deleted [3 times]. ENDL 141 PDF pg 40, pg 19, 4.13.1.1.2 Update write operation <> [s/b] <> ENDL 142 PDF pg 40, pg 19, 4.13.1.1.2 Update write operation <> [s/b] <> ENDL 143 PDF pg 40, pg 19, 4.13.1.1.3 Regenerate operation <> [s/b] <> ENDL 144 PDF pg 40, pg 19, 4.13.1.1.3 Regenerate operation <> [s/b] <> ENDL 145 PDF pg 40, pg 19, 4.13.1.1.3 Regenerate operation <> [s/b] <> ENDL 146 PDF pg 40, pg 19, 4.13.1.1.3 Regenerate operation <> [s/b] <> ENDL 147 PDF pg 40, pg 19, 4.13.1.1.3 Regenerate operation <> [s/b] <> for consistency with the sentence that follows the 1,2,3 list. ENDL 148 PDF pg 40, pg 19, 4.13.1.1.3 Regenerate operation <> [s/b] <> ENDL 149 PDF pg 40, pg 19, 4.13.1.1.3 Regenerate operation Start a new paragraph not in the 1,2,3 list for <> This information is not part of step 4. It is not part of the step-wise algorithm. ENDL 150 PDF pg 40, pg 19, 4.13.1.1.3 Regenerate operation <> [s/b] <> ENDL 151 PDF pg 40, pg 19, 4.13.1.1.4 Rebuild operation <> [s/b] <> ENDL 152 PDF pg 40, pg 19, 4.13.1.1.4 Rebuild operation <> [s/b] <> for consistency with the sentence that follows the 1,2,3 list. ENDL 153 PDF pg 40, pg 19, 4.13.1.1.4 Rebuild operation <> [s/b] <> ENDL 154 PDF pg 40, pg 19, 4.13.1.1.4 Rebuild operation <> [s/b] <> for consistency with the sentence that follows the 1,2,3 list. ENDL 155 PDF pg 40, pg 19, 4.13.1.1.4 Rebuild operation <> [s/b] <> ENDL 156 PDF pg 41, pg 20, 4.13.1.2.1 Additional array subsystem considerations overview <> [s/b] <> ENDL 157 Technical PDF pg 41, pg 20, 4.13.1.2.2 Buffer full status handling <> [s/b] <> ENDL 158 PDF pg 41, pg 20, 4.13.1.2.2 Buffer full status handling <> [s/b] <> ENDL 159 PDF pg 41, pg 20, 4.13.1.2.2 Buffer full status handling <> [s/b] <> ENDL 160 PDF pg 41, pg 20, 4.13.1.2.2 Buffer full status handling <> [s/b] <> ENDL 161 PDF pg 41, pg 20, 4.13.1.2.2 Buffer full status handling <> [s/b] <> ENDL 162 PDF pg 41, pg 20, 4.13.1.2.2 Buffer full status handling <> [s/b] <> ENDL 163 PDF pg 41, pg 20, 4.13.1.2.2 Buffer full status handling <> [s/b] <> ENDL 164 PDF pg 41, pg 20, 4.13.1.2.2 Buffer full status handling <> [s/b] <> The storage array controller may issue multiple XDWRITE commands too. There is nothing special about XDWRITEREAD in that regard. ENDL 165 PDF pg 41, pg 20, 4.13.1.2.3 Access to an inconsistent stripe <> [s/b] <> ENDL 166 PDF pg 41, pg 20, 4.13.1.2.3 Access to an inconsistent stripe <> If the change in the previous sentence is accepted, then there is no need for a definition of strip and this sentence/definition can be deleted. ENDL 167 PDF pg 41, pg 20, 4.13.1.2.3 Access to an inconsistent stripe Insert a paragraph break before <> to separate the definitions from the model. ENDL 168 PDF pg 41, pg 20, 4.13.1.2.3 Access to an inconsistent stripe <> [s/b] <> ENDL 169 PDF pg 41, pg 20, 4.13.1.2.3 Access to an inconsistent stripe Insert a paragraph break before <> to separate the model from the requirements. ENDL 170 PDF pg 41, pg 20, 4.13.1.2.3 Access to an inconsistent stripe <> [s/b] <> ENDL 171 PDF pg 41, pg 20, 4.13.1.2.3 Access to an inconsistent stripe <> [s/b] <> ENDL 172 PDF pg 41, pg 20, 4.13.1.2.3 Access to an inconsistent stripe Insert a paragraph break before <> because I think one is needed. ENDL 173 PDF pg 41, pg 20, 4.13.1.2.3 Access to an inconsistent stripe <> [s/b] <> This change is suggested in an attempt to clarify what is different between WRITE and XDWRITE. ENDL 174 PDF pg 41, pg 20, 4.13.1.3 Error handling considerations <> [s/b] <> Note 'extent' a glossary defined term in this standard. ENDL 175 PDF pg 41, pg 20, 4.13.1.3 Error handling considerations <> [s/b] <> ENDL 176 Technical PDF pg 42, pg 21, 4.13.1.3 Error handling considerations <> This entire subclause belongs in SCC-2 not in SBC-2. If the subclause is not remove then something must be done about > because there is no second class of errors. Next, 'primary target' and 'secondary command' must be defined. Then, 'parity error' should be changed to 'service delivery subsystem error'. Finally, it must be explained how the result of a service delivery subsystem error during the transfer of a status byte can effectively be changed to a CHECK CONDITIONS status. If the subclause is deleted (as recommended), then the heading for <<4.13.1.3.1 Error handling considerations overview>> should also be removed. ENDL 177 PDF pg 42, pg 21, 4.13.1.4 XOR data retention requirements <> [s/b] <> ENDL 178 PDF pg 42, pg 21, 4.13.1.4 XOR data retention requirements <> [s/b] <> ENDL 179 PDF pg 42, pg 21, 4.13.1.4 XOR data retention requirements <> [s/b] <> ENDL 180 PDF pg 42, pg 21, 4.13.1.4 XOR data retention requirements <> [s/b] <> ENDL 181 PDF pg 42, pg 21, 4.14.1 START STOP UNIT and power conditions overview <> [s/b] <> ENDL 182 PDF pg 42, pg 21, 4.14.1 START STOP UNIT and power conditions overview <> [s/b] <> ENDL 183 PDF pg 44, pg 23, 4.14.2.3.3 Transition SSU_PC1:Active to SSU_PC3:Standby <> [s/b] <> ENDL 184 PDF pg 45, pg 24, 4.14.2.4.3 Transition SSU_PC2:Idle to SSU_PC3:Standby <> [s/b] <> for consistency with 4.14.2.3.3 ENDL 185 PDF pg 45, pg 24, 4.14.2.6.1 SSU_PC4:Stopped state description <> [s/b] <> ENDL 186 PDF pg 45, pg 24, 4.14.2.6.1 SSU_PC4:Stopped state description <> [s/b] <> ENDL 187 PDF pg 46, pg 25, 4.15.1 Protection information overview <> [s/b] <> ENDL 188 PDF pg 46, pg 25, 4.15.1 Protection information overview <<(e.g., write to medium, store in non-volatile memory, recalculate on read back)>> [s/b] <<(e.g., written to medium, stored in non-volatile memory, recalculated on read back)>> to maintain consistency with the modified verb 'retained' ENDL 189 PDF pg 46, pg 25, 4.15.1 Protection information overview <> This is not an example. Change to <> ENDL 190 PDF pg 46, pg 25, 4.15.1 Protection information overview <> [2 instances in this paragraph] is not a glossary defined term, either in this standard or in SAM-3. Either A) change to <>, B) change the first occurrence to <> ENDL 191 PDF pg 47, pg 26, 4.15.2 Protection information format <> add CRC to the abbreviations list in 3.2 ENDL 192 PDF pg 47, pg 26, 4.15.2 Protection information format <> [2 instances in this paragraph] is not a glossary defined term, either in this standard or in SAM-3. Either A) change to <>, B) change the first occurrence to <> and change the two occurrences of <> to < ENDL 193 PDF pg 48, pg 27, 4.15.2 Protection information format <> [s/b] <> ENDL 194 PDF pg 50, pg 29, 4.15.3.4 CRC test cases <> [s/b] <> ENDL 195 PDF pg 50, pg 29, 4.15.4 Application of protected data <> [s/b] <> ENDL 196 PDF pg 50, pg 29, 4.15.4 Application of protected data <> [s/b] <> [2 times in this 1,2,3 list ENDL 197 PDF pg 50, pg 29, 4.15.4 Application of protected data The first letters of entries in this 1,2,3 list should be lower case for consistency with previous lists in this standard. ENDL 198 PDF pg 50, pg 29, 4.15.5 Protected data commands <> [s/b] <> ENDL 199 PDF pg 50, pg 29, 4.15.5 Protected data commands <> [s/b] <> ENDL 200 Technical PDF pg 50, pg 29, 4.16 Grouping function <> [s/b] <> ... assuming that the transmission of collected information is not the only thing that is outside the scope of this standard ENDL 201 PDF pg 50, pg 29, 4.16 Grouping function <> [s/b] <> Note: any change the removes the 'if' is acceptable. This example not close to an if/then situation. ENDL 202 PDF pg 51, pg 30, 5.1 Commands for direct-access devices overview <> [s/b] <> ENDL 203 Technical PDF pg 54, pg 33, 5.1 Commands for direct-access devices overview <> This requirement belongs in SPC-3, not in SBC-2. ENDL 204 PDF pg 54, pg 33, 5.1 Commands for direct-access devices overview Per the ISO style guide, the table footnotes and notes must appear at the bottom of every page containing table 9. The table notes must appear first and must be separated from the table footnotes by the same separator (i.e., a double line) that separates the header from the body of the table. Also the double line must be present on the bottom of each continuation page of the table. ENDL 205 PDF pg 54, pg 33, 5.1 Commands for direct-access devices overview <> [s/b] <> ENDL 206 Technical PDF pg 55, pg 34 <<5.2 Variable length CDBs>> This subclause belongs in an informative annex. ENDL 207 Technical PDF pg 55, pg 34 <<5.3 Service action CDBs>> This subclause belongs in an informative annex. ENDL 208 PDF pg 56, pg 35, 5.4.1 FORMAT UNIT command overview <> [s/b] <> ENDL 209 PDF pg 56, pg 35, 5.4.1 FORMAT UNIT command overview <> [s/b] <> ENDL 210 PDF pg 56, pg 35, 5.4.1 FORMAT UNIT command overview <> [s/b] <> ENDL 211 PDF pg 56, pg 35, 5.4.1 FORMAT UNIT command overview <> [s/b] <> ENDL 212 PDF pg 57, pg 36, 5.4.1 FORMAT UNIT command overview <> [s/b] <> [2 times on this page] ENDL 213 PDF pg 57, pg 36, 5.4.1 FORMAT UNIT command overview <> [s/b] <> ENDL 214 PDF pg 57, pg 36, 5.4.1 FORMAT UNIT command overview <> [s/b] <> ENDL 215 PDF pg 57, pg 36, 5.4.1 FORMAT UNIT command overview <> [s/b] <> i.e., application client ownership of the logical block reference tag is discussed only in 4.15.2 and a more direct reference will be more valuable to the reader ENDL 216 PDF pg 57, pg 36, 5.4.1 FORMAT UNIT command overview <> [s/b] <> as per the conventions stated in 3.4. ENDL 217 Technical PDF pg 57, pg 36, 5.4.1 FORMAT UNIT command overview [add at the end of the LONGLIST paragraph] If the FMTDATA bit is set to zero, the contents of the LONGLIST bit shall be ignored. ENDL 218 Technical PDF pg 57, pg 36, 5.4.1 FORMAT UNIT command overview <> [s/b] <> ENDL 219 Technical PDF pg 57, pg 36, 5.4.1 FORMAT UNIT command overview [add a new paragraph after the second a,b list]] If the FMTDATA bit is set to zero, the contents of the CMPLST bit shall be ignored. ENDL 220 PDF pg 58, pg 37, 5.4.1 FORMAT UNIT command overview <> add N/A to the abbreviations in 3.2 ENDL 221 PDF pg 58, pg 37, 5.4.1 FORMAT UNIT command overview << > >> add > to the list of symbols in 3.2 ENDL 222 PDF pg 58, pg 37, 5.4.1 FORMAT UNIT command overview In table 14 <> [s/b] <> The 'All others' in the same row takes care of everything else. See table 21 for an example. ENDL 223 PDF pg 58, pg 37, 5.4.1 FORMAT UNIT command overview Per the ISO style guide, in table 14 the table notes must appear first and must be separated from the table footnotes by a double line. The word Notes: must not appear above the table footnotes (per the ISO style guide) because that might cause the reader to believe that the table footnotes are not normative. ENDL 224 PDF pg 59, pg 38, 5.4.2.1 FORMAT UNIT parameter list overview <> [s/b] <> ENDL 225 PDF pg 59, pg 38, 5.4.2.2 Parameter list header <> [s/b] <> There is at least one field in each parameter list header format. ENDL 226 PDF pg 59, pg 38, 5.4.2.2 Parameter list header <> [s/b] <> ENDL 227 PDF pg 59, pg 38, 5.4.2.2 Parameter list header <> [s/b] <> ENDL 228 PDF pg 60, pg 39, 5.4.2.2 Parameter list header <> [s/b] <> ENDL 229 PDF pg 60, pg 39, 5.4.2.2 Parameter list header <> [s/b] <> ENDL 230 PDF pg 60, pg 39, 5.4.2.2 Parameter list header <> [s/b] <> ENDL 231 PDF pg 60, pg 39, 5.4.2.2 Parameter list header <> [s/b] <> ENDL 232 Technical PDF pg 60, pg 39, 5.4.2.2 Parameter list header <> [s/b] <> ENDL 233 PDF pg 60, pg 39, 5.4.2.2 Parameter list header <> If this phrase is necessary here, then it is necessary in every sentence defining every behavior of the FORMAT UNIT command. Better to delete it once than to add it everywhere. ENDL 234 PDF pg 60, pg 39, 5.4.2.2 Parameter list header <> [s/b] <> for consistency with the DPRY bit definition ENDL 235 PDF pg 60, pg 39, 5.4.2.2 Parameter list header <> [s/b] <> [2 times, in this paragraph and the next] ENDL 236 PDF pg 60, pg 39, 5.4.2.2 Parameter list header <> [s/b] <> [2 times, in this paragraph and the next] ENDL 237 PDF pg 61, pg 40, 5.4.2.2 Parameter list header <> All the fields described in this subclause are in the parameter list header. Delete this redundant information. ENDL 238 Technical PDF pg 61, pg 40, 5.4.2.2 Parameter list header <> This statement is not appropriate in a T10 standard and should be deleted. ENDL 239 PDF pg 61, pg 40, 5.4.2.2 Parameter list header <> [s/b] <> ENDL 240 PDF pg 61, pg 40, 5.4.2.2 Parameter list header <> [s/b] <> ENDL 241 PDF pg 62, pg 41, 5.4.2.3 Initialization pattern descriptor <<(i.e.>> [s/b] <<(i.e.,>> ENDL 242 PDF pg 63, pg 42, 5.4.2.3 Initialization pattern descriptor <> [s/b] <
> The IP modifier field is described in table 19, not table 20. ENDL 243 PDF pg 63, pg 42, 5.4.2.3 Initialization pattern descriptor <> This line should be deleted since (per the ISO style guide) in indicates that the paragraphs below it are not normative, as opposed to the normative paragraphs that appear as table footnoted in table 20. ENDL 244 PDF pg 64, pg 43, 5.4.2.4.2 Short block format address descriptor <> [s/b] <> [2 times, once here and once in 5.4.2.4.3] ENDL 245 PDF pg 64, pg 43, 5.4.2.4.4 Bytes from index format address descriptor <> should be deleted for consistency with the description of the SECTOR NUMBER field in 5.4.2.4.5. ENDL 246 PDF pg 65, pg 44, 5.4.2.4.5 Physical sector format address descriptor <> [s/b] <> for consistency with the 'one track' phrase earlier in this sentence. ENDL 247 PDF pg 65, pg 44, 5.5 LOCK UNLOCK CACHE (10) command <> [s/b] <> ENDL 248 PDF pg 66, pg 45, 5.5 LOCK UNLOCK CACHE (10) command <> [s/b] <> [4 times in this subclause on this page] ENDL 249 PDF pg 66, pg 45, 5.5 LOCK UNLOCK CACHE (10) command <> [s/b] <> ENDL 250 PDF pg 66, pg 45, 5.5 LOCK UNLOCK CACHE (10) command <> [s/b] <> ENDL 251 Technical PDF pg 66, pg 45, 5.5 LOCK UNLOCK CACHE (10) command <> [s/b] <> and <> [s/b] < [3 instances total in this paragraph] ENDL 252 PDF pg 66, pg 45, 5.6 LOCK UNLOCK CACHE (16) command <> [s/b] <> ENDL 253 PDF pg 66, pg 45, 5.6 LOCK UNLOCK CACHE (16) command <> [s/b] <> Standards describe things, but more importantly they define things. ENDL 254 PDF pg 67, pg 46, 5.7 PRE-FETCH (10) command <> [s/b] <> Unless there is more than one definition of the LOGICAL BLOCK ADDRESS field ENDL 255 PDF pg 67, pg 46, 5.7 PRE-FETCH (10) command <> What other cache memory would it be? Delete this unnecessary phrase. ENDL 256 PDF pg 67, pg 46, 5.7 PRE-FETCH (10) command <> [s/b] <> ENDL 257 PDF pg 67, pg 46, 5.7 PRE-FETCH (10) command <> [s/b] <> ENDL 258 PDF pg 68, pg 47, 5.8 PRE-FETCH (16) command <> [s/b] <> ENDL 259 PDF pg 68, pg 47, 5.9 READ (6) command <> [s/b] <> ENDL 260 PDF pg 69, pg 48, 5.9 READ (6) command <> [s/b] <> ENDL 261 PDF pg 69, pg 48, 5.9 READ (6) command It seem like note 10 should precede note 9 so that it is closer to the definition of the TRANSFER LENGTH field. ENDL 262 PDF pg 70, pg 49, 5.9 READ (6) command <> [s/b] <> ENDL 263 PDF pg 70, pg 49, 5.9 READ (6) command <> [s/b] <> ENDL 264 PDF pg 71, pg 50, 5.10 READ (10) command <> [s/b] <> ENDL 265 PDF pg 71, pg 50, 5.10 READ (10) command <> [s/b] <> ENDL 266 PDF pg 71, pg 50, 5.10 READ (10) command Per the ISO style guide, there should be a double line at the bottom of each page in table 33. ENDL 267 Technical PDF pg 72, pg 51, 5.10 READ (10) command <> Does 'Shall not' in the Extended INQUIRY VPD page column of table 33 mean that the page shall not be supported? ENDL 268 PDF pg 73, pg 52, 5.10 READ (10) command <> [s/b] <> [2 times on this page] ENDL 269 PDF pg 73, pg 52, 5.10 READ (10) command <> [s/b] <> ENDL 270 PDF pg 73, pg 52, 5.10 READ (10) command <> [s/b] <> [2 times on this page] ENDL 271 PDF pg 73, pg 52, 5.10 READ (10) command <> [s/b] <> ENDL 272 PDF pg 73, pg 52, 5.10 READ (10) command <> [s/b] <> ENDL 273 PDF pg 73, pg 52, 5.10 READ (10) command <> [s/b] <> ENDL 274 PDF pg 74, pg 53, 5.10 READ (10) command <> [s/b] <> [9 times in this table] ENDL 275 PDF pg 74, pg 53, 5.10 READ (10) command <> is redundant with <> Delete one or the other. [2 times in table 34] ENDL 276 PDF pg 75, pg 54, 5.11 READ (12) command <> [s/b] <> for consistency with table 93. ENDL 277 PDF pg 75, pg 54, 5.11 READ (12) command <> [s/b] <> ENDL 278 PDF pg 75, pg 54, 5.12 READ (16) command <> [s/b] <> for consistency with table 93. ENDL 279 PDF pg 75, pg 54, 5.12 READ (16) command <> [s/b] <> ENDL 280 PDF pg 76, pg 55, 5.13 READ (32) command <> [s/b] <> ENDL 281 PDF pg 76, pg 55, 5.13 READ (32) command <> [s/b] <> ENDL 282 PDF pg 76, pg 55, 5.13 READ (32) command <> [s/b] <> [2 times in this paragraph] ENDL 283 PDF pg 76, pg 55, 5.13 READ (32) command <> [s/b] <> ENDL 284 PDF pg 76, pg 55, 5.13 READ (32) command <> [s/b] <> ENDL 285 PDF pg 76, pg 55, 5.13 READ (32) command <> What other command would it be? Delete this. [2 times on this page] ENDL 286 PDF pg 76, pg 55, 5.13 READ (32) command <> [s/b] <> ENDL 287 PDF pg 77, pg 56, 5.14 READ CAPACITY (10) command <> [s/b] <> ENDL 288 PDF pg 78, pg 57, 5.14 READ CAPACITY (10) command <> [s/b] <> ENDL 289 PDF pg 78, pg 57, 5.15 READ CAPACITY (16) command <> [s/b] <> ENDL 290 PDF pg 78, pg 57, 5.15 READ CAPACITY (16) command <> [s/b] <> ENDL 291 PDF pg 79, pg 58, 5.15 READ CAPACITY (16) command <> [s/b] <> per the conventions in 3.4. ENDL 292 PDF pg 80, pg 59, 5.16 READ DEFECT DATA (10) command <> [s/b] <> ENDL 293 PDF pg 80, pg 59, 5.16 READ DEFECT DATA (10) command <> The intent of this sentence is unclear. Maybe <> is correct, maybe not. ENDL 294 PDF pg 81, pg 60, 5.16 READ DEFECT DATA (10) command <> [s/b] <> ENDL 295 PDF pg 81, pg 60, 5.16 READ DEFECT DATA (10) command <> [s/b] <> ENDL 296 Technical PDF pg 81, pg 60, 5.16 READ DEFECT DATA (10) command Is note 16 really useful? If the size of the defect list exceeds the allocation length of a READ DEFECT DATA(12) command, there is no way at all to determine the number of defects. ENDL 297 PDF pg 81, pg 60, 5.16 READ DEFECT DATA (10) command <> [s/b] <> since READ DEFECT DATA(12) has a greater probability of being able to represent the defect list in the capacity of the ALLOCATION LENGTH field. ENDL 298 PDF pg 81, pg 60, 5.16 READ DEFECT DATA (10) command <> [s/b] <> ENDL 299 PDF pg 81, pg 60, 5.16 READ DEFECT DATA (10) command <> Delete this. The second sentence on this page says the same thing better. ENDL 300 PDF pg 81, pg 60, 5.17 READ DEFECT DATA (12) command <> [s/b] <> ENDL 301 PDF pg 82, pg 61, 5.17 READ DEFECT DATA (12) command <> [s/b] <> ENDL 302 PDF pg 82, pg 61, 5.18 READ LONG (10) command <> [s/b] <> ENDL 303 PDF pg 82, pg 61, 5.18 READ LONG (10) command <> The LOCK UNLOCK CACHE(10) defines the LOGICAL BLOCK ADDRESS field as specifying the first logical block in an extent. That definition is correct for READ LONG only if the length of the extent is one. It is not clear that a reference to 5.5 is appropriate here. ENDL 304 PDF pg 83, pg 62, 5.19 READ LONG (16) command <> [s/b] <> ENDL 305 PDF pg 83, pg 62, 5.20 REASSIGN BLOCKS command [Global in REASSIGN BLOCKS] Since the address descriptors used by REASSIGN BLOCKS are not the same as the address descriptors used by FORMAT UNIT, READ DEFECT DATA, and elsewhere, consideration should be given to changing the name. 'Logical block address descriptor' is one possibility. ENDL 306 PDF pg 83, pg 62, 5.20 REASSIGN BLOCKS command <> [s/b] <> ENDL 307 PDF pg 83, pg 62, 5.20 REASSIGN BLOCKS command <> [s/b] <> or <> ENDL 308 PDF pg 83, pg 62, 5.20 REASSIGN BLOCKS command <> This sentence is not strictly correct. Since 'each address descriptor' can specify exactly one logical block, it seems unlikely that more than one logical block can be relocated in response to one (aka 'each') address descriptor. ENDL 309 PDF pg 83, pg 62, 5.20 REASSIGN BLOCKS command <> [s/b] <> ENDL 310 PDF pg 84, pg 63, 5.20 REASSIGN BLOCKS command <> [s/b] <> for consistency with 3.4 ENDL 311 PDF pg 85, pg 64, 5.20 REASSIGN BLOCKS command <> [s/b] <> ENDL 312 PDF pg 85, pg 64, 5.20 REASSIGN BLOCKS command <> [s/b] <> for consistency with 3.4 ENDL 313 PDF pg 85, pg 64, 5.20 REASSIGN BLOCKS command <> [s/b] <> Surely, errors are never expected. ENDL 314 PDF pg 85, pg 64, 5.20 REASSIGN BLOCKS command It seems like note 18 should not be a note. ENDL 315 PDF pg 85, pg 64, 5.21 START STOP UNIT command <> [s/b] <> Also delete the paragraph immediately before table 52 ... for consistency with the other command descriptions in this standard ENDL 316 PDF pg 85, pg 64, 5.21 START STOP UNIT command <> [s/b] <> ENDL 317 PDF pg 85, pg 64, 5.21 START STOP UNIT command <> [s/b] <> ENDL 318 PDF pg 85, pg 64, 5.21 START STOP UNIT command <> [s/b] <> for consistency with 4.12 ENDL 319 PDF pg 86, pg 65, 5.21 START STOP UNIT command <> [s/b] <> as per table 52 and other text in this subclause ENDL 320 PDF pg 86, pg 65, 5.21 START STOP UNIT command Delete <> Since I can find no way to omit the POWER CONDITIONS field from the CDB, it seems unlikely that the field is optional. ENDL 321 PDF pg 86, pg 65, 5.21 START STOP UNIT command <> [s/b] <> ENDL 322 PDF pg 86, pg 65, 5.21 START STOP UNIT command <<0h then>> [s/b] <<0h, then>> i.e., add a comma between the if clause and the then clause ENDL 323 PDF pg 86, pg 65, 5.21 START STOP UNIT command The device server may ignore the contents of the POWER CONDITIONS field. ENDL 324 PDF pg 86, pg 65, 5.21 START STOP UNIT command <> [s/b] <> ENDL 325 PDF pg 86, pg 65, 5.21 START STOP UNIT command <> [s/b] <> or <> [3 times on this page] ENDL 326 PDF pg 86, pg 65, 5.21 START STOP UNIT command <> [s/b] <> ENDL 327 PDF pg 86, pg 65, 5.21 START STOP UNIT command <> [s/b] <> ENDL 328 PDF pg 86, pg 65, 5.21 START STOP UNIT command <> [s/b] <> ENDL 329 PDF pg 86, pg 65, 5.21 START STOP UNIT command <> [s/b] <> for consistency with item a) in the same list ENDL 330 PDF pg 86, pg 65, 5.21 START STOP UNIT command <> [s/b] <> [2 times in list entry a)] ENDL 331 PDF pg 86, pg 65, 5.21 START STOP UNIT command <> [s/b] <> ENDL 332 PDF pg 87, pg 66, 5.21 START STOP UNIT command Either add <<(see SPC-3)>> here Or delete (see SPC-3) from the paragraph defining LU_CONTROL on the previous page. ENDL 333 PDF pg 87, pg 66, 5.21 START STOP UNIT command <> [s/b] <> for consistency with terminology used everywhere else in this subclause ENDL 334 PDF pg 87, pg 66, 5.22 SYNCHRONIZE CACHE (10) command <> [s/b] <> ENDL 335 PDF pg 87, pg 66, 5.22 SYNCHRONIZE CACHE (10) command <> [s/b] <> ENDL 336 PDF pg 87, pg 66, 5.22 SYNCHRONIZE CACHE (10) command <> [s/b] <> ENDL 337 PDF pg 87, pg 66, 5.22 SYNCHRONIZE CACHE (10) command <> [s/b] <> ENDL 338 PDF pg 87, pg 66, 5.22 SYNCHRONIZE CACHE (10) command <> [s/b] <> strictly speaking the current wording prohibits requirements for SYNCHRONIZE CACHE usage from appearing anywhere in this clause, i.e., clause 5, and invalidates statements made in the START STOP UNIT command definition. ENDL 339 PDF pg 87, pg 66, 5.22 SYNCHRONIZE CACHE (10) command <> [s/b] <> ENDL 340 PDF pg 87, pg 66, 5.22 SYNCHRONIZE CACHE (10) command <> [s/b] <> ENDL 341 PDF pg 87, pg 66, 5.22 SYNCHRONIZE CACHE (10) command <> [s/b] <> and Delete <> at the end of the sentence. ENDL 342 PDF pg 87, pg 66, 5.22 SYNCHRONIZE CACHE (10) command <> Remove period from the end of this non-sentence. ENDL 343 PDF pg 88, pg 67, 5.22 SYNCHRONIZE CACHE (10) command <> [s/b] <> ENDL 344 PDF pg 88, pg 67, 5.22 SYNCHRONIZE CACHE (10) command <> [s/b] <> [3 times on this page] ENDL 345 PDF pg 88, pg 67, 5.22 SYNCHRONIZE CACHE (10) command Why is the LOCK UNLOCK CACHE (10) command (see 5.5) definition for the NUMBER OF BLOCKS field incorrect for SYNCHRONIZE CACHE(10)? ENDL 346 PDF pg 88, pg 67, 5.23 SYNCHRONIZE CACHE (16) command <> [s/b] <> ENDL 347 PDF pg 88, pg 67, 5.23 SYNCHRONIZE CACHE (16) command <> [s/b] <> ENDL 348 PDF pg 88, pg 67, 5.23 SYNCHRONIZE CACHE (16) command <> [s/b] <> ENDL 349 PDF pg 88, pg 67, 5.23 SYNCHRONIZE CACHE (16) command <> [s/b] <> ENDL 350 PDF pg 88, pg 67, 5.23 SYNCHRONIZE CACHE (16) command <> [s/b] <> ENDL 351 PDF pg 88, pg 67, 5.24 VERIFY (10) command <> [s/b] <> ENDL 352 PDF pg 89, pg 68, 5.24 VERIFY (10) command <> [s/b] <> ENDL 353 PDF pg 89, pg 68, 5.24 VERIFY (10) command <> [s/b] <> ENDL 354 PDF pg 89, pg 68, 5.24 VERIFY (10) command <> [s/b] <> for consistency with table 93. ENDL 355 PDF pg 89, pg 68, 5.24 VERIFY (10) command <> [s/b] <> [2 times in this paragraph] ENDL 356 PDF pg 89, pg 68, 5.24 VERIFY (10) command <> [s/b] <> ENDL 357 PDF pg 89, pg 68, 5.24 VERIFY (10) command <> [s/b] <> for consistency with the rest of the standard ENDL 358 PDF pg 89, pg 68, 5.24 VERIFY (10) command <> [s/b] <> to match the number of the sentence subject (i.e., settings) ENDL 359 Technical PDF pg 90, pg 69, 5.24 VERIFY (10) command <> Does 'Shall not' in the Extended INQUIRY VPD page column of table 58 and table 59 mean that the page shall not be supported? ENDL 360 PDF pg 90, pg 69, 5.24 VERIFY (10) command There should be a double line at the bottom of each page of table 58, table 60, and table 61. ENDL 361 PDF pg 92, pg 71, 5.24 VERIFY (10) command <> [s/b] <> [2 times on this page] ENDL 362 PDF pg 92, pg 71, 5.24 VERIFY (10) command <> [s/b] <> ENDL 363 PDF pg 92, pg 71, 5.24 VERIFY (10) command <> [s/b] <> ENDL 364 PDF pg 92, pg 71, 5.24 VERIFY (10) command <> [s/b] <> ENDL 365 PDF pg 93, pg 72, 5.24 VERIFY (10) command <> [s/b] <> [2 times on this page] ENDL 366 PDF pg 93, pg 72, 5.24 VERIFY (10) command <> [s/b] <> ENDL 367 PDF pg 93, pg 72, 5.24 VERIFY (10) command <> [s/b] <> ENDL 368 PDF pg 93, pg 72, 5.24 VERIFY (10) command <> [s/b] <> ENDL 369 PDF pg 95, pg 74, 5.24 VERIFY (10) command <> [s/b] <> [2 times on this page] ENDL 370 PDF pg 95, pg 74, 5.24 VERIFY (10) command <> [s/b] <> ENDL 371 PDF pg 95, pg 74, 5.24 VERIFY (10) command <> [s/b] <> ENDL 372 PDF pg 97, pg 76, 5.24 VERIFY (10) command <> [s/b] <> [2 times on this page] ENDL 373 PDF pg 97, pg 76, 5.24 VERIFY (10) command <> [s/b] <> ENDL 374 PDF pg 97, pg 76, 5.24 VERIFY (10) command <> [s/b] <> ENDL 375 PDF pg 97, pg 76, 5.25 VERIFY (12) command <> [s/b] <> ENDL 376 PDF pg 98, pg 77, 5.25 VERIFY (12) command <> [s/b] <> ENDL 377 PDF pg 98, pg 77, 5.25 VERIFY (12) command <> [s/b] <> ENDL 378 PDF pg 98, pg 77, 5.25 VERIFY (12) command <> [s/b] <> for consistency with table 93. ENDL 379 PDF pg 98, pg 77, 5.25 VERIFY (12) command <> [s/b] <> ENDL 380 PDF pg 98, pg 77, 5.26 VERIFY (16) command <> [s/b] <> ENDL 381 PDF pg 98, pg 77, 5.26 VERIFY (16) command <> [s/b] <> ENDL 382 PDF pg 98, pg 77, 5.26 VERIFY (16) command <> [s/b] <> ENDL 383 PDF pg 98, pg 77, 5.26 VERIFY (16) command <> [s/b] <> for consistency with table 93. ENDL 384 PDF pg 98, pg 77, 5.26 VERIFY (16) command <> [s/b] <> ENDL 385 PDF pg 99, pg 78, 5.27 VERIFY (32) command <> [s/b] <> ENDL 386 PDF pg 99, pg 78, 5.27 VERIFY (32) command The text describing the RTO_EN bit appears between the subclause heading and the CDB format table for all the other VERIFY(n) commands. Be consistent. ENDL 387 PDF pg 99, pg 78, 5.27 VERIFY (32) command <> [s/b] <> ENDL 388 PDF pg 99, pg 78, 5.27 VERIFY (32) command <> [s/b] <> [2 times in this paragraph] ENDL 389 PDF pg 99, pg 78, 5.27 VERIFY (32) command <> [s/b] <> ENDL 390 PDF pg 99, pg 78, 5.27 VERIFY (32) command <> [s/b] <> ENDL 391 PDF pg 99, pg 78, 5.27 VERIFY (32) command <> [s/b] <> [on this page and the next] ENDL 392 PDF pg 100, pg 79, 5.27 VERIFY (32) command Delete <> ENDL 393 PDF pg 100, pg 79, 5.28 WRITE (6) command <> [s/b] <> ENDL 394 PDF pg 100, pg 79, 5.28 WRITE (6) command <> [s/b] <> ENDL 395 PDF pg 100, pg 79, 5.28 WRITE (6) command <> [s/b] <> ENDL 396 PDF pg 100, pg 79, 5.28 WRITE (6) command <> [s/b] <> ENDL 397 PDF pg 100, pg 79, 5.28 WRITE (6) command <> [s/b] <> ENDL 398 PDF pg 100, pg 79, 5.28 WRITE (6) command <> [s/b] <> This text was what surfaced the need for commas, as I read it. ENDL 399 PDF pg 101, pg 80, 5.29 WRITE (10) command <> [s/b] <> ENDL 400 PDF pg 101, pg 80, 5.29 WRITE (10) command <> [s/b] <> [2 times in this paragraph] ENDL 401 PDF pg 101, pg 80, 5.29 WRITE (10) command <> [s/b] <> ENDL 402 PDF pg 101, pg 80, 5.29 WRITE (10) command <> [s/b] <> Note: the suggested replacement text is a cut and paste from the READ(10) command, with only the table number changed. ENDL 403 PDF pg 101, pg 80, 5.29 WRITE (10) command <> [s/b] <> [4 times in this table] ENDL 404 PDF pg 102, pg 81, 5.29 WRITE (10) command In READ(10), the RDPROTECT field is described before the FUA bits. Here the reverse is true. Be consistent. ENDL 405 PDF pg 102, pg 81, 5.29 WRITE (10) command There should be a double line at the bottom of each page of table 68. ENDL 406 PDF pg 103, pg 82, 5.29 WRITE (10) command <> [s/b] <> [2 times on this page] ENDL 407 PDF pg 103, pg 82, 5.29 WRITE (10) command <> [s/b] <> [2 times on this page] ENDL 408 PDF pg 103, pg 82, 5.29 WRITE (10) command <> [s/b] <> [2 times on this page] ENDL 409 PDF pg 104, pg 83, 5.30 WRITE (12) command <> [s/b] <> [2 times on this page] ENDL 410 PDF pg 104, pg 83, 5.30 WRITE (12) command <> [s/b] <> for consistency with table 93. ENDL 411 PDF pg 104, pg 83, 5.30 WRITE (12) command <> [s/b] <> ENDL 412 PDF pg 104, pg 83, 5.31 WRITE (16) command <> [s/b] <> for consistency with table 93. ENDL 413 PDF pg 104, pg 83, 5.31 WRITE (16) command <> [s/b] <> ENDL 414 PDF pg 105, pg 84, 5.32 WRITE (32) command <> [s/b] <> ENDL 415 PDF pg 105, pg 84, 5.32 WRITE (32) command <> [s/b] <> ENDL 416 PDF pg 105, pg 84, 5.32 WRITE (32) command <> [s/b] <> [2 times in this paragraph] ENDL 417 PDF pg 105, pg 84, 5.32 WRITE (32) command <> [s/b] <> ENDL 418 PDF pg 105, pg 84, 5.32 WRITE (32) command <> [s/b] <> ENDL 419 PDF pg 105, pg 84, 5.32 WRITE (32) command <> What other command would it be? Delete this. [2 times on this page] ENDL 420 PDF pg 105, pg 84, 5.32 WRITE (32) command <> [s/b] <> ENDL 421 PDF pg 106, pg 85, 5.33 WRITE AND VERIFY (10) command <> [s/b] <> [2 times on this page] ENDL 422 PDF pg 106, pg 85, 5.33 WRITE AND VERIFY (10) command <> [s/b] <> ENDL 423 PDF pg 106, pg 85, 5.33 WRITE AND VERIFY (10) command <> [s/b] <> [2 times in this paragraph] ENDL 424 PDF pg 106, pg 85, 5.33 WRITE AND VERIFY (10) command <> [s/b] <> [2 times in this paragraph] ENDL 425 PDF pg 106, pg 85, 5.33 WRITE AND VERIFY (10) command <> [s/b] <> ENDL 426 PDF pg 107, pg 86, 5.34 WRITE AND VERIFY (12) command <> [s/b] <> ENDL 427 PDF pg 107, pg 86, 5.34 WRITE AND VERIFY (12) command <> [s/b] <> for consistency with table 93. ENDL 428 PDF pg 107, pg 86, 5.34 WRITE AND VERIFY (12) command <> [s/b] <> ENDL 429 PDF pg 107, pg 86, 5.35 WRITE AND VERIFY (16) command <> [s/b] <> ENDL 430 PDF pg 107, pg 86, 5.35 WRITE AND VERIFY (16) command <> [s/b] <> ENDL 431 PDF pg 107, pg 86, 5.35 WRITE AND VERIFY (16) command <> [s/b] <> for consistency with table 93. ENDL 432 PDF pg 107, pg 86, 5.35 WRITE AND VERIFY (16) command <> [s/b] <> ENDL 433 PDF pg 108, pg 87, 5.36 WRITE AND VERIFY (32) command <> [s/b] <> ENDL 434 PDF pg 108, pg 87, 5.36 WRITE AND VERIFY (32) command <> [s/b] <> ENDL 435 PDF pg 108, pg 87, 5.36 WRITE AND VERIFY (32) command <> [s/b] <> [2 times in this paragraph] ENDL 436 PDF pg 108, pg 87, 5.36 WRITE AND VERIFY (32) command <> [s/b] <> ENDL 437 PDF pg 108, pg 87, 5.36 WRITE AND VERIFY (32) command <> [s/b] <> ENDL 438 PDF pg 108, pg 87, 5.36 WRITE AND VERIFY (32) command <> What other command would it be? Delete this. [2 times on this page] ENDL 439 PDF pg 108, pg 87, 5.36 WRITE AND VERIFY (32) command <> [s/b] <> ENDL 440 PDF pg 109, pg 88, 5.37 WRITE LONG (10) command <> The LOCK UNLOCK CACHE(10) defines the LOGICAL BLOCK ADDRESS field as specifying the first logical block in an extent. That definition is correct for WRITE LONG only if the length of the extent is one. It is not clear that a reference to 5.5 is appropriate here. ENDL 441 PDF pg 109, pg 88, 5.37 WRITE LONG (10) command <> [s/b] <> ENDL 442 PDF pg 110, pg 89, 5.38 WRITE LONG (16) command <> [s/b] <> ENDL 443 PDF pg 110, pg 89, 5.39 WRITE SAME (10) command <> [s/b] <> ENDL 444 PDF pg 110, pg 89, 5.39 WRITE SAME (10) command <> [s/b] <> ENDL 445 PDF pg 110, pg 89, 5.39 WRITE SAME (10) command <> [s/b] <> ENDL 446 PDF pg 110, pg 89, 5.39 WRITE SAME (10) command <> [s/b] <> ENDL 447 PDF pg 111, pg 90, 5.39 WRITE SAME (10) command <> [s/b] <> Otherwise the use of 'into' is redundant in the sentence. ENDL 448 Technical PDF pg 111, pg 90, 5.39 WRITE SAME (10) command <> If the least significant four bytes are written, how can the most significant <> be written first? ENDL 449 PDF pg 111, pg 90, 5.39 WRITE SAME (10) command <> [s/b] <> for consistency with 3.4 ENDL 450 PDF pg 111, pg 90, 5.39 WRITE SAME (10) command <> [s/b] <> ENDL 451 PDF pg 111, pg 90, 5.40 WRITE SAME (16) command <> [s/b] <> ENDL 452 PDF pg 112, pg 91, 5.40 WRITE SAME (16) command <> [s/b] <> ENDL 453 PDF pg 112, pg 91, 5.40 WRITE SAME (16) command <> [s/b] <> ENDL 454 PDF pg 113, pg 92, 5.41 WRITE SAME (32) command <> [s/b] <> ENDL 455 PDF pg 113, pg 92, 5.41 WRITE SAME (32) command <> [s/b] <> ENDL 456 PDF pg 113, pg 92, 5.41 WRITE SAME (32) command <> [s/b] <> [2 times in this paragraph] ENDL 457 PDF pg 113, pg 92, 5.41 WRITE SAME (32) command <> [s/b] <> ENDL 458 PDF pg 113, pg 92, 5.41 WRITE SAME (32) command <> [s/b] <> ENDL 459 PDF pg 113, pg 92, 5.41 WRITE SAME (32) command <> What other command would it be? Delete this. [2 times on this page] ENDL 460 PDF pg 113, pg 92, 5.41 WRITE SAME (32) command <> [s/b] <> ENDL 461 PDF pg 114, pg 93, 5.42 XDREAD (10) command <> [s/b] <> ENDL 462 PDF pg 114, pg 93, 5.42 XDREAD (10) command <> [s/b] <> ENDL 463 PDF pg 114, pg 93, 5.42 XDREAD (10) command <> [s/b] <> ENDL 464 PDF pg 114, pg 93, 5.42 XDREAD (10) command <> [s/b] <> ENDL 465 PDF pg 114, pg 93, 5.42 XDREAD (10) command Insert new paragraph for consistency with XPWRITE command definition. [2 times on this page] ENDL 466 PDF pg 114, pg 93, 5.42 XDREAD (10) command <> [s/b] <> [2 times in this paragraph] ENDL 467 PDF pg 114, pg 93, 5.42 XDREAD (10) command <> [s/b] <> [2 times in this paragraph] ENDL 468 PDF pg 114, pg 93, 5.42 XDREAD (10) command <> [s/b] <> ENDL 469 PDF pg 115, pg 94, 5.43 XDREAD (32) command <> [s/b] <> ENDL 470 PDF pg 115, pg 94, 5.43 XDREAD (32) command <> [s/b] <> ENDL 471 PDF pg 115, pg 94, 5.43 XDREAD (32) command <> [s/b] <> ENDL 472 PDF pg 115, pg 94, 5.44 XDWRITE (10) command <> [s/b] <> [2 times in this paragraph] ENDL 473 PDF pg 115, pg 94, 5.44 XDWRITE (10) command <> [s/b] <> ENDL 474 PDF pg 116, pg 95, 5.44 XDWRITE (10) command <> [s/b] <> ENDL 475 PDF pg 116, pg 95, 5.44 XDWRITE (10) command <> [s/b] <> [1 time each in this paragraph and the next] ENDL 476 PDF pg 116, pg 95, 5.44 XDWRITE (10) command The first three paragraphs after table 84 should be agglomerated into one paragraph, for consistency with the rest of the standard. ENDL 477 PDF pg 116, pg 95, 5.44 XDWRITE (10) command <> [s/b] <> ENDL 478 PDF pg 116, pg 95, 5.45 XDWRITE (32) command <> [s/b] <> [2 times in this paragraph] ENDL 479 PDF pg 116, pg 95, 5.45 XDWRITE (32) command <> [s/b] <> ENDL 480 PDF pg 117, pg 96, 5.45 XDWRITE (32) command <> [s/b] <> ENDL 481 PDF pg 117, pg 96, 5.45 XDWRITE (32) command <> [s/b] <> ENDL 482 PDF pg 117, pg 96, 5.46 XDWRITEREAD (10) command <> [s/b] <> ENDL 483 PDF pg 117, pg 96, 5.46 XDWRITEREAD (10) command <> [s/b] <> ENDL 484 PDF pg 118, pg 97, 5.46 XDWRITEREAD (10) command <> [s/b] <> ENDL 485 PDF pg 118, pg 97, 5.46 XDWRITEREAD (10) command <> [s/b] <> ENDL 486 PDF pg 118, pg 97, 5.47 XDWRITEREAD (32) command <> [s/b] <> ENDL 487 PDF pg 118, pg 97, 5.47 XDWRITEREAD (32) command <> [s/b] <> ENDL 488 PDF pg 119, pg 98, 5.47 XDWRITEREAD (32) command <> [s/b] <> ENDL 489 PDF pg 119, pg 98, 5.47 XDWRITEREAD (32) command <> [s/b] <> ENDL 490 PDF pg 119, pg 98, 5.48 XPWRITE (10) command <> [s/b] <> ENDL 491 PDF pg 119, pg 98, 5.48 XPWRITE (10) command <> [s/b] <> ENDL 492 PDF pg 120, pg 99, 5.48 XPWRITE (10) command <> [s/b] <> [2 times in this paragraph] ENDL 493 PDF pg 120, pg 99, 5.48 XPWRITE (10) command <> [s/b] <> ENDL 494 PDF pg 120, pg 99, 5.48 XPWRITE (10) command <> [s/b] <> [3 times on this page] ENDL 495 PDF pg 120, pg 99, 5.48 XPWRITE (10) command <> [s/b] <> [3 times on this page] ENDL 496 PDF pg 120, pg 99, 5.48 XPWRITE (10) command <> [s/b] <> ENDL 497 PDF pg 120, pg 99, 5.48 XPWRITE (10) command <> [s/b] <> [2 times on this page] ENDL 498 PDF pg 121, pg 100, 5.49 XPWRITE (32) command <> [s/b] <> ENDL 499 PDF pg 121, pg 100, 5.49 XPWRITE (32) command <> [s/b] <> ENDL 500 PDF pg 122, pg 101, 6.1.2 Translate Address Output diagnostic page <> [s/b] <> I would have said <> but <> is more consistent with the usage of <> later in this paragraph. ENDL 501 PDF pg 122, pg 101, 6.1.2 Translate Address Output diagnostic page <> [s/b] <> for consistency with the next sentence. ENDL 502 PDF pg 123, pg 102, 6.1.2 Translate Address Output diagnostic page <
> [s/b] <
> ENDL 503 PDF pg 124, pg 103, 6.1.2 Translate Address Output diagnostic page Delete <> because it adds no value. ENDL 504 PDF pg 126, pg 105, 6.2.2 Format Status log page <> [s/b] <> ENDL 505 PDF pg 126, pg 105, 6.2.2 Format Status log page <> [s/b] <> ENDL 506 PDF pg 126, pg 105, 6.2.2 Format Status log page <> [s/b] <> ENDL 507 PDF pg 126, pg 105, 6.2.2 Format Status log page <> [s/b] <> ENDL 508 Technical PDF pg 126, pg 105, 6.2.2 Format Status log page <> [s/b] <> ENDL 509 PDF pg 126, pg 105, 6.2.2 Format Status log page <> [s/b] <> ENDL 510 PDF pg 126, pg 105, 6.2.2 Format Status log page <> [s/b] <> ENDL 511 Technical PDF pg 126, pg 105, 6.2.2 Format Status log page Some text appears to be missing. 1) There is no description for the total new blocks reallocated parameter. 2) The existing definition of the total blocks reallocated during format parameter would fit better as the definition of the total new blocks reallocated parameter. ENDL 512 PDF pg 126, pg 105, 6.2.2 Format Status log page <> [s/b] <> ENDL 513 PDF pg 126, pg 105, 6.2.2 Format Status log page <> [s/b] <> ENDL 514 PDF pg 126, pg 105, 6.2.2 Format Status log page <> [s/b] <> ENDL 515 PDF pg 127, pg 106, 6.2.3 Non-volatile Cache log page <> [s/b] <<(see table 95)>> ENDL 516 PDF pg 127, pg 106, 6.2.3 Non-volatile Cache log page <<(either permanently or temporarily, e.g.,>> [s/b] <<(either permanently or temporarily (e.g.,>> ENDL 517 PDF pg 132, pg 111, 6.3.2.3 Long LBA mode parameter block descriptor <> [s/b] <> for consistency with 3.4 [2 times on this page] ENDL 518 PDF pg 133, pg 112, 6.3.3 Caching mode page <> [s/b] <> to help the 'dependent on' phrase make more sense ENDL 519 PDF pg 134, pg 113, 6.3.3 Caching mode page <> [s/b] <> ENDL 520 PDF pg 134, pg 113, 6.3.3 Caching mode page <> [s/b] <> [2 times in this paragraph] ENDL 521 Technical PDF pg 134, pg 113, 6.3.3 Caching mode page <> Is this statement true for any of the following fields: FSW, LBCSS, DRA, NUMBER OF CACHE SEGMENTS, CACHE SEGMENT SIZE, NV_DIS, or NON CACHE SEGMENT SIZE? It appears that fields have been appended to the page contents since the time that this statement was written. ENDL 522 PDF pg 134, pg 113, 6.3.3 Caching mode page <> [s/b] <> ENDL 523 PDF pg 134, pg 113, 6.3.3 Caching mode page <> [s/b] <> ENDL 524 Technical PDF pg 134, pg 113, 6.3.3 Caching mode page <> Is this statement true for any of the following fields: FSW, LBCSS, DRA, NUMBER OF CACHE SEGMENTS, CACHE SEGMENT SIZE, NV_DIS, or NON CACHE SEGMENT SIZE? It appears that fields have been appended to the page contents since the time that this statement was written. ENDL 525 PDF pg 134, pg 113, 6.3.3 Caching mode page <> [s/b] <> ENDL 526 PDF pg 135, pg 114, 6.3.3 Caching mode page Delete <> I do not think any replacement wording is needed, but if some is it would be <> ENDL 527 PDF pg 135, pg 114, 6.3.3 Caching mode page <> [s/b] <> ENDL 528 PDF pg 135, pg 114, 6.3.3 Caching mode page Delete <> ENDL 529 PDF pg 135, pg 114, 6.3.3 Caching mode page <> [s/b] <> ENDL 530 PDF pg 135, pg 114, 6.3.3 Caching mode page <> [s/b] <> ENDL 531 PDF pg 135, pg 114, 6.3.3 Caching mode page Delete <> Pre-fetch is defined to use cache memory (see previous page) and besides 'cache' s/b 'cache memory' ENDL 532 PDF pg 135, pg 114, 6.3.3 Caching mode page Perhaps note 24 should not be a note. ENDL 533 Technical PDF pg 135, pg 114, 6.3.3 Caching mode page <> [s/b] <> ENDL 534 PDF pg 135, pg 114, 6.3.3 Caching mode page <> [s/b] <> ENDL 535 PDF pg 135, pg 114, 6.3.3 Caching mode page <> [s/b] <> ENDL 536 PDF pg 135, pg 114, 6.3.3 Caching mode page <> [s/b] <> ENDL 537 PDF pg 136, pg 115, 6.3.4 Read-Write Error Recovery mode page <> [s/b] <> for consistency with table 93. ENDL 538 PDF pg 136, pg 115, 6.3.4 Read-Write Error Recovery mode page <> [s/b] <> to be consistent with the previous sentence ENDL 539 PDF pg 137, pg 116, 6.3.4 Read-Write Error Recovery mode page <> [s/b] <> ENDL 540 PDF pg 137, pg 116, 6.3.4 Read-Write Error Recovery mode page In keeping with the usage elsewhere in this standard, note 26 should be a table footnote in table 108. ENDL 541 PDF pg 138, pg 117, 6.3.4 Read-Write Error Recovery mode page <> [s/b] <> [2 times on this page] ENDL 542 PDF pg 139, pg 118, 6.3.4 Read-Write Error Recovery mode page <> [s/b] <> ENDL 543 PDF pg 140, pg 119, 6.3.4 Read-Write Error Recovery mode page <> [s/b] <> ENDL 544 PDF pg 140, pg 119, 6.3.4 Read-Write Error Recovery mode page Delete <> The last sentence in the paragraph says the same thing. ENDL 545 PDF pg 140, pg 119, 6.3.4 Read-Write Error Recovery mode page <> [s/b] <> ENDL 546 PDF pg 140, pg 119, 6.3.5 Verify Error Recovery mode page Delete <> The last sentence of the next paragraph says the same thing. ENDL 547 PDF pg 140, pg 119, 6.3.5 Verify Error Recovery mode page Insert a paragraph break for consistency with the Read-Write Error Recovery mode page definition. ENDL 548 PDF pg 140, pg 119, 6.3.5 Verify Error Recovery mode page Is note 27 necessary? If it is, does it belong 6.3.4 too? ENDL 549 PDF pg 141, pg 120, 6.3.6 XOR Control mode page <> [s/b] <> ENDL 550 PDF pg 141, pg 120, 6.3.6 XOR Control mode page <> [s/b] <> [2 times on this page] ENDL 551 PDF pg 141, pg 120, 6.3.6 XOR Control mode page <> [s/b] <> ENDL 552 PDF pg 142, pg 121, 6.4.2 Block Limits VPD page <> [s/b] <> ENDL 553 PDF pg 142, pg 121, 6.4.2 Block Limits VPD page <> [s/b] <> ************************************************************** Comments attached to Abs ballot from Elwood Parsons of Foxconn Electronics: Abstain due to lack of expertise. ************************************************************** Comments attached to Yes ballot from Rob Elliott of Hewlett Packard Co.: HPQ #1 PDF Page iv ANSI page Change 2003 to 2004 HPQ #2 PDF Page 10 4 Models Need to mention format corrupted and the additional sense code that results from media access commands when the medium is in that state. HPQ #3 PDF Page 11 4.4 Logical blocks Add a unit attention and a new additional sense code CAPACITY DATA HAS CHANGED whenever any of the READ CAPACITY data changes (number of blocks, block size, or various protection information settings). HPQ #4 PDF Page 18 4.13 Model for XOR commands Move 4.13.1.nn to 4.13.2, 4.13.3, etc. to eliminate hanging paragraph and unnecessary nesting HPQ #5 PDF Page 30 5.1 Commands overview To require that SBC-2 devices support long LBAs and deprecate usage of short LBAs... a) change READ (6) and READ (10) from mandatory to optional in table 9 and in note 9 b) change READ CAPACITY (10) from mandatory to optional in table 9 c) mark each of these commands as mandatory in table 9 if any of their smaller counterparts are implemented (the same rule that currently exists for WRITE (16)): LOCK UNLOCK CACHE (16) PRE-FETCH (16) READ DEFECT DATA (12) READ LONG (16) SYNCHRONIZE CACHE (16) VERIFY (16) WRITE AND VERIFY (16) WRITE LONG (16) WRITE SAME (16) XDREAD (32) XDWRITE (32) XDWRITEREAD (32) XPWRITE (32) HPQ #6 PDF Page 30 5.1 Commands overview To not require that SBC-2 devices support long LBAs... a) change READ (16) from mandatory to optional in table 9 b) change WRITE (16) from mandatory (if any WRITE is supported) to optional in table 9 and remove footnote d HPQ #7 PDF Page 88 5.37 WRITE LONG (10) command Delete "NOTE 21 - Any other bytes that can be corrected by ECC should be included (e.g., a data synchronization mark within the area covered by ECC). A READ LONG command may be issued before issuing a WRITE LONG command." Let READ LONG define the complete contents of the data. HPQ #8 PDF Page 108 6.3.1 Mode parameters overview Table 102 after "Management" add "mode page" HPQ #9 PDF Page 109 6.3.2.1 Mode parameter block descriptors overview After: A unit attention condition add: with an additional sense code set to MODE PARAMETERS CHANGED ************************************************************** Comments attached to No ballot from George O. Penokie of IBM Corp.: IBM #1 PDF Page vi Revision Information This section needs to be removed before going to public review. IBM #2 PDF Page 1 1 Scope This information is not necessary << (this bit was formerly reserved for direct-access device types, so is just marked reserved in this standard); >> IBM #3 PDF Page 3 2.1 Normative references overview This should be << (e.g., including BSI, JIS, and DIN). IBM #4 PDF Page 3 2.1 Normative references overview This should be << (i.e., ISO, IEC, CEN/CENELEC, ITUT); >> IBM #5 PDF Page 5 3.1.8 command descriptor block (CDB): This should be << server. See SPC-3. >>. The space is missing. IBM #6 PDF Page 6 3.1.25 logical unit reset event: This << logical unit as described in SAM-3. >> should be << logical unit. See SAM-3. >> IBM #7 PDF Page 6 3.1.28 non-volatile medium: This << cycles. An example of this is a disk within a device that stores data as magnetic field changes that do not require device power to exist. >> should be << cycles (e.g., a disk within a device that stores data as magnetic field changes that do not require device power to exist). >> IBM #8 PDF Page 7 3.1.44 volatile medium: This << power cycles. An example of this is a silicon memory device that loses data written to it if device power is lost. >> should be << power cycles (e.g.,a silicon memory device that loses data written to it if device power is lost). >> IBM #9 PDF Page 10 4.2 Direct-access device type model overview This << and can be read without >> should be << and are able to be read without >> IBM #10 PDF Page 10 4.3.1 Removable medium overview This << cartridge (or jacket) to prevent >> should be << cartridge or jacket to prevent >> 4.3.2 Removable medium with an attached medium changer This << Only one medium transport element is permitted (element 0) and only one data transfer element is permitted. >> should be deleted as it not relevant to this standard and should instead be in SMC-2. IBM #12 PDF Page 11 4.4 Logical blocks This << A READ CAPACITY command may be issued to determine the value of [n-1]. >> should be changed to << A READ CAPACITY command should be issued to determine the value of [n-1]. >>. The reason is that we are trying to encourage the use of READ CAPACITY. IBM #13 PDF Page 11 4.4 Logical blocks This << The READ CAPACITY data (see 5.14) describes the block lengths that are used on the medium. >> should be << The parameter data returned by the READ CAPACITY command (see 5.14) describes the block lengths that are used on the medium. >> IBM #14 PDF Page 11 4.5 Ready state This << commands can be processed. >> should be << commands are able to be processed. >> IBM #15 PDF Page 12 4.6 Initialization This << Parameters related to the geometry and performance characteristics may be set with the MODE SELECT command prior to the format operation. >> is not really true now that we have eliminated the geometry and the format device mode pages, therefore it should be deleted. IBM #16 PDF Page 12 4.6 Initialization This << After changing the mode parameter block descriptor with MODE SELECT, the new values do not become effective until FORMAT UNIT command completes. >> is a gross oversimplification that is not accurate in all cases. The best thing to do is delete it as everything is correctly defined in the mode page header descriptions. IBM #17 PDF Page 12 4.8 Medium defects This << defects that can cause user data >> should be << defects that may cause user data >> IBM #18 PDF Page 12 4.8 Medium defects This << (to reference while formatting) >> states no useful information and should be deleted or at least made into an (e.g., ...) to indicate an example of what it could be used for. IBM #19 PDF Page 13 4.8 Medium defects This << this manner can be specified in the >> should be << this manner may be specified in the >> IBM #20 PDF Page 13 4.9 Cache memory This << block and can increase the overall data >> should be << block and may increase the overall data >> IBM #21 PDF Page 13 4.9 Cache memory This << VERIFY and WRITE AND VERIFY commands >> should be << VERIFY command or WRITE AND VERIFY command >> IBM #22 PDF Page 14 4.9 Cache memory This << VERIFY command or WRITE AND VERIFY >> should be << VERIFY command or WRITE AND VERIFY command >> IBM #23 PDF Page 14 4.10 Reservations This << enters the current task state for the first time. >> should be << enters the current task state (see SAM-3) for the first time. >> IBM #24 PDF Page 15 Table 3 Global Having tables with footnotes were the footnote do no occur at every page break makes it difficult to the reader to find the footnote or even know if there are any footnote. The footnotes should be changed to appear on every page. Also, not have a line at the end of each page break is confusing. It looks like something is missing from the table. IBM #25 PDF Page 16 4.11 Error reporting This << a READ LONG or WRITE LONG command did >> should be << a READ LONG command or WRITE LONG command did >> IBM #26 PDF Page 17 4.11 Error reporting Table 5 All the command names should have << command >> after them. IBM #27 PDF Page 17 4.12.1 Examples overview The statement << The following examples show some typical variations >> should be << This clause describes examples of some typical variations >> IBM #28 PDF Page 17 4.12.2 Rotating media This << tracks that can be accessed without >> should be << tracks that are accessed without >> IBM #29 PDF Page 17 4.12.2 Rotating media This <> should be << formatted by the manufacture. >> IBM #30 PDF Page 18 4.12.2 Rotating media This <> should be << some aspects may be evaluated and controlled >> IBM #31 PDF Page 18 4.12.2 Rotating media This << The READ LONG and WRITE LONG commands >> should be << The READ LONG command and WRITE LONG command >> IBM #32 PDF Page 18 4.12.2 Rotating media This << when generating mirror copies.o >> should be << when generating mirror copies. >> as there is no reason to quote that phrase. IBM #33 PDF Page 18 4.13.1.1.1 Storage array controller supervised XOR operations overview This << XDWRITE, XPWRITE, and XDREAD.>> should be an a,b,c list or at least << XDWRITE command, XPWRITE command, and XDREAD command.>> IBM #34 PDF Page 18 4.13.1.1.1 Storage array controller supervised XOR operations overview This << XDWRITE followed by XDREAD.>> should be << XDWRITE command followed by XDREAD command.>> IBM #35 PDF Page 18 4.13.1.1.1 Storage array controller supervised XOR operations overview This << uses READ and WRITE commands for >> should be << uses READ commands and WRITE commands for >> IBM #36 PDF Page 19 4.13.1.1.2 Update write operation (storage array controller supervised) This << XOR data (received in the previous XDREAD command) to the >> should be << XOR data (i.e., XOR data received in the previous XDREAD command) to the >> 4.13.1.1.3 Regenerate operation (storage array controller supervised) This << all devices (except the failed device) in the redundancy >> should be << all devices, except the failed device, in the redundancy IBM #38 PDF Page 19 4.13.1.1.4 Rebuild operation (storage array controller supervised) This << all devices (except the failed device) in the redundancy >> should be << all devices, except the failed device, in the redundancy IBM #39 PDF Page 20 4.13.1.2.1 Additional array subsystem considerations overview This << to any array subsystem, but describes how use of the XOR >> does not read very well. I think it should be << to any array subsystem, and describes how use of the XOR >> IBM #40 PDF Page 20 4.13.1.2.2 Buffer full status handling This << This locks up part or all (depending on the size of the devices buffer and the size of the XOR data) of the deviceAs buffer space. >> should be << Depending on the size of the deviceAs buffer and the size of the XOR data, this locks up part or all of the deviceAs buffer space. >> IBM #41 PDF Page 20 4.13.1.2.3 Access to an inconsistent stripe This << updated (making the stripe consistent again). >> should be << updated (i.e., making the stripe consistent again). >> IBM #42 PDF Page 21 4.13.1.3.2 Primary errors - errors resulting directly from the primary command This << primary command (primary target) and are not due >> should be << primary command (i.e., primary target) and are not due >> IBM #43 PDF Page 22 4.14.2.1 START STOP UNIT and power conditions state machine overview Figure 2 This << but may have additional characteristics unique to this standard. >> sounds like we don't know what our own standards have in them. Either SPC-3 have additional characteristics or not. I think the <> should be deleted. IBM #44 PDF Page 22 4.14.2.2.1 SSU_PC0:Powered_on state description This << This logical unit shall enter this >> should be << The logical unit shall enter this >> IBM #45 PDF Page 30 5.1 Commands for direct-access devices overview This << indicated by the oProtection informationo column. >> should be << indicated by the protection information column. >> IBM #46 PDF Page 33 5.1 Commands for direct-access devices overview Table 9 This << If either PERSISTENT RESERVE IN or PERSISTENT RESERVE OUT is implemented, >> should be << If either PERSISTENT RESERVE IN command or PERSISTENT RESERVE OUT command is implemented, >> IBM #47 PDF Page 33 5.1 Commands for direct-access devices overview Table 9 This << If any of WRITE (6)/(10)/(12) is implemented, WRITE (16) shall also be >> should be << If any of WRITE (6)/(10)/(12) command is implemented, the WRITE (16) command shall also be >> IBM #48 PDF Page 35 5.4.1 FORMAT UNIT command overview This << new commands except INQUIRY, REPORT LUNS, and REQUEST SENSE with a CHECK >> should be << new commands except INQUIRY command, REPORT LUNS command , and REQUEST SENSE command with a CHECK >> IBM #49 PDF Page 42 5.4.2.4.1 Address descriptor formats overview This << SEND DIAGNOSTIC and RECEIVE DIAGNOSTIC RESULTS commands. >> should be << SEND DIAGNOSTIC command and RECEIVE DIAGNOSTIC RESULTS command. >> IBM #50 PDF Page 42 5.4.2.4.1 Address descriptor formats overview This << the FORMAT UNIT and READ DEFECT DATA commands; >> should be << the FORMAT UNIT command and READ DEFECT DATA command; >> IBM #51 PDF Page 45 5.5 LOCK UNLOCK CACHE (10) command This << memory are actually locked. >> should be << memory are locked. >> IBM #52 PDF Page 48 5.9 READ (6) command This << READ (6) command; however, no default values are >> should be << READ (6) command. However, no default values are >> IBM #53 PDF Page 52 5.10 READ (10) command Table 33 This << READ (10), READ (12),and READ (16) commands with >> should be << a READ (10) command, READ (12) command, or READ (16) command with >> IBM #54 PDF Page 52 5.10 READ (10) command Table 33 This << READ (10), READ (12), and READ (16) commands >> should be << the READ (10) command, READ (12) command, or READ (16) command >> IBM #55 PDF Page 55 5.13 READ (32) command For the LOGICAL BLOCK APPLICATION TAG and the LOGICAL BLOCK REFERENCE TAG there is no indication as to whether the same value that is received is checked against all the LBAs that are read or if it is incremented in some fashion for each LBA. The current definition seems to imply there is only one value for all the LBAs that are read. This needs to be made clear. IBM #56 PDF Page 59 5.16 READ DEFECT DATA (10) command This is a bad reference as I have no idea where to look for this << (see the DEFECT LIST FORMAT field in the defect list header). >> it should be << (see table x.x.x for the DEFECT LIST FORMAT field). >> IBM #57 PDF Page 60 5.16 READ DEFECT DATA (10) command This << SCSI device >> should be << logical unit >>. IBM #58 PDF Page 60 5.16 READ DEFECT DATA (10) command This << SCSI device >> should be << logical unit >>. IBM #59 PDF Page 62 5.18 READ LONG (10) command This << the data bytes; however, they should be in the same >> should be << the data bytes. However, they should be in the same >> IBM #60 PDF Page 62 5.20 REASSIGN BLOCKS command This << protection information, if present, >> should be << protection information, if any, >> as it is everywhere else. IBM #61 PDF Page 63 5.20 REASSIGN BLOCKS command This << protection information, if present, >> should be << protection information, if any, >> as it is everywhere else. IBM #62 PDF Page 64 5.20 REASSIGN BLOCKS command NOTE 18 This note looks like it should be part of the main text instead of a note. IBM #63 PDF Page 72 5.24 VERIFY (10) command Table 59; 000b value The REF_CHK = 1 should have footnote reference to the following footnote: If the RTO_EN bit is set to zero in the long read capacity data (see 5.15), the device server checks the logical block reference tag by comparing it to the lower 4 bytes of the LBA associated with the logical block. If the RTO_EN bit is set to one, the device server checks the logical block reference tag only if it has knowledge of the contents of the LOGICAL BLOCK REFERENCE TAG field. The method for acquiring this knowledge is not defined by this standard. IBM #64 PDF Page 78 5.27 VERIFY (32) command For the LOGICAL BLOCK APPLICATION TAG and the LOGICAL BLOCK REFERENCE TAG there is no indication as to whether the same value that is received is checked against all the LBAs that are verified or if it is incremented in some fashion for each LBA. The current definition seems to imply there is only one value for all the LBAs that are verified. This needs to be made clear. IBM #65 PDF Page 82 5.29 WRITE (10) command This <> should be << <> IBM #66 PDF Page 82 5.29 WRITE (10) command This << terminate WRITE (10), WRITE (12), and WRITE (16) commands >> should be << terminate the WRITE (10) command, WRITE (12) command, or WRITE (16) command >> IBM #67 PDF Page 84 5.32 WRITE (32) command For the LOGICAL BLOCK APPLICATION TAG and the LOGICAL BLOCK REFERENCE TAG there is no indication as to whether the same value that is received is checked against all the LBAs that are read or if it is incremented in some fashion for each LBA. The current definition seems to imply there is only one value for all the LBAs that are read. This needs to be made clear. IBM #68 PDF Page 85 5.33 WRITE AND VERIFY (10) command This << data and includes protection information as >> should be << data and protection information, if any, as >> 5.34 WRITE AND VERIFY (12) command This << data and includes protection information as >> should be << data and protection information, if any, as >> IBM #70 PDF Page 86 5.35 WRITE AND VERIFY (16) command This << data and includes protection information as >> should be << data and protection information, if any, as >> IBM #71 PDF Page 87 5.36 WRITE AND VERIFY (32) command This << protection information, if any. >> should be << protection information, if any, as required by the WRPROTECT field and the medium format. The data is only transferred once from the application client to the device server. >> IBM #72 PDF Page 87 5.36 WRITE AND VERIFY (32) command This paragraph << If the RTO_EN bit is set to zero in the long read capacity data (see 5.15), the device server shall terminate the command with CHECK CONDITION status with a sense key of ILLEGAL REQUEST and an additional sense code set to INVALID COMMAND OPERATION CODE. If the RTO_EN bit is set to one, the device server may process the command. >> should be move to above table 75. That would make the WRITE AND VERIFY (32) the same format as the other WRITE AND VERIFY commands. IBM #73 PDF Page 87 5.36 WRITE AND VERIFY (32) command For the LOGICAL BLOCK APPLICATION TAG and the LOGICAL BLOCK REFERENCE TAG there is no indication as to whether the same value that is received is checked against all the LBAs that are verified or if it is incremented in some fashion for each LBA. The current definition seems to imply there is only one value for all the LBAs that are verified. This needs to be made clear. IBM #74 PDF Page 88 5.37 WRITE LONG (10) command Note 21 This << Any other bytes that can be corrected by ECC >> should be << Any other bytes that are able to be corrected by ECC >> IBM #75 PDF Page 88 5.37 WRITE LONG (10) command The term << exactly >> should be deleted as it adds nothing. IBM #76 PDF Page 89 5.39 WRITE SAME (10) command This << data and includes protection information as >> should be << data and protection information, if any, as >> IBM #77 PDF Page 90 5.39 WRITE SAME (10) command table 79, row 0 0 There is nothing said about what value should be placed in the Data Block Guard field. This should be fixed. A statement like this should be added <> IBM #78 PDF Page 90 5.40 WRITE SAME (16) command This << data and includes protection information as >> should be << data and protection information, if any, as >> IBM #79 PDF Page 92 5.41 WRITE SAME (32) command This << protection information, if any. >> should be << protection information, if any, as required by the WRPROTECT field and the medium format. >> IBM #80 PDF Page 92 5.41 WRITE SAME (32) command This paragraph << If the RTO_EN bit is set to zero in the long read capacity data (see 5.15), the device server shall terminate the command with CHECK CONDITION status with a sense key of ILLEGAL REQUEST and an additional sense code set to INVALID COMMAND OPERATION CODE. If the RTO_EN bit is set to one, the device server may process the command. >> should be move to above table 81. That would make the WRITE AND VERIFY (32) the same format as the other WRITE AND VERIFY commands. IBM #81 PDF Page 92 5.41 WRITE SAME(32) command For the LOGICAL BLOCK APPLICATION TAG and the LOGICAL BLOCK REFERENCE TAG there is no indication as to whether the same value that is received is written to all the LBAs to be written or if it is incremented in some fashion for each LBA. The current definition seems to imply there is only one value for all the LBAs that are written. This needs to be made clear. IBM #82 PDF Page 93 5.42 XDREAD (10) command This << data and includes protection information as >> should be << data and protection information, if any, as >> IBM #83 PDF Page 94 5.4.3 XDREAD (32) command This << data and includes protection information as >> should be << data and protection information, if any, as >> IBM #84 PDF Page 94 5.44 XDWRITE (10) command This << data and includes protection information as >> should be << data and protection information, if any, as >> IBM #85 PDF Page 95 5.45 XDWRITE (32) command This << data and includes protection information as >> should be << data and protection information, if any, as >> IBM #86 PDF Page 96 5.46 XDWRITEREAD (10) command This << data and includes protection information as >> should be << data and protection information, if any, as >> IBM #87 PDF Page 97 5.47 XDWRITEREAD (32) command This << data and includes protection information as >> should be << data and protection information, if any, as >> IBM #88 PDF Page 98 5.48 XPWRITE (10) command This << data and includes protection information as >> should be << data and protection information, if any, as >> IBM #89 PDF Page 100 5.49 XPWRITE (32) command This << data and includes protection information as >> should be << data and protection information, if any, as >> IBM #90 PDF Page 101 6.1.2 Translate Address Output diagnostic page This << command (see 5.4.2.4) - a short block format address, a long block format address, a physical sector format address, or a bytes from index format address - into any >> should be << command (see 5.4.2.4) (i.e., a short block format address, a long block format address, a physical sector format address, or a bytes from index format address) into any >> IBM #91 PDF Page 102 6.1.3 Translate Address Input diagnostic page This << vendor reserved area, etc.). >> should be << vendor reserved area). >> IBM #92 PDF Page 104 6.2.2 Format Status log page This << the most recent successful FORMAT UNIT command >> should be << the last successful FORMAT UNIT command >> IBM #93 PDF Page 105 6.2.2 Format Status log page This << the most recent FORMAT UNIT command >> should be << the last FORMAT UNIT command >> IBM #94 PDF Page 105 6.2.2 Format Status log page This << the most recently successful FORMAT UNIT command >> should be << the last successful FORMAT UNIT command >> IBM #95 PDF Page 105 6.2.2 Format Status log page This << the most recently successful FORMAT UNIT command >> should be << the last successful FORMAT UNIT command >> 6.2.3 Non-volatile Cache log page Table 98 This << Non-volatile cache is volatile (either permanently or temporarily, e.g., if batteries need to be recharged). >> should be << Non-volatile cache is either permanently or temporarily volatile (e.g., if batteries need to be recharged). >> IBM #97 PDF Page 109 6.3.2.1 Mode parameter block descriptors overview This << If it returns a mode >> should be << If the device server returns a mode >> IBM #98 PDF Page 109 6.3.2.1 Mode parameter block descriptors overview This << If it returns a mode >> should be << If the device server returns a mode >> IBM #99 PDF Page 109 6.3.2.1 Mode parameter block descriptors overview This << If it returns a mode >> should be << If the device server returns a mode >> IBM #100 PDF Page 109 6.3.2.1 Mode parameter block descriptors overview This << If it returns a mode >> should be << If the device server returns a mode >> IBM #101 PDF Page 110 6.3.2.3 Long LBA mode parameter block descriptor This << the MODE SELECT (10) and MODE SENSE (10) commands when >> should be << the MODE SELECT (10) command and MODE SENSE (10) command when >> IBM #102 PDF Page 112 6.3.3 Caching mode page This << use the NUMBER OF CACHE SEGMENTS field or the CACHE SEGMENT SIZE field, dependent upon the SIZE bit, to control the caching algorithm >> should be << use the NUMBER OF CACHE SEGMENTS field if SIZE is set to zero or the CACHE SEGMENT SIZE field if SIZE is set to one to control the caching algorithm >> IBM #103 PDF Page 112 6.3.3 Caching mode page This << upon Caching mode page bytes 4 through 11 and is operation and/or vendor-specific. >> makes no sense. Change to << upon Caching mode page bytes 4 through 11. >> IBM #104 PDF Page 112 6.3.3 Caching mode page This << A CAP bit set to zero specifies that caching analysis be disabled to reduce overhead time or to prevent nonpertinent operations from impacting tuning values. >> should be << A CAP bit set to zero specifies that caching analysis be disabled. >> IBM #105 PDF Page 113 6.3.3 Caching mode page Table 106 u Demand read retention priority and write retention priority This table combines the description of the DEMAND RETENTION PRIORITY field and the WRTIE RETENTION PRIORITY field into one table. This is very confusing. Make two tables, one for each field. IBM #106 PDF Page 113 6.3.3 Caching mode page table 106 This << WRITE or WRITE AND VERIFY command >> should be << WRITE command or WRITE AND VERIFY command >> IBM #107 PDF Page 113 6.3.3 Caching mode page table 106 This << WRITE or WRITE AND VERIFY command >> should be << WRITE command or WRITE AND VERIFY command >> 6.3.3 Caching mode page This << The MINIMUM PRE-FETCH field specifies either a number of blocks or a scalar multiplier of the TRANSFER LENGTH, depending upon the setting of the MF bit. >> should be << The MINIMUM PRE-FETCH field specifies a number of blocks if the MF bit set set to zero or a scalar multiplier of the TRANSFER LENGTH if the MF bit is set to one.>> IBM #109 PDF Page 114 6.3.3 Caching mode page This << according to the rules for reporting deferred errors. >> needs a reference as to were the rules are defined. IBM #110 PDF Page 114 6.3.3 Caching mode page This << The MAXIMUM PRE-FETCH field specifies either a number of blocks or a scalar multiplier of the TRANSFER LENGTH, depending upon the setting of the MF bit. >> should be << The MINIMUM PRE-FETCH field specifies a number of blocks if the MF bit set set to zero or a scalar multiplier of the TRANSFER LENGTH if the MF bit is set to one.>> IBM #111 PDF Page 114 6.3.3 Caching mode page This << reorder the sequence of writing addressed logical blocks in order to achieve a faster command completion. >> should be << reorder the sequence of writing addressed logical blocks. >> IBM #112 PDF Page 114 6.3.3 Caching mode page This << perform the SCSI buffer function. >> needs a reference as to were the SCSI buffer function is defined. IBM #113 PDF Page 115 6.3.4 Read-Write Error Recovery mode page This << medium (e.g., READ, WRITE, WRITE AND VERIFY, etc.). >> should be << medium (e.g., READ commands, WRITE commands , WRITE AND VERIFY commands). >> IBM #114 PDF Page 115 6.3.4 Read-Write Error Recovery mode page This << bits (EER, PER, DTE, and DCR) >> should be << bits (i.e., EER, PER, DTE, and DCR) >> IBM #115 PDF Page 115 6.3.4 Read-Write Error Recovery mode page This << bits (TB, EER, PER, DTE, and DCR) >> should be << bits (i.e., TB, EER, PER, DTE, and DCR) >> IBM #116 PDF Page 116 6.3.4 Read-Write Error Recovery mode page note 25 This << This bit may be used in image processing, audio, or video applications. >> should be << The TB (?) bit may be used in image processing, audio, or video applications. >> IBM #117 PDF Page 116 6.3.4 Read-Write Error Recovery mode page This << bits (EER, DCR, DTE, and PER) within >> should be << bits (i.e., EER, DCR, DTE, and PER) within >> IBM #118 PDF Page 116 6.3.4 Read-Write Error Recovery mode page This << definitions for EER, PER, DTE and DCR are contained >> should be << definitions for the EER bit, PER bit, DTE bit and DCR bit are contained >> IBM #119 PDF Page 117 6.3.4 Read-Write Error Recovery mode page Table 109 the 0 0 0 0 row This description needs to be rewritten. I suggest it be changed to: << Error correction and the full number of retries as specified in the READ RETRY COUNT field, WRITE RETRY COUNT field or VERIFY RETRY COUNT field (see 6.3.5) shall be attempted to recover the data. A CHECK CONDITION is not reported at the completion of the command for recovered errors. The command terminates with CHECK CONDITION status before the transfer count is exhausted only if an unrecoverable error is detected. If an unrecoverable data error occurred on a read operation, the data in the block with the unrecoverable error shall only be transferred if the TB bit is set to one. >> IBM #120 PDF Page 117 6.3.4 Read-Write Error Recovery mode page Table 109 the 0 0 0 1 row This description needs to be rewritten. I suggest it be changed to: <> IBM #121 PDF Page 117 6.3.4 Read-Write Error Recovery mode page Table 109 the 0 1 0 0 row This description needs to be rewritten. I suggest it be changed to: << Error correction and the full number of retries as specified in the READ RETRY COUNT field, WRITE RETRY COUNT field or VERIFY RETRY COUNT field (see 6.3.5) shall be attempted to recover the data. The command terminates with CHECK CONDITION status before the transfer count is exhausted only if an unrecoverable error is detected. If an unrecoverable data error occurred on a read operation, the data in the block with the unrecoverable error shall only be transferred if the TB bit is set to one. A CHECK CONDITION with a sense key of RECOVERED ERROR is reported at the completion of the command for any recoverable error that occurs. The INFORMATION field in the sense data shall contain the LBA of the last recovered error that occurred during the transfer.>> IBM #122 PDF Page 117 6.3.4 Read-Write Error Recovery mode page Table 109 the 0 1 0 1 row This description needs to be rewritten. I suggest it be changed to: <> IBM #123 PDF Page 118 6.3.4 Read-Write Error Recovery mode page Table 109 the 0 1 1 0 row This description needs to be rewritten. I suggest it be changed to: << Error correction and the full number of retries as specified in the READ RETRY COUNT field, WRITE RETRY COUNT field or VERIFY RETRY COUNT field (see 6.3.5) shall be attempted to recover the data. The command terminates with CHECK CONDITION status before the transfer count is exhausted if any recoverable or unrecoverable error is detected. The INFORMATION field in the sense data shall contain the LBA of the block in error. If an unrecoverable data error occurred on a read operation, the data in the block with the unrecoverable error shall only be transferred if the TB bit is set to one.>> IBM #124 PDF Page 118 6.3.4 Read-Write Error Recovery mode page Table 109 the 0 1 1 1 row This description needs to be rewritten. I suggest it be changed to: <> IBM #125 PDF Page 118 6.3.4 Read-Write Error Recovery mode page Table 109 the 1 0 0 0 row This description needs to be rewritten. I suggest it be changed to: << Error correction and the fewest possible number of retries shall be attempted to recover the data. A CHECK CONDITION is not reported at the completion of the command for recovered errors. The command terminates with CHECK CONDITION status before the transfer count is exhausted only if an unrecoverable error is detected. If an unrecoverable data error occurred on a read operation, the data in the block with the unrecoverable error shall only be transferred if the TB bit is set to one. >> IBM #126 PDF Page 118 6.3.4 Read-Write Error Recovery mode page Table 109 the 1 1 0 0 row This description needs to be rewritten. I suggest it be changed to: << Error correction and the fewest possible number of retries shall be attempted to recover the data. The command terminates with CHECK CONDITION status before the transfer count is exhausted only if an unrecoverable error is detected. If an unrecoverable data error occurred on a read operation, the data in the block with the unrecoverable error shall only be transferred if the TB bit is set to one. A CHECK CONDITION with a sense key of RECOVERED ERROR is reported at the completion of the command for any recoverable error that occurs. The INFORMATION field in the sense data shall contain the LBA of the last recovered error that occurred during the transfer. >> IBM #127 PDF Page 118 6.3.4 Read-Write Error Recovery mode page Table 109 the 1 1 1 0 row This description needs to be rewritten. I suggest it be changed to: << Error correction and the fewest possible number of retries shall be attempted to recover the data. The command terminates with CHECK CONDITION status before the transfer count is exhausted if any recoverable or unrecoverable error is detected. The INFORMATION field in the sense data shall contain the LBA of the block in error. If an unrecoverable data error occurred on a read operation, the data in the block with the unrecoverable error shall only be transferred if the TB bit is set to one. >> IBM #128 PDF Page 119 6.3.4 Read-Write Error Recovery mode page This paragraph combines two fields into one description which makes it unclear as to what the field or fields are that are being defined. It needs to be make into two paragraphs such as: << The READ RETRY COUNT field specifies the number of times that the device server shall attempt its recovery algorithm during a read operation. If the READ RETRY COUNT field and the RECOVERY TIME LIMIT field are both specified in a MODE SELECT command, the field that requires the least time for data error recovery actions shall have priority. The WRITE RETRY COUNT fields specifies the number of times that the device server shall attempt its recovery algorithm during a write operation. If the WRITE RETRY COUNT field and the RECOVERY TIME LIMIT field are both specified in a MODE SELECT command, the field that requires the least time for data error recovery actions shall have priority. >> IBM #129 PDF Page 119 6.3.5 Verify Error Recovery mode page This << The EER, PER, DTE, and DCR bits are defined in 6.3.4. >> should be << The EER bit, PER bit, DTE bit, and DCR bit are defined in 6.3.4. >> IBM #130 PDF Page 119 6.3.5 Verify Error Recovery mode page Note 27 This << set the EER bit to zero, the PER, DTE, and DCR bits to one and the number of retries and recovery time limit to zero. >> should be << set the EER bit to zero, the PER bit to one, the DTE bit to one, and DCR bit to one and the number of retries to zero and recovery time limit to zero. >> IBM #131 PDF Page 120 6.3.6 XOR Control mode page This << within a device. >> should be << within the logical unit. >> IBM #132 PDF Page 120 6.3.6 XOR Control mode page This << parameters of the target. >> should be << parameters of the logical unit. >> IBM #133 PDF Page 120 6.3.6 XOR Control mode page This << within a device. >> should be << within the logical unit. >> IBM #134 PDF Page 120 6.3.6 XOR Control mode page This << the target accepts >> should be << the device server accepts >> IBM #135 PDF Page 121 6.4.2 Block Limits VPD page This << a single PRE-FETCH, READ, VERIFY, WRITE, WRITE AND VERIFY, XDREAD, XDWRITE, XDWRITEREAD, or XPWRITE command. >> should be << a single PRE-FETCH command, READ command, VERIFY command, WRITE command, WRITE AND VERIFY command, XDREAD command, XDWRITE command, XDWRITEREAD command, or XPWRITE command. >> IBM #136 PDF Page 121 6.4.2 Block Limits VPD page This << a single PRE-FETCH, READ, VERIFY, WRITE, WRITE AND VERIFY, XDREAD, XDWRITE, XDWRITEREAD, or XPWRITE command. >> should be << a single PRE-FETCH command, READ command, VERIFY command, WRITE command, WRITE AND VERIFY command, XDREAD command, XDWRITE command, XDWRITEREAD command, or XPWRITE command. >> IBM #137 PDF Page 121 6.4.2 Block Limits VPD page This << a single PRE-FETCH, READ, VERIFY, WRITE, WRITE AND VERIFY, XDREAD, XDWRITE, XDWRITEREAD, or XPWRITE command. >> should be << a single PRE-FETCH command, READ command, VERIFY command, WRITE command, WRITE AND VERIFY command, XDREAD command, XDWRITE command, XDWRITEREAD command, or XPWRITE command. >> IBM #138 PDF Page 122 A.2 Update write operation This << Figure A.1 illustrates a read-modify-write >> should be << Figure A.1 shows a read-modify-write >> IBM #139 PDF Page 122 A.2 Update write operation This << a data disk device (holding protected user data), and a parity disk device (holding check data) >> should be << a data disk device that holds protected user data, and a parity disk device that holds check data >> IBM #140 PDF Page 122 A.2 Update write operation This << XDWRITE, XDREAD, and XPWRITE. XDWRITEREAD may be used in place of any sequence of XDWRITE followed by XDREAD. >> should be << XDWRITE command, XDREAD command, and XPWRITE command. XDWRITEREAD command may be used in place of any sequence of an XDWRITE command followed by an XDREAD command. >> IBM #141 PDF Page 122 A.2 Update write operation This << disk device (the supervising storage array controller does not yet have the intermediate XOR data for this command; the purpose of issuing the XPWRITE command at this time is to cause the parity disk device to begin reading XOR data from its medium to its buffer memory). >> should be << disk device (i.e., the supervising storage array controller does not yet have the intermediate XOR data for this command. The purpose of issuing the XPWRITE command at this time is to cause the parity disk device to begin reading XOR data from its medium to its buffer memory). >> IBM #142 PDF Page 122 A.2 Update write operation This << XOR data (read with the XDREAD command) available >> should be << XOR data (i.e., read with the XDREAD command) available >> IBM #143 PDF Page 123 A.3 Regenerate operation This << Figure A.2 illustrates a regenerate >> should be << Figure A.2 shows a regenerate >> IBM #144 PDF Page 123 A.3 Regenerate operation This << used: READ, XDWRITE, and XDREAD. XDWRITEREAD may be used in place of any sequence of XDWRITE followed by XDREAD. >> should be << used: READ command, XDWRITE command, and XDREAD command. XDWRITEREAD command may be used in place of any sequence of an XDWRITE command followed by an XDREAD command. >> IBM #145 PDF Page 123 A.2 Update write operation Figure A.1 This << XDWRITE >> should be << XDWRITE command >> IBM #146 PDF Page 123 A.2 Update write operation Figure A.1 This << XDREAD >> should be << XDREAD command>> IBM #147 PDF Page 123 A.2 Update write operation Figure A.1 This << XPWRITE >> should be << XPWRITE command>> IBM #148 PDF Page 124 A.3 Regenerate operation Figure A.2 This << READ >> should be << READ command >>. IBM #149 PDF Page 124 A.3 Regenerate operation Figure A.2 This << XDWRITE or XDWRITEREAD >> should be << XDWRITE command or XDWRITEREAD command>> IBM #150 PDF Page 124 A.3 Regenerate operation Figure A.2 This << XDREAD or XDWRITEREAD >> should be << XDREAD command or XDWRITEREAD command>> IBM #151 PDF Page 124 A.3 Regenerate operation Figure A.2 This << XDWRITE or XDWRITEREAD >> should be << XDWRITE command or XDWRITEREAD command>> IBM #152 PDF Page 124 A.3 Regenerate operation Figure A.2 This << XDREAD or XDWRITEREAD >> should be << XDREAD command or XDWRITEREAD command>> IBM #153 PDF Page 124 A.4 Rebuild operation This << Figure A.3 illustrates a rebuild >> should be << Figure A.3 shows a rebuild >> IBM #154 PDF Page 124 A.4 Rebuild operation This << used: READ, XDWRITE, XDREAD, and WRITE. XDWRITEREAD may be used in place of any sequence of XDWRITE followed by XDREAD. >> should be << used: READ command, XDWRITE command, XDREAD command, and WRITE command. XDWRITEREAD command may be used in place of any sequence of an XDWRITE command followed by an XDREAD command. >> IBM #155 PDF Page 125 A.4 Rebuild operation This << is the "rebuilt" data and is sent >> should be << is the rebuilt data and is sent >> IBM #156 PDF Page 125 A.4 Rebuild operation Figure A.3 This << READ >> should be << READ command >>. IBM #157 PDF Page 125 A.4 Rebuild operation Figure A.3 This << XDWRITE or XDWRITEREAD >> should be << XDWRITE command or XDWRITEREAD command>> IBM #158 PDF Page 125 A.4 Rebuild operation Figure A.3 This << XDREAD or XDWRITEREAD >> should be << XDREAD command or XDWRITEREAD command>> IBM #159 PDF Page 125 A.4 Rebuild operation Figure A.3 This << WRITE >> should be << WRITE command >>. ************************************************************** Comments attached to Abs ballot from David Hawks of Iomega Corp.: Iomega has not chosen to carefully review SBC-2. Instead, Iomega believes SBC-3 should obsolete the Removable Media portion of SBC. Actual massively distributed boot BIOS and operating systems have never fully supported Removable SBC as specified by t10.org. Consequently, designs for broad compatibility in read, boot, and write are moving to removable MMC UDF for more than 32 GiB/disc, as in the example of the Iomega REV drive, else to detachable but not removable SBC, as in the example of USB adapters for small Compact Flash. ************************************************************** Comments attached to Yes ballot from John Lohmeyer of LSI Logic Corp.: LSI #1 PDF Page vi Pages vi -- x Remove revision information from copy that goes to public review. LSI #2 PDF Page 2 1 Scope In item l) of the last list in this subclause, "this" s/b "the". LSI #3 PDF Page 5 3.1.6 check data "allows" s/b "may allow" LSI #4 PDF Page 6 3.1.26 media: Either remove this definition or add a definition of medium. LSI #5 PDF Page 6 3.1.27 non-volatile cache memory: It is probably better to define non-volatile separately from cache memory (and medium). We would then have separate definitions for non-volatile, volatile, cache memory, and medum. LSI #6 PDF Page 7 3.2 Symbols and abbreviations Please add: CRC cyclic redundancy check LSI #7 PDF Page 12 4.5 Ready state In the last paragraph, replace "set to" with "set to one to". LSI #8 PDF Page 12 4.8 Medium defects First paragraph: Some block devices provide the application client control (of what?) through use of the mode parameters. LSI #9 PDF Page 13 4.8 Medium defects In list item d), replace "theprevious" with "the previous". LSI #10 PDF Page 13 4.9 Cache memory In the fourth paragraph, last sentence, consider adding "without power" to the end of the sentence, "There may be a limit on the amount of time a non-volatile cache is able to retain data." LSI #11 PDF Page 13 4.9 Cache memory Fifth paragraph, first sentence. Change "is written" to "is to be written". LSI #12 PDF Page 14 4.9 Cache memory Seventh paragraph, last sentence. Delete "specifies". Alternatively, re-word this sentence as follows: "An FUA_NV bit of one permits the device server to access a non-volatile cache memory rather than the medium." LSI #13 PDF Page 14 4.9 Cache memory Ninth paragraph, first sentence should read, "When a VERIFY command or a WRITE AND VERIFY command is processed,..." LSI #14 PDF Page 33 5.1 Commands for direct-access block devices Table 9, Note e Change 'of'' to 'if'. LSI #15 PDF Page 36 5.4.1 FORMAT UNIT command overview Note 7 ahould be deleted. The information is already covered in the paragraph above table 13. LSI #16 PDF Page 65 5.21 START STOP UNIT command third paragraph from the end of this subclause Replace this paragraph with "It is not an error to specify that the logical unit transition to its current power condition." ************************************************************** Comments attached to No ballot from Mark Evans of Maxtor Corp.: Maxtor #1 PDF Page 1 Change the text in item (a) to, "Permit an application client to communicate over a SCSI service delivery subsystem with a logical unit that declares itself to be a direct-access device, write-once device, or optical memory device in the device type field of the INQUIRY command response data;" Maxtor #2 PDF Page 1 Change "devices" to "device". Maxtor #3 PDF Page 5 Add, "The set of contiguous logical blocks may be all of logical blocks on the device." Maxtor #4 PDF Page 7 This definition is not clear. I don't understand it well enough to offer a recommendation. Maxtor #5 PDF Page 10 Should "write" be changed to "data-out"? One way or the other, we should have a definition or at least be consistent. Maxtor #6 PDF Page 10 Should "read" be changed to "data-in"? One way or the other, we should have a definition or at least be consistent. Maxtor #7 PDF Page 10 Change to, "A block device containing a removable medium may require receipt of a START STOP UNIT command to become accessible for data-out or data-in operations." Maxtor #8 PDF Page 11 Change "need" to "may". Maxtor #9 PDF Page 13 Change to, "Cache memory is an area of temporary storage with fast access time that is implemented in most block devices to enhance performance. Cache memory exists separately from the user data stored on the medium and is not uniquely accessible by the application client. Use of cache memory for data-out or data-in operations may reduce the access time to a logical block and can increase the overall data throughput." Maxtor #10 PDF Page 13 Change to, "During data-in operations, the block devices use..." Maxtor #11 PDF Page 13 Change to, "During data-out operations, block devices use..." Maxtor #12 PDF Page 13 Change to, "...to be written..." Maxtor #13 PDF Page 13 I think this is supposed to be something like, "When the cache memory is filled with blocks of data that are being stored for possible future access, new blocks of data that are to be stored replace those currently in cache memory. The disable page out (DPO) bit allows the application client to influence the replacement of logical blocks in the cache. For data-out operations, setting this bit to one advises the device server to not replace existing blocks in the cache memory with the new data-out data. For data-in operations, setting the DPO bit to one advises the device server to not replace existing blocks in the cache memory with the new data-in data." Maxtor #14 PDF Page 13 Change to, "...READ and WRITE..." Maxtor #15 PDF Page 14 Delete, "effectively". Maxtor #16 PDF Page 18 Delete, "traditionally". Maxtor #17 PDF Page 18 Delete, "virtually". Maxtor #18 PDF Page 20 This is a huge paragraph that is not clear to me. I recommend making this be more paragraphs something like: A stripe is a set of corresponding strips of consecutively addressed storage from two or more block devices. A strip is an equal division of the storage capacity in a set of consecutively addressed LBAs on a single block device. When the storage array controller issues an update write to a device, the data in the device has been updated when successful status is returned for the command. Until the device containing check data has been updated, however, the associated stripe in the redundancy group is not consistent (e.g., performing an XOR operation on the protected data does not produce the check data). The storage array controller shall keep track of this window of inconsistency and make sure that a regenerate or rebuild operation for any data extent within the stripe is not attempted until after the device containing check data has been updated (making the stripe consistent again). For multi-initiator systems, tracking the updates may be more complex because each storage array controller needs to ensure that a second storage array controller is not writing to a stripe that the first storage array controller is regenerating or rebuilding. The coordination between storage array controllers is system specific and is beyond the scope of this standard. A storage array controller needs to prevent data corruption due to a temporarily inconsistent stripe in one case. When an XDWRITE or XDWRITEREAD command has been issued and completed, the device containing protected data has been updated but the device containing check data has not. The stripe is inconsistent until the XPWRITE command to the device containing check data returns completion status. Maxtor #19 PDF Page 21 Delete "directly" unless there are other errors to be defined that result "indirectly" from the primary command. Maxtor #20 PDF Page 29 Delete, "primarily". Maxtor #21 PDF Page 29 Delete, "primarily". Maxtor #22 PDF Page 30 Change to "O". Maxtor #23 PDF Page 31 Change to "O". Maxtor #24 PDF Page 33 Delete note. Maxtor #25 PDF Page 45 Delete, "actually" Maxtor #26 PDF Page 48 Delete, "directly". Maxtor #27 PDF Page 66 Change to, "The synchronize cache function is also performed by the device server as the result of other commands defined in other clauses of this standard (e.g., when the FUA bit is set to one for a write command, contiguous cached data may be written to the media)." Maxtor #28 PDF Page 115 Change to, "(e.g., READ, WRITE, and WRITE AND VERIFY). Maxtor #29 PDF Page 115 Change to, "The device server shall report any failures that occur during the reallocation operation. Error reporting as specifed by the error recovery bits (EER, PER, DTE, and DCR) shall be performed only after completion of the reallocation." Maxtor #30 PDF Page 115 Change to, "The device server shall report any failures that occur during the reallocation operation. Error reporting as specifed by the error recovery bits shall be performed only after completion of the reallocation." Maxtor #31 PDF Page 115 Add, "The data returned in this case is vendor specific." Maxtor #32 PDF Page 116 There needs to be a tie-in to "fabricated data". This note could be deleted with the recommended insertion in the previous paragraph. Maxtor #33 PDF Page 116 Change to, "an". Maxtor #34 PDF Page 120 Change to, "shall enable". ************************************************************** Late Comments from Jeff Mastro of Microsoft Corp.: MSFT #1 4.4 Logical blocks Add notification of a change to block size / number of blocks Modification of READ CAPACITY data (block size or number) shall result in a unit attention condition with a sense code indicating such change. ************************************************************** Comments attached to Abs ballot from Jim Jones of Quantum Corp.: Not materially affected by this proposal. ************************************************************** Comments attached to No ballot from Gerald Houlder of Seagate Technology: Seagate #1 PDF Page 10 SPC-3, under table 237, has this statement: "For a list of commands affected by the SWP bit and details of the WP bit see the command standard (see 3.1.18) for the specific device type." I would expect SBC-2 to have such a list in the model section, but it doesn't even mention a write protected mode. We have SPC-3 making a promise that SBC-2 is not keeping. A write protect section listing affected commands should be added to the model. Seagate #2 PDF Page 52 Change to: This knowledge may be obtained by use of the READ (32) command (see 5.13) or by a method not defined by this standard. Seagate #3 PDF Page 52 Change to: This knowledge may be obtained by use of the READ (32) command (see 5.13) or by a method not defined by this standard. Seagate #4 PDF Page 71 Change to: This knowledge may be obtained by use of the VERIFY (32) command (see 5.27) or by a method not defined by this standard. Seagate #5 PDF Page 71 Change to: This knowledge may be obtained by use of the VERIFY (32) command (see 5.27) or by a method not defined by this standard. Seagate #6 PDF Page 81 A note is needed: If the RTO_EN bit is set to zero in the long read capacity data (see 5.15) and the device server checks the logical block reference tag, the device server checks it with the lower 4 bytes of the LBA associated with the logical block. If the RTO_EN bit is set to one and the device server checks the logical block reference tag, the device server checks the logical block reference tag only if it has knowledge of the contents of the LOGICAL BLOCK REFERENCE TAG field. This knowledge may be obtained by use of the Write (32) command (see 5.32) or by a method not defined by this standard. Seagate #7 PDF Page 81 A note is needed: If the RTO_EN bit is set to zero in the long read capacity data (see 5.15), the device server checks the logical block reference tag with the lower 4 bytes of the LBA associated with the logical block. If the RTO_EN bit is set to one, the device server checks the logical block reference tag only if it has knowledge of the contents of the LOGICAL BLOCK REFERENCE TAG field. This knowledge may be obtained by use of the Write (32) command (see 5.32) or by a method not defined by this standard. Seagate #8 PDF Page 81 A note is needed: If the RTO_EN bit is set to zero in the long read capacity data (see 5.15) and the device server checks the logical block reference tag, the device server checks it with the lower 4 bytes of the LBA associated with the logical block. If the RTO_EN bit is set to one and the device server checks the logical block reference tag, the device server checks the logical block reference tag only if it has knowledge of the contents of the LOGICAL BLOCK REFERENCE TAG field. This knowledge may be obtained by use of the Write (32) command (see 5.32) or by a method not defined by this standard. Seagate #9 PDF Page 81 A note is needed: If the RTO_EN bit is set to zero in the long read capacity data (see 5.15), the device server checks the logical block reference tag with the lower 4 bytes of the LBA associated with the logical block. If the RTO_EN bit is set to one, the device server checks the logical block reference tag only if it has knowledge of the contents of the LOGICAL BLOCK REFERENCE TAG field. This knowledge may be obtained by use of the Write (32) command (see 5.32) or by a method not defined by this standard. ************************************************************** Comments attached to Yes ballot from Vit Novak of Sun Microsystems, Inc.: Sun 1 page 10, 4.1, 3rd paragraph The first 3 sentences of this paragraph are misleading. There is no provision in SBC or SPC for each block to have a unique block length. This should simply say something like, "The block length is constant for the entire LU." Sun 2 page 12, 4.6, 1st paragraph, 3rd sentence. Do any of the mode page parameters still allow setting of "geometry" since we obsoleted the format and geometry pages? Sun 3 page 30, table 9, Inquiry row. Protection information column should say 'yes'. Sun 4 page 30, table 9, Read (6) row. Protection information column needs a note. The Read (6) command has no protection information in the CDB nor the data, yet protection information may be a factor in the processing of the command by the device. Sun 5 page 31, table 9, Read Capacity (16) row. Protection information column should say 'yes'. Sun 6 page 31, table 9, Synchronize Cache (10 & 16) rows. Protection information columns should say 'no'. Sun 7 page 32, table 9, Write (6) row. Protection information column needs a note. The Write (6) command has no protection information in the CDB nor the data, yet protection information may be a factor in the processing of the command by the device. Sun 8 p 45, 5.5, last paragraph. Suggested rewording to deal with the "initiator port" vs. I_T nexus issue (see 04-088) "Multiple locks may be in effect from initiator ports associated with more than one I_T nexus. Locks associated with different I_T nexus may overlap. An unlock of an overlapped area does not release the lock associated with another I_T nexus." Sun 9 page 55, 5.13, 5th and 6th paragraphs. "When checking of the LOGICAL BLOCK APPLICATION TAG is enabled" should also include "and the ATO bit is 1". If the App tag is owned by the device server these two fields in Read(32) should be ignored. Sun 10 page 67, 5.24, first paragraph "log read" s/b "long read" Sun 11 page 81, table 68, row 1 (000b/Yes) notes g and h. Is a Write(32) with wrprotect=000b allowed? If so, then notes g and h should be modified to allow the values from the Write(32) CDB to be used. If not, then we need to address this in the Write(32) text. Sun 12 page 81, table 81, 001b/yes/app-tag Why "Shall not" check? Seems to me is should be "May" with note c. But, it should also be qualified with whether or not ATO is set. If ATO, and Write(32) then the device should be checking.... If not ATO, then the device is generating or ignoring, but not checking. Sun 13 Page 84, 5.32, paragraphs 5 & 6. "When checking of the LOGICAL BLOCK APPLICATION TAG is enabled" needs to be qualified with the setting of ATO. If ATO is zero, these fields are ignored. Sun 14 page 89, 5.39, 1st paragraph. "WRPROTECT" should be followed with a reference back to table 68 Sun 15 page 109, 6,3.2.2, paragraphs 3 & 4. The two paragraphs starting with "On a MODE SENSE command,...." confuse me. The previous paragraph states that the Number of Blocks field on a Mode Sense returns a certain value, yet these two paragraphs seem to contradict that. It would make more sense if these rules were on a Mode Select, but I'm still not sure I see the subtle difference - the last phrase of each paragraph is exactly the same. Same comment applies below to the Long LBA format also. ************************************************************** Comments attached to Abs ballot from Ashlie Fan of Tyco Electronics: lack of expertise to comment ******************** End of Ballot Report ********************