DDS/4 tape drive firmware update v912L for Microsoft Windows - IBM System x

Download Description

• New IBM xSeries firmware and device driver file naming convention

Severity: Suggested

Change history

Version 912L

This release note summarizes the firmware changes made to the IBM DDS/4 DAT drive firmware from version V910B to version V912L.

Changes detailed in this document are grouped into the following sections:

1. Reliability Improvements/Tape Handing (4 entries)
   
2. Reliability Improvements/Media and Drive Errors (9 entries)
   
3. Miscellaneous (4)
   
4. Reliability Improvements/SCSI (1)
   
5. Autoloader (4)

The changes detailed in this firmware difference list represent an estimated total 27597 DPPM improvement in product quality for the DDS4 drive.

Reliability Improvements/Tape handling

1. A large number of 03/52/00 Tape Jam errors have been seen in drive return EEPROM data. A hardware change to the hub Reel Cap addressing the root cause of this issue.
2. Failure analysis noticed a case where when a Space to End Of Data was issued at Beginning Of Tape, the drive started re-reading using a calibration start point just before the append point. Calibration should not occur at this point. Made a correction to the recovery step involving recalibration to compensate for this case.
3. Very rare cases have been seen in drive returns where the cylinder has been left spinning with the tape stationary, but fully engaged for very long periods of time, causing tape and cylinder damage. A change has been made so the drive will now detect the condition and send a reset to the servo firmware.
4. Ported over firmware changes from DAT 72 that strengthen the reliability of tape cartridge hole detection pattern. This helps prevent invalid Cartridge Fault (03/52/00) error codes from being reported.
   

Reliability Improvements/Media Errors/Drive Errors

1. Testing found that writing a small block size with a fast host could result in a Media Error (03/0C/00/FC). An incorrect append frame count setup on a missed interrupt retry could result in the Media Error. Modified the missed interrupt retry logic to follow the same path as excessive rewrite retries.
2. A stress test that writes/spaces/overwrites/reads returned a Media Error (03/50/00) during an attempt to append data. In this case, an append table for the frames was incorrect. Modified the firmware to prevent invalid append frame numbers from being entered into the append frame table.
3. A timeout occurred during stress testing when a Space to End Of Data (EOD) command was issued where the position was very near the physical end of tape. In this
   case, the cylinder was left spinning with tape engaged, causing possible head and media damage. The firmware was modified to disable interrupts around critical areas of code where the Tape Task is updating some MCP variables that control tape motion.
4. Testing which uses Read, Space, and Locate commands sometimes results in a Media Error (03/3B/08/99). Modified firmware to strengthen subcode validation to prevent incorrectly detecting filemark/setmark and returning the wrong group to the DC task.
5. A Media Error (03/3B/BF) occurred during a search/read type test. In this case the firmware was setting a register in the Kukai ASIC to an incorrect value during the search mode. Modified the firmware to set correct values during the particular search mode.
6. Testing revealed a Media Error (03/3B/00/E5) occurred during a Space to End Of Data on a full tape, starting close to End Of Data. In this case an internal command timeout was too short and caused the drive to stop the operation and report the error seen. Modified the firmware to remove the command timeout for this case.
7. During a Stepp-a AutoRandom test, a sample DDS4 unit returned a Hardware Error. Though the condition was not duplicated, code review determined that the error was detected by the ring buffer code, where the firmware SCSI task enqueued the new DATA descriptor, when the data for the last DATA descriptor had not yet transferred into the ring buffer. Modified the firmware to increase a polling time for the current write_burst_done flag from ~1 ms to 24 ms. Removed clearing of the write_burst_done flag which relied on the AIC chip’s DMA counter going to zero.
8. DAT72 testing found a case where a Reset issued while the drive was writing would cause a Hardware Error (04/44/00) to be reported. Corrected a case where a state transition was not properly recorded in the firmware during Reselection Timeout handling, triggering the failure behavior when the Reset was processed.
9. Enhanced FW to save sense data to the EEPROM, and trace information in flash, for Aborted Command (0xB) error codes.

Miscellaneous

1. After code review, a possible condition was identified where the mode motor could be left on longer than anticipated after a tape is ejected. The firmware was modified to provide additional functionality ensuring the mode motor is turned off as intended.
2. The behavior of the cleaning LED did not match the manual. Corrected.
3. During research on 03/52/00 errors seen in drive return EEPROM data, it was noted that EEPROM updates were only occurring if a SCSI command was sent to the drive when it was in the error state, and that resulted in a check condition . A change was made so that the error will now be saved, even if there is no SCSI activity.
4. A particular hardware failure on a returned drive allowed that drive to write data that could not be read – reporting a 03/31/00 0x9F (DC Detected Group out of sequence) error. The returned drive had an open trace in the PCB causing the problem. Modified firmware to perform verification is correct while writing data. If not, post 04/44/BF/00D7 error. This will prevent the read side failure.

Reliability Improvements/SCSI

1. Enhancement: Implemented Receive Diagnostics Results Page 0x81, containing current SCSI transfer speed information.