1. 31r3020.exe Driver
2. VS160 Firmware V35.txt File Contents

VS160 Firmware V35.txt Driver File Contents (31r3020.exe)

VS160 
V35 Firmware Release Notes 
OEM Version 
This document describes the changes in the specified release of the VS160 firmware. 
September 21, 2004 

VS160 V35 Firmware Release Notes, OEM version 
1. INTRODUCTION 
This document describes the changes since V34.0 for the V35.0 release of VS160 firmware. 
The firmware consists of code that supports both Bluelite and Morpheus VS160 subsystems. 
2. SOFTWARE RELEASE IDENTIFICATION 
Control Revision: 35 (23h) 00 (00h) (See INQUIRY data bytes 20h, 21h, 22h, 23h) 
3. RELATED DOCUMENTS 
VS160 Product Specification (81-81275-01 Rev. A01)
VS160 SCSI Interface Guide (81-81268-01 Rev. A01) 
4. COMPATIBILITY 
There is not any compatibility restriction with this level of firmware. 
Quantum Corporation, 1650 Technology Drive, Suite 800 
San Jose, CA 95110 
Page 2 

VS160 V35 Firmware Release Notes, OEM version 
5. ENHANCEMENTS AND NEW FUNCTIONALITY 
5.1 Description: If a Locate command terminates on EOD, a subsequent write type command will 
create a bugcheck B0300037. 
Root Cause: The read-to-write transition is failing due to incorrect position update at the end of 
the Locate command. 
Corrective Action: Properly update position at the completion of the Locate command. 
5.2 Description: If the cartridge ejector arm is pressed in and then quickly released it can create a 
bugcheck B0001100. For example, a person uses a finger to press the arm without a cartridge 
being present. 
Root Cause: The servo processor begins the load sequence and then upon recognition of the 
cartridge present sensor status change, begins issuing messages for which the controller 
processor is not prepared. 
Corrective Action: Change made to recognize the condition and notify the controller processor 
so that it can command the servo processor back into an appropriate state. 
5.3 Description: A non-zero count field (bytes 2-4) on a Space EOD command will cause drive to 
respond with a check condition. 
Root Cause: Not properly ignoring count field on a Space EOD command. 
Corrective Action: Ignore count field on a Space EOD command. 
5.4 Description: If the drive has entered Dump to LEDs mode and an attempt is made to re-enter 
the FrontPanel Wizard, it can create a bugcheck B0830029. 
Root Cause: Not clearing a timer upon exit of Dump to LEDs mode. 
Corrective Action: Properly clear timer upon exit of Dump to LEDs mode. 
5.5 Description: Send Diagnostics command with PF=1 (byte 1, bit 4) and Parameter List 
Length=0 (bytes 3-4) will create a check condition. 
Root Cause: Send Diagnostics command is documented as not supporting the Page Format bit 
and requiring it to be set to 0. This is an unforeseen exception condition. 
Corrective Action: Ignore command and return good status. 
5.6 Description: FrontPanel Wizard can allow multiple modes to be entered simultaneously which 
can create erroneous data and/or bugchecks B0830000A and B0830029. 
Root Cause: Unforeseen exception condition. 
Corrective Action: Implement zone exclusion properties within FrontPanel Wizard. 
5.7 Description: If an SDLT cartridge is loaded and the drive is powered off just as the load 
process begins and proper tension has not yet been achieved, upon power-on the drive will 
enter into an infinite tension/de-tension loop. The only way out of this loop is power-cycle the 
drive at which point a dropped leader will occur. 
Root Cause: The load ring motion on power-on exceeds the allowable range to insure the 
leader stays engaged. 
Corrective Action: Added code to learn the position of the load ring during power-on and only 
move it in a direction that will insure the leader stays engaged. Also, the correct drive state was 
added for an error during a mid-tape load on an exception cartridge. 
5.8 Description: A timing condition exists where if the takeup reel is stopped just after the leader 
is unhooked it may create a bugcheck B0500009. 
Quantum Corporation, 1650 Technology Drive, Suite 800 
San Jose, CA 95110 
Page 3 

VS160 V35 Firmware Release Notes, OEM version 
Root Cause: The servo processor is entering the unloading state before signaling the controller 
processor that an error has occurred. The subsequent unload command from the controller 
processor is rejected because servo believes it is already unloading. 
Corrective Action: Create a new drive state to signify an unload is necessary (due to the error 
condition) but not yet in progress. 
5.9 Description: A multi-initiator emulation test suite is failing due to strategically placed bus 
resets creating race conditions. 
Root Cause: Improper detection and handling of bus resets during large data transfers created 
situations where the drive did not always properly flush the partial record from the buffer. 
Corrective Action: Proper bus reset detection and write abort handling. 
5.10 Enhancement: Made the Unload Button Mode Zone 7 (Dump Logs to Tape) non-operational. 
5.11 Enhancement: Ensure tape is not loaded before entering Unload Button Mode Zone 6 
(Read/Write Diagnostics) to avoid potentially overwriting current media in the drive which 
might have been part of a recent customer data operation. 
5.12 Description: Bugcheck B0300043 on read of 256KB record when buffer is nearly empty. 
Root Cause: If the read queue is not full and a DMA burst is started, a certain alignment of 
data within the entities and timing of loading read data into the buffer will cause firmware to 
lose track of its location in the data buffer. 
Corrective Action: Insure the read queue is full before starting a DMA burst. 
5.13 Description: Bugcheck B0500009 on an ERASE command on a write-protected tape. 
Root Cause: Unforeseen exception condition. 
Corrective Action: Check for write-protect status in erase function. 
5.14 Description: If the drive is in a condition where on a power loss during loading the takeup 
reel can stop prior to the supply reel, the continuing supply reel motion can allow the cartridge 
buckle and drive leader to separate. 
Root Cause: At the proper location in the tape path when power is lost, the supply reel is 
capable of staying in motion for a short distance (1/8”-1/4”) because of its large inertial mass. 
The takeup reel stops quickly, because of its small inertial mass. The additional movement of 
the supply reel causes the drive leader end to slip out of the cartridge buckle. When power is 
re-applied, the tip and buckle are already separated and cannot be recovered. 
Corrective Action: Control both reel motors during the tape path loading process which 
produces improved tape position handling and speed control. 
5.15 Enhancement: Added function to learn load ring position before moving it to a new position. 
Also added function as part of the midtape load process to detect and recover from condition 
where takeup reel flange is located on the wrong side of the buckler flange. 
5.16 Enhancement: Only write evflash dump if at least 8 tape motion hours have accumulated 
since last evflash dump (or if evflash is empty) so that original error dump is not overwritten 
due to secondary error conditions. 
5.17 Enhancement: Compression can be inadvertently enabled if a MODE SELECT is issued 
with default density. Change was made to not affect compression mode when default density is 
issued. The only valid method to enable/disable compression is via MODE SELECT Pages 
0Fh (Data Compression Page) and 10h (Device Configuration Page). 
Quantum Corporation, 1650 Technology Drive, Suite 800 
San Jose, CA 95110 
Page 4 

VS160 V35 Firmware Release Notes, OEM version 
6. ENHANCEMENTS AND NEW FUNCTIONALITY FOR LIBRARY DRIVE 
APPLICATIONS 
6.1 Description: Set File Protect [0x18] is only functioning when tape is loaded, which is opposite 
of its intended function. 
Root Cause: Specification misinterpretation. 
Corrective Action: Allow Set File Protect [0x18] to only function when tape is unloaded. 
6.2 Description: Media Type (byte 0, bits 3-0) in the SEND TAPE DATA 3 response packet 
shows 0x0 No Tape/Unknown Tape during a cleaning operation. 
Root Cause: Two different possible cleaning tapes created a mistake in coding of the response 
packet. 
Corrective Action: Correct Media Type to appropriately show 0x1 Cloth Cleaning Tape or 0x9 
VS Cleaning Tape during a cleaning operation. 
6.3 Description: Clean Requested (byte 3, bit 3) in the General Status Packet is not being set upon 
a permanent read error or 250 tape motion hour threshold. 
Root Cause: The Cleaning Required bit was instead mistakenly being set. 
Corrective Action: Properly set Clean Requested bit. 
6.4 Description: Library interface TapeAlert flags are not always set synchronously with the host 
interface TapeAlert flags. 
Root Cause: Specification misinterpretation. 
Corrective Action: Insure all TapeAlerts flags are set synchronously on both the library 
interface and host interface. 
.2004 Quantum Corporation. All rights reserved. Quantum and the Quantum logo are registered trademarks 
of Quantum Corporation, registered in the U.S.A. and other countries. DLTtape, and the DLT logo are 
trademarks of Quantum Corporation. Any products mentioned herein are for identification purposes only and 
may be registered trademarks or trademarks of their respective companies. 
Printed in the U.S.A. 
Quantum Corporation, 1650 Technology Drive, Suite 800 
San Jose, CA 95110 
Page 5