.fill .justify .rm 65 .m1 2 .m2 1 .m3 1 .m4 2 .in 5 .ls 1 .ul all .nr a 0 .autoparagraph .define chapter .bp .nr a +1 .nr b 0 .fo /User's Manual/Chapter @na/Page #/ .ce $1 $2 $3 $4 $5 $6 $7 $8 $9 .cl 0 Chapter @na $1 $2 $3 $4 $5 $6 $7 $8 $9 .sp .en .define section .sp .nr b +1 .sp .ti -@nb Section @na.@nb $1 $2 $3 $4 $5 $6 $7 $8 $9 .cl 1 Section @na.@nb $1 $2 $3 $4 $5 $6 $7 $8 $9 .sp .en .he /Custom Services/System Test Utility/Version 6.6/ .fo /User's Manual/©1988-91 /Page #/ .chapter General Information .section Introduction STU is an Amiga system diagnostic. Its purpose is to test disk drives and memory on any model Amiga using the operating system functions. With this program you can test all available memory and the disk functions of the system. The user has the ability to test individual drives for errors. The tests performed may be random or serial in nature. All I/O is performed in an asynchronous manner allowing the system to schedule each drive as needed. This allows faster drives to get ahead of the slower ones and create the best test possible. Each Manufacturer supplies diagnostics that work with their own products. One Manufacturer's memory diagnostic will not work with another's memory board. This means that if you have three different memory boards in your Amiga, you will have three different test disks, one for each expansion device. While STU does not replace all of these diagnostics, it does allow you to test all of your expansion memory boards at the same time! You can run one test that will check all memory on your system. The same idea applies to disk drives. Each disk drive has its own special requirements and device drivers. STU will perform all the basic I/O operations checking each one for correct operation. Bad Media, misaligned disk heads, bad blocks on a hard disk drive can all be found with STU. Any disk device that AmigaDos can recognize and use can be tested with STU! STU has another advantage in that it tests your system in a manner similar to normal operation. This tends to find those special problems the average manufacturer's diagnostic cannot detect. It helps to isolate system problems and ensures that a fixed problem is really fixed. .section Policy This is not a public domain utility or Shareware. This is a fully copyrighted commercial program. You are not allowed to make copies of this disk for other than your own personal use. You may not distrubute this program, it is against federal law. Each disk has a hidden serial number embedded on the disk to identify the owner. The programs and documentation are provided "as is" without any expressed or implied warranty including, but not limited to, the implied warranties of merchantability and fitness for a particular purpose. The entire risk as to the results and performance of any program is assumed by you. Custom Services has no liability which shall include damages for loss of use, anticipated profits, benefits, or any other incidental or consequential costs, expenses, or damages, including with limitation any data or information which may be lost or rendered inaccurate. Your usage of this program signifies your agreement to these terms and conditions. If you lose it because you mis-used it, we are sorry, but no matter what your lawyer may say or lead you to believe, we are not liable. .section Copyright These programs and the related documentation are copyrighted. You may not use, copy, modify, or transfer the programs, documentation, or any copy except as expressly provided for in this document. .section License You have the non-exclusive right to use STU on your own computer. You may not decompile, disassemble, or in any way reverse engineer the program. You are not allowed to modify the serial numbers embedded on the program for any reason. All other rights and uses not specifically granted in this license are reserved by Custom Services. You may transfer your license to another person, for their exclusive use, only if you give them all existing copies and documentation. You may not keep a copy in this case. .section Warranty Registration and Reward The required form in in the appendix. You will receive one free upgrade with your warranty registration. Future upgrades will be supplied to registered users only. Additional upgrades will be available for a small charge for postage and handling. Registered Users will be notified of all updates as they become available. If you obtained this program through other than a normal purchase, please send in a registration form and for a small charge, you will be able to upgrade to the latest version. If you recieved this program through an illegal means, you may still send in the purchase price and obtain the latest version. If you supply us with information leading to the procecution of the pirate(s) responsible, you will be eligable for a reward. For additional information please contact: .nf .in +5 .sp Custom Services P. O. Box 254 Moorestown, NJ., 08057 .fi .in -5 .chapter Program Operation .section General STU is a system diagnostic. It will test your disk drives and memory using the operating system functions. This not only gives you a "confidence" test on the hardware, it also checks the basic operating system I/O functions. This program is designed to test the basic I/O functions of the system. It will read, write, test, and verify all system I/O. The user has the ability to test individual drives for errors. The tests performed may be random or serial in nature. All I/O is performed in an asynchronous manner allowing the system to schedule each drive as needed. This will allow faster drives to get ahead of the slower ones and create the best test possible. .section Limitations To use STU, please set the preferences to 80 column. You do not need external memory to run STU, it will test and use whatever memory you have. The address and data displays on errors are in Hexidecimal. This program is designed to handle up to four 3 1/2 inch AmigaDos drives. It has been expanded to not only test floppies but to test any device mounted except PAR:, SER:, RAW:, and CON:. This means that any Hard drive mounted with the name "DH0:" for example can be tested. A drive with the mounted name of "QH0:" can also be tested. The IBM 5 1/4 drives are not supported and will be ignored during the testing. STU will present you with a list of devices available for testing. At the current time, you may test up to seven of these devices at a time. These may be partitions on a Hard Drive, floppies, or any combinations of these testable devices. You must select which devices in the initial setup screen. Custom Services has not tested every possible device or combinations. Our direct testing has been with with A500s, A1000s, A2000, A2500, and A3000 using Quatum and Seagate drives and Microbotics controllers(StarDrive and HardFrames). We have also tested with the Commodore A2091, A2090 and GVP Series II controllers. We feel confident that if AmigaDos can use it, STU can test it. We will attempt to help any registered user match STU to their hardware. This is one of the benefits of registration! If we have your product registration on file, you can contact us for direct support. Because STU uses AmigaDos to do the actual I/O to the devices, being able to use STU assumes that you have at least one drive or partition(a bootable one) that works well enough to load the system and STU. There is little difference between media errors and device errors except that errors will not usually go away if you switch disks. Errors from the device can cause media problems(such as a head scratching the disk surface). It is important that disks which have errors be used only after you have visually inspected the surface(if removable). In any case, do not use a floppy diskette which has scratches or indentations on the surface. It is also important to only use Double Sided, Double Density diskettes in your drives. STU does not fix problems, it just detects them. You should always obtain service from qualified service centers. After your system is fixed, run STU to be sure that the problems were fully corrected. Timings performed during tests are accurate only to the accuracy of the system clock(plus or minus 0.001 seconds). Other tasks running during a test will cause errors in the form of increased run times. Running tests on multiple drives while a good system test, gives lousy timing information. Since each drive runs asynchronously, the timing overlaps. To get accurate timing of a device, run the tests on only that drive and make sure you shut down all other system activity. See the section on device timings for more information on how to time a drive. This program is designed to run from either the CLI or the WorkBench. The procedure for running from cli is to type in the command STU, From the Workbench just double click on the STU icon. All operations are identical regardless of how you start up STU. Due to the nature of the tests performed, multitasking while not disabled, will be very limited to programs running from RAM: or drives not under test. This program will effectively take over the system due to the testing performed and the amounts of cpu time needed. For best results, this program should be run with no other tasks running. .section Functions STU can perform single or combined drive tests. Testing may be sequential or random in nature. Random testing will jump all over the disk selecting random blocks to read/write(depending on the tests selected). Sequential testing will start at track zero and cycle up to the last track on the device. The tests are based on the functions available to the drives which are: 1) Formatting(always sequential), 2) Seeks, 3) writes, and reads. You are allowed to select the type of testing on an individual drive basis. Each drive is independent of others. All tests are controlled by command options. Invalid options will be ignored. All I/O operations are fully checked for errors. Options are toggled(except for "G" which runs the test). STU will use the device information displayed on the initialization screen to control the "logical" drive or partition it tests. .chapter Operation .section Getting Started To Start with you will need to backup the disks supplied with the normal system command DISKCOPY. Make several backups of the original disk to save you from accidental writing on the program disk. After you have make your backups then proceed to create your data disks. Warning, STU can and will write on your disk devices. Even with the requesters warning you that you are about to write on the disk you may make mistakes. Please be careful! You CANNOT recover data from an STU written disk. STU will write over all information(in sequential testing, and most during random testing). This can be rather deadly to a large hard drive since STU will write data patterns(much like a destructive search for bad spots) over every byte of data. STU will "pop" up a requester whenever a format or write test is selected for a drive asking you if you are sure you want to destroy your data. Please be careful. If you accidently start a destructive test abort it with a ^C, ^D, or a Q as soon as possible, you may be able to salvage "some" of your data. Make sure you backup your hard drive on a regular basis to protect against "human" errors like this. .section General Instructions When you startup STU(either from workbench or CLI) you will get a window which contains the copyright notice. You are obligated to read this and then hit any character to continue. Next you will see the initialization screen. STU will display all disk devices it recognizes( any device mounted on the system with a name that starts with the letter 'D') and let you select the drives to test. The selection is made by clicking the appropriate gadget or entering the number of the device. The first two gadgets are "Test"(0) and "Reset(1)". Test will stop the device selection and continue on with the display of the selected devices. The system characteristics of each highlighted device will be displayed. Reset will clear each highlighted device, deselecting it from testing. Clicking on the gadget with the mouse will accomplish the same effect. When you have finished viewing the selected devices, you can continue by entering any character as instructed. At this point you can select the desired test options. See the sections later in this manual for information on the test options. STU will first test memory and then test the disk devices. You must enable the all desired tests, for safety sake, all tests are turned off. STU will wait at the completion of memory testing for you to enter a character to continue. When disk testing is done and you hit a character the display window will close. The purpose of this is so you can see what happened before continuing. You may enter a Control-C, Q, or q to quit during the testing and STU will stop upon the completion of the current test. STU will still require you to enter a character before proceeding. The result is that you must hit two character to stop testing, a "Q" for example and any other character. This allows you to examine the window to look at errors before exiting the test. You may select to have errors held until you hit a character. The "E" option will allow you to toggle whether or not STU should pause or just continue when an error is detected. .section Making a Bootable Diagnostic Disk STU is not distributed as a bootable disk. You can make a diagnostic testing disk by making a copy of your WorkBench disk and copying the STU files over the Workbench disk. .section Data Disks Taking a backup of your STU program disk insert it into a drive and double click the STU Icon. To create the special Data Disks STU will use in testing your drives you will need to first do a STU version of the FORMAT command. STU allows you to do this on all floppies at the same time. Select the device. It doesn't make sense to create data disks information on your hard drive since it is a destructive test and all normal amigados data will be lost. Select the disk drives and enable the format test. This is done by selecting the "D" class from either the menu or the keyboard. The "D" class selects the disk drives. Toggle the testing option with the "T" command(or the menu) until you see a test with the option "FORMAT" in front of it. This should appear for all you drives. This option will take any disk and convert it into the special STU data disk format. This disk will no longer be usable by AmigaDos until you reformat it with the system command "FORMAT" or the workbench menu item to Initialize the disk. It is recommended that you create a data disk for each drive and keep them separate from your regular AmigaDos disks. AmigaDos may think the drive is empty or Non-Dos when an STU disk is present in the drive. After setting all drives to format, you must enable the testing by using the "X" command. All drives will be labeled as "Yes" under the enabled column. When you enter the "G" option to start the creation of the special data disks. STU will ask with system requesters to 1) place a DS/DD disk in each drive(or you can abort). Click on the "CONTINUE" box after you have the disks in each drive. Now you will get a requester which will warn you that you are about to write on each Drive. For each of these you must click on the "CONTINUE" box to proceed. If you have a WRITE PROTECTED disk you will be asked to un-write protect it, try it so you will see how the program will act! After you have started up each drive the program will write the special test patterns to each disk. These disks will all be identical, and may be used in any drive. Save these disks for future testing after STU informs you that each drive is done(it will display a "CLOSED" message and turn off the drive light). Do not remove the disks until the light on the drive goes off! Removing a disk while the motor is on may damage the disk and the drive! You will want to label each disk with the drive that wrote it for future reference, see the section about test scenarios. For your Hard Drive, you cannot(unless you have a removable cartridge drive) save a data disk. It is suggested that you do a full backup of a Hard Drive before doing any testing. That means a full backup of ALL partitions. While STU does test on a partition basis, that is, it will only test the area defined by the partition, leaving the rest of the physical disk undisturbed, a hardware problem could cause a total failure of your drive and a backup may be the only way to recover. When testing with a Hard Drive, you should always use the Format or Write test if you wish to do a surface test. DO NOT USE the Format or Write unless you have the data backed up. STU will do a destructive pattern test which will overwrite all sectors on a partition. You only have to backup the partition under test, STU will only access the areas specified in your mountlist for the partitions under test. If you do not wish the data overwritten, use only the READ test and turn off the data verify. With a Hard Drive, you have to create the data disk each time you wish to do intensive testing. Using the FORMAT or WRITE tests will write over any data on the disk. When you format the disk drive using AmigaDos after testing, you will be removing the STU test data patterns. If you are having problems with a particular area of a Hard Drive, mount a partition that covers the area in question and test only that area. In any case, expect STU to destroy any data on that area(or a larger partition that includes that area). Always backup a hard drive prior to testing with STU. If you accidentally write test patterns on a partition, you will lose all data that was there. .section Input Summary The testing requires you to define the type of test to perform. Testing is specified with the command line options. The options available are: .nf Option Tells STU to: ------ ------------- A Test All Classes B Toggle the buffer size for current units and class C Toggle verify each sector for STU data patterns, D Select the Disk test class E Toggle pause on errors during testing F Toggle Memory Testing Forever G Start the Testing, I Toggle the I/O ordering for current units and class J Enable Memory Test: Walking Zero Test K Enable Memory Test: Walking One Test L Enable Memory Test: Address Sums Test M Select the Memory test class P Log results to the printer(PRT:), Q Quit the Testing, or Quit STU, Control-D, Control-C also work. S Select the Audio test class(currently not available) T Toggle test selection option(enables memory test) X Toggle the enable flag for current units and class V Be more verbose with output, ? Display a little bit about Custom Services, ! Display a little bit about STU, + Display a little bit about the Future, 0-7 Test device number, .fill The options selected are locked in when you use the "G" option. You are allowed to toggle the "P", "V", "E", and "C" options during the testing. The "Q, Control-D, or Control-C option will cause STU to stop testing and shut down. STU may not stop immediately since it will complete the current test operation before stopping. STU will display the options selected prior to the start of the testing. During the testing the current system status will be displayed. Note, If you do not have enough memory on your system to support the buffer size you requested for the disks, STU will exit without performing any test. If this occurs just restart STU and select a smaller buffer option. .chapter Reference Section .section G - Start the Testing, This command will cause STU start the selected tests. Memory Tests are ran if selected and then the selected drives are initialized and tested. .section F - Forever This command will cause STU to repeat the memory test selected and not stop when the test is completed. This means if you select the memory tests, they will be repeated over and over again until you hit ^C, ^D, or Q. If you select the "Forever" option, you will not get to the disk tests until you manually stop the memory tests with a ^C, ^D or Q. The Memory test will not stop until the current test is completed. .section Q - Quit the Test This command will stop STU during a test. Control-D, Control-C also work in the same manner. Testing will stop at the next logical point. In the case of memory tests, the current block of memory will continue to be tested and will not stop until the current test is completed. .section P - printer log This command will toggle the echoing of the testing results to the printer. STU will use the PRT: device to output test results. When this option is used timing data will be printed on the device PRT:. This command may be used during the testing. Note: if you are testing all you floppies and attempt to open the printer output AFTER YOU START, the system will request the system disk(which STU is actively using) and you will be caught in a catch 22. If you do this the only solution is to abort the test by rebooting. .section V - verbose output This command will toggle the verbose output option. This option will display information during the tests. This option will slow down the testing if the printer option is used since printers cannot print as fast as STU can generate output. This command may be used during the testing. .section E - Pause on errors This command will toggle the option to pause whenever an error is detected. STU will freeze until any character is hit before continuing. .section ? - About Custom Services This command will display a little bit about Custom Services. .section ! - About STU This command will display a little bit about STU. .section + - The Future This command will display a little bit about what is planned. .section 0 1 2 or 3 This command will select a particular device for testing. .section C - verify data on reads This command will toggle the read verification logic on and off. When STU is writing or formatting a disk under test this option should be ON. If you are checking a NON-STU disk(like your workbench disk) this option must be off or you will generate a false error for each read. .section T - test selection This command will toggle the selected class's test. If Memory is selected the type of memory is toggled from: None to Chip to Chip and External, to just External. For Disk devices it toggles thru the various combinations of Format, Write, Read, and Seek. You must use this command to select the type of memory to test or STU will not have a type to test and will finish testing in a hurry! .section X -Enable Test This command will toggle the current class/device combination enabling and disabling the test selection. For example, when memory is selected, it will allow you to select the combinations of chip and fast memory types. .section B - Buffer size This command will toggle the current devices buffer size for i/o. This only affects the disk devices. Care must be taken since some devices require the use of Chip memory and your system may not have a large enough block of memory for the requested buffer. .section I - I/O ordering This command will toggle the method of I/O used during the test from sequential to random and back. This only affects disk devices. .section A - Select All Classes This command will select all classes of devices to be affected by the commands. .section D - Select Disk Devices This command will select only the disk devices to be affected by the commands. .section M - Select Memory Devices This command will select only the memory devices to be affected by the commands. .section J - Walking Zero Test This command will toggle the walking zeros memory Tests. .section K - Walking One Test This command will toggle the walking ones memory Tests. .section L - Address Sums Test This command will toggle the Address Sums Memory Tests. .chapter Test Scenarios All of these tests require that you first create the STU Data Disks before running any test not using the "FORMAT" option. Testing Hard Drive devices should not be formatted unless you do not care for any of the data on the partitions under test. STU will do a total surface check when doing sequential I/O. .section Memory Tests STU will do a basic pattern sensitivity test on any memory selected for tests. In addition you may select up to three additional test. The Walking Zeros test will take each memory location and move a zero thru a word of ones shifting it from right to left. Each location is tested with this rapid test. The Walking Ones test is identical to the above test except that a one is shifted thru a word of zeroes. The Address Sum test is another pattern sensitivity test that will make a value from the current address and store it in that address. This gives memory adjacent memory locations at least one bit change. All memory tests are done using a 16 bit test pattern. The Long option will use 128 test patterns, the short will use 4 test patterns. The regular memory test will alternate a test pattern and it's complement. .section Basic Drive Test A basic Drive test would be the "WRITE/READ/SEEK" test with random I/O. This will give a general test of all the basic functions of each drive. Make sure the drives are write enabled! .section Random Seeks and Reads By using the Random I/O option you create I/O that will jump around over the disk in a random manner. Drives will operate at different speeds due to the randomness of the seeks(some drives will get shorter seeks than others). If you do not have a data disk(or you are testing a hard drive) make sure you disable the data verify(or it will generate false errors). .section Writes Writing occurs when either the FORMAT or WRITE option is selected during the testing. The FORMAT option will always recreate the STU data Disk and may be applied to any disk. This option always performs sequential I/O. The buffer size will be the size of one track for a format. The WRITE option will allow either sequential or random I/O. STU always does a read after the format or write operation to verify that it was correct. .section Individual Testing Individual units may be tested with any of the tests described in this chapter by using the device number as a command option. You have to select the devices to test on the initial display. Selection(highlighted devices or partitions) restricts the drives/partitions that are testable. Each device can be setup with it's own test(sequential or random i/o, different buffer sizes, and different tests). By creating partitions, you can limit the area of a drive under test. All units may be selected either by default or by using the "A" option. You can vary the tests from unit to unit by changing each drive individually. .section Random Seeks and Head Cleaning When a problem occurs the first defense, before you attempt to service the disk drive is to clean the heads with a good double sided cleaner. STU will actually help you clean your drive by moving the heads around. This has the effect of extending the life of the cleaning disk and your drives. A cleaning type operation may be done with the command options: "D", "C", and the drive number. You have to select the drive for random I/O and use only the Seek test with no data verify. The drive will cycle with random seeks for awhile. The "C" option will allow STU to ignore the data on the disk(or the lack of it!). This should be enough to completely clean the head on the selected drive. Enter a Q or ^C at any time after the Seeks stop and the sequential Read test starts. It is a good idea to clean the heads on all drives on a regular schedule. While cleaning is important it normally is not needed more than once a month. Any time read or write errors occur it is a good idea to clean the drive just to be sure. If errors occur on a regular basis you may need to either change the types of diskettes you are using or replace the drive. Drives do wear out, even the best ones will eventually have to be replaced. .section Fast I/O Tests For quick tests of your drives the options specifying a larger buffer can be used to decrease the time taken for the test. When a particular drive shows errors it is a good idea to rerun the test with the smallest buffer size to obtain an accurate idea of where the error is. Note, certain forms of copy protection will deliberately place bad tracks or sectors on the disk to prevent coping of the disk. These disks will generate various errors when tested. See the section on testing other disks for information on checking non-stu data disks in your drives. .section Device Timing You may wish to time the access to a device. There are two ways to do so. You may run other normal tasks on your system and let STU multi-task getting a timing that represents the device speed under loaded conditions, or you can shut down all the utilities, PD software and other tasks and run the test getting the most accurate fastest time possible. The buffer size for a test will vary the time it takes to do an I/O, the smaller the buffer, the faster the I/O but the more time it takes to read a device from beginning to end. For the fastest possible test you should shut down all other activity and test a single device at a time with the largest buffer that is a multiple of sectors that will exactly read the whole device or partition. One track's worth of sectors is a good choice, or two tracks if there are an even number of tracks in the device. A change in the characteristics of a device timing may indicate a problem. .section Testing other disks You may test any disk not just the data disks written by STU. What this will tell you is where a disk has bad spots. STU may be used in this manner with the command option "C" and the number of the drive. This command will cause STU to read every sector on the drive, displaying all sectors it cannot read. Just because STU can read a sector does not indicate that it is correct, it just says that it is a readable sector. Disks with deliberate bad spots will generate read errors during this test. Make sure you have the "READ" option showing when testing a non STU disk. If you accidentally select the "WRITE" or "FORMAT" option you will get a requester making sure you want to write on the disk. Any disk written on by STU will TOTALLY lose the original data, not even DiskDoctor can recover it. This applies to Hard Drives as well as floppies! .chapter Other items of interest .section Future Enhancements This program will not remain static. Just as new devices are added to the system for you to use, Custom Services will be adding new capabilities to STU help you find problems. The areas which will be addressed will depend on what is asked for from you the user. Please fill out the warranty registration and suggestion form found in Appendix A. .chapter Appendix A - Registration .nf PRODUCT: STU REGISTRATION: [yes] [no] (check one) DATE RECEIVED:_________/_________/_________ RECEIVED FROM:______________________________________________ _____________________________________________________________ _____________________________________________________________ Your Name:_________________________________________________________ Address:______________________________________________________________ ______________________________________________________________ ______________________________________________________________ Who told you about this product:_______________________________________ _______________________________________________________________________ _______________________________________________________________________ _______________________________________________________________________ What improvements would you like:_______________________________________ _______________________________________________________________________ _______________________________________________________________________ _______________________________________________________________________ .fill .chapter Appendix E - Errors .section STU Messages .nf Error Text Unable to allocate a buffer Not enough memory Observed xxxxx Expected at offset ddd data error, disk read problem or was not a data disk. Closed, finished The drive has finished the requested tests. .fill .section SYSTEM errors .nf Unable to get a Port System problem getting started Unable to get a pointer System problem getting started Unable to set break control System problem getting started I/O request Aborted Control C was used to abort ***These are the various disk errors which may occur*** Error couldn't be determined Couldn't find sector header Error in sector preamble Error in sector identifier Header field has bad checksum Sector data field has bad checksum Incorrect number of sectors on track Bad sector header Disk is write-protected Disk has been changed or not in drive Seek error verifying position Out of Memory for this operation Bad Unit number for drive Bad drive type, not an AMIGA drive! Drive is busy, another task has it ***Drive: dd Unknown Error Code=dd*** .fill With any of the above errors the general procedure would be to clean the drive, retest, switch media, retest before suspecting a hardware problem! If a cleaning does not eliminate the problem, you should seek an authorized service center to help you correct the problem. .page .he / / / / .pn roman .fo / /-#-/ / .nf .sp .center T a b l e o f C o n t e n t s .sp .printcont