; ****************** List Program Parameters *******************
%list
The "LIST" keyword provides a display of the currently enabled options
as well as various program parameters.
/
; ***************** Short Command Summary ******************
%commands
COMMANDS displays a short summary of the available commands.
/
; ****************** Error Summary *******************
%es
%ses
During the execution of the various tests and commands, all
errors are noted and counted individually. At any time you may get a
detailed error summary by using the "es" keyword.
A separate summary is provided for use with the SRA commands. The
statistical error summary is displayed when the "ses" keyword is entered.
/
; ****************** Read the Task File *******************
%rtf
The "rtf" command reads all of the WD1010 task file registers
and displays their contents.
/
; ****************** Drive Selection *******************
%d
%drive
Most hard disk controllers support more than one physical drive.
To select a drive you must set the desired drive number with the "DRIVE"
parameter keyword. The drive is not physically selected until a command
is issued to the controller. A typical sequence is as follows:
d:1 ;select drive 1
read ;read from drive 1
d:0 ;select drive 0
read ;read from drive 0
/
; ****************** Read Command **********************
%r
%read
%rl
%read.long
The read command initiates a disk read command sequence. One or more
sectors may be read with one read command. Use the BLOCKS keyword to set the
number of sectors to read or write. To specify where on disk you would like to
read, you must specify a disk address. A physical disk address is specified
in terms of cylinder, sector and head. To read, you would have to specify
the disk address in physical terms using the C, S and H keywords followed
by the read ( ie c:5,s:0,h:1,read ). A typical read sequence would look as
follows:
b:1
c:100,s:5,h:3,r
/
; ****************** Write Command **********************
%w
%write
%wl
%write.long
The Write command initiates a disk write command sequence. One or
more sectors may be written with one write command. Use the BLOCKS keyword
to set the number of sectors to write. To select a location on disk for the
write operation, you must pre-specify a disk address, in terms of cylinder,
sector, and head. In other words, to write, you first specify the disk
address in physical terms using the C, S and H keywords, then select the
desired write command (i.e. c:5,s:0,h:1,write). A typical write sequence
would look as follows:
b:1
c:100,s:5,h:3,w
/
; ******************** Seek Command ********************
%sk
%seek
The seek command allows the user to seek to a specified cylinder. The
cylinder would be specified as a physical cylinder number. The command syntax
for this command is as follows:
sk
or
seek
/
; **************** Disk Address Specification ******************
%c
%s
%h
Whenever you read or write to a disk drive you must specify to the
controller the exact place to begin. This place is known as a disk address.
A disk address is comprised of a cylinder number, a head number and a sector
number. To specify these parameters use the "C", "H" and "S" keywords.
Physical Disk Address Specificaiton Keywords
c:cylinder-number
h:head-number
s:sector-number
/
; *************** Controller Diagnostics *****************
%diag
This command instructs the controller to perform various levels of
self-test. At the end of this command, the error register will be set to
indicate whether or not any errors were detected. A value of "1" indicates
no errors have been detected.
/
; ************ Repeat the Last Command Line Entered ***************
%rep
%repeat
The "REPEAT" keyword allows you to repeat the previously entered and
executed line of commands. The accompanying numeric parameter is the repeat
count. If you wanted to read sector zero 1000 times you would use the
following sequence of commands.
read
repeat:1000
NOTE that this is not at all the same as:
read,repeat:1000
Remember that the repeat keyword repeats the previously executed line,
not the current one!!!!!!!
/
; ************** Production Test Procedure ****************
%ta
The "TA" keyword initiates the production test procedure. It is best
if you are in the "LOGERR" logging mode to reduce the amount of extraneous
output. This test procedure takes approximately two minutes to complete.
The individual tests and commands that are executed are:
1) Recalibrate Drive
2) Format last "N" Cylinders( FD )
3) Write/Read/Compare Test
4) Error Correction Test
/
; ***************** Read/Write All Test ********************
%ra
%wa
The "RA" and "WA" commands start a continuous read or write test.
"RA" will read every sector of the disk unless it is interrupted by pressing
the enter key. While doing this it will monitor the I/O timing ( see RETCNT )
and indicate on the screen whenever it thinks the controller has performed
a retry on a sector read. A normal read will appear as a period ( . ) on
the screen. A retry on a read will appear as an exclamation point ( ! )
on the screen. Every retry will be counted and a cumulative total will be
maintained. Note that the retry evaluation only works correctly if the block
size is set to one (i.e., B:1). The "RA" test does not verify the data read
into memory. It is only looking for timing and controller errors.
The "WA" command will continuously write the currently selected data
pattern ( see PATTERN ) onto every sector of the disk unless interrupted by
pressing the ENTER key.
The starting disk address is specified by using the C, S, and H
commands prior to execution.
/
; *************** Read/Write/Compare Data Test *****************
%rt
This test is included in the production test procedure. It can be
executed directly by using the "RT" keyword. The test procedure performs the
following command sequence:
1) Write current data pattern to current sector
2) Read sector to establish timing position
3) Read sector and time operation( retry evaluation )
4) Compare "should be" and "was" data.
Any errors will be reported. As with the "RA" and "WA" tests, this
test can be terminated by pressing the ENTER key. The starting disk address
is specified using the C, S, and H commands.
/
; ****************** Seek Exercisor ********************
%st
The "ST" keyword intiates a series of seek commands to the controller
The disk will seek between the extremes ( inner cyl to outer cyl ) in a
butterfly or crescendo pattern. No data reads or writes take place.
/
; ************** Error Correction Test ****************
%tecc
The error correction test is started by the "TECC" keyword. This test
forces correctable and uncorrectable data errors on the disk. It makes sure
that the controller can correct these errors, and monitors the expected error
status indications. Any correctable errors that were not noticed will be
flagged as controller errors. Improper error responses in general are flagged
as controller errors.
/
; *************** Bad Track Handling ****************
%bad.track
%bt
%btlist
%bton
%btoff
The BT command allows the specification of a bad track list. This list
of undesirable tracks will be skipped in the RA, WA, RT, LTT, DRA and FD
commands. A bad track specification consists of two numbers per line separated
by a comma. Each number must be in decimal format. The first number of the pair
is the cylinder number. The second number is the head. Up to 50 bad tracks may
be specified. The bad track list can come from a normal ascii disk file or
directly from the keyboard. A typical file might look like this....
100,1
200,0
201,3
The BTLIST command displays the current bad track list. It will be
displayed onto the crt screen as well as a log file if logging is enabled.
The BTOFF command is a mode command. It disables bad track skipping.
Its counter part is BTON which enables bad track skipping.
/
; ******* IDC - Initialize Drive Characteristics *******
%idc
The IDC command is used to inform the controller about the physical
structure of the particular disk that is being used. Prior to using this
command, you should have set the drive parameters to the proper values. These
parameters are set using the following keyword commands:
cyl: max number of cylinders
heads: number of read/write heads
sectors: sectors/track
precomp: cylinder where write precompensation starts
reduce: cylinder where reduce write current starts
Note: THE IDC COMMAND MUST BE ISSUED PRIOR TO THE FIRST DISK ACCESS
UNLESS THE DEFAULT PARAMETERS AT POWER UP ARE SUITABLE FOR
YOUR PARTICULAR DISK !!!!!!
/
; ******* CYL - Set Maximum Cylinders per Drive *******
%cyl
The "CYL" keyword is used for setting the maximum cylinder drive
parameter. This parameter and some others are given to the controller whenever
the IDC command is issued. A typical use of this parameter keyword is as
follows:
cyl:306,idc ;sets max cylinders, send to controller
The cylinder parameter is different for each style of hard disk and must be set
prior to any disk activity. See IDC command.
/
; ****************** Sectors per Track ********************
%sectors
The "sectors" command establishes the number of sectors on each
track. The sectors per track informs the controller how to perform
multi-sector reads and writes and also how to interpret the interleave table
at format time. Command syntax is as follows:
sectors:n
/
; ******* HEADS - Set Number of heads per Drive *******
%heads
The "HEADS" keyword is used for setting the number of heads per drive
parameter. This parameter and some others are given to the controller whenever
the IDC command is issued. A typical use of this parameter keyword is as
follows:
cyl:306,heads:4,idc ;sets max cylinders, send to controller
The heads parameter is different for each style of hard disk and must be set
prior to any disk activity. See IDC command.
/
; ******* Write Precompensation *******
%precomp
The Precomp keyword is used to set the cylinder number at which write
precompensation will be enabled. This keyword is used in conjuntion with the
IDC command.
precomp:128,cyl:306,heads:4,idc
The precomp parameter is different for each style of hard disk and must be set
prior to any disk activity. See IDC command.
/
; ******* Reduced Write Current *******
%reduce
The REDUCE keyword is used to set the cylinder number at which reduced
write current is enabled. This parameter keyword is used prior to the IDC
command. A typical use of this parameter keyword is as follows:
reduce:64,precomp:128,cyl:306,heads:4,idc
The REDUCE parameter is different for each style of hard disk and must be set
prior to any disk activity. See IDC command.
/
; ******* Step Rate Options *******
%step
The stepping rate can be specified with the STEP keyword. The tandon
controller requires that the IDC command be issued after setting the STEP
option. All other controllers do not require that the IDC command be issued
after setting a new step option.
Step Rate Table
STEP
0
1
2
3
4
5
6
7
8
Step Rate Table ( continued )
STEP
9
10
11
12
13
14
15
/
; ******* Format Drive - Initializes Headers *******
%fd
%format.drive
The format drive command is responsible for formatting the entire drive
starting from the specified cylinder or logical disk address in the case of
SHD. All or a portion of the disk can be formatted using this command. For
example, if you set c:100 and issued a "fd" command all cylinders from 100 to
the end of the disk would be formatted. Before this command is issued, the
drive parameters should have been set and the IDC command issued if necessary.
The interleave factor must also have been specified with the "i" parameter
keyword. To format the entire drive you would normally do the following command
sequence.
recal,fd
See "idc" and "i".
/
; ************** Format Track Commands ********************
%ft
%format.track
The format track commands are responsible for formating one track.
Note that the "fd" command is really many "ft" commands. Prior to formatting
each track, the sector buffer is filled with an interleave table indicating
to the controller the correct order for the individual sectors.
ft
format.track
/
; ******* Specify Sector Interleave Factor *******
%i
%interleave
The interleave parameter keyword specifies what interleave factor is
to be used when formating the disk. This parameter must be set to the desired
value prior to issuing a "fd" or "ft" command. Typical useage of this keyword
is as follows:
i:3,recal,fd
A value of 9 to 17 will create a table with an interleave table of
1 to 8 starting with track 0.
/
; **************** Controller Retries *******************
%enable.retries
%er
%disable.retries
%dr
Retries of two types can be enabled or disabled simultaneously. Upon
execution of this command the WD1010 automatic retries are en/disabled as
requested, as well as retries on ECC errors.
er
enable.retries
dr
disable.retries
/
; ***************** Logging Option Keywords *********************
%lognorm
%logerr
%logall
%logoff
The logging options keywords allow you to specify the level of detail
of controller protocol that is displayed and optionally saved in a log file.
The nominal or default logging mode is the "normal" mode. This mode can be
entered by using the "LOGNORM" keyword. This mode produces an abreviated
protocol display.
At the other extreme is the most detailed logging mode which is enabled
by using the "LOGALL" keyword. This mode produces the most detailed and
therefore most voluminous output. You normally use this mode when debugging
controllers that are somewhat ill.
Another useful mode is the errors only logging mode. This particular
mode of logging only produces some output to the screen or logfile when there
is an error situation. This mode is used mostly during production testing,
and for long term testing where you want to minimize the amount of output.
This mode is entered by using the "LOGERR" keyword.
All logging modes can be directed to a disk file or a printer. This
allows you save testing results for review at a later time. To direct the
output of the HDT program to a log file or printer use the following syntax:
lognorm:filename or lognorm:lpt1
The same syntax applies to all logging keywords. To terminate a logfile
session use the "LOGOFF" keyword. This keyword just terminates logfile output
without changing the logging mode.
/
; **************** Sectors to Read or Write ******************
%blocks
%b
All of the controllers that are supported by HDT are capable of
multisector reads and writes. By using the "BLOCKS" keyword, you can tell the
controller how many disk blocks (sectors) to transfer for each read or write
operation. The syntax for this key word is as follows:
blocks:sectors-to-transfer
b:sectors-to-transfer
The maximum of physical buffers allocated by this program is 36.
Any value greater than this will result in repeated use of the first 512 byte
buffer.
/
; *************** Decimal to Hex Conversion ******************
%cvt
The "CVT" keyword allows you to convert hex numbers to decimal and
decimal numbers to hex. This is a calculator type of feature and is intended as
a convenience.
cvt:100 will display the hex equivalent of decimal 100
cvt:100h will display the decimal equivalent of hex 100
This particular keyword serves no other function.
/
; **************** Getting Help ********************
%help
%commands
HDT supports a help feature. There are two keywords that allow access
to support information. These keywords are "COMMANDS" and "HELP". The
"COMMANDS" keyword allows you to browse through a brief command glossary.
This level of help is intended as a memory jogger in case you forget the name
of a command. The glossary is alphabetically sorted and includes a brief
command description.
The "HELP" keyword provides a detailed description of most keywords supported
by the HDT program. If you want to know something about the "IDC" command,
you would merely enter the following command.
help:idc
The "IDC" command will be looked up in the command help file and the
appropriate text will be displayed on the crt screen. If you merely enter
the "HELP" keyword with no parameter, a short narative on how to get more help
is displayed.
/
; ************* Clear All of the Error Counters **************
%zero
The "ZERO" keyword clears all of the various error and retry counters
that are used internally within the program. These counters are cumulative.
/
; **************** Buffer Manipulation ******************
%buf
%o
%offset
%data
%v
%dump
%du
%pattern
%pat
%cmp
%compare
%fill
This version of HDT has a new buffer scheme. There are two data
buffers, one dedicated to reads and related operations, and the other dedicated
to writes and related operations. They are called the "Read" and "Write"
buffers. The read buffer contains the data most recently read from the disk.
The write buffer contains the data most recently written to the disk.
The following keyword commands allow you to manipulate these sector buffers:
1) DUR - Displays the contents of the READ Buffer (DUmp Read)
2) DUW - Displays the contents of the WRITE Buffer (DUmp Write)
3) BUF - toggles between buffers for the "DU" command only
4) DUMP, DU - allows you to view the most recently used buffer
5) OFFSET, O - manual buffer modification offset (Write Buffer)
6) DATA, V - 8 bit value to be stored at current Write Buf. offset
7) PATTERN, PAT - data pattern generation in Write Buffer
8) COMPARE, CMP - compare both data buffers for equality
9) FILL - Specify Write Buffer fill pattern (8 or 16 bits)
The "BUF" command allows you to change between displays of the Read or
Write Buffer when the "DU" command is used. This command was left in HDT for
compatibility with previous versions.
To display the contents of the most recently used buffer you simply
enter: "DU:length" where length is the optional number of bytes you
wish to see. If length is not specified, the valid data in the buffer will be
displayed. The amount of data in the buffer varies from command to command.
The DU command will also tell you which buffer you are viewing.
To fill the Write buffer with a desired data pattern prior to executing
a write command, the "PATTERN" keyword may be used. There are currently 11
data patterns available. They are numbered from 0 to 10. Pattern 0 (zero)
disables pattern generation ( i.e., pat:0 ). It is important to note that when
automatic data pattern generation takes place, the contents of the write buffer
are over-written (lost). By using the DUW or DU commands you can see the data
pattern in the buffer. The available data patterns are as follows:
0) pat:0 disable pattern generation
1) pat:1 all zeros in buffer
2) pat:2 all ones in buffer
3) pat:3 count 0 thru FFh in buffer
4) pat:4 fill with 6DB6h
5) pat:5 fill with A5's
6) pat:6 fill with 55's
7) pat:7 random data fill
8) pat:8 interleave table
9) pat:9 fill with 18C63h
10) pat:10 User-selected fill pattern (from FILL command)
To write a sector with manually created data, requires the use of the
"OFFSET" and "DATA" keywords. These keywords modify the contents of the write
data buffer. Then the appropriate write command is issued. It is important to
note that when you manually modify the write data buffer using the "OFFSET" and
"DATA" keywords, the automatic pattern generation is disabled. If this were
not the case, your manually modified buffer would be over-written right before
the write command takes place. The following gives a typical sequence for
manually creating a sector buffer and writing it to disk.
pat:1 ; fill the write buffer with zeros
o:0 ; specify offset = 0 ( i.e., first byte )
v:1 ; put a 01h into first byte
v:2 ; 02h to second byte ( o is incremented by v )
write ; write to disk
You should have noticed in the above example that the buffer offset
is incremented every time a value is stored by the "DATA" or "V" keywords.
This allows you to more easily build a custom pattern in the write buffer.
If you wish to compare two data buffers you must use the "COMPARE"
keyword. This command causes a data compare operation to take place and the
differences between the contents of the two data buffers to be displayed.
This is useful for write/read/compare sequences.
/
; ***************** Execute a Command File *******************
%exec
%execute
HDT can execute command sequences that have been previously prepared
and saved in a disk file. In this mode, disk file input substitutes for
keyboard input. Execution of these command files can be started in two
different ways.
The first way is to specify the names of the files on the command
line when you first load the HDT program. One or more command files can be
specified on the command line. The general syntax is as follows:
HDT file1 file1 file3 .....
The second way to execute command files presumes that you are already
running HDT from the keyboard. In this case you can use the "EXECUTE" keyword
to perform the same thing as demonstrated above. For example to execute the
the command file "file1" you would enter in the following command:
exec:file1
Once that file has been processed, keyboard input will resume, at which
time you may execute another file if you choose to do so. Command files can
have "execute" commands within them. They can be nested, but this practice is
recommended.
/
; ******************* Error Stop Control ********************
%stop
%go
The keywords "stop" and "go" control the action of the program whenever
an error occurs. STOP changes the mode to stop and wait for keyboard response
on each error. GO terminates the stop mode. GO is the default mode.
/
; ************** Long Term Test ( multiple pass RT ) ****************
%ltt
This test is simply a re-occuring read/write/compare test. Multiple
passes over the entire disk take place. This test is interruptable and may be
resumed by merely re-issuing the ltt keyword again. You might stop it briefly
to do an error summary command, for example.
/
; ***************** Recalibrate the Drive ********************
%t0
%recal
%recalibrate
This command forces the drive back to track 0. The controller will
issue 1024 step pulses backwards until track zero is sensed.
/
; ******************* Reset the Controller ********************
%reset
The "reset" command issues a reset pulse to the controller and
monitors the results of the controller self diagnostics. See the "diag"
command for details.
/
; ********************** Retry Evaluation ***********************
%retcnt
The "ra" and "rt" commands are designed to figure out if the controller
had to retry on a read operation. The determination of whether or not a retry
has occurred is based solely upon timing. A sector is read. The same sector
is read again, and the time to the first DRQ is measured using the "retcnt"
value as a reference. If the time to the first DRQ exceeds "retcnt" then a
retry is assumed to have occurred. The "retcnt" value can be set either
manually or automatically. The proper value must be determined empirically.
Manual Method: retcnt:value
Automatic Method: retcnt:0,retcnt
Note that the automatic mode writes a sector and reads it 50 times.
The minimum timer value is assumed to be 1 rotation time. A retry is flagged
if 1.5 rotations have occurred. The timer value for "retcnt" is displayed
after it has been deduced automatically. Note that the Block size must be set
to 1 (one) for these commands to work properly.
/
; ********** Random Number Generator Initialization ************
%seed
The "seed" parameter keyword allows the user to initialize the starting
reference point of the random number generator used for writing random data.
The "pattern:7" generates random data into each sector. To insure repeatable
test sequences requires setting the random number seed to the same value.
seed:n
/
; ****************** Status Register Read ******************
%sr
The "sr" command causes the WD1010 register to be read and its contents
to be displayed onto the crt screen. If the error bit is on, the error register
will also be read.
Command: sr
/
; ***************** Statistical Read All *******************
%sra
The "SRA" test is designed to allow large amounts of data to be
read in the shortest amount of time. This test is usually used to characterize
data separators. If an error occurs, the test allows for multiple re-reads of
the error sector. The errors occurring during the re-reads are accumulated and
displayed. In this way it is hoped that the error can be attributed to actual
media related problems or data separator problems. See "sret" for multiple
re-reads. A typical sequence of commands to use the "sra" test is as follows:
b:17
sret:100
recal,sra
It should be noted that sra does not do retry evaluation as does the
"ra" test, and furthermore the data field is not checked for validity against
a pattern. This test does read the entire disk, however.
/
; ****************** Re-Reads for SRA Test ********************
%sret
The "sret" parameter keyword allows the user to specify how many
re-reads of each error sector will be performed during the SRA test. The
command syntax follows:
sret:n
/
; *********** Cylinders to Test During Production Test ************
%test
The "test" parameter keyword allows the number of cylinders that are
to be tested by the production test "ta" to be set. Only the innermost
cylinders of the disk are tested since they are the ones most likely to cause
problems. To test the last 100 cylinders on the disk, merely enter test:100.
The command syntax is as follows:
test:n
/
; ****************** Error Summary Titles *******************
%title
The "title" command allows you to specify a title that will be
displayed at the top of each "list" or error summary. The command syntax
is as follows:
title:text
Note that the text field cannot have embedded blanks.
/
; **************** Clear Error Counters ******************
%zero
The "zero" command clears out all error counters. Command syntax is
as follows:
zero
/
; ****************** Read Verify All ********************
%rva
The "rva" command performs a read-verify disk operation on all sectors
of the disk under test. This operation begins at the currently specified
cylinder location, and continues until the maximum cylinder number is reached.
At this point the test stops. When each cylinder is completed, a "+" mark is
displayed on the screen, thus a "rva" will seem to go more slowly than a read-
all operation (ra) which displays the "." mark for each sector. However, the
rva is actually about a factor of three faster than a standard read.
Be sure to set the desired cylinder location prior to starting the
"rva" command. Also, enable or disable retries as desired. The command
syntax for the "rva" command is:
rva
/
; ****************** Format Bad Track ********************
%fbt
The "fbt" command is used to generate a bad track in a known location.
The currently set cylinder location will be modified by this command to
generate an error each time that location is read or written. The "fbt"
command can be reversed by performing a "format track" ("ft") on the same
cylinder.
Be sure to set the desired cylinder location prior to starting the
"fbt" command. The command syntax for the "fbt" command is:
fbt
/
; ****************** Dump Read Buffer ********************
%dur
The "dur" command causes a display of the contents of the Read Buffer.
At any time, this buffer will contain the data from the most recent read
operation. The command syntax for the "dur" command is:
dur
/
; ****************** Dump Write Buffer ********************
%duw
The "duw" command causes a display of the contents of the Write
Buffer. At any time, this buffer will contain the data from the most recent
write, offset, value, or format operation. The command syntax for the "duw"
command is:
duw
/
; ************** Copy Read Buffer to Write Buffer *********
%copy
The "copy" command copies the read buffer in its entirity to the
write buffer. This is usefull to create ECC errors. Do a Read long, copy,
then modify the write buffer with the "offsets" and "value" commands, then
do a write long. This will cause the wrong ECC bytes to be written to the
disk.
/
; ************** Write the ID Field in the Write Buffer *********
%idon
The "idon" command causes the cylinder, head, and sector values to
be written in the 1st 4 bytes of the write buffer. This is only valid when
blocks are set to 1. This command is useful to validate logical to physical
translation schemes.
/
; *******Disable Writing the ID Field in the Write Buffer *********
%idoff
The "idoff" command cancels the "idon" command.
/
