File: gateSim.html

package info (click to toggle)
tkgate 2.1%2Brepack-7
links: PTS, VCS
area: main
in suites: trixie
size: 28,808 kB
sloc: ansic: 62,300; tcl: 20,345; xml: 2,731; yacc: 1,177; lex: 839; sh: 701; makefile: 187; perl: 39
file content (1088 lines) | stat: -rw-r--r-- 45,387 bytes
parent folder | download | duplicates (5)
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML>
<HEAD>
   <TITLE>TKGate User Documentation (Simulation)</TITLE>
   <link rel="stylesheet" href="../tkgate.css" type="text/css">
</HEAD>
<BODY>

<h2>5. Using the Simulator</h2>  

The TkGate simulator, Verga (VERilog simulator for GAte), is a
discrete time simulator based on Verilog.  All modules, including
those designed graphically, are converted to Verilog for simulation.
Verga is a discrete event simulator.  Time advances in discrete time
units called "epochs".  When simulating, time remains at the current
time step (or epoch) until all work scheduled for the epoch as been
completed.  The simulator will then advance the current simulation to
the next epoch that has work scheduled for it.


<a name="start"></a>
<h3>5.1 Starting the Simulator</h3>

<div class=rfig>
<a name=hdleditor>
<IMG SRC="../fig/simulatetab.gif">
<br>
<b>Figure 5.1: Simulator Mode Tab</b>
</div>
To start the simulator, press the "Simulate" tab at the top of the
main window.  If you have not enabled "auto-start" in the circuit
properties, the simulator will compile your design and wait in paused
mode.  If you have enabled "auto-start", the simulation will begin
immediately.
<p style="clear: right;">

<div class=rfig>
<a name=simmodhier>
<IMG SRC="../fig/simmodhier.gif">
<br>
<b>Figure 5.2: Simulator Mode Module Hierarchy</b>
</div>

The simulation will be performed with the designated root module at
the top-level.  The simulator internally expands any module instances
in your circuit.  While the simulator is active, the Module Hierarchy
View <img src=../fig/modtree.gif> changes to show the hierarchy of module
instances as shown in <a href="#simmodhier">Figure 5.2</a>.  In this
mode, the display shows each instance of each module, rather than
simply one entry per module.  The module name for the instance are
shown in angle brackets after the instance name.  For example, the
instance "RF" under "eunit" is an instance of a "REG16" module.  That
module contains an instance "RF0" of a "ZREG" module, and three
instances "RF1", "RF2" and "RF3" of a "REG4" module.  Double clicking
on an instance in this hierarchy view will move the simulator to that
module, and ensure that any probes or signals viewed are for the
instance you selected.
<p>
While the simulator is active you must navigate using either the
Module Hierarchy <img src=../fig/modtree.gif>, or by right clicking on a
module and using "<img src="../fig/blk_open.gif"> Open" and "<img
src="../fig/blk_close.gif"> Close".  You can not use the Module List <img
src=../fig/modlist.gif> view to navigate since TkGate can not tell which
instance of a module to open.

<br style="clear: right;">

<h4>5.1.1 Simulator Console</h4>

<div class=rfig>
<a href="../fig/simcon_msg.gif"><IMG SRC="../fig/small-simcon_msg.gif"><br>
(Click to Enlarge)</a>
<br><br>
<b>Figure <![figure:hdleditor]>5.2: Simulator Console</b>
</div>

At the bottom of the TkGate main window, is a pull up tab for the
simulator console.  When you start the simulator, the console will
open automatically, but you can view it in edit mode as well by
manually pulling up the tab.
<p>
The simulator console has four tabs with different pages of information.
These four pages are:
<p style="clear: right;">
<table class=display style="padding-right: 50;">
<tr><th width=60 align=left>Page</th><th align=left>Description</th></tr>
<tr><td colspan=2><hr></td></tr>

<tr><td><img src=../fig/log.gif class=tool></td><td>
<b>Message Page</b> - This page displays messages from the
console.  Any output from <tt>$display</tt> or other system tasks
generating output will be displayed here.</td></tr> 

<tr><td><img src=../fig/sim_break.gif class=tool></td><td>
<b>Breakpoint Page</b> - Displays breakpoints that have been set for
the simulation and the status of those breakpoints.</td></tr> 

<tr><td><img src=../fig/sim_script.gif class=tool></td><td>
<b>Script Page</b> - Shows the loaded simulation scripts and their status.</td></tr> 

<tr><td><img src=../fig/simoptions.gif class=tool></td><td>
<b>Simulation Control Page</b> - Allows you to set options controlling
the simulation such as the number of epochs to advance at a time when
stepping through a simulation.  </td></tr>
</table>

<br style="clear: right;">

<h4>5.1.2 Simulator Control</h4>


A set of five buttons controls the advance of time in the simulator.
These buttons are:
<p>


<table class=display style="padding-right: 50;" >
<tr><th align=left width=60>Button</th><th align=left>Description</th></tr>
<tr><td colspan=2><hr></td></tr>

<tr><td><img src="../fig/sim_go.gif" class=tool></td><td>

<b>Run</b> - Enters continuous simulation mode.  The simulation will
continue as long as there are events in the event queue.  If there are
any clock gates in your circuit, this will mean the simulation will
continue indefinitely.  If the circuit is combinational, the
simulation will continue until the circuit reaches quiescence.</td></tr>


<tr><td><img src="../fig/sim_pause.gif" class=tool></td><td>
<b>Pause</b> - Causes a continuously running simulation to stop.</td></tr>

<tr><td><img src="../fig/sim_step.gif" class=tool></td><td>
<b>Step</b> - Causes the simulation to advance a fixed number of epochs.  The number of
epochs to advance can be set on the simulation options menu.  You can also
invoke this command with the space-bar.</td></tr>


<tr><td><img src="../fig/sim_clock.gif" class=tool></td><td>
<b>Clock Step</b> - Causes the simulation to advance to the rising edge of a clock.
You can set the number of clock cycles to simulate and the number of
epochs past the designated cycle to step (to allow time for registers
to change value).  The default is to trigger on any clock, but you can
designate a specific clock in the simulator options menu.
You can also invoke this command with the tab key.</td></tr>

<tr><td><img src="../fig/sim_stop.gif" class=tool></td><td>
<b>Stop</b> - Causes the simulation to be terminated and all probes to be deleted.</td></tr>

</table>

<div class=rfig>
<a name=simcont>
<a href="../fig/simcon_opts.gif"><IMG SRC="../fig/small-simcon_opts.gif"><br>
(Click to Enlarge)</a><br><br>
<b>Figure 5.3: Simulator Console Control Page</b>
</div>

You can use the control page of the simulator console shown in <a
href="#simcont">Figure 5.3</a> to control the effects of the step <img
src="../fig/sim_step.gif" class=tool> and clock step <img
src="../fig/sim_clock.gif" class=tool> buttons.  Enter the number of
epochs to advance for each press of <img src="../fig/sim_step.gif"
class=tool> (or space-bar) into the "Epoch Step Size" box.  The clock
step button <img src="../fig/sim_clock.gif" class=tool> (or tab) will
advance by the number of clock steps indicated in "Clock Cycle Step
Size", plus an additional number of epochs entered in the "Clock
Overstep" entry.



<a name="output"></a>
<a name="verilog"></a>
<h3 style="clear: right; ">5.2 Observing the Output</h3>

Except when otherwise noted, Verilog syntax is used to specify and
display values.  A Verilog syntax number contains a prefix to specify
the bit width, a quote character, a radix character and the digits of
the number.  The radix characters using in TkGate are "b" for binary,
"h" for hexadecimal, "o" for octal and "d" for decimal.  For example
"8'h3e" is the 8-bit hexadecimal number "3e".

<h4>5.2.1 Current Value Display</h4>

<div class=rfig>
<a name=valueDisplay>
<IMG SRC="../fig/showval.gif">
<br>
<b>Figure 5.3: Value Display</b>
</div>

To display the current value of a signal in a circuit, click and hold
the mouse button on a wire.  This will display the current value
driven on the wire in Verilog syntax as shown in <a
href="#valueDisplay">Figure 5.3</a>.  The value will disappear when
you release the mouse button.  This feature can be used both when the
simulator is paused and when it is in continuous simulation mode.
When the simulator is in continuous simulation mode, the value
displayed will be the value at the time the mouse button was first
pressed.
<p>
There is a slight delay between pressing the mouse button and
displaying the value.  This is because double clicks on a wire are
used to set probes, so the delay must be long enough to determine that
the mouse press is not a double click.  By default, this delay is
333ms (1/3 of a second), but the delay may be changed through the <a
href="gateOptions.html#simulate">Simulate Options</a> dialog box.

<br style="clear: right;">

<h4>5.2.2 Scope View</h4>

<div class=rfig>
<a name=scopeWin>
<a href="../fig/scope.gif"><IMG SRC="../fig/small-scope.gif"><br>
(Click to Enlarge)</a><br><br>
<b>Figure 5.4: Scope Window</b>
</div>

To set a permanent probe on a signal, double click on a wire.  This
will add or remove a probe.  When a probe is set on a wire, its value
will be continuously displayed in the scope window.  The values of
multi-bit signals will be displayed as hex numbers.  The scope window
(<a href="#scopeWin">Figure 5.4</a>) is viewable any time you have at
least one signal with a probe on it.  When there are no current
probes, the scope window is hidden.
<p>
You can slide the range of time displayed in the scope window by using
the scroll bar for coarse changes, or by clicking and dragging in the
trace part of the window for fine grain control.  To zoom in, you can
press the left mouse button while holding the shift key, press the '>'
key, or press the <img src=../fig/zoom_in.gif class=tool> button on the
toolbar.  To zoom out, you can press the right mouse button while
holding the shift key, press the '<' key or press the <img
src=../fig/zoom_out.gif class=tool> button on the toolbar.
<p style="clear: right;">

<div class=rfig>
<a name=traceValues>
<table border width=400>
<TH>Scope</TH><TH >Description</TH>
<TR>
<TD ALIGN=CENTER VALIGN=CENTER BGCOLOR="#FFFFFF"><img src="../fig/logic0.gif"></TD>
<TD>&nbsp;<b>0</b> - Logic 0 or false (1 bit signals only)</TD>
</TR>

<TR>
<TD ALIGN=CENTER VALIGN=CENTER BGCOLOR="#FFFFFF"><img src="../fig/logic1.gif"></TD>
<TD>&nbsp;<b>1</b> - Logic 1 or true (1 bit signals only)</TD>
</TR>

<TR>
<TD ALIGN=CENTER VALIGN=CENTER BGCOLOR="#FFFFFF"><img src="../fig/logicx.gif"></TD>
<TD>&nbsp;<b>x</b> - Unknown value</TD>
</TR>

<TR>
<TD ALIGN=CENTER VALIGN=CENTER BGCOLOR="#FFFFFF"><img src="../fig/logicz.gif"></TD>
<TD>&nbsp;<b>z</b> - Floating or high impedance</TD>
</TR>

<TR>
<TD ALIGN=CENTER VALIGN=CENTER BGCOLOR="#FFFFFF"><img src="../fig/logicL.gif"></TD>
<TD>&nbsp;<b>L</b> - Low (the signal is either floating or zero)</TD>
</TR>

<TR>
<TD ALIGN=CENTER VALIGN=CENTER BGCOLOR="#FFFFFF"><img src="../fig/logicH.gif"></TD>
<TD>&nbsp;<b>H</b> - High (the signal is either floating or one)</TD>
</TR>

<TR>
<TD ALIGN=CENTER VALIGN=CENTER BGCOLOR="#FFFFFF"><img src="../fig/logicdata.gif"></TD>
<TD>&nbsp;<b>Data</b> - Data value on a multi-bit wire.</TD>
</TR>
</table>
<br>
<b>Figure 5.5: Logic Trace Values</b>
</div>

<p>
The time-line on the scope window is displayed as a base value shown in
the lower left corner, and an offset value shown along the bottom.  In
the example shown in <a href="#scopeWin">Figure 5.4</a>, the base
value is 50016916ns (i.e., 0.05 simulated seconds).  Each tick mark on
the x-axis represents an additional 5000ns past the base time value.

<p>

The scope includes a cross hair that follows the mouse whenever it is
in the scope window to help you correlate events across different
traces.  You can enable or disable the cross hairs with the <img
src=../fig/show_xhair.gif class=tool> button on the toolbar.  The scope
window toolbar also includes some of the same simulation control and
other simulation-related buttons that are on the main TkGate window.
<p>

<a href="#traceValues">Figure 5.5</a> shows how the various logic
values are displayed in the scope windows.  On multi-bit wires, the
value of the wire is displayed in hexadecimal, unless the scale is
such that there is no room to display the value between when it starts
and the next logic transition.  In this case, it will first compress
the value to a "#", and if there is no room to display that either, no
value or symbol will be displayed.  The colors used in the scope
traces can be configured through the <a
href="gateOptions.html#color">Color Options</a> dialog box.

<p style="clear: right;">


<h4>5.2.3 Printing Scope Traces</h4>

<div class=rfig>
<IMG SRC="../fig/scopeSelection.gif"><br>
<b>Figure 5.6: Print Region Selection</b>
</div>


To print a scope trace, first use the right mouse button to select a
region to be printed.  Click and drag with the right mouse to select
the region.  The region will be highlighted as shown in Figure 5.6.
Once you have an active region, you can press the right mouse button
again while holding the Shift key to expand or shrink the selected
region.  If you print without making a selection, then the area of the
trace visible in the scope window will be taken as the selection.
<p>
After you chosen a region, push the <img src=../fig/file_print.gif
class=tool> button on the toolbar of the scope window.  This will
bring up the Scope Print dialog box having an "Output" and a "Content"
page.  The "Output" page of this dialog box is the same as the
"Output" page of the <a href="gateEdit.html#print">dialog box used to
print circuits</a>.
<p style="clear: right;">
If you choose the "Save as Encapsulated Postscript" option, the
"Content" page will be disabled, and the selected region of the trace
will be written as a single Encapsulated Postscript figure suitable
for inclusion in a document (e.g., by Latex).
<p>

<div class=rfig>
<a href="../fig/simprint.gif"><IMG SRC="../fig/small-simprint.gif"><br>
(Click to Enlarge)</a><br><br>
<b>Figure 5.7: Scope Print Dialog Box</b>
</div>

The "Content" page (Figure 5.7) is divided into a "Range Selection"
and a "Scale Selection" portion.  The Range Selection portion has a
selector to modify the start time of the trace plot, and the range of
the selected region in the trace.  A diagram showing the total amount
of trace data available (in green), and the region selected for
printing (in gray) is shown to the right of the selectors.
<p>
The "Scale Selection" allows you to set the amount of time to plot per
line when generating output.  You must be cautious to set this
carefully, as too low a setting can result in trace output with a huge
number of pages.  To help you in setting this value, an estimate of
the number of pages needed is displayed next to this selector.  The
default value for the "Line Length" is one full scale of the scope
window at the current zoom setting.

<p style="clear: right;">


<a name="input"></a>
<h3>5.3 Controlling the Input</h3>

There are two types of built-in circuit elements that can be used to
control your circuit: the single-bit switch, the multi-bit dip switch.
You can also provide input through VPDs (Virtual Peripheral Devices).
An important VPD that is included with TkGate is the TTY which will
also be described in this section.

<h4>5.3.1 Switches and Dip Switches</h4>

<div class=rfig>
<center>
<IMG SRC="../fig/dipval.gif"><br><br>
<b>Figure 5.8: Dip Value<br>Dialog Box</b>
<br><br>

<a name=tty>
<img src=../fig/tty.gif><br><br>
<b>Figure 5.9<br>TTY Device</b><br><br>
</center>
</div>

Switches can be manipulated by simply clicking on them to toggle their
values.  To change a dip switch value, click on the dip switch to open
the dip value dialog box for setting the value of that dip switch.
Then, enter a value in the entry area, and press the "Apply" button to
set the switch.  The value should be a hexadecimal number.  The dip
value dialog box will remain open until you hit "Close", and you can
open as many at once as you like.

<a name="ttys"></a>
<h4>5.3.2 TTYs </h4>


In versions of TkGate before version 2.0, the TTY device was built
into TkGate as a primitive device.  As of version 2.0, the TTY is now
implemented as a VPD (Virtual Peripheral Device).  In order to use the
TTY device, you must open the <a href="gateEdit.html#loadlib">Library
Manager</a> and load the "tty" library.  You can then create a module
of type "TTY" in your design in the same way that you would create a
module instance for any other module.  <a href="#tty">Figure 5.9</a>
shows an example of a TTY device.  Note that it is displayed in the
magenta color of a module, rather than the blue color of built-in
devices.  See the section on the <a href="gateLibrary.html#tty"> TTY Library
Component</a> for information on how to use this device.

<p style="clear: right;">
<a name="breakpoint">
<h3>5.4 Setting Breakpoints</h3>

<div class=rfig>
<a name=simconbreak>
<a href="../fig/simcon_break.gif"><IMG SRC="../fig/small-simcon_break.gif"><br>
(Click to Enlarge)</a><br><br>
<b>Figure 5.10: Simulator Console Breakpoint Page</b>
</div>

Breakpoints can be used to set conditions which will cause a
continuously running simulation to pause.  To display the breakpoints,
press the <img src=../fig/sim_break.gif class=tool> tab on the simulator
console as shown in <a href="#simconbreak">Figure 5.10</a>.  To add a
new breakpoint, press the "Add..." button or double click on the next
empty slot under "Condition".  You can enter any valid Verilog
expression as the breakpoint condition.  You can edit the expression
of a breakpoint by double clicking on the expression.
<p>
Breakpoints will be activated for any non-zero value of their
condition expression.  When one of the registered breakpoints is
activated, the simulation will stop, and a <img src=../fig/bp_stop.gif>
symbol will appear in the "S" column of the breakpoint.  Breakpoints
are only triggered on transitions.  If you press the <img
src="../fig/sim_go.gif" class=tool> button to resume the simulation, the
simulation will continue until the breakpoint expression makes a new
transition to a non-zero value.
<p>
The "ID" column of the breakpoint list indicates an identifying number
for the breakpoint.  The column labeled "S" indicates the state of the
breakpoint.  The state is indicated by one of the following symbols:
<p>
<table class=display>
<tr><th align=left width=50>State</th><th align=left>Description</th></tr>
<tr><td colspan=2><hr></td></tr>
<tr><td><img src=../fig/bp_bad.gif></td><td><b>Error</b> - There is an error such as a syntax error or undefined variable error in the breakpoint expression.</td></tr>
<tr><td><img src=../fig/bp_go.gif></td><td><b>Go</b> - The breakpoint is active, but has not been tripped.</td></tr>
<tr><td><img src=../fig/bp_stop.gif></td><td><b>Stop</b> - The breakpoint has tripped, and simulation is stopped at this breakpoint.</td></tr>
<tr><td><img src=../fig/bp_disabled.gif></td><td><b>Disabled</b> - The breakpoint has been temporarily disabled.</td></tr>
<tr><td><img src=../fig/bp_standby.gif></td><td><b>Stand By</b> - The breakpoint is ready and will be active when the simulator starts.</td></tr>
</table>
<p>
The "Value" column shows the current value of the breakpoint
expression.  If the expression is binary, the value will be a 1 or 0,
but if it is a multi-bit expression, then it could be an arbitrary value.
<p>
There are two buttons to "Enable" and "Disable" a breakpoint.  While a
breakpoint is disabled, the breakpoint will remain in the breakpoint
list, but its value will be ignored until you re-enable it.

<br clear=right>

<a name="memory">
<h3>5.5 Initializing Memories</h3>

A circuit can contain one or more memories (ROM and RAM gates).  You
can initialize memories from a file, or dump the contents of a memory
to a file.  The following toolbar buttons can be used to load or dump
memories.
<p>

<table class=display>
<tr><th width=75>Button</th><th align=left>Description</th></tr>
<tr><td colspan=2><hr></td></tr>

<TR>
<TD align=center><img src="../fig/sim_load.gif" class=tool></TD>
<TD><b>Load Memory</b> - Load memories from the selected file.  If a memory gate is selected, that memory
will be the default memory to load.  If the memory file contains one or more "memory" keywords,
the specified memory(ies) will be loaded with the contents of the file.  When loading a file,
the current directory, the directory of the current circuit file, and the user's home directory
will be searched.</TD>
</TR>

<TR>
<TD align=center><img src="../fig/sim_dump.gif" class=tool></TD>
<TD><b>Dump Memory</b> - Dumps the contents of the selected memory to a file.</TD>
</TR>
</table>

<br>

Memory files have the extension ".mem" be default.  The default format
for memory files is slightly different between Version 2.0 memory
files, and pre-2.0 memory files, but TkGate 2.0 can read files from
either format.
<p>
A memory file is composed lines that can contain commands, or memory
data.  Blank lines and lines beginning with '#' are ignored.  The
supported commands are:
<p>
<table class=display>
<tr><th align=left width=130>Command</th><th align=left>Description</th></tr>
<tr><td colspan=2><hr></td></tr>

<tr><td><tt>@memory</tt> <i>name</i></td><td>Memory data after this
line will be loaded into the memory given by <i>name</i>.  <i>Name</i>
should be the fully qualified Verilog name of the memory comprised of
a "." separated list of the modules down to the module in which the
memory is located.  The "<tt>@</tt>" for this command only can be
omitted, since that is the pre-2.0 style for this command.</td></tr>

<tr><td><tt>@radix</tt> <i>radix</i></td><td>Specify the radix to use
for memory data after this line.  The radix can be 2, 8, 10 or 16 with
16 being the default.</td></tr>

<tr><td><tt>@</tt><i>addr</i></td><td>Specifies the address at which
to begin loading data.  The address should be a hexadecimal number.</td></tr>

</table>
<p>

Lines that do not contain one of the above commands are assumed to be
data in the current radix.  For files loaded through the interface,
the radix is assumed to be hexadecimal, but can be changed with the
<tt>@radix</tt> command.  For files loaded through the Verilog
<tt>$readmemh</tt> system task, the assumed radix is hexadecimal, and
for files loaded through the <tt>$readmemb</tt> system task the
assumed radix is binary.
<p>
Here is an example of a memory file:
<pre>
@100
e1 f0 0 0 e1 e0 0 0
81 0 0 0 12 1 bd 0

@200
e 1 e1 d0 dc 7 85 0
</pre>
This file will load 16 bytes starting at address 100, then an
additional 8 bytes at address 200 (assuming the memory is declared
as an 8-bit wide memory).
<p>
Old style memory files use a slash after an address to indicate where
to load data.  For example:
<pre>
130/ 2 0 ed 0 60 6 62 6
138/ ed 0 5e 6 1 0 85 0
140/ 81 0 0 0 26 4 69 f0
</pre>
Here the 130, 138 and 140 are the address in hexadecimal.  You can use
either syntax in your memory files.
<p>
In the above examples, there was no explicit mention of the target
memory.  For that reason, those file can only be loaded when you
explicitly indicate the target memory either by clicking on before
pressing <img src="../fig/sim_load.gif" class=tool>, or by specifying the
memory as an argument in the <tt>$readmemh</tt> or <tt>$readmemb</tt>
system task.
<p>
Using the <tt>@memory</tt> command, you can specify the memory (or
memories) to be loaded within the file itself.  The memory keyword
requires a single argument specifying the name of a memory.  For
example:

<pre>
@memory memory.m1
@100
e1 f0 0 0 e1 e0 0 0
81 0 0 0 12 1 bd 0

@memory memory.m2
@100
62 65 61 6b 20 69 73 0a
</pre>
This memory file will load 16 bytes of data into the memory "m1" in
the instance named "memory" which is a sub-module of the root module.
It will also load 8 bytes into the memory "m2" in the same module as
"m1".
<p>
You can also use the "<tt>x</tt>" and "<tt>z</tt>" characters in any
digit of data values specified in a memory file to indicate unknown or
floating values.  For example:
<pre>
@100
10 x 8x 3e z9 3a zx 9x
</pre>
will load a memory that includes unknown and floating values.
<p>
For RAM memories, the contents of the loaded memory are frozen until
first time the write line transitions to a stable value (logic 0 or
1).  This prevents the data loaded from being destroyed due to unknown
values on the write and address lines until the circuit has time to
initialize these signals.

<h4>5.5.1 Using GMAC to Create Memory Files</h4>
<p>
In many cases, memory files are initialized to act as microstores or
to contain machine instructions for a user designed processor.  In
these cases, it is very tedious and error prone to explicitly specify
the contents of the memory.  For this reason, TkGate includes a tool,
gmac, for compiling microcode and macrocode to TkGate-compatible memory
files.  Complete documentation on <a href="gateGmac.html">Gmac</a> is
given in a later section of this document.

<a name="script">
<h3>5.6 Simulation Scripts</h3>

<div class=rfig>
<a href="../fig/simcon_script.gif"><IMG SRC="../fig/small-simcon_script.gif"><br>
(Click to Enlarge)</a><br><br>
<b>Figure 5.11: Simulator Console Script Manager Page</b>
</div>

Simulation scripts are useful for setting up a simulation before
starting, or for running a simulation in batch mode.  You can perform
most of the operations you can do manually through the interface
including setting and remove probes, changing switch values, loading
memories, setting breakpoints, and stepping the simulator.
<p>
Simulator script files use Verilog syntax.  This section will give
some simple examples of how to write scripts.  More detailed
information on Verilog format can be found in the chapter on <a
href=gateHDL.html>Verilog Modules</a>.

<h4>5.6.1 Loading Simulator Scripts</h4>

To load a simulator script, first press the <IMG
SRC="../fig/sim_script.gif" class=tool> tab on the simulator console.
You can then press the "Add..." button or double click on the next
empty slot.  The default extension for simulation scripts is ".vs".
When loading a file, the current directory, the directory of the
current circuit file, and the user's home directory will be searched.
<p>
You can also arrange for simulation scripts to be automatically
executed when you start the simulator by adding one or more simulation
scripts in the <a href="gateEdit.html#circopt">circuit options</a>
dialog box.
<p>
The "S" column next to each simulator script indicates the state of
the script.  The state is one of the following:
<p>
<table class=display>
<tr><th align=left width=50>State</th><th align=left>Description</th></tr>
<tr><td colspan=2><hr></td></tr>
<tr><td><img src=../fig/bp_bad.gif></td><td><b>Error</b> - There is an error such as a syntax error or undefined variable error in the script.</td></tr>
<tr><td><img src=../fig/bp_go.gif></td><td><b>Go</b> - The script is active and running.</td></tr>
<tr><td><img src=../fig/ss_done.gif></td><td><b>Stop</b> - The script has completed executing.</td></tr>
<tr><td><img src=../fig/bp_disabled.gif></td><td><b>Disabled</b> - Execution of the script is disabled.</td></tr>
<tr><td><img src=../fig/bp_standby.gif></td><td><b>Stand By</b> - The script is ready and will execute when the simulator is started.</td></tr>
</table>


<h4 style="clear: right;">5.6.2 Simulator Script Format</h4>

Simulator scripts are fragments of Verilog code assumed to be defined
in the context of the body of the top-level module.  You can create
one or more parallel threads as well as declare local variables for
use within the script.

<p>
Most scripts are defined as a Verilog <tt>initial</tt> block.  For
example:
<pre>
  initial
    begin
      $readmemh("test.mem");    // Load memory file "test.mem".
      $tkg$probe(a, b, c);      // Place probes on the signals a, b and c.
      repeat (5)                // Advance the simulator 5 steps of the
        @ (posedge clock);      //    clock named 'clock'.
      # 10;                     // Advance the simulator 10 epochs.
      $stop;                    // Stop the simulator and put it in "pause" mode.
    end
</pre>

If you define multiple <tt>initial</tt> blocks in your script file,
each block will execute in parallel.  You can also use <tt>always</tt>
blocks which execute their bodies in an infinite loop.
<p>
You can also define local variables in a simulator script.  For example:
<pre>
  integer i;

  initial
    begin
      for (i = 0;i < 10;i = i + 1)        // Loop ten times
        begin
          @ (posedge clock);              // Advance to rising edge of "clock".
          $display("%t: x=%h",$time,x);   // Print time and value of x signal.
        end
      $stop;                              // Pause the simulator.
    end
</pre>
This example will step for 10 clock periods and print the value of the
<tt>x</tt> signal in the simulator console at each of those clock periods.
<p>

<h4>5.6.3 Setting Signal Values</h4>


<div class=rfig>
<a name=adder>
<IMG SRC="../fig/adder_circuit2.gif">
<br>
<b>Figure 5.12: Full Adder Circuit</b>
</div>

You can set the value of any register variables in a module using an
assignment statement with the "<tt>=</tt>" operator.  For netlist
modules, any nets that are attached to a switch or a dip element are
considered register variables.  However, unlike in TkGate 1.8 script
files, you must use the name of the wire, and not the name of the
switch itself.  For example, the following script will set probes on
the inputs and output of the circuit shown in Figure 5.12, then step
through each combination of inputs with a delay of 50 epochs after we
set the input values:
<p>
<pre>
  initial
    begin
      $tkg$probe(a,b,ci,s,co);
      ci = 0; a = 0; b = 0;
      #50 ci = 0; a = 0; b = 0;
      #50 ci = 0; a = 0; b = 1;
      #50 ci = 0; a = 1; b = 0;
      #50 ci = 0; a = 1; b = 1;
      #50 ci = 1; a = 0; b = 0;
      #50 ci = 1; a = 0; b = 1;
      #50 ci = 1; a = 1; b = 0;
      #50 ci = 1; a = 1; b = 1;
    end
</pre>
Note that the left hand side of each assignment statement is a wire
name.  For example, <tt>a</tt> is used instead of the name of the
switch (<tt>g25</tt>) to which it is attached.  The right-hand side of
assignment statements can be arbitrary expressions referencing
variables declared in the simulation script and any nets in the
design.
<p>
You can use fully qualified path names to set the value of switches/nets
at levels other than the top level.  For example:
<pre>
    top.memctl.ttyreg = 8'h
</pre>
will set the value of <tt>ttyreg</tt> in the module instance
<tt>memctl</tt> which is instantiated in the top-level module.

<h4 style="clear: right;">5.6.4 Commonly Used System Tasks</h4>

Scripts can make calls to system tasks to perform various useful
functions.  Systems tasks begin with a "<tt>$</tt>" and are used
somewhat like function calls.  A complete description of system tasks
are given in the <a href=systemTasks.html>System Tasks Appendix</a>.
Some of the system tasks that are frequently used in scripts are:


<table class=display>
<tr><th width=400 align=left>Task</th><th align=left>Description</th></tr>
<tr><td colspan=2><hr></td></tr>

<tr><td><tt>$display(</tt><i>arg1</i><tt>, </tt><i>arg2</i><tt>, </tt>...<tt>)</tt></td><td>
Display messages to the simulator console.  Similar to the C <tt>printf</tt> in functionality.</td></tr>

<tr><td><tt>$readmem(</tt><i>filename</i><i> [, memory [, start [, stop]]]</i><tt>)</tt></td><td>
Loads the contents of a memory file to one or more memories.
</td></tr>

<tr><td><tt>$tkg$probe(</tt><i>sig1</i><tt>, </tt><i>sig2</i><tt>, </tt>...<tt>)</tt></td><td>
Places probes on the specified signals.
</td></tr>

<tr><td><tt>$tkg$unprobe(</tt><i>sig1</i><tt>, </tt><i>sig2</i><tt>, </tt>...<tt>)</tt></td><td>
Removes probes from the specified signals.
</td></tr>

<tr><td><tt>$stop</tt></td><td>
Stop the simulation and put it into "paused" mode.
</td></tr>

<tr><td><tt>$finish</tt></td><td>
Terminate the simulation and return to "edit" mode.
</td></tr>

<tr><td><tt>$random</tt><i>[(seed)]</i></td><td>
Return a random number, or set the seed if an argument is given.
</td></tr>

<tr><td><tt>$time</tt></td><td>
Return the simulation time in epochs as a 64-bit integer.
</td></tr>

<tr><td><tt>$tkg$systime</tt></td><td>
Returns the actual system time in milliseconds since January 1, 1970 as a 64-bit integer.
</td></tr>


</table>
<p>

<h4>5.6.5 Using Scripts for Test Vector Generation</h4>

Another method for using scripts is to generate random test vectors.
For example, suppose we wish to test a 32-bit adder that we have
designed.  The following script will apply 10,000 random vectors to
the design and check them against the answer calculated in the script.
When a mismatch between the circuit output and the expected output is
found, an error message is displayed, and the simulation is paused so
the user can view internal signal values.

<pre>
  reg tempCout;                                    // Declare 1-bit variable for carry out
  reg [31:0] tempS;                                // Declare 32-bit variable for sum
  integer i;                                       // Declare index variable

  initial
    begin
      $random($tkg$systime);                       // Use system clock to set random seed
      for (i = 0;i < 10000; i = i + 1)             // Loop 10,000 times
        begin
          a = $random;                             // Pick random value for A
          b = $random;                             // Pick random value for B
          cin = $random;                           // Pick random value for carry-in

          # 50;                                    // Step 50 epochs

          tempCout = (33'h0 + a + b + cin) >> 32;  // Get correct carry out value
          tempS = a + b + cin;                     // Get correct sum value

          //
          // If output from the design is not as expected, print an error message
          // and stop the simulator. 
          //
          if ({cout,s} != {tempCout,tempS})
            begin
              $display("error: got 31'h%h + 31'h%h + 31'h%h = {1'h%h, 31'h%h}",a,b,cin,cout,s);
              $display(" -- should have gotten {1'h%h, 31'h%h}", tempCout, tempS);
              $stop;
            end
        end

      $display("test completed.");
      $stop;
    end
</pre>

<a name="error">
<h3>5.7 Using the Error Reporter</h3>


<div class=rfig>
<a name=errbox>
<a href="../fig/simerr.gif"><IMG SRC="../fig/small-simerr.gif"><br>
(Click to Enlarge)</a>
<br><br>
<b>Figure <![figure:hdleditor]>5.13: Simulator Error List</b>
</div>

If there are any errors in the circuit when you start the simulator,
an error list box will appear as shown in <a href="#errbox">Figure
5.13</a>.  Click on an error message to gave TkGate show you the
location of the error.  If the error is in a netlist module, TkGate
will navigate to the module and place cross-hairs over the as shown in
<a href="#xhair">Figure 5.14</a>.  If the error is in an HDL module,
then the line with the error will be highlighted.

<div class=lfig>
<a name=xhair>
<img src=../fig/simerr_circ.gif><br><br>
<b>Figure <![figure:hdleditor]>5.14: Error Location Cross-hairs</b>
</div>

<p>
In the example shown here, the module "ADD32" had an internal port
name of "a", but the name "A" was erroneously used on the interface.
The first error message indicates that the external port name "A" does
not have a corresponding port inside, and the second error message
indicates that the internal port "a", does not have a corresponding
port on the outside.

<br style="clear: both;">

<a name="delay">
<h3>5.8 Gate Delay Files</h3>

Gate delay, area and power (power specifications are not used)
parameters can be specified through a collection of gdf (Gate Delay
File) specification files.  The default file "gdf/default.gdf" in the
TkGate home directory is always loaded, but the definitions may be
replaced by loading additional delay files through the <a
href="gateOptions.html#simulate">Simulate Options</a> dialog box.
<p>
Each gate delay file should contain one or more technology blocks
having the form:
<pre>
technology cmos {
  ...
}
</pre>
Technology blocks implement a new set of delay parameters and the
specified technology name becomes a name that can be specified as a
technology through the <a href="gateEdit.html#delay">"Delay" page of
the gate properties box</a>.  The body of the technology block should
consist of a set of gate primitive declarations.  Each gate primitive
should be described in a block defining all the delay and possible
area and power parameters of the block in terms of number of inputs,
bit-widths of inputs and existence of inverters on inputs.  For
example, the block for the "mux" primitive might look like:

<pre>
primitive mux {
  delay&lt;I-Z> = (2*(num(S)+1) + 2*num(I)) + 2*(inv(I) || inv(Z)); 
  delay&lt;S-Z> = (2*(num(S)+1) + 2*num(I));
  area = bits(Z)*((2*(num(S)+1) + 2*num(I)) + 2*inv(I)); 
}
</pre>

The two "delay" lines define the delay from the input to the output
(I-Z) and from the select to the output (S-Z), respectively.  The area
line defines the estimated area of the gate.  Expressions may include
the C-style operators <tt>"+", "-", "*", "/", "&&", "||", "==", "!=", ">", ">=", "&lt;",
"&lt;="</tt>, and <tt>"!"</tt>, the "power of" operator <tt>"**"</tt> and
the functions listed in the table below.


<p>
<table border>
<tr><th>Function</th><th>Description</th></tr>

<tr><td>num(p)</td><td> Normally, "p" specifies a group of related
ports in which case this function returns the number of ports in the
group.  For example, on an n-input AND gate with inputs I1 through In,
the expression num(I) would return n.  </td></tr>

<tr><td>bits(p)</td><td> Returns the number of bits on the specified
port.  If "p" represents a group of ports, the highest bit-width of
the group is returned.  </td></tr>

<tr><td>inv(p)</td><td> When "p" is specific port, a 1 is returned if
there is an inverter on the port, 0 otherwise.  When "p" specifies a
group of ports, the number of ports with inverters is returned.
</td></tr>

<tr><td>log(expr)</td><td>Returns the ceiling of the base-2 log of an expression.</td></tr>
</td></tr>

</table>
<p>

<p>
It is also possible to write procedural delay/area definitions.  For example, consider
the parameter specification for the "and" primitive.
<pre>
primitive and {
  delay&lt;I-Z> = {
    if ((inv(I) == num(I)))     // Determine if an inverter is necessary.  An
      d = inv(Z);               //   inverter is not required if the output is
    else if ((inv(I) == 0))     //   inverting and all inputs are non-inverting
      d = !inv(Z);              //   (i.e., it is an AND gate), or if all inputs
    else                        //   are inverting and the output is non-inverting
      d = 1;                    //   (i.e., it is a NOR gate). 

    if (num(I) == 1) {          // If one input, treat this as a reduction gate    
      return 2*bits(I0) + 2*d;  //   one Tr. delay per bit plus inverter delay.    
    } else {                    // If multiple inputs, treat this as a normal gate 
      return 2*num(I) + 2*d;    //   one Tr. delay per input plus inverter delay.  
    }
  }

  area = {
    if ((inv(I) == num(I)))     // Estimate number of inverters required.  If all
      d = inv(Z);               //   inputs are inverted, an inverter is required
    else if ((inv(I) == 0))     //   iff the output is inverted.  If all inputs
      d = !inv(Z);              //   are non-inverted, an inverter is required iff
    else                        //   the output is non-inverted.  Otherwise we need
      d = inv(I);               //   an inverter for each inverted input.

    if (num(I) == 1) {          // If one input, treat this as a reduction gate    
      a = 2*bits(I0) + 2*d;	//   one Tr. per bit plus inverter Trs.    
    } else {			// If multiple inputs, treat this as a normal gate 
      a = 2*num(I) + 2*d;	//   one Tr. per input plus inverter Trs.  
    }
    return bits(Z)*a;           // Multiply by number of bit slices.
  }
}
</pre>

In a procedural specification, statements are executed sequentially
until a "<tt>return</tt>" statement sets the value for the parameter.
C-style <tt>if</tt> and <tt>switch</tt> statements may be used in
procedural specifications, but there are no looping constructs.
<p>

A technology block need not specify every single primitive type.  For
example, suppose we have a technology definition for CMOS which
includes a definition for a buffer as shown here:

<pre>
technology CMOS {
  primitive buf {
    delay&lt;I-Z> = 2 + 2*(inv(I) == inv(Z));
    area = bits(Z)*(2 + 2*(inv(I) == inv(Z)));
  }
  ...rest of CMOS definition...
}
</pre>

We might create a special technology definition "HP_CMOS" which
contains high-power versions of a subset of the standard CMOS gates.
In the example here, we provide a special high-power buffer that has
 the delay of a standard buffer, but twice the area (and power).

<pre>
/*
 * A high-power buffer with half the delay and double the area/power.
 */
technology HP_CMOS {
  primitive buf {
    delay&lt;I-Z> =  1 + (inv(I) == inv(Z));
    area = 2*bits(Z)*(2 + 2*(inv(I) == inv(Z)));
  }
}
</pre>
<p>
If you assign the technology type HP_CMOS to any gates for which there
is no primitive definition, delay values from the technology "default"
will be used.
<p>

The table below lists all of the primitive gates for which delay may
be specified and their delay parameters.  The default values given are
the values for a "basic" gate created from the "Make" with no changes
to the number of inputs, input/output inverters, or bit widths of any
of the ports.

<br><br>
<center>
<table border>
<tr><th>Gate Type</th><th>Parameter</th><th>Default</th><th align=left>Description</th></tr>
<tr><td>and </td><td> I-Z </td><td> 6</td><td>Delay from input to output.</td></tr>
<tr><td>or </td><td> I-Z </td><td> 6</td><td>Delay from input to output.</td></tr>
<tr><td>xor </td><td> I-Z </td><td> 8</td><td>Delay from input to output.</td></tr>
<tr><td>buf </td><td> I-Z </td><td> 4</td><td>Delay from input to output.</td></tr>
<tr><td rowspan=2>bufif1 </td><td> E-Z </td><td> 4</td><td>Delay from enable to output.</td></tr>
<tr><td> I-Z </td><td> 6</td><td>Delay from data input to output.</td></tr>
<tr><td rowspan=2>nmos </td><td> I-Z </td><td> 2</td><td>Delay from data input to output.</td></tr>
<tr><td> G-Z </td><td> 1</td><td>Delay from gate to output.</td></tr>
<tr><td rowspan=2>pmos </td><td> I-Z </td><td> 2</td><td>Delay from data input to output.</td></tr>
<tr><td> G-Z </td><td> 1</td><td>Delay from gate to output.</td></tr>
<tr><td rowspan=4>add </td><td> A/B-S </td><td> 68</td><td>Delay from operand input to sum.</td></tr>
<tr><td> A/B-CO </td><td> 70</td><td>Delay from operand input to carry out.</td></tr>
<tr><td> CI-S </td><td> 62</td><td>Delay from carry in to sum.</td></tr>
<tr><td> CI-CO </td><td> 64</td><td>Delay from carry in to carry out.</td></tr>
<tr><td rowspan=3>register </td><td> setup </td><td> 10</td><td>Time before clock edge data-in must be stable.</td></tr>
<tr><td> hold </td><td> 10</td><td>Time after clock edge data-in must remain stable.</td></tr>
<tr><td> CK-Q </td><td> 20</td><td>Time from clock edge until output changes.</td></tr>
<tr><td rowspan=2>mux </td><td> S-Z </td><td> 8</td><td>Delay from select line to output.</td></tr>
<tr><td> I-Z </td><td> 8</td><td>Delay from data input to output.</td></tr>
<tr><td rowspan=2>demux </td><td> E-Z </td><td> 6</td><td>Delay from enable line to output.</td></tr>
<tr><td> I-Z </td><td> 6</td><td>Delay from data input to output.</td></tr>
<tr><td>mult </td><td> A/B-P </td><td> 252</td><td>Delay from operand input to output.</td></tr>
<tr><td rowspan=2>div </td><td> A/B-Q </td><td> 236</td><td>Delay from operand input to quotient.</td></tr>
<tr><td> A/B-R </td><td> 236</td><td>Delay from operand input to remainder.</td></tr>
<tr><td rowspan=7>ram </td><td> OE-D </td><td> 10</td><td>Delay from output enable to data-out.</td></tr>
<tr><td> CS-D </td><td> 10</td><td>Delay from chip select to data out.</td></tr>
<tr><td> A-D </td><td> 70</td><td>Delay from the address line to the data out.</td></tr>
<tr><td> addr_setup </td><td>10</td><td>Time before write is asserted that address must be stable.</td></tr>
<tr><td> data_setup </td><td> 10</td><td>Time before write is asserted that data-in must be stable.</td></tbr>
<tr><td> addr_hold </td><td> 10</td><td>Time after write is unasserted that address must remain stable.</td></tr>
<tr><td> data_hold </td><td> 10</td><td>Time after write is unasserted that data-in must remain stable.</td></tr>
<tr><td rowspan=2>rom </td><td> OE-D </td><td> 10</td><td>Delay from output enable to data-out.</td></tr>
<tr><td> A-D </td><td> 50</td><td>Delay from address to data-out.</td></tr>

<tr><td rowspan=2>lshift </td><td> S-Z </td><td> 16</td><td>Delay from shift select to output.</td></tr>
<tr><td> I-Z </td><td> 18</td><td>Delay from data-in to output.</td></tr>
<tr><td rowspan=2>rshift </td><td> S-Z </td><td> 16</td><td>Delay from shift select to output.</td></tr>
<tr><td> I-Z </td><td> 18</td><td>Delay from data-in to output.</td></tr>
<tr><td rowspan=2>arshift </td><td> S-Z </td><td> 16</td><td>Delay from shift select to output.</td></tr>
<tr><td> I-Z </td><td> 18</td><td>Delay from data-in to output.</td></tr>
<tr><td rowspan=2>roll </td><td> S-Z </td><td> 16</td><td>Delay from shift select to output.</td></tr>
<tr><td> I-Z </td><td> 18</td><td>Delay from data-in to output.</td></tr>
</table>
</center>


</BODY>
</HTML>