File:  ~cs552-2/public/UsefulFiles/mentorOverview

 
 * Designs/Components/Sheets/Symbols/Interfaces
--------------------------------------------------------------------------------

A component is a collection of sheets whereas a symbol is an abstraction
of component. An interface is a map between a symbol and a component,
that is it describes the correspodence between the ports of the symbol
and the component.  Ofcourse a component can include another one and
this helps in creating a hierachical structure.

Finaly a design is a collection of components with connections between them.

Let's see at an example:

Say we want to create a 4-by-1 multiplexor. To do this we are going
first to create a 2-to-1 multiplexor. This is by itself a component.

So we choose File>Open sheet. Let's name our component as "mux2to1".  To
do so we append this to the path that appears in the component box.
Note that with our set up this path should point to your ~/mentor
directory.  We have also set "$PROJ_PARTS_namehere" to point there too.
So we can also type in "$PROJ_PARTS_namehere/mux2to1".

We can also change the value of the "Sheet" box to something more
meaningful although this is not necessary; "sheet1" will do fine.

We draw the mux2to1 and save it. Then we can create a symbol for it
using "Misc>Create Symbol". We have to check, save and then check again
this so that Mentor will create the appropriate interface.  This will
create a symbol named "mux2by1".

It is important to note that the 4-to-1 mux is a separate component
although it uses a 2-to-1 mux. So to create it first open a new
component.  If we didn't do this then we are just adding a sheet to
component "mux2to1".  Use FIle>open and now enter the mux4to1 name into
the component box.

To place instances of the 2-to-1 mux just click on the "Choose Symbol"
button of the session window. Ofcourse we need two 3-to-1 multiplexors.
The hierarchy we created look as follows:

			mux4to1
		       /   |   \
		      /    |    \
		     /     |     \
		 mux2to1 mux2to1  mux2to1

A final note of caution: a component cannot be referenced from itself!
So even if you have to seperate sheets that belong to the same component
and create a symbol using only one of them you CANNOT reference the
symbol from the second sheet. Mentor will complain saying that this will
be recursive.

* Location Maps

Mentor applications use a special text file called a "location map" to
determine the location in the filesystem of libraries and other components
that may need.

These location maps allow symbolic name, like for example the
"PROJ_PARTS_YourAccountNameHere" in your .location file, to be used in
place of full pathnames (eg. ~yourHomeDir/mentor).  In Mentor's
terminology these are called "soft prefixes" and "hard pathnames"
respectively.

You can come up with your own soft prefixes and enter the corresponding
hard pathnames in your .location file. If you do this, you can move your
designs around without worrying about the actual physical location of
the design on the disk.

* Properties

DA component symbols and schematic objects (wires, component instances,
etc.) can have information attached to them in the form of properties.
Properties are used by DA and other applications (like QS) and indicate
information about the diesign beyond that of the logical structure of
the schematic. Properties are used to indicate the delay through a gate,
for instance, or the name of a wire.

Properties are "Owned" by the object they are attached to. An object can
own many different properties. Properties themselves are composed of a
NAME, a VALUE and a TYPE. For example, the text "NET" on a newly created
"portin" component - which you later change to a meaningful name using
Properties>Change Text Values.. - is an example of a property.  Say you
change the text from "NET" to "clock". The name of this property happens
to be "NET", the value is "clock" and the type is "string". The fact
that the value of the NET property was "NET" before you changed it is
incidental.

By default, most objects do not have any properties that aren't needed
for normal operation. For instance, wires normaly do not have a NET
property attached to them; DA and QS refer to nets not connects to a
portin or portout symbol by automatically generated symbolic names
unless you add a NET property with a value that means something to you.

* Component symbols

A schematic sheet can have a symbol associated with it. This symbol
would be used in other schematics when you wish to include the
functionality of the first schematic in that schematic. A symbol can be
automatically generated from a schematic by choosing
Miscallaneous>Create Symbol while your schematic window is active in DA.
This symbol then represents the schematic "underneath" it. Portin's and
portout's become pins on the symbol, which are indicated by the purple
diamonds on the symbol.

If you check a symbol before saving it, you will get an error message
indicating that "the pins are not in the interface". This means that
there is no interface for the part. An interface is a description of
what portin's and portout's are on the schematic sheets, what pins are
on the symbol and what is the correspodence between them.  In our
example saving the SYMBOL will automatically create a new interface and
a subsequent check will indicate no error.

Note that if you change the schematic you'll have to update the symbol
as well.  Choose Miscellaneous>Create Symbol to do so.

While the symbol is on your screen, it is a good idea to attach a COMP
property to the outline. The value of this property should be something
that describes what the part does. The comp property will appear near to
any instance of this symbol.

After creating a symbol, be sure to check it, save it, and check it
again!

To add one of your one symbols to a schematic return to the main
schematic palette and choose "Choose Symbol". You will get a navigator
window, which you can use to find your component and select ADD.

* Buses

A Bus is a collection of wires. The width of a bus is the number of its
wires.  Bus widths are specified in DA by assigning a NET property value
of the form "NAME(X:Y)" to a Bus, where "NAME" is the name of the BUS.
Note also thta X must be greater that Y and usualy Y is set to 0.  For
example "In(3:0)" describes a BUS named "In" that has 4 wires.

To draw a BUS select the ADD>Bus option. Drawing it is similar to
drawing a wire.

While in DA use this option to draw a vertical bus. Place a AND gate to
its right and 2 portin's to its left. In order to gain access to the
individual wires of the bus you have to use a component from "gen_lib"
called "bus ripper" AKA "rip". Fortunately Mentor places instances of
this component when you draw a wire to a bus. So choose Add>Wire to
connect the 2 inputs of the AND and the 2 portins to your Bus.

After doing it you should notice the magenta R's that appear near each
"rip".  Changing these to numbers specifies which wire of the bus is
being accessed through the rip. Be sure to add a name to the bus...

Finaly there are also parts to access more than one bits at a time (look
at "rip8x1").

--- ACK ---

Again this originates from a handout by Rob Pfile.