Heading:qjk40(635) SimMOSy756qjk40\6f6 Page Numbers: Yes X: 527 Y: 10.5"qjk40 SimMOSy624j\f9b by Martin Newelll3680d2999e12j\f1 2f0b Version of December 5, 1980 12:27 PMz18697e24j\f1b [Ivy]SimMOS>SimMOS.bravo,.pressz18697e24j\35b z18697e24j\b PREFACEz18697x2e64j\b SimMOS is a mixed mode switch level and functional simulator for n-channel MOS circuits. No timing information is represented in the simulator, although sequential circuits, and even some asynchronous circuits, can be simulated. Parts of a circuit can be replaced with user-written procedures that model the behaviour of those parts. This permits the simulation of critically timing dependent subcircuits, enhances efficiency, and permits top down development of circuits by starting with only their functional simulation, and substituting implementations of parts of the circuit as they are developed.l3634e12jk40(2116) XEROXz18697l10583y184e12(635)\f2 PALO ALTO RESEARCH CENTER 3333 Coyote Hill Road Palo Alto California 94304z20355l10592(2116)\b e24\b CONTENTSe24c\b1f1 7f0 e24jk72(635)\b 1. INTRODUCTIONe24jk72\b4f1 11f0 1.1 Backgrounde12jk72 1.2 Acknowledgementse12jk72 2. OPERATIONe24jk72\b4f1 8f0 2.1 Input Formatse12jk72 2.2 Usagee12jk72 2.3 Where to find SimMOSe12jk72 3. COMMANDSe24jk90\b4f1 7f0 3.1 Initializatione12jk72 3.2 Circuit Definitione12jk72 3.3 Setting and Reading Valuese12jk72 3.4 Running a Simulatione12jk72 3.5 Saving and Restoring Statee12jk72 3.6 Othere12jk72 4.0 EXAMPLESe24jk72\b5f1 7f0 4.1 Nor Gatee12jk72 4.2 Shift Registere12jk72 4.3 Two Phase Clock Generatore12jk72 4.4 Functional Simulatione12jk72 5.0 REFERENCESe24jk72\b5f1 9f0 6.0 INDEX OF COMMANDSe24jk72\b5f1 8f0 1f1 7f0 e12j(1792)\b 1. INTRODUCTIONe24jk72(635)\b4f1 11f0 SimMOS is a mixed mode switch level and functional simulator for n-channel MOS circuits. No timing information is represented in the simulator, although sequential circuits, and even some asynchronous circuits, can be simulated. Parts of a circuit can be replaced with user-written procedures that model the behaviour of those parts. This permits the simulation of critically timing dependent subcircuits, enhances efficiency, and permits top down development of circuits by starting with only their functional simulation, and substituting implementations of parts of the circuit as they are developed. However, use of this technique is not yet documented.l3634e12jk40 1.1 Backgrounde12jk72 SimMOS is based on the switch level simulator MOSSIM from MIT. MOSSIM was originally written by Randy Bryant to support the VLSI course given at MIT by Lynn Conway in the Fall of 1978. Since then it has been further developed by Randy Bryant and has been adapted by several people into various versions in several languages. A version by Chris Terman, written in Lisp, was used to diagnose the design of the Scheme79 microprocessor. It is this version that formed the basis for the Mesa implementation, SimMOS.l3634e12jk40 1.2 Acknowledgementse12jk72 To Randy Bryant and Chris Terman at MIT, and to Alan Bell, for several long discussions and help in sorting out the basic algorithms. Voluminous feedback from the first heavy user, Jim Cherry, was instrumental in getting SimMOS into a usable state.l3634e12jk40 2. OPERATIONe24jk72\b4f1 8f0 SimMOS is implemented as a set of programs in Mesa. Several versions are available for running on different configurations of machines, ranging from a single bank Alto version, through an XM Alto version using high memory for data storage, to D0 and Dorado versions that make efficient use of high memory. The D version on the Dorado is about 60 times faster than the XM Alto version. Only the XM Alto and D machine versions are distributed.l3634e12jk40 While the basic SimMOS functions are implemented in Mesa, the user interface is provided by running SimMOS under the interactive programming system JaM [JaM]. JaM provides a simple interpreter that supports the JaM programming language, and from which Mesa programs can be called. This programmibility of the user interface proves to be invaluable for developing test procedures for circuits being simulated. (It also provides the basic mechanism by which functional simulation can be implemented along with the switch level simulation). For the purposes of this document the term SimMOS will be used to cover the combination of the basic implementation and the JaM user interface.l3634e12jk40 2.1 Input Formatse12jk72 Circuits can be specified to SimMOS in one of two formats: the .sim format or as a JaM program, or as a combination of both. The .sim format is provided to allow input from Clark Baker's circuit extractor, developed at MIT. This is a simple keyword format in which every node and every transistor is explicitly listed. Apart from that the format is fairly compact. It is unlikely that anyone will want to create a specification of a circuit in .sim format, except from the circuit extractor.l3634e12jk40 The other format, a JaM program, is typically used for making incremental changes to a circuit that has been read in from .sim format. Since JaM programs can be executed from files, this format also provides a way of storing such changes, or even whole circuits. An advantage of using the JaM format is that procedural descriptions are available, so that repeated parameterized structures can be efficiently represented.l3634e12jk40 2.2 Usagee12jk72 While many different scenarios are possible, one is given here to indicate one way in which SimMOS can be used. SimMOS is first obtained and initialized by executing the relevant command file for the machine being used. A circuit specification, typically obtained from the circuit extractor is read from a .sim file into the internal data structures of SimMOS. Initially, to check out some small part of the circuit, certain named nodes might be set to chosen values (hi or lo) using JaM commands, and the circuit solved. (This can take anywhere between instantaneous for a small circuit, through a few seconds for a large circuit on a Dorado, to a few minutes for a large circuit on an Alto). Other node values can now be examined individually to determine whether the right behaviour was obtained. If anomolous behaviour is observed, the cause can usually be found by running other small tests and looking at other nodes, and then the circuit can be reconfigured interactively.l3634e12jk40 It soon gets tedious having to set and examine individual node values. Consequently a few utilities are provided to deal with sequences of values for sets of nodes. One of these allows a vector of input node names to be defined, together with a string of values for each node name. A similar vector of output node names can be defined. The procedure arraysim takes these vectors and for each set of values in the input vector it sets up those inputs, solves the circuit, and collects the outputs named in the output vector. This mechanism allows sets of test patterns to be applied to a circuit. However, depending on the size of circuit and the length of the test strings each such test can take a considerable amount of time. To relieve some of this burden, it is possible to break such a test from the keyboard, examine the output strings collected thus far, modify the circuit/input vectors/node values/etc., and then to resume the test where it left off. Of course, the test can be restarted from the beginning if desired.l3634e12jk40\354i8I When a session is over, the entire state of SimMOS can be saved, including the state of an interrupted test for resumption at some later date. Much of this state is saved automatically in JaM's virtual memory which is always saved and restored from run to run. However, the circuit is held in the SimMOS data structure and if it has been modified then it must be saved separately. This can be done in a number of ways. First, a new .sim file can be written. Alternatively, a JaM file whose subsequent execution will recreate the circuit, can be written. Both of these methods can consume large amounts of disk space, and are inefficient if only a small part of the circuit has been modified. With foresight it is possible to automatically log all JaM commands that modify the circuit, using the logchanges command. Given this log the modified circuit can be reconstructed by reading whatever definition was read on the previous session, and then replaying the log of commands that modified the circuit. This method has the added advantage of allowing backing up to previous versions of the circuit, by removing the last part of the log file using a text editor.l3634e12jk40\802i10I Other scenarios are possible. For example, using the ability to mix switch level and functional simulation it is possible to read in a JaM specification of the functional behaviour of a system. Selected parts of the system can then be disabled, and replaced with circuit specifications that implement the disabled parts. In this way, individual pieces of a system can be diagnosed independently and efficiently.l3634e12jk40 2.3 Where to find SimMOSe12jk72 SimMOS can be obtained and initialized by executing one of the command files:l3634e12jk40 [Ivy]Alto>SimMOS>SimMOS.cm [Ivy]DStar>SimMOS>SimMOS.cml7056e12k40 for the XM Alto and D machine versions respectively. Also on that directory is a file:l3634e12jk40 [Ivy]SimMOS>cleanupSimMOS.cml7056e12k40 that will delete most of the files created during a run. This is useful for cleaning up after using a Dorado.l3634e12jk40 3. COMMANDSe24jk90\b4f1 7f0 The available commands are from three sources - commands directed to JaM, primitive SimMOS commands, and JaM procedures. While users need not be concerned with these distinctions, it may be helpful to note that online documentation of most commands that are implemented as JaM procedures can be obtained using the JaM utility command "?", e.g. (arraysim)?. Moreover, these commands can be used as a starting point for new commands tailored to a particular need.l3634e12jk40 The commands are presented here in six categories: Initialization, Circuit Definition, Setting and Reading Values, Running a Simulation, Saving and Restoring State, and Other.l3634e12jk40 3.1 Initializatione12jk72 First get and initialize SimMOS as described under 2.3 "Where to find SimMOS". From here on the operating procedures for both Alto and D machine versions are the same. SimMOS is normally started by typing:l3634e12jk40 jam simmosl4269e12jk40\f8i When loaded SimMOS should prompt with the message:l3634e12jk40 SiMMOS >>l4269e12jk40\f8i at which time it is ready for commands. The initial getting and initializing if SimMOS will leave SimMOS in this same state.l3634e12jk40 Note that commands in JaM are executed when a carriage return is typed. Several commands can be put on one line.l3634e12jk40 simreset Usage: simreset Initialize the stored circuit to null. The circuit is automatically initialized on start-up, so simreset need be called only to clean out an existing circuit. e.g.l4269d3634e12jk40\b8B8b8B simresetl4904e12jk40\f8i8I Fine point: while simreset is implemented in all versions, it doesn't currently release the space used in all versions, so its use for large circuits is to be discouraged - get out of SimMOS and start again.l4269e12jk40\f1 3.2 Circuit Definitione12jk72 simread Usage: simread Read the definition of a circuit in .sim format from the indicated file. e.g.l4269d3634e12jk40\b7B19b7B (mycircuit.sim)simreadl4904e12jk40\f8i22I .run Usage: .run Run the indicated file as a JaM program. This JaM primitive is included here since it is used for creating a circuit specified as a JaM program. e.g.l4269d3634e12jk40\b4B19b4B (mycircuit.jam).runl4904e12jk40\f8i19I simnode Usage: simnode Create a node having the given name. If a node of the given name already exists this does nothing. Note that the nodes (VDD) and (GND) are always defined in SimMOS, and have the expected values. Node names are stored as originally specified, but their recognition is case shift independent. For example (vdd), (Vdd), and (VDD) are all equivalent. e.g.l4269d3634e12jk40\b7B19b7B (newnode)simnodel4904e12jk40\f8i16I would create a node having the name "newnode".l4269e12jk40 etrans Usage: etrans Create an enhancement mode transistor connecting the nodes having the specified names. Any node that doesn't already exist is created. e.g.l4269d3634e12jk40\b6B30b6B (input)(GND)(pulldown)etransl4904e12jk40\f8i28I dtrans Usage: dtrans Create a depletion mode transistor connecting the nodes having the specified names. Any node that doesn't already exist is created. e.g. to create a pullup for node bus0:l4269d3634e12jk40\b6B30b6B (bus0)(bus0)(VDD)dtransl4904e12jk40\f8i23I listetrans Usage: listetrans List all enhancement mode transistors having the specified gate, and source and drain being term1 and term2 in either order. Output is in the form of a list of etrans commands. A null string for a node name acts as a wild card, e.g. to list all enhancement mode transistors (not to be used indiscriminately):l4269d3634e12jk40\b10B29b10B ()()()listetransl4904e12jk40\f8i16I or to list all enhancement mode transistors with gates connected to the node PHI1:l4269e12jk40 (PHI1)()()listetransl4904e12jk40\f8i20I listdtrans Usage: listdtrans Same as listetrans except that only depletion mode transistors matching the given names are listed.l4269d3634e12jk40\b10B29b10B Note that depletion mode pullups are not represented explicitly in the SimMOS data structure. Consequently super buffer pullups will be reported as regular depletion mode pullups. Other types of depletion mode transistors, such as in function blocks, are correctly reported.l4269e12jk40 listtrans Usage: listtrans A combination of listetrans and listdtrans.l4269d3634e12jk40\b9B29b9B redefinetrans Usage: redefinetrans Redefine the transistor specified by the first three parameters to be connected as given by the last three parameters. The first three parameters are used as in listtrans, using wild cards etc. If the first three parameters do not define a unique transistor then no action is taken. e.g. to reconnect the drain of a given transistor from node (n1) to (n2):l4269d3634e12jk40\b13B51b13B (g0)(n0)(n1) (g0)(n0)(n2) redefinetransl4904e12jk40\f8i39I logchanges Usage: logchanges Arrange to log on all JaM commands that modify the circuit, for later replaying of such modifications. The stream must be set up using the JaM command .bytestream, e.g.l4269d3634e12jk40\b10B17b10B (changesstream)(changes.log) 4 .bytestream .def changesstream logchangesl4904e12jk40\f8i72I This opens the file changes.log and sets it up for logging changes. The parameter, 4, to .bytestream creates the stream for appending. This is useful for adding to an existing log of changes. Use of 6 instead would rewrite or create the file afresh.l4269e12jk40\20i11I unlogchanges Usage: unlogchanges Switch off logging of changes.l4269d3634e12jk40\b12B8b12B 3.3 Setting and Reading Valuese12jk72 circuitreset Usage: circuitreset Charge every node that is connected to at least one gate to , for = (0) | (1) | (X). (X) indicates "undefined". e.g.l4269d3634e12jk40\b12B16b12B (0)circuitresetl4904e12jk40\f8i15I would charge all gate nodes to low.l4269e12jk40 hi Usage: hi Tie the indicated node to VDD. e.g.l4269d3634e12jk40\b2B19b2B (input1)hil4904e12jk40\f8i10I lo Usage: lo Tie the indicated node to GND. e.g.l4269d3634e12jk40\b2B19b2B (input1)lol4904e12jk40\f8i10I x Usage: x Remove VDD or GND from the indicated node. The value of the node will not change even if it is not a storage node. This is useful for initialization. e.g.l4269d3634e12jk40\b1B19b1B (trinode)hi (trinode)xl4904e12jk40\f8i22I chhi Usage: chhi Charge the indicated node to high. The node must be connected to the gate of at least one transistor for this to have any effect. e.g.l4269d3634e12jk40\b4B19b4B (bit5)chhil4904e12jk40\f8i10I chlo Usage: chlo Charge the indicated node to low. The node must be connected to the gate of at least one transistor for this to have any effect. e.g.l4269d3634e12jk40\b4B19b4B (bit5)chlol4904e12jk40\f8i10I gatestorage Usage: gatestorage Set mode for charge storage. If switch=.true then charge can be stored only on the gates of transistors. If switch=.false then charge can be stored on any node. In either case, charge cannot be stored on a node that is either an input or is pulled up.l4269d3634e12jk40\b12B16b11B getnodevalue Usage: getnodevalue => Get the current value of the indicated node as a string. The is left on the JaM stack as a string = (1) | (0) | (X) . e.g. (input1)getnodevalue might leave the string (1) on the stack. This string could be printed with the JaM command "=", e.g.l4269d3634e12jk40\b13B18b12B153i20I26i3I (input1)getnodevalue =l4904e12jk40\f8i22I might printl4269e12jk40 0l4904e12jk40\f8i1I getbit Usage: getbit => Get the current value of the indicated node as an integer. The is left on the JaM stack as an integer = (-1 | 0 | 0) for (hi | lo | x), respectively (notice that x results in 0). e.g. (input1)getbit might leave the integer -1 on the stack. This integer could be printed with the JaM command "=", i.e.l4269d3634e12jk40\b7B18b6B214i14I27i2I (input1)getbit =l4904e12jk40\f8i16I would printl4269e12jk40 -1l4904e12jk40\f8i2I This value is useful for subsequent JaM logical operations, such as .bitand, .bitor, .bitnot, .bitxor, for functional simulation.l4269e12jk40 putbit Usage: putbit Set the indicated node to the indicated value. The node is set to lo | hi for = zero | non-zero, respectively. e.g.l4269d3634e12jk40\b6B26b6B (input1) 1 putbitl4904e12jk40\f8i17I would set input1 to hi.l4269e12jk40 getword Usage: getword => Construct a word from the current values of the nodes defined by , and leave that word on the JaM stack. The is a JaM array of node names, typically set up using the JaM utilities "[", "]", and ".def", e.g.l4269d3634e12jk40\b8B23b7B (aluoutput) [ (out4)(out3)(out2)(out1)(out0) ] .defl4904e12jk40\f8i51I Spaces around the "[" and "]" are necessary. Thenl4269e12jk40 aluoutput getwordl4904e12jk40\f8i17I will leave an integer on the JaM stack made up of the values of the indicated nodes, right justified in the word, i.e. in the above example, out0 will be represented as the lsb of the result.l4269e12jk40 putword Usage: putword Set the nodes defined in to the corresponding bits in the integer . The is as defined in getword. e.g.l4269d3634e12jk40\b7B31b7B (bus) [ (in2)(in1)(in0) ] .defl4904e12jk40\f8i30I thenl4269e12jk40 bus 6 putwordl4904e12jk40\f8i13I would set: in2 to hi, in1 to hi, in0 to lo.l4269e12jk40 arraynames Usage: arraynames List the names of the nodes given in . This command is for use with arraysim and for looking at wordspecifiers (see getword, putword). e.g.l4269d3634e12jk40\b10B25b10B86i8I20i14I6i7I2i7I outarray arraynamesl4904e12jk40\f8i19I might generatel4269e12jk40 (out1)(out2)l4904e12jk40\f8i12I arraystrings Usage: arraystrings List the current values of all the node strings in . This command is for use with the command arraysim. More than one array can be listed by giving two commands consecutively, e.g.l4269d3634e12jk40\b12B25b12B112i8I inarray arraystrings outarray arraystringsl4904e12jk40\f8i43I might generatel4269e12jk40 01010101 = in0 00110011 = in1 00001111 = in2 01101001 = out1 10110101 = out2l4904e12jk40\f8i77I If each string will not fit on one line then the whole set is continued below. This is to make it easier to establish correspondence between values.l4269e12jk40 arrayvalues Usage: arrayvalues List the current values of the nodes given in . This command is for use with arraysim and for looking at wordspecifiers (see getword, putword). e.g.l4269d3634e12jk40\b11B25b11B95i8I20i14I6i7I2i7I outarray arrayvaluesl4904e12jk40\f8i20I might generatel4269e12jk40 out1 = 1 out2 = 0l4904e12jk40\f8i17I 3.4 Running a Simulatione12jk72 simstep Usage: simstep Run the simulator to solve the circuit for the current set of input values. See also maxmicrosteps. In case of long or unterminating computations, simstep can be interrupted using the JaM break key (rightshift-swat). Fine point: currently it is inadvisable to hold the break key down for an extended period, or you will get a stack overflow. Better to strike the key firmly, then have patience, since the break will not actually happen until the next microstep has completed.l4269d3634e12jk40\b7B8b7B87i13I121f1 259f0 simsolve Usage: simsolve => Same as simstep except that the number of iterations needed to achieve convergence is left on the JaM stack. Zero iterations means that the state of the circuit did not change. This is useful for including functional simulation. See also maxmicrosteps.l4269d3634e12jk40\b9B7b8B268i13I maxmicrosteps Usage: maxmicrosteps Set maximum number of iterations that will be attempted in simstep and simsolve. The default is 100.l4269d3634e12jk40\b13B17b13B microstep Usage: microstep => <#changes> Run one iteration of the simulator, and leave the number of nodes that changed on the stack.l4269d3634e12jk40\b10B7b9B reportshorts Usage: reportshorts Control reporting of shorts in microstep. Default is .true.l4269d3634e12jk40\b13B17b12B reportchanges Usage: reportchanges Control reporting of changes in microstep. Default is .false. If .true then nodes that change during a microstep are listed. This is useful for diagnosing circuits that do not stabilize.l4269d3634e12jk40\b14B17b13B arraysim Usage: arraysim For each set of values in the input vector, set up the inputs, solve the circuit, and collect the outputs named in the output vector. The input and output arrays are JaM arrays of node names, typically set up using the JaM utilities "[", "]", and ".def", e.g.l4269d3634e12jk40\b8B34b8B (inarray) [ (in0)(in1)(in2) ] .def (outarray) [ (out1)(out2) ] .defl4904e12jk40\f8i67I Each of the names specified in the input array must be defined as JaM strings representing the sequence of values required for each input node, e.g.l4269e12jk40 (in0)(01010101).def (in1)(00110011).def (in2)(00001111).defl4904e12jk40\f8i59I The node names specified in the output array will be associated with the collected output strings and need not be initialized. Given the above definitions, the command:l4269e12jk40 inarray outarray arraysiml4904e12jk40\f8i25I will solve the circuit 8 times, having set up the inputs in2, in1, in0 to 000, 001, 010, 011, 100, 101, 110, 111, and will collect the outputs at each step as strings in the JaM variables out1 and out2. These strings can be examined using the JaM command =, e.g.l4269e12jk40 out1 =l4904e12jk40\f8i6I might generatel4269e12jk40 01101001l4904e12jk40\f8i8I It is often convenient to have all the inputs and outputs listed together. See the commands arraystrings, arrayvalues, and arraynames under 3.3.l4269e12jk40\93i12I2i11I6i10I While various clocking schemes etc. can be implemented using arraysim as described, it is sometimes more convenient to be able to do several things between setting up each set of inputs and collecting the corresponding outputs. For this reason, arraysim actually works as follows. For each set of values in the inputarray arraysim does the following 4 steps:l4269e12jk40 1) set up the input values, 2) invoke the JaM procedure clock1, 3) collect the outputs, and, 4) invoke the JaM procedure clock2.l4269e12jk40\56i6I59i6I At initialization the procedure clock1 is set up to simply call simstep, and the procedure clock2 is a nop. These procedures can be redefined to implement more elaborate schemes. For example, to implement a 2 phase, non-overlapping clock, PHI1 and PHI2, to be cycled after each input is set up, clock1 could be redefined as follows:l4269e12jk40 (clock1) ((PHI1)hi simstep (PHI1)lo simstep (PHI2)hi simstep (PHI2)lo simstep ).cvx .defl4904e12jk40\f8i91I Incidentally, for this example it is normally adequate, and more efficient, to define clock1 to be:l4269e12jk40 (clock1) ((PHI1)hi (PHI2)lo simstep (PHI1)lo (PHI2)hi simstep ).cvx .defl4904e12jk40\f8i73I arraysim can be interrupted during operation by pressing the JaM break key (rightshift-swat). Fine point: currently it is inadvisable to hold the break key down for an extended period, or you will get a stack overflow. Better to strike the key firmly, then have patience, since the break will not actually happen until the next simulation cycle has completed. Having interrupted arraysim, almost all of the commands can be safely used for e.g. examining node values, setting node values, modifying the circuit, changing the input array strings (but not the list of names in the input and output arrays), and even for leaving SimMOS completely. The simulation can be restarted using resumearraysim (see below).l4269e12jk40\95f1 266f0 325i14I resumearraysim Usage: resumearraysim Resume the running of an arraysim after it has been interrupted.l4269d3634e12jk40\b14B8b14B 3.5 Saving and Restoring Statee12jk72 simwrite Usage: simwrite Write a .sim file describing the circuit currently stored in the SimMOS data structure. e.g.l4269d3634e12jk40\b8B19b8B (newcircuit.sim)simwritel4904e12jk40\f8i24I A circuit thus saved can be restored later using simread.l4269e12jk40\49i7I writestate Usage: writestate Write a JaM file describing the current state of the circuit stored in the SimMOS data structure. This file takes the form of a list of invocations of hi,lo,x,chhi,chlo. e.g.l4269d3634e12jk40\b10B19b10B (circuitstate.jam)writestatel4904e12jk40\f8i28I Subsequent use of the .run command on this file will cause the state of all the nodes to be restored. After such a restoration, one call to simstep should be made, to drive non-storage nodes to their correct values.l4269e12jk40 3.6 Othere12jk72 This section contains miscellaneous commands from JaM and SimMOS that might be found useful.l3634e12jk40 inverter Usage: inverter Create a pair of transistors implementing an inverter for the given pair of nodes.l4269d3634e12jk40\b8B32b8B nand Usage: nand Create 3 transistors to implement a nand circuit.l4269d3634e12jk40\b4B24b4B nor Usage: nor Create 3 transistors to implement a nor circuit.l4269d3634e12jk40\b3B24b3B pullup Usage: pullup Create a depletion mode pullup transistor to pull up the indicated node. e.g.l4269d3634e12jk40\b6B19b6B (bus0) pullupl4904e12jk40\f8i13I gensym Usage: gensym => Leaves a unique node name as a string on the JaM stack. This is useful for generating names of interior nodes in procedurally defined circuits.l4269d3634e12jk40\b7B7b6B gennames Usage: gennames Declares JaM variables having the given names, and assigns to them their own names concatenated with the result of a single common call to gensym. This command is useful in procedural definitions of circuits in which local names that can later be identified are needed. e.g.l4269d3634e12jk40\b9B25b8B (in out int)gennamesl4904e12jk40\f8i20I generates the 3 JaM variables, in, out, and int, with the string values (in#), (out#), and (int#), respectively, where # is the result of a single call to gensym.l4256e12jk40 /getargs Usage: ...( ...) /getargs This JaM command expects a string of names separated by spaces. It creates JaM variables having the given names, and assigns to them the values preceding the string on the stack in a one to one correspondence. e.g.l4269d3634e12jk40\b8B51b8B 3 1.5 (hello) (i r s) /getargsl4904e12jk40\f8i31I would create the JaM variables i,r, and s, and assign to them the values 3, 1.5, and the string hello, respectively. This command is useful as the first command in a JaM procedure, to get the arguments from the stack.l4256e12jk40 .print Usage: .print This JaM command prints the given . e.g.l4269d3634e12jk40\b6B17b6B (hello).printl4904e12jk40\f8i13I /print Usage: /print This JaM command prints the given followed by a carriage return.l4269d3634e12jk40\b6B17b6B .def Usage: .def This is the JaM assignment command, and assigns to . .def is used to set up both ordinary variables and procedures. e.g.l4269d3634e12jk40\b4B22b4B (numbervariablename) 23 .def (stringvariablename)(this is a string, not a procedure) .def (procedurename)((hello).print).cvx .defl4904e12jk40\f8i129I .cvx Usage: