User Tools

Site Tools


geda:icarus_vpi_within_vvp

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
geda:icarus_vpi_within_vvp [2007/05/24 22:27]
127.0.0.1 external edit
geda:icarus_vpi_within_vvp [2012/02/20 15:14] (current)
Line 1: Line 1:
 +====== VPI_within_VVP ======
 +<​code>/​*
 + * Copyright (c) 2001 Stephen Williams (steve@icarus.com)
 + *
 + ​* ​ $Id: vpi.txt,v 1.7 2003/02/09 23:33:26 steve Exp $
 + */
  
 +
 +VPI WITHIN VVP
 +
 +System tasks and functions in Verilog are implemented in Icarus
 +Verilog by C routines written with VPI. This implies that the vvp
 +engine must provide at least a subset of the Verilog VPI
 +interface. The minimalist concepts of vvp, however, make the method
 +less then obvious.
 +
 +Within a Verilog design, there is a more or less fixed web of
 +vpiHandles that is the design database as is available to VPI
 +functions. The Verilog standard defines quite a lot of types, but the
 +vvp only implements the ones it needs. The VPI web is added into the
 +design using special pseudo-ops that create the needed objects.
 +
 +
 +LOADING VPI MODULES
 +
 +The vvp runtime loads VPI modules at runtime before the parser reads
 +in the source files. This gives the modules a chance to register tasks
 +and functions before the source is compiled. This allows the compiler
 +to resolve references to system tasks and system functions to a
 +vpiHandle at compile time. References to missing tasks/​function can
 +thus be caught before the simulation is run.
 +
 +     NOTE: This also, miraculously,​ allows for some minimal support of
 +     the compiletf call. From the perspective of VPI code, compilation
 +     of the VVP source is not unlike compilation of the original
 +     ​Verilog.
 +
 +The handle that the vvp threads have to the VPI are the vpiHandles of
 +the system tasks and functions. The %vpi_call instruction,​ once compiled,
 +carries the vpiHandle of the system task.
 +
 +
 +SYSTEM TASK CALLS
 +
 +A system task call invokes a VPI routine, and makes available to that
 +routine the arguments to the system task. The called routine gets
 +access to the system task call by calling back the VPI requesting the
 +handle. It uses the handle, in turn, to get hold of the operands for
 +the task.
 +
 +All that vvp needs to know about a system task call is the handle of
 +the system task definitions (created by the vpi_register_systf
 +function) and the arguments of the actual call. The arguments are
 +tricky because the list has no bound, even though each particular call
 +in the Verilog source has a specific set of parameters.
 +
 +Since each call takes a fixed number of parameters, the input source
 +can include in the statement the list of arguments. The argument list
 +will not fit in a single generated instruction,​ but a vpiHandle that
 +refers to a vpiSysTfCall does. Therefore, the compiler can take the
 +long argument list and form a vpiSysTaskCall object. The generated
 +instruction then only needs to be a %vpi_call with the single parameter
 +that is the vpiHandle for the call.
 +
 +
 +SYSTEM FUNCTION CALLS
 +
 +System function calls are similar to system tasks. The only
 +differences are that all the arguments are input only, and there is a
 +single magic output that is the return value. The same %vpi_call can
 +even be used to call a function.
 +
 +System functions, like system tasks, can only be called from thread
 +code. However, they can appear in expressions,​ even when that
 +expression is entirely structural. The desired effect is achieved by
 +writing a wrapper thread that calls the function when inputs change,
 +and that writes the output into the containing expression.
 +
 +
 +SYSTEM TASK/​FUNCTION ARGUMENTS
 +
 +The arguments to each system task or call are not stored in the
 +instruction op-code, but in the vpiSysTfCall object that the compiler
 +creates and that the %vpi_call instruction ultimately refers to. All
 +the arguments must be some sort of object that can be represented by a
 +vpiHandle at compile time.
 +
 +Arguments are handled at compile time by the parser, which uses the
 +argument_list rule to build a list of vpiHandle objects. Each argument
 +in the argument_list invokes whatever function is appropriate for the
 +token in order to make a vpiHandle object for the argument_list. When
 +all this is done, an array of vpiHandles is passed to code to create a
 +vpiSysTfCall object that has all that is needed to make the call.
 +
 +
 +SCOPES
 +
 +VPI can access scopes as objects of type vpiScope. Scopes have names
 +and can also contain other sub-scopes, all of which the VPI function
 +can access by the vpiInternalScope reference. Therefore, the run-time
 +needs to form a tree of scopes into which other scoped VPI objects are
 +placed.
 +
 +A scope is created with a .scope directive, like so:
 +
 + <​label>​ .scope "​name"​ [, <​parent>​];​
 + .timescale <​units>;​
 +
 +The scope takes a string name as the first parameter. If there is an
 +additional parameter, it is a label of the directive for the parent
 +scope. Scopes that have no parent are root scopes. It is an error to
 +declare a scope with the same name more then once in a parent scope.
 +
 +The name string given when creating the scope is the basename for the
 +scope. The vvp automatically constructs full names from the scope
 +hierarchy, and runtime VPI code can access that full name with the
 +vpiFullname reference.
 +
 +The .timescale directive changes the scope units from the simulation
 +precision to the specified precision. The .timescale directive affects
 +the current scope.
 +
 +Objects that place themselves in a scope place themselves in the
 +current scope. The current scope is the one that was last mentioned by
 +a .scope directive. If the wrong scope is current, the label on a
 +scope directive can be used to resume a scope. The syntax works like
 +this:
 +
 + .scope <​symbol>;​
 +
 +In this case, the <​symbol>​ is the label of a previous scope directive,
 +and is used to identify the scope to be resumed. A scope resume
 +directive cannot have a label.
 +
 +
 +VARIABLES
 +
 +Reg vectors (scalars are vectors of length 1) are created by .var
 +statements in the source. The .var statement includes the declared
 +name of the variable, and the indices of the MSB and LSB of the
 +vector. The vpiHandle is then created with this information,​ and the
 +vpi_ipoint_t pointer to the LSB functor of the variable. VPI programs
 +access the vector through the vpiHandle and related functions. The VPI
 +code gets access to the declared indices.
 +
 +The VPI interface to variable (vpiReg objects) uses the MSB and LSB
 +values that the user defined to describe the dimensions of the
 +object.
 +
 +/*
 + * Copyright (c) 2001 Stephen Williams (steve@icarus.com)
 + *
 + ​* ​   This source code is free software; you can redistribute it
 + ​* ​   and/or modify it in source code form under the terms of the GNU
 + ​* ​   General Public License as published by the Free Software
 + ​* ​   Foundation; either version 2 of the License, or (at your option)
 + ​* ​   any later version.
 + *
 + ​* ​   This program is distributed in the hope that it will be useful,
 + ​* ​   but WITHOUT ANY WARRANTY; without even the implied warranty of
 + ​* ​   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. ​ See the
 + ​* ​   GNU General Public License for more details.
 + *
 + ​* ​   You should have received a copy of the GNU General Public License
 + ​* ​   along with this program; if not, write to the Free Software
 + ​* ​   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 + ​*/</​code>​
geda/icarus_vpi_within_vvp.txt ยท Last modified: 2012/02/20 15:14 (external edit)