diff options
author | Steve Reinhardt <stever@eecs.umich.edu> | 2004-04-06 10:02:00 -0700 |
---|---|---|
committer | Steve Reinhardt <stever@eecs.umich.edu> | 2004-04-06 10:02:00 -0700 |
commit | d4069233eb8cc875aa443cb7c6a16b748e3acc42 (patch) | |
tree | 1c492289fff76c43a00aea660889eecd040b95bd /util/config/m5config.py | |
parent | 82837b1c3690084ce3cd6ae2f6a1700caf23be00 (diff) | |
download | gem5-d4069233eb8cc875aa443cb7c6a16b748e3acc42.tar.xz |
Beta version of Python configuration tool. Generates .ini files from
Python script description.
arch/alpha/alpha_memory.cc:
dev/io_device.cc:
Add DEFINE_SIM_OBJECT_CLASS_NAME for intermediate SimObjects.
test/paramtest.cc:
Fix stupid spelling.
--HG--
extra : convert_revision : dc020208cb6507c1afb1ed771a7218daba678e09
Diffstat (limited to 'util/config/m5config.py')
-rw-r--r-- | util/config/m5config.py | 743 |
1 files changed, 743 insertions, 0 deletions
diff --git a/util/config/m5config.py b/util/config/m5config.py new file mode 100644 index 000000000..ea5d68504 --- /dev/null +++ b/util/config/m5config.py @@ -0,0 +1,743 @@ +from __future__ import generators + +import os +import re +import sys + +##################################################################### +# +# M5 Python Configuration Utility +# +# The basic idea is to write simple Python programs that build Python +# objects corresponding to M5 SimObjects for the deisred simulation +# configuration. For now, the Python emits a .ini file that can be +# parsed by M5. In the future, some tighter integration between M5 +# and the Python interpreter may allow bypassing the .ini file. +# +# Each SimObject class in M5 is represented by a Python class with the +# same name. The Python inheritance tree mirrors the M5 C++ tree +# (e.g., SimpleCPU derives from BaseCPU in both cases, and all +# SimObjects inherit from a single SimObject base class). To specify +# an instance of an M5 SimObject in a configuration, the user simply +# instantiates the corresponding Python object. The parameters for +# that SimObject are given by assigning to attributes of the Python +# object, either using keyword assignment in the constructor or in +# separate assignment statements. For example: +# +# cache = BaseCache('my_cache', root, size=64*K) +# cache.hit_latency = 3 +# cache.assoc = 8 +# +# (The first two constructor arguments specify the name of the created +# cache and its parent node in the hierarchy.) +# +# The magic lies in the mapping of the Python attributes for SimObject +# classes to the actual SimObject parameter specifications. This +# allows parameter validity checking in the Python code. Continuing +# the example above, the statements "cache.blurfl=3" or +# "cache.assoc='hello'" would both result in runtime errors in Python, +# since the BaseCache object has no 'blurfl' parameter and the 'assoc' +# parameter requires an integer, respectively. This magic is done +# primarily by overriding the special __setattr__ method that controls +# assignment to object attributes. +# +# The Python module provides another class, ConfigNode, which is a +# superclass of SimObject. ConfigNode implements the parent/child +# relationship for building the configuration hierarchy tree. +# Concrete instances of ConfigNode can be used to group objects in the +# hierarchy, but do not correspond to SimObjects themselves (like a +# .ini section with "children=" but no "type=". +# +# Once a set of Python objects have been instantiated in a hierarchy, +# calling 'instantiate(obj)' (where obj is the root of the hierarchy) +# will generate a .ini file. See simple-4cpu.py for an example +# (corresponding to m5-test/simple-4cpu.ini). +# +##################################################################### + +##################################################################### +# +# ConfigNode/SimObject classes +# +# The Python class hierarchy rooted by ConfigNode (which is the base +# class of SimObject, which in turn is the base class of all other M5 +# SimObject classes) has special attribute behavior. In general, an +# object in this hierarchy has three categories of attribute-like +# things: +# +# 1. Regular Python methods and variables. These must start with an +# underscore to be treated normally. +# +# 2. SimObject parameters. These values are stored as normal Python +# attributes, but all assignments to these attributes are checked +# against the pre-defined set of parameters stored in the class's +# _param_dict dictionary. Assignments to attributes that do not +# correspond to predefined parameters, or that are not of the correct +# type, incur runtime errors. +# +# 3. Hierarchy children. The child nodes of a ConfigNode are stored +# in the node's _children dictionary, but can be accessed using the +# Python attribute dot-notation (just as they are printed out by the +# simulator). Children cannot be created using attribute assigment; +# they must be added by specifying the parent node in the child's +# constructor or using the '+=' operator. + +# The SimObject parameters are the most complex, for a few reasons. +# First, both parameter descriptions and parameter values are +# inherited. Thus parameter description lookup must go up the +# inheritance chain like normal attribute lookup, but this behavior +# must be explicitly coded since the lookup occurs in each class's +# _param_dict attribute. Second, because parameter values can be set +# on SimObject classes (to implement default values), the parameter +# checking behavior must be enforced on class attribute assignments as +# well as instance attribute assignments. Finally, because we allow +# class specialization via inheritance (e.g., see the L1Cache class in +# the simple-4cpu.py example), we must do parameter checking even on +# class instantiation. To provide all these features, we use a +# metaclass to define most of the SimObject parameter behavior for +# this class hierarchy. +# +##################################################################### + +# The metaclass for ConfigNode (and thus for everything that dervies +# from ConfigNode, including SimObject). This class controls how new +# classes that derive from ConfigNode are instantiated, and provides +# inherited class behavior (just like a class controls how instances +# of that class are instantiated, and provides inherited instance +# behavior). +class MetaConfigNode(type): + + # __new__ is called before __init__, and is where the statements + # in the body of the class definition get loaded into the class's + # __dict__. We intercept this to filter out parameter assignments + # and only allow "private" attributes to be passed to the base + # __new__ (starting with underscore). + def __new__(cls, name, bases, dict): + priv_keys = [k for k in dict.iterkeys() if k.startswith('_')] + priv_dict = {} + for k in priv_keys: priv_dict[k] = dict[k]; del dict[k] + # entries left in dict will get passed to __init__, where we'll + # deal with them as params. + return super(MetaConfigNode, cls).__new__(cls, name, bases, priv_dict) + + # initialization: start out with an empty param dict (makes life + # simpler if we can assume _param_dict is always valid). Also + # build inheritance list to simplify searching for inherited + # params. Finally set parameters specified in class definition + # (if any). + def __init__(cls, name, bases, dict): + super(MetaConfigNode, cls).__init__(cls, name, bases, {}) + # initialize _param_dict to empty + cls._param_dict = {} + # __mro__ is the ordered list of classes Python uses for + # method resolution. We want to pick out the ones that have a + # _param_dict attribute for doing parameter lookups. + cls._param_bases = \ + [c for c in cls.__mro__ if hasattr(c, '_param_dict')] + # initialize attributes with values from class definition + for (pname, value) in dict.items(): + try: + setattr(cls, pname, value) + except Exception, exc: + print "Error setting '%s' to '%s' on class '%s'\n" \ + % (pname, value, cls.__name__), exc + + # set the class's parameter dictionary (called when loading + # class descriptions) + def set_param_dict(cls, param_dict): + # should only be called once (current one should be empty one + # from __init__) + assert not cls._param_dict + cls._param_dict = param_dict + # initialize attributes with default values + for (pname, param) in param_dict.items(): + try: + setattr(cls, pname, param.default) + except Exception, exc: + print "Error setting '%s' default on class '%s'\n" \ + % (pname, cls.__name__), exc + + # Lookup a parameter description by name in the given class. Use + # the _param_bases list defined in __init__ to go up the + # inheritance hierarchy if necessary. + def lookup_param(cls, param_name): + for c in cls._param_bases: + param = c._param_dict.get(param_name) + if param: return param + return None + + # Set attribute (called on foo.attr_name = value when foo is an + # instance of class cls). + def __setattr__(cls, attr_name, value): + # normal processing for private attributes + if attr_name.startswith('_'): + object.__setattr__(cls, attr_name, value) + return + # no '_': must be SimObject param + param = cls.lookup_param(attr_name) + if not param: + raise AttributeError, \ + "Class %s has no parameter %s" % (cls.__name__, attr_name) + # It's ok: set attribute by delegating to 'object' class. + # Note the use of param.make_value() to verify/canonicalize + # the assigned value + object.__setattr__(cls, attr_name, param.make_value(value)) + + # generator that iterates across all parameters for this class and + # all classes it inherits from + def all_param_names(cls): + for c in cls._param_bases: + for p in c._param_dict.iterkeys(): + yield p + +# The ConfigNode class is the root of the special hierarchy. Most of +# the code in this class deals with the configuration hierarchy itself +# (parent/child node relationships). +class ConfigNode(object): + # Specify metaclass. Any class inheriting from ConfigNode will + # get this metaclass. + __metaclass__ = MetaConfigNode + + # Constructor. Since bare ConfigNodes don't have parameters, just + # worry about the name and the parent/child stuff. + def __init__(self, _name, _parent=None): + # Type-check _name + if type(_name) != str: + if isinstance(_name, ConfigNode): + # special case message for common error of trying to + # coerce a SimObject to the wrong type + raise TypeError, \ + "Attempt to coerce %s to %s" \ + % (_name.__class__.__name__, self.__class__.__name__) + else: + raise TypeError, \ + "%s name must be string (was %s, %s)" \ + % (self.__class__.__name__, _name, type(_name)) + # if specified, parent must be a subclass of ConfigNode + if _parent != None and not isinstance(_parent, ConfigNode): + raise TypeError, \ + "%s parent must be ConfigNode subclass (was %s, %s)" \ + % (self.__class__.__name__, _name, type(_name)) + self._name = _name + self._parent = _parent + self._children = {} + if (_parent): + _parent.__addChild(self) + # Set up absolute path from root. + if (_parent and _parent._path != 'Universe'): + self._path = _parent._path + '.' + self._name + else: + self._path = self._name + + # When printing (e.g. to .ini file), just give the name. + def __str__(self): + return self._name + + # Catch attribute accesses that could be requesting children, and + # satisfy them. Note that __getattr__ is called only if the + # regular attribute lookup fails, so private and parameter lookups + # will already be satisfied before we ever get here. + def __getattr__(self, name): + try: + return self._children[name] + except KeyError: + raise AttributeError, \ + "Node '%s' has no attribute or child '%s'" \ + % (self._name, name) + + # Set attribute. All attribute assignments go through here. Must + # be private attribute (starts with '_') or valid parameter entry. + # Basically identical to MetaConfigClass.__setattr__(), except + # this handles instances rather than class attributes. + def __setattr__(self, attr_name, value): + if attr_name.startswith('_'): + object.__setattr__(self, attr_name, value) + return + # not private; look up as param + param = self.__class__.lookup_param(attr_name) + if not param: + raise AttributeError, \ + "Class %s has no parameter %s" \ + % (self.__class__.__name__, attr_name) + # It's ok: set attribute by delegating to 'object' class. + # Note the use of param.make_value() to verify/canonicalize + # the assigned value + object.__setattr__(self, attr_name, param.make_value(value)) + + # Add a child to this node. + def __addChild(self, new_child): + # set child's parent before calling this function + assert new_child._parent == self + if not isinstance(new_child, ConfigNode): + raise TypeError, \ + "ConfigNode child must also be of class ConfigNode" + if new_child._name in self._children: + raise AttributeError, \ + "Node '%s' already has a child '%s'" \ + % (self._name, new_child._name) + self._children[new_child._name] = new_child + + # operator overload for '+='. You can say "node += child" to add + # # a child that was created with parent=None. An early attempt + # at # playing with syntax; turns out not to be that useful. + def __iadd__(self, new_child): + if new_child._parent != None: + raise AttributeError, \ + "Node '%s' already has a parent" % new_child._name + new_child._parent = self + self.__addChild(new_child) + return self + + # Print instance info to .ini file. + def _instantiate(self): + print '[' + self._path + ']' # .ini section header + if self._children: + # instantiate children in sorted order for backward + # compatibility (else we can end up with cpu1 before cpu0). + child_names = self._children.keys() + child_names.sort() + print 'children =', + for child_name in child_names: + print child_name, + print + self._instantiateParams() + print + # recursively dump out children + if self._children: + for child_name in child_names: + self._children[child_name]._instantiate() + + # ConfigNodes have no parameters. Overridden by SimObject. + def _instantiateParams(self): + pass + +# SimObject is a minimal extension of ConfigNode, implementing a +# hierarchy node that corresponds to an M5 SimObject. It prints out a +# "type=" line to indicate its SimObject class, prints out the +# assigned parameters corresponding to its class, and allows +# parameters to be set by keyword in the constructor. Note that most +# of the heavy lifting for the SimObject param handling is done in the +# MetaConfigNode metaclass. + +class SimObject(ConfigNode): + # initialization: like ConfigNode, but handle keyword-based + # parameter initializers. + def __init__(self, _name, _parent=None, **params): + ConfigNode.__init__(self, _name, _parent) + for param, value in params.items(): + setattr(self, param, value) + + # print type and parameter values to .ini file + def _instantiateParams(self): + print "type =", self.__class__._name + for pname in self.__class__.all_param_names(): + value = getattr(self, pname) + if value != None: + print pname, '=', value + +##################################################################### +# +# Parameter description classes +# +# The _param_dict dictionary in each class maps parameter names to +# either a Param or a VectorParam object. These objects contain the +# parameter description string, the parameter type, and the default +# value (loaded from the PARAM section of the .odesc files). The +# make_value() method on these objects is used to force whatever value +# is assigned to the parameter to the appropriate type. +# +# Note that the default values are loaded into the class's attribute +# space when the parameter dictionary is initialized (in +# MetaConfigNode.set_param_dict()); after that point they aren't +# used. +# +##################################################################### + +# Force parameter value (rhs of '=') to ptype (or None, which means +# not set). +def make_param_value(ptype, value): + # nothing to do if None or already correct type + if value == None or isinstance(value, ptype): + return value + # this type conversion will raise an exception if it's illegal + return ptype(value) + +# Regular parameter. +class Param(object): + # Constructor. E.g., Param(Int, "number of widgets", 5) + def __init__(self, ptype, desc, default=None): + self.ptype = ptype + self.desc = desc + self.default = default + + # Convert assigned value to appropriate type. + def make_value(self, value): + return make_param_value(self.ptype, value) + +# The _VectorParamValue class is a wrapper for vector-valued +# parameters. The leading underscore indicates that users shouldn't +# see this class; it's magically generated by VectorParam. The +# parameter values are stored in the 'value' field as a Python list of +# whatever type the parameter is supposed to be. The only purpose of +# storing these instead of a raw Python list is that we can override +# the __str__() method to not print out '[' and ']' in the .ini file. +class _VectorParamValue(object): + def __init__(self, list): + self.value = list + + def __str__(self): + return ' '.join(map(str, self.value)) + +# Vector-valued parameter description. Just like Param, except that +# the value is a vector (list) of the specified type instead of a +# single value. +class VectorParam(object): + # Constructor. The resulting parameter will be a list of ptype. + def __init__(self, ptype, desc, default=None): + self.ptype = ptype + self.desc = desc + self.default = default + + # Convert assigned value to appropriate type. If the RHS is not a + # list or tuple, it generates a single-element list. + def make_value(self, value): + if value == None: return value + if isinstance(value, list) or isinstance(value, tuple): + # list: coerce each element into new list + val_list = [make_param_value(self.ptype, v) for v in + iter(value)] + else: + # singleton: coerce & wrap in a list + val_list = [ make_param_value(self.ptype, value) ] + # wrap list in _VectorParamValue (see above) + return _VectorParamValue(val_list) + +##################################################################### +# +# Parameter Types +# +# Though native Python types could be used to specify parameter types +# (the 'ptype' field of the Param and VectorParam classes), it's more +# flexible to define our own set of types. This gives us more control +# over how Python expressions are converted to values (via the +# __init__() constructor) and how these values are printed out (via +# the __str__() conversion method). Eventually we'll need these types +# to correspond to distinct C++ types as well. +# +##################################################################### + +# Integer parameter type. +class Int(object): + # Constructor. Value must be Python int or long (long integer). + def __init__(self, value): + t = type(value) + if t == int or t == long: + self.value = value + else: + raise TypeError, "Int param got value %s %s" % (repr(value), t) + + # Use Python string conversion. Note that this puts an 'L' on the + # end of long integers; we can strip that off here if it gives us + # trouble. + def __str__(self): + return str(self.value) + +# Counter, Addr, and Tick are just aliases for Int for now. +class Counter(Int): + pass + +class Addr(Int): + pass + +class Tick(Int): + pass + +# Boolean parameter type. +class Bool(object): + + # Constructor. Typically the value will be one of the Python bool + # constants True or False (or the aliases true and false below). + # Also need to take integer 0 or 1 values since bool was not a + # distinct type in Python 2.2. Parse a bunch of boolean-sounding + # strings too just for kicks. + def __init__(self, value): + t = type(value) + if t == bool: + self.value = value + elif t == int or t == long: + if value == 1: + self.value = True + elif value == 0: + self.value = False + elif t == str: + v = value.lower() + if v == "true" or v == "t" or v == "yes" or v == "y": + self.value = True + elif v == "false" or v == "f" or v == "no" or v == "n": + self.value = False + # if we didn't set it yet, it must not be something we understand + if not hasattr(self, 'value'): + raise TypeError, "Bool param got value %s %s" % (repr(value), t) + + # Generate printable string version. + def __str__(self): + if self.value: return "true" + else: return "false" + +# String-valued parameter. +class String(object): + # Constructor. Value must be Python string. + def __init__(self, value): + t = type(value) + if t == str: + self.value = value + else: + raise TypeError, "String param got value %s %s" % (repr(value), t) + + # Generate printable string version. Not too tricky. + def __str__(self): + return self.value + + +# Enumerated types are a little more complex. The user specifies the +# type as Enum(foo) where foo is either a list or dictionary of +# alternatives (typically strings, but not necessarily so). (In the +# long run, the integer value of the parameter will be the list index +# or the corresponding dictionary value. For now, since we only check +# that the alternative is valid and then spit it into a .ini file, +# there's not much point in using the dictionary.) + +# What Enum() must do is generate a new type encapsulating the +# provided list/dictionary so that specific values of the parameter +# can be instances of that type. We define two hidden internal +# classes (_ListEnum and _DictEnum) to serve as base classes, then +# derive the new type from the appropriate base class on the fly. + + +# Base class for list-based Enum types. +class _ListEnum(object): + # Constructor. Value must be a member of the type's map list. + def __init__(self, value): + if value in self.map: + self.value = value + self.index = self.map.index(value) + else: + raise TypeError, "Enum param got bad value '%s' (not in %s)" \ + % (value, self.map) + + # Generate printable string version of value. + def __str__(self): + return str(self.value) + +class _DictEnum(object): + # Constructor. Value must be a key in the type's map dictionary. + def __init__(self, value): + if value in self.map: + self.value = value + self.index = self.map[value] + else: + raise TypeError, "Enum param got bad value '%s' (not in %s)" \ + % (value, self.map.keys()) + + # Generate printable string version of value. + def __str__(self): + return str(self.value) + +# Enum metaclass... calling Enum(foo) generates a new type (class) +# that derives from _ListEnum or _DictEnum as appropriate. +class Enum(type): + # counter to generate unique names for generated classes + counter = 1 + + def __new__(cls, map): + if isinstance(map, dict): + base = _DictEnum + keys = map.keys() + elif isinstance(map, list): + base = _ListEnum + keys = map + else: + raise TypeError, "Enum map must be list or dict (got %s)" % map + classname = "Enum%04d" % Enum.counter + Enum.counter += 1 + # New class derives from selected base, and gets a 'map' + # attribute containing the specified list or dict. + return type.__new__(cls, classname, (base,), { 'map': map }) + + +# +# "Constants"... handy aliases for various values. +# + +# For compatibility with C++ bool constants. +false = False +true = True + +# Some memory range specifications use this as a default upper bound. +MAX_ADDR = 2 ** 63 + +# For power-of-two sizing, e.g. 64*K gives an integer value 65536. +K = 1024 +M = K*K +G = K*M + +##################################################################### +# +# Object description loading. +# +# The final step is to define the classes corresponding to M5 objects +# and their parameters. These classes are described in .odesc files +# in the source tree. This code walks the tree to find those files +# and loads up the descriptions (by evaluating them in pieces as +# Python code). +# +# +# Because SimObject classes inherit from other SimObject classes, and +# can use arbitrary other SimObject classes as parameter types, we +# have to do this in three steps: +# +# 1. Walk the tree to find all the .odesc files. Note that the base +# of the filename *must* match the class name. This step builds a +# mapping from class names to file paths. +# +# 2. Start generating empty class definitions (via def_class()) using +# the OBJECT field of the .odesc files to determine inheritance. +# def_class() recurses on demand to define needed base classes before +# derived classes. +# +# 3. Now that all of the classes are defined, go through the .odesc +# files one more time loading the parameter descriptions. +# +##################################################################### + +# dictionary: maps object names to file paths +odesc_file = {} + +# dictionary: maps object names to boolean flag indicating whether +# class definition was loaded yet. Since SimObject is defined in +# m5.config.py, count it as loaded. +odesc_loaded = { 'SimObject': True } + +# Find odesc files in namelist and initialize odesc_file and +# odesc_loaded dictionaries. Called via os.path.walk() (see below). +def find_odescs(process, dirpath, namelist): + # Prune out SCCS directories so we don't process s.*.odesc files. + i = 0 + while i < len(namelist): + if namelist[i] == "SCCS": + del namelist[i] + else: + i = i + 1 + # Find .odesc files and record them. + for name in namelist: + if name.endswith('.odesc'): + objname = name[:name.rindex('.odesc')] + path = os.path.join(dirpath, name) + if odesc_file.has_key(objname): + print "Warning: duplicate object names:", \ + odesc_file[objname], path + odesc_file[objname] = path + odesc_loaded[objname] = False + + +# Regular expression string for parsing .odesc files. +file_re_string = r''' +^OBJECT: \s* (\w+) \s* \( \s* (\w+) \s* \) +\s* +^PARAMS: \s*\n ( (\s+.*\n)* ) +''' + +# Compiled regular expression object. +file_re = re.compile(file_re_string, re.MULTILINE | re.VERBOSE) + +# .odesc file parsing function. Takes a filename and returns tuple of +# object name, object base, and parameter description section. +def parse_file(path): + f = open(path, 'r').read() + m = file_re.search(f) + if not m: + print "Can't parse", path + sys.exit(1) + return (m.group(1), m.group(2), m.group(3)) + +# Define SimObject class based on description in specified filename. +# Class itself is empty except for _name attribute; parameter +# descriptions will be loaded later. Will recurse to define base +# classes as needed before defining specified class. +def def_class(path): + # load & parse file + (obj, parent, params) = parse_file(path) + # check to see if base class is defined yet; define it if not + if not odesc_loaded.has_key(parent): + print "No .odesc file found for", parent + sys.exit(1) + if not odesc_loaded[parent]: + def_class(odesc_file[parent]) + # define the class. + s = "class %s(%s): _name = '%s'" % (obj, parent, obj) + try: + # execute in global namespace, so new class will be globally + # visible + exec s in globals() + except Exception, exc: + print "Object error in %s:" % path, exc + # mark this file as loaded + odesc_loaded[obj] = True + +# Munge an arbitrary Python code string to get it to execute (mostly +# dealing with indentation). Stolen from isa_parser.py... see +# comments there for a more detailed description. +def fixPythonIndentation(s): + # get rid of blank lines first + s = re.sub(r'(?m)^\s*\n', '', s); + if (s != '' and re.match(r'[ \t]', s[0])): + s = 'if 1:\n' + s + return s + +# Load parameter descriptions from .odesc file. Object class must +# already be defined. +def def_params(path): + # load & parse file + (obj_name, parent_name, param_code) = parse_file(path) + # initialize param dict + param_dict = {} + # execute parameter descriptions. + try: + # "in globals(), param_dict" makes exec use the current + # globals as the global namespace (so all of the Param + # etc. objects are visible) and param_dict as the local + # namespace (so the newly defined parameter variables will be + # entered into param_dict). + exec fixPythonIndentation(param_code) in globals(), param_dict + except Exception, exc: + print "Param error in %s:" % path, exc + return + # Convert object name string to Python class object + obj = eval(obj_name) + # Set the object's parameter description dictionary (see MetaConfigNode). + obj.set_param_dict(param_dict) + + +# Walk directory tree to find .odesc files. +# Someday we'll have to make the root path an argument instead of +# hard-coding it. For now the assumption is you're running this in +# util/config. +root = '../..' +os.path.walk(root, find_odescs, None) + +# Iterate through file dictionary and define classes. +for objname, path in odesc_file.iteritems(): + if not odesc_loaded[objname]: + def_class(path) + +# Iterate through files again and load parameters. +for path in odesc_file.itervalues(): + def_params(path) + +##################################################################### + +# The final hook to generate .ini files. Called from configuration +# script once config is built. +def instantiate(*objs): + for obj in objs: + obj._instantiate() |