1 Name usage types: as parameters, as base classes, as callables. This potentially restricts
2 attribute usage effects because names mentioned as base classes are not propagated and
3 made freely available for use in attribute accesses.
4
5 Low-Level Instructions and Macro Instructions
6 =============================================
7
8 Have contexts and values stored separately in memory. This involves eliminating DataValue
9 and storing attributes using two words.
10
11 Migrate macro instructions such as the *Index instructions to library code implemented
12 using low-level instructions.
13
14 Consider introducing classic machine level instructions (word addition, subtraction, and
15 so on) in order to implement all current RSVP instructions.
16
17 Move common code sequences to a library routine, such as the context checking that occurs
18 in functions and methods.
19
20 Dataflow Optimisations
21 ======================
22
23 Assignments, particularly now that no result register exists, may cause StoreTemp/LoadTemp
24 instruction pairs to be produced and these could be eliminated.
25
26 Class and Module Attribute Assignment
27 =====================================
28
29 Verify that the context information is correctly set, particularly for the unoptimised
30 cases.
31
32 Update docs/assignment.txt.
33
34 Prevent assignments within classes, such as method aliasing, from causing the source of an
35 assignment from being automatically generated. Instead, only external references should be
36 registered.
37
38 Prevent "from <module> import ..." statements from registering references to such local
39 aliases such that they cause the source of each alias to be automatically generated.
40
41 Consider attribute assignment observations, along with the possibility of class and module
42 attribute assignment.
43
44 (Note direct assignments as usual, indirect assignments via the attribute usage
45 mechanism. During attribute collection and inference, add assigned values to all
46 inferred targets.)
47
48 (Since class attributes can be assigned, StoreAttrIndex would no longer need to reject
49 static attributes, although this might still be necessary where attribute usage analysis
50 has not been performed.)
51
52 Potentially consider changing static attribute details to use object-relative offsets in
53 order to simplify the instruction implementations. This might allow us to eliminate the
54 static attribute flag for attributes in the object table, at least at run-time.
55
56 Dynamic Attribute Access
57 ========================
58
59 Consider explicit accessor initialisation:
60
61 attr = accessor("attr")
62 getattr(C, attr)
63
64 Attribute Usage
65 ===============
66
67 Consolidate interface observations by taking all cached table accesses and determining
68 which usage patterns lead to the same types. For example, if full usage of {a, b} and
69 {a, b, c} leads to A and B in both cases, either {a, b} can be considered as partial usage
70 of the complete interface {a, b, c}, or the latter can be considered as an
71 overspecification of the former.
72
73 Make the gathering of usage parameterisable according to the optimisation level so that a
74 choice can be made between control-flow-dependent observations and the simple collection
75 of all attributes used with a name (producing a more static interface observation).
76
77 Usage of self to restrict attribute usage observations and coverage.
78
79 Perform attribute usage on attributes of self as names, potentially combining observations
80 across methods.
81
82 Loop entry points and other places where usage becomes more specific might be used as
83 places to impose guards. See tests/attribute_access_type_restriction_loop_list.py for an
84 example.
85
86 Consider attribute usage observations being suspended inside blocks where AttributeError
87 may be caught (although this doesn't anticipate such exceptions being caught outside a
88 function altogether).
89
90 Consider type deduction and its consequences where types belong to the same hierarchy
91 and where a guard could be generated for the most general type.
92
93 Consider permitting multiple class alternatives where the attributes are all identical.
94
95 Support class attribute positioning similar to instance attribute positioning, potentially
96 (for both) based on usage observations. For example, if __iter__ is used on two classes,
97 the class attribute could be exposed at a similar relative position to the class (and
98 potentially accessible using a LoadAttr-style instruction).
99
100 **** Constant attribute users need not maintain usage since they are already resolved. ****
101
102 Consider handling CallFunc in micropython.inspect in order to produce instances of specific classes.
103 Then, consider adding support for guard removal/verification where known instances are involved.
104 Consider handling branches of values within namespaces in order to support more precise value usage.
105
106 Frame Optimisations
107 ===================
108
109 Stack frame replacement where a local frame is unused after a call, such as in a tail call
110 situation.
111
112 Local assignment detection plus frame re-use. Example: slice.__init__ calls
113 xrange.__init__ with the same arguments which are unchanged in xrange.__init__. There is
114 therefore no need to build a new frame for this call, although in some cases the locals
115 frame might need expanding.
116
117 Reference tracking where objects associated with names are assigned to attributes of other
118 objects may assist in allocation optimisations. Recording whether an object referenced by
119 a name is assigned to an attribute, propagated to another name and assigned to an
120 attribute, or passed to another function or method might, if such observations were
121 combined, allow frame-based or temporary allocation to occur.
122
123 Instantiation
124 =============
125
126 Specific instances could be produced, providing type information and acting somewhat like
127 classes during inspection.
128
129 Inlining
130 ========
131
132 Where a function or method call can always be determined, the body of the target could be
133 inlined - copied into place - within the caller. If the target is only ever called by a
134 single caller it could be moved into place. This could enhance deductions based on
135 attribute usage since observations from the inlined function would be merged into the
136 caller.
137
138 Function Specialisation
139 =======================
140
141 Specialisation of certain functions, such as isinstance(x, cls) where cls is a known
142 constant.
143
144 Structure and Object Table Optimisations
145 ========================================
146
147 Fix object table entries for attributes not provided by any known object, or provide an
148 error, potentially overridden by options. For example, the augmented assignment methods
149 are not supported by the built-in objects and thus the operator module functions cause
150 the compilation to fail. Alternatively, just supply the methods since something has to do
151 so in the builtins.
152
153 Consider attribute merging where many attributes are just aliases for the same underlying
154 definition.
155
156 Consider references to defaults as occurring only within the context of a particular
157 function, thus eliminating default value classes if such functions are not themselves
158 invoked.
159
160 Scope Handling
161 ==============
162
163 Consider merging the InspectedModule.store tests with the scope conflict handling.
164
165 Consider labelling _scope on assignments and dealing with the assignment of removed
166 attributes, possibly removing the entire assignment, and distinguishing between such cases
167 and unknown names.
168
169 Check name origin where multiple branches could yield multiple scope interpretations:
170
171 ----
172 try:
173 set # built-in name
174 except NameError:
175 from sets import Set as set # local definition of name
176
177 set # could be confused by the local definition at run-time
178 ----
179
180 Object Coverage
181 ===============
182
183 Support __init__ traversal (and other implicit names) more effectively.
184
185 Other
186 =====
187
188 Consider a separate annotation phase where deductions are added to the AST for the
189 benefit of both the reporting and code generation phases.
190
191 Support self attribute visualisation in the reports and/or provide a function or
192 annotations which can provide the eventual optimisation directly to such components.
193
194 Check context_value initialisation (avoiding or handling None effectively).
195
196 Consider better "macro" support where new expressions need to be generated and processed.
197
198 Detect TestIdentity results involving constants, potentially optimising status-affected
199 instructions:
200
201 TestIdentity(x, y) # where x is always y
202 JumpIfFalse(...) # would be removed (never false)
203 JumpIfTrue(...) # changed to Jump(...)
204
205 Status-affected blocks could be optimised away for such constant-related results.