1 | Status of matita port of CompCert 1.6 C semantics |
---|
2 | ================================================= |
---|
3 | |
---|
4 | Last seen working with matita svn r11115. |
---|
5 | |
---|
6 | - Most of the memory model has been ported (common/Mem.v -> Mem.ma). |
---|
7 | The main exception is minor arithmetic results than were proven with |
---|
8 | omega, which we hope to prove using matita's automation at some |
---|
9 | point. For the C semantics deliverable these (unnecessary) properties |
---|
10 | have been commented out. |
---|
11 | |
---|
12 | - The memory model is executable; some small tests can be found in |
---|
13 | test/memorymodel.ma. |
---|
14 | |
---|
15 | - The Clight syntax file has been ported, minus a few lemmas that are not used |
---|
16 | by the semantics. |
---|
17 | |
---|
18 | - The small-step inductively defined Clight semantics has been ported. |
---|
19 | |
---|
20 | - A collection of definitions and lemmas that probably ought to be in |
---|
21 | matita's standard library can be found in extralib.ma. |
---|
22 | |
---|
23 | - Some of the theory about machine integers (i.e., integers modulo) are |
---|
24 | given as axioms in Integers.ma. Proving them should be routine now that |
---|
25 | we have a binary representation of integers. The CompCert version is given |
---|
26 | as a functor on the word size - we don't have an equivalent of this yet. |
---|
27 | |
---|
28 | * The executable semantics can be found in Cexec.ma, which uses the |
---|
29 | error monad in Errors.ma and the resumption monad defined in IOMonad.ma. |
---|
30 | |
---|
31 | * The patch in acc-0.1.spaces.patch adds a -ma option to the prototype |
---|
32 | compiler which outputs a matita file containing the Clight program as |
---|
33 | a definition usable by the semantics. It also adds parsing support for |
---|
34 | the memory space extensions (__data and friends). |
---|
35 | |
---|
36 | * The patch in compcert-1.7.1-matita.patch adds a -dmatita option to |
---|
37 | output a matita file containing a definition for the C program. |
---|
38 | Note that there must be a "main" function. (This was made against |
---|
39 | 1.7.1 for convenience - the Clight syntax is the same as 1.6.) |
---|
40 | This is an older version of the prototype compiler patch, and lacks the |
---|
41 | memory extensions. |
---|
42 | |
---|
43 | There is a more recent version of compcert available with a revised memory |
---|
44 | model that may be more closely suited to our needs. |
---|
45 | |
---|
46 | matita issues and workarounds |
---|
47 | ----------------------------- |
---|
48 | |
---|
49 | No unfold tactic yet; use reduction or rewriting instead. |
---|
50 | |
---|
51 | Normalization of term containing large definitions (such as some of the |
---|
52 | Integers library) can be prohibitive. This is a particular problem because |
---|
53 | ndestruct currently normalizes the goal and every hypothesis that is rewritten. |
---|
54 | For ndestruct, getting rid of irrelevant information (e.g., replacing the goal |
---|
55 | with False) sometimes helps, an extra lemma can be used, or the definitions |
---|
56 | can make careful use of 'nlet rec' to prevent reduction. |
---|
57 | |
---|
58 | Some of the pattern matching in the C syntax has had to be unfolded in |
---|
59 | a rather unpleasant way (where matching on pairs was used in the original). |
---|
60 | |
---|
61 | The use of records in place of modules is a little delicate: large records |
---|
62 | tend to use a lot of memory and processor at the moment, and the names of the |
---|
63 | fields tend to clash a lot. |
---|
64 | |
---|
65 | The lack of sections has been replaced by duplicating the variables in some |
---|
66 | places, and adding a record in others. |
---|
67 | |
---|
68 | The meminj_no_overlap definition in Mem.ma uses the logical Or explicitly |
---|
69 | rather than the ∨ notation to regain reasonable performance. (This is |
---|
70 | presumably due to disambiguation - the notation is also used for booleans.) |
---|
71 | |
---|
72 | There are a couple of places in Mem.ma where rewriting had to be done carefully |
---|
73 | to avoid assertions during unification; I'm not really sure what the problem |
---|
74 | is. |
---|
75 | |
---|
76 | Detail per-file |
---|
77 | --------------- |
---|
78 | |
---|
79 | AST.ma |
---|
80 | |
---|
81 | Program transformation sections mostly left alone. |
---|
82 | |
---|
83 | CexecComplete.ma [new] |
---|
84 | |
---|
85 | Completeness of steps of executable semantics wrt inductive |
---|
86 | |
---|
87 | CexecEquiv.ma [new] |
---|
88 | |
---|
89 | Equivalence of complete executions to inductive semantics |
---|
90 | |
---|
91 | Cexec.ma [new] |
---|
92 | |
---|
93 | Definition of executable semantics |
---|
94 | |
---|
95 | CexecSound.ma |
---|
96 | |
---|
97 | Soundness of individual steps of executable semantics wrt inductive |
---|
98 | |
---|
99 | Coqlib.ma |
---|
100 | |
---|
101 | Only essential parts: |
---|
102 | - align |
---|
103 | - some 2^n lemmas (which could go into the binary libraries) |
---|
104 | - option_map |
---|
105 | - parts of list disjointness and non-repetition |
---|
106 | |
---|
107 | CostLabel.ma [new] |
---|
108 | |
---|
109 | Definition of cost labels |
---|
110 | |
---|
111 | Csem.ma |
---|
112 | |
---|
113 | The small step semantics have been ported. None of the big step semantics or |
---|
114 | the equivalence proof has been ported. |
---|
115 | |
---|
116 | Csyntax.ma |
---|
117 | |
---|
118 | Everything except for a few minor definitions / lemmas that are not required |
---|
119 | by the semantics. |
---|
120 | |
---|
121 | Errors.ma |
---|
122 | |
---|
123 | Only a basic definition without error messages. |
---|
124 | |
---|
125 | Events.ma |
---|
126 | |
---|
127 | Most of the definitions; haven't needed many of the lemmas yet. |
---|
128 | |
---|
129 | extralib.ma [new] |
---|
130 | |
---|
131 | Definitions which probably belong in the matita libraries, if anywhere. |
---|
132 | |
---|
133 | Floats.ma |
---|
134 | |
---|
135 | Axiomatised, but so was the original! Likely to be removed because we do |
---|
136 | not intend to support floating point. |
---|
137 | |
---|
138 | Globalenvs.ma |
---|
139 | |
---|
140 | Basic definition only; sufficient for animating the semantics. |
---|
141 | |
---|
142 | Integers.ma |
---|
143 | |
---|
144 | The operations have been ported, although the results about them generally |
---|
145 | have not. The equality results have been left axiomatised because they |
---|
146 | require proof irrelevance because each integer carries a proof that it is |
---|
147 | in range. The use of the Coq module system to abstract over the word size |
---|
148 | has been left out for now. |
---|
149 | |
---|
150 | IOMonad.ma [new] |
---|
151 | |
---|
152 | Definition and properties of the resumption+error monad for I/O. |
---|
153 | |
---|
154 | Maps.ma |
---|
155 | |
---|
156 | Only a basic definition. |
---|
157 | |
---|
158 | Mem.ma |
---|
159 | |
---|
160 | Everything except for omega in proofs. |
---|
161 | |
---|
162 | Smallstep.ma |
---|
163 | |
---|
164 | Ported definitions for multiple steps and program behaviour; left |
---|
165 | alternative definitions, some lemmas, plus most of the simulations sections |
---|
166 | for later. |
---|
167 | |
---|
168 | Values.ma |
---|
169 | |
---|
170 | The definitions about operations relevant to the Clight semantics have been |
---|
171 | ported. The other definitions and most of the lemmas still remain to be done. |
---|
172 | |
---|
173 | |
---|