David Bühler authored 1 year ago and Allan Blanchard committed 1 year ago

- Only generates assigns clauses when required:
  - when the analysis uses the specification to analyze a function
    (decided in function_calls.ml)
  - for Frama_C_* builtins, such as Frama_C_assert. Others builtins should
    already have a specification.
  - when interpreting a Frama_C_show_each directive, so other plugins known
    these functions have no effect.
  - in [Logic_inout.valid_behaviors], used by inout and from plugins to
    interpret a function specification.

- Also generates assigns clauses for functions with a body: this is required
  for the analysis of recursive functions and of functions specified in the
  -eva-use-spec parameter (in both case, a specific Eva warning is emitted).

- In recursion.ml, renames get_spec to check_spec.
  The generation of assigns clauses is now uniformly done in function_calls.

[accept_base] is used by the Inout and From plugins, and its implementation
in callgraph is very slow on programs with many callsites.

Intented to replace calls to Db.Properties.Interp.loc_to_loc*, used by
the inout and from plugins.

Uses a reference to a function provided by Eva to compute the dependency
of a term.

Exports these functions in Eva.mli.

Used by the from plugin when option -from-verify-assigns is enabled.

This file will contain all functions used by the inout and from plugins to
interpret ACSL predicates and assigns clauses.
It is now in the legacy/ directory.

Fixes a crash when the location of an \assigns clause cannot be computed.

- shares and moves the functions [reduce_offset_by_validity] of Locations and
  Precise_locs into base.ml.
- replaces the boolean argument ~for_writing into the new type access, that
  represents Read, Write or No_access. Without any access, offsets must point
  into or just beyond the base validity.
- fixes the support of accesses of size 0: they are now invalid:
  + in bases with Invalid validity;
  + one past a base validity unless the base ends with an empty struct.

(blind make headers from specifications)

- Fix Value/Value#5
- Fix Value/Value#14

Some case studies (e.g. dyad) use some ugly casts from fd_set_t which lead
to the analysis stopping too early.
Changing the representation of fd_set_t should also help it better conform to
the standard (since a fd_set_t should be able to hold FD_SETSIZE elements).

[Ptests] preserve LOG after STDOPT directive

See merge request frama-c/frama-c!2073