Frama-C-discuss mailing list archives

This page gathers the archives of the old Frama-C-discuss archives, that was hosted by Inria's gforge before its demise at the end of 2020. To search for mails newer than September 2020, please visit the page of the new mailing list on Renater.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Frama-c-discuss] Verifications of calls to unannotated functions

Hollas Boris (CR/AEY1) wrote:
> Consider a function that calls another function:
>
> Void f1(...) {
>   ...
>   K = f2(...); // (*)
>   ...
> }
>
> I assume that if f2 has a contract (ie, pre- and postcondition as given by the user), this contract will be used by the verifier at the place of the function call. That is, the prover shows that
> - the postcondition of f2 implies the wp derived so far
> - the precondition of f2 is the new precondition for line (*).
>
> What happens if f2 isn't annotated? Will the call to f2 just be ignored by the verifier?
>
>   
the call to f2 cannot be "ignored". A default contract will be 
considered, with "true" as both pre- and post-condition.

The main issue is for the assigns clause. in principle it should default 
to "everything" whatever it means.
In the Jessie plugin, there are two cases:

. If f2() is given an implementation, then an assigns clause which 
over-approximate the correct one is built from the implementation. In 
that situation you are safe.

. If f2() has no implementation, then "assigns \nothing" is assumed. 
Caution should of course be taken.

Hope this helps,

- Claude

-- 
Claude March?                          | tel: +33 1 72 92 59 69           
INRIA Saclay - ?le-de-France           | mobile: +33 6 33 14 57 93 
Parc Orsay Universit?                  | fax: +33 1 74 85 42 29   
4, rue Jacques Monod - B?timent N      | http://www.lri.fr/~marche/
F-91893 ORSAY Cedex                    |