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] Proof conditions with simple pointer assignment

Hello,
Just from an ACSL point of view, the proof of the "assigns"
and "ensures" clauses of your spec requires locations "p"
and "*p" to be \separated.
Patrick Baudin.
-- 

Patrick Baudin,
CEA, LIST, SOL, F-91191 Gif-sur-Yvette cedex, France.
tel: +33 (0)1 6908 2072

> Hello,
>
> Consider the following spec,
>
> /*@ requries \valid(p);
>    @ assigns *p;
>    @ ensures *p == 1;
> */
> void f(int *p)
> {
>     *p = 1;
> }
>
> This generates the proof condition (in consice terms)
> H2:  integer_of_int32(result) = 1
> int_P_int_M_p_1_0: int(int_P, int32) memory
> H3: int_P_int_M_p_1_0 = store(int_P_int_M_p_1, p, result)
> -------------------------
> integer_of_int32(select (int_P_int_M_p_1_0, p)) = 1
>
> The conditions is not discharged although axiom select_store_eq seems
> quite sufficient to discharge it. Is integer_of_int32 somehow getting in
> the way?
>
> Thank you
> Damien
>