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] Proper usage of -wp-unalias-vars?

Hello,

2017-08-04 1:53 GMT+02:00 Yifan Lu <me at yifanlu.com>:
> It appears that if I remove the const
>
> extern int * const globalx;
> extern int * const globaly;
>
> I can use -wp-model ref and all is well...
>
> but I ran into another problem
>
> /*@ predicate something(int *x) = \true; */
>
> /*@ requires something(x); */
> int test7(int *x);
>
> /*@ requires *x == 5;
>   @*/
> int test6(int *x) {
>   //@ assert *x == 5;
>   test7(x);
>   //@ assert *x == 5;
>   return 0;
> }
>
> The second assertion cannot be proved. If I remove the predicate, it
> works. Is this a bug?


Could you provide the exact file and command line that you used to
obtain this result? The second assert should not be provable (unless
the call to test7 is guaranteed to never return successfully, in which
case the code below the call is dead and any assertion on it holds) as
long as there is no assigns for test7. In fact, Frama-C should have
emitted a warning on a missing assigns clause. The kernel generates a
clause based on the prototype of test7, which is absolutely not
guaranteed to be sound and must at least be carefully reviewed. In the
case at hand, as test7 takes a pointer as argument, Frama-C assumes
that *x may be assigned by test7. Hence, no hypothesis on the content
of *x after the call can be made.

Best regards,
-- 
E tutto per oggi, a la prossima volta
Virgile