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] NON TERMINATING FUNCTION when adding specification

  • Subject: [Frama-c-discuss] NON TERMINATING FUNCTION when adding specification
  • From: anne.pacalet at free.fr (Anne Pacalet)
  • Date: Fri, 16 Nov 2012 11:51:34 +0100
Dear List,

I have a strange problem getting a :
[value] Values at end of function main:
           NON TERMINATING FUNCTION
on an application.

I tried to get the simplest example from it, and there it is :
-------------------------------------------------------
//@ assigns dest[0..]; ensures \result == dest;
extern char *str_add (char *dest, int val);

int main (void) {
   int t[10];
   char str[256];
   for (int i = 0; i < 10; i++) t[i] = i;
   for (int i = 0; i < 10; i++) str_add (str, t[i]);
   return 0;
}
-------------------------------------------------------
The strange thing is that, when I remove the `ensures` property
from `str_add` specification, the value analysis gives :
[value] Values at end of function main:
           t[0..9] ? [0..9] or UNINITIALIZED
           str[0..255] ? [--..--] or UNINITIALIZED
           __retres ? {0}
as expected.

So my question is : why is this property disturbing the analysis ?
Should I always give the smallest specification ?

Thanks a lot for you help and advises.
-- 
Anne.