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] Frama-C labels/States

  • Subject: [Frama-c-discuss] Frama-C labels/States
  • From: barbaraisabelvieira at gmail.com (Bárbara Vieira)
  • Date: Fri Nov 21 17:43:10 2008
Hi everyone!

 

I?m using Frama-C new version.

 

I was writing a predicate to a very simple function, and I don?t understand
why it doesn?t distinguish the states when the proof obligations are
generated.  That is, when I apply the axiom, I should have all the variables
in two different states to prove the proof obligation generated by the
assert clause, and I have the variables in the same state.

The code is the following:

 

********************************************************************

 

/*@ axiomatic InterInv1 {

  @ predicate interInv{L1,L2}

  @               (int i, unsigned char *indata_i, unsigned char
*outdata_i,unsigned char x_i, 

  @               unsigned char y_i, unsigned char *d_i, unsigned char tx_i,
unsigned char ty_i,

  @               unsigned char *dkey_i,

  @               unsigned char *indata_f, unsigned char *outdata_f,unsigned
char x_f, 

  @               unsigned char y_f, unsigned char *d_f, unsigned char tx_f,
unsigned char ty_f,

  @               unsigned char *dkey_f,int j, unsigned long len); }

  @ axiomatic InterInv2 {

  @ axiom interInv_ind{L1,L2}:

  @         \forall int i, unsigned char *indata_i, unsigned char
*outdata_i, unsigned char x_i,

  @                     unsigned char y_i, unsigned char *d_i,unsigned char
tx_i, unsigned char ty_i,unsigned char *dkey_i,

  @                 unsigned char *indata_f, unsigned char *outdata_f, 

  @                     unsigned char x_f, unsigned char y_f, unsigned char
*d_f,

  @                     unsigned char tx_f, unsigned char ty_f,unsigned char
*dkey_f,

  @                     int j, unsigned long len;

  @     (\at(x_f,L2) == ((\at(x_i,L1)+1)&0xff) &&

  @         \at(tx_f,L2)== \at(d_i[x_f],L1) && 

  @         \at(y_f,L2) == ((\at(tx_f,L2)+ \at(y_i,L1))&0xff) &&

  @         \at(ty_f,L2)== \at(d_i[y_f],L1) &&

  @         \at(d_f[x_f],L2)== \at(ty_f,L2) &&

  @         \at(d_f[y_f],L2) == \at(tx_f,L2) && 

  @         \at(dkey_f[((len>>3L)-j)*i],L2) == \at(d_f[(tx_f+ty_f)&0xff],L2)
&&

  @         \at(outdata_f[i],L2) == (\at(d_f[(tx_f+ty_f)&0xff],L2) ^
\at(indata_f[i],L2)) &&

  @         \at(indata_f[i],L2) == \at(indata_i[i],L2)) ==>

  @
interInv{L1,L2}(i,indata_i,outdata_i,x_i,y_i,d_i,tx_i,ty_i,dkey_i,

  @
indata_f,outdata_f,x_f,y_f,d_f,tx_f,ty_f,dkey_f,j,len);

  @ }

  @*/

   

 /*@ requires \valid_range(indata,0,len) &&

   @          \valid_range(outdata,0,len) &&

   @        0<=i<=len &&

   @          \valid_range(d,0,255) &&

   @          \valid_range(dkey,0,len);

   @ ensures

   @
interInv{Old,Here}(i,indata,outdata,x,y,d,tx,ty,dkey,indata,outdata,x,y,d,tx
,ty,dkey,j,len);

   @*/

 void LOOP(int i, unsigned char x, unsigned char y,unsigned char tx,
unsigned char ty,

               unsigned char *d, unsigned char *outdata, unsigned char
*indata, unsigned char *dkey,

               int j, const unsigned long len)

 {

  

      x=((x+1)&0xff); 

      tx=d[x]; 

      y=((tx+y)&0xff); 

      ty=d[y]; 

      d[x]= ty;

      d[y]=tx; 

      dkey[((len>>3L)-j)*i] = d[(tx+ty)&0xff];

      (outdata[i]) = (d[(tx+ty)&0xff] ^ (indata[i]));

      

   /*@ ghost goto L; */

   /*@ ghost L:*/ 

   

      /*@ assert

        @
interInv{Pre,L}(i,indata,outdata,x,y,d,tx,ty,dkey,indata,outdata,x,y,d,tx,ty
,dkey,j,len);

        @*/

 }

********************************************************************

 

 

I made a post some days ago asking if I can specify the axioms for
predicates in the way I did in the example above. Does anybody knows the
answer? It is equivalent to the definition of an ?axiomatic? with the
predicates and axioms, or the generated proof obligations, aren?t the same?

 

 

Thanks.

Best regards,

B?rbara 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.gforge.inria.fr/pipermail/frama-c-discuss/attachments/20081121/a5a050f7/attachment.html