This is the mail archive of the gcc-bugs@gcc.gnu.org mailing list for the GCC project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: a.out gives different results when compiled by g++ 3.1 and g++ 3.1 -O

Falk Hueffner wrote:


enum {N = 624,              // length of state vector
      M = 397,              // period parameter
      MAGIC = 0x9908B0DFU}; // magic constant

ui32 state[N];              // internal state
[...]
*p++ = twist( p[M-N], p[0], p[1] );


You're accessing out of the array bounds here, so anything might
happen.

Dear Falk,

I have found the latest version of the code, and the guy goes:
v0.7 - Fixed operator precedence ambiguity in reload()

which gives:
for( i = N - M; i--; ++p )
*p = twist( p[M], p[0], p[1] );
for( i = M; --i; ++p )
*p = twist( p[M-N], p[0], p[1] );

I suppose it's what your were talking about. So there is no standard and the behaviour may depend on optmization?

Regards,
Erick.






Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]