[Bug c/19056] New: 4.0 optimizes away a SEGV; 3.4.1 does not

[Bug c/19056] New: 4.0 optimizes away a SEGV; 3.4.1 does not

[bh at techhouse dot brown dot edu]

A minimized test case:

int main() {
    int y = 1;
    int *x = &y;
    volatile int sum = 0;
    while(1) {
        sum += *x;
        x++;
    }
    return 0;
}

Clearly, this should crash; under 3.4.1 it does.  Under my checkout of a few
hours ago, however, it does not: instead, it enters an infinite loop (everything
inside the while(1) gets optimized away and we're left with just a jmp).

Workaround: make either x or sum volatile.  This forces it to actually do the
computations inside the loop.

I'm not sure whether to view this as a compiler bug: this code is, at heart,
morally unbalanced.  But it is a change in behaviour so I figured I'd report it.

-- 
           Summary: 4.0 optimizes away a SEGV; 3.4.1 does not
           Product: gcc
           Version: 4.0.0
            Status: UNCONFIRMED
          Severity: minor
          Priority: P2
         Component: c
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: bh at techhouse dot brown dot edu
                CC: gcc-bugs at gcc dot gnu dot org
 GCC build triplet: i686-pc-linux-gnu
  GCC host triplet: i686-pc-linux-gnu
GCC target triplet: i686-pc-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19056

[Bug c/19056] 4.0 optimizes away a SEGV; 3.4.1 does not

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-12-17 15:18 -------
The code is undefined, once you get into that, anything can happen so closing as invalid.

We just optimize away the add and the load since the load is no longer needed.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |INVALID


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19056