Running gcc-4.3 -O2 on PR1386.c I get different results than with -O1.

$ gcc-4.3 -O2 PR1386.c && ./a.out
PR1386.c: In function ‘main’:
PR1386.c:15: warning: large integer implicitly truncated to unsigned type
PR1386.c:17: warning: large integer implicitly truncated to unsigned type
bad bits: fffffffffffffff0
$ gcc-4.3 -O2 PR1386.c && ./a.out
PR1386.c: In function ‘main’:
PR1386.c:15: warning: large integer implicitly truncated to unsigned type
PR1386.c:17: warning: large integer implicitly truncated to unsigned type
bad bits: fffffffffffffff0

gcc-4.2 is fine:
$ gcc-4.2 -O2 PR1386.c && ./a.out
PR1386.c: In function ‘main’:
PR1386.c:15: warning: large integer implicitly truncated to unsigned type
PR1386.c:17: warning: large integer implicitly truncated to unsigned type
bad bits: 0

$ gcc-4.2 -O3 PR1386.c && ./a.out
PR1386.c: In function ‘main’:
PR1386.c:15: warning: large integer implicitly truncated to unsigned type
PR1386.c:17: warning: large integer implicitly truncated to unsigned type
bad bits: 0

The file is from an LLVM testcase http://llvm.org/bugs/show_bug.cgi?id=1386

#include <stdio.h>
#include <stdint.h>

struct X {
  unsigned char pad : 4;
  uint64_t a : 64;
  uint64_t b : 60;
} __attribute__((packed));

int main (void)
{
  struct X x;
  uint64_t bad_bits;

  x.pad = 255;
  x.a = -1ULL;
  x.b = -1ULL;

  bad_bits = ((uint64_t)-1ULL) ^ *(1+(uint64_t *) &x);
  printf("bad bits: %llx\n", bad_bits);
  return bad_bits != 0;
}