On Wednesday 16 May 2007 20:32, Jeremy Allison wrote:
= > uni_p = fl3->path ? (uint32)fl3->path : 0;
= > if(!prs_uint32("ptr", ps, depth, &uni_p))
= > return False;
= > Although the 32-bit systems will not notice, the code does not seem
= > right for 64-bit ones, where pointers are 64-bit wide. Can someone,
= > please, comment? Silencing the warning is easy, but there may be a
= > genuine problem pointed out here.
= Actually it should be putting "1" here as all that matters is
= that the pointer is non-zero.
The way things are now, it could be getting it wrong every once in a while,
when the lower 32-bits of the pointer happen to be zeros. :-(
uni_p = (fl3->path != NULL);