Re: proc changes for rusage

[Roland McGrath]

> There are two callers of REAP, one at wait.c:206 and the second at
> wait.c:218.  The first expects *PID_STATUS to be set to be set to PID
> and the second expects it to be set to CHILD->p_pid.  Therefore, your
> patch does not fix the problem.

Please look a little deeper.  CHILD->p_pid is in fact always the right value.
In one instance PID always has the same value.

> Secondly, the if statement at the beginning of REAP adds nothing to the
> code except to make it a good deal harder to see the intension.

I rewrote the function to make it concise without redundant tests.