Arithmetic overflow in may_expand_vm()

[Earl Chew <earl_chew@agilent.com>]

This code currently reads:

> int may_expand_vm(struct mm_struct *mm, unsigned long npages)
> {
>         unsigned long cur = mm->total_vm;       /* pages */
>         unsigned long lim;
> 
>         lim = current->signal->rlim[RLIMIT_AS].rlim_cur >> PAGE_SHIFT;
> 
>         if (cur + npages > lim)
>                 return 0;
>         return 1;
> }

If npages is stupendously large, the failure predicate may
return a false negative due to (cur + npages) overflowing and
wrapping.

I think it's more robustly written as:

          if (npages > lim - cur)
                  return 0;