[PATCH 07/28] gss_krb5: Use random value to initialize confounder

[Kevin Coffman]

On Mon, Mar 31, 2008 at 12:24 PM, Kevin Coffman <kwc at citi.umich.edu> wrote:
> On Mon, Mar 31, 2008 at 11:26 AM, Chuck Lever <chuck.lever at oracle.com> wrote:
>  > On Mar 31, 2008, at 10:31 AM, Kevin Coffman wrote:

>  >  >  {
>  >  >       static u64 i = 0;
>  >  >       u64 *q = (u64 *)p;
>  >  > @@ -102,8 +102,19 @@ make_confounder(char *p, int blocksize)
>  >  >        * uniqueness would mean worrying about atomicity and rollover,
>  >  > and I
>  >  >        * don't care enough. */
>  >  >
>  >  > -     BUG_ON(blocksize != 8);
>  >  > -     *q = i++;
>  >  > +     /* initialize to random value */
>  >  > +     if (i == 0) {
>  >  > +             i = random32();
>  >  > +             i = (i << 32) | random32();
>  >  > +     }
>  >
>  >  Nit: The RPC layer usually uses net_random() instead of calling
>  >  random32() directly.
>
>  O.K.

Hmmm.  Is it worth including <linux/net.h> here to get the definition
of net_random()?  (Which is currently a #define to random32).