Re: start unlocking kbind(2)

[Theo de Raadt]

I guess there is another question -- should the ps_kbind_* variables
be stored inside the uvmspace, rather than inside pr?