On Tue, May 29, 2007 at 05:20:33PM -0500, David Young wrote:
> 
> This seems like an awful lot of #ifdef'age to achieve very limited
> protection against stack smashing.  Suppose the kernel copies to ifreq
> a sockaddr whose sa_len > sizeof(struct sockaddr_storage) ?

The kernel won't: sockaddr_storage is, by definition, large enough to
contain any protocol-specific sockaddr.  That's what it's for.

The issue with kernel->user copies was truncation of addresses.  The
stack-smashing issue involved legitimate programming practices like
trying to zero the entire sockaddr_dl "contained" in an ifreq...

-- 
Thor Lancelot Simon	                               tls@rek.tjls.com
  "All of my opinions are consistent, but I cannot present them all
   at once."	-Jean-Jacques Rousseau, On The Social Contract