Re: [CHECKER] potential dereference of user pointer errors

[Alan Cox]

On Fri, 2003-03-21 at 22:15, Chris Wright wrote:
> on first pass of the cmd.  However, this is inconsistent with the rest
> of the file, so here is a patch to use kcmd.resbuf.  I also added a NULL
> check, as done in similar funcitons in this file.  Alan, this look ok?

Looks slightly wrong to me

#1 ->resbuf = NULL is a completely acceptable if odd user choice. If invalid
its covered

#2 - We copy to the users nominated cmd->resbuf. You are correct there, 
that we should be using the kernel side copy. Fixed in my tree.