Re: how to deny reading of several sysctls (for a set of uids, f.e.)

On Wednesday 24 January 2007 14:23, you wrote:
On Tue, 23 Jan 2007 14:10:19 +0100
Max Laier <max@xxxxxxxxxxxxxx> wrote:

[..]

td->td_proc->p_ucred has the user credentials. You
probably want to do
your checks in userland_sysctl() according to the
comment just above.

Thanks, it is really what I need.

Now I have once more question.
I made the kernel object with one check-function and all
works fine from userland via syscall().

Is there a documented possibility to use syscalls _inside_
kernel code?
In other words, I need to call the function located in
loadable kernel object from kernel, doesn't matter how
this would be done (syscall, etc).

My goal is to avoid kernel rebuilding each time after
function modification.

Is it possible?

You should use a function pointer. Look for example at "ip_dn_io_ptr" in
netinet. This is a function pointer that is populated from the dummynet
module and (if non-NULL) called from the kernel (or ip_fw_pfil.c). Note
that you might need some locking around the pointer access if you want to
be able to load and unload the module at will. OTOH, sysctl seems to be
protected by Giant anyways.

--
/"\ Best regards, | mlaier@xxxxxxxxxxx
\ / Max Laier | ICQ #67774661
X http://pf4freebsd.love2party.net/ | mlaier@EFnet
/ \ ASCII Ribbon Campaign | Against HTML Mail and News

Attachment: pgphMbTHGLp1J.pgp
Description: PGP signature

Relevant Pages