Date: Wed, 13 Mar 2002 12:00:47 -0500 From: "Ugen Antsilevitch" <ugen@undp.org> To: questions@freebsd.org Subject: Kernel module and syscall locking. Message-ID: <897afe89645d.89645d897afe@undp.org>
next in thread | raw e-mail | index | archive | help
I am getting a little rusty on my syscall skills. So here is the
situation:
- I have a kernel module that loads and initializes a function pointer
or two. Once initialized, these function pointers will be used in
a system call (such as execve for instance) to do some evil magic.
- With ipfw as an example locking for the initialization is easy -
splnet, that shuts down network processing for as long as init runs.
- Now with a system call here is what i am wondering:
1) Do i have to lock at all? Can a user initiate a system call
while something else happens in kernel? (That is - can the context
get switched to a user and then another kernel?)
2) If i do have to lock is it by means of splsomething? I think
these are pretty much interrupt levels, so for user calls that
won't do any good?
3) 4.2 (and 4.X in general) are still single "thread" kernels,
right? what about 5.0 ? (that may be an ignorance question but
i can't download the source tree now, with the modem connection
being the only one available).
Thanx!
--Ugen
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?897afe89645d.89645d897afe>