Re: [PATCH 1/10] Trivial: /dev/kvm interface is no longer experimental.

[Arnd Bergmann <arnd-r2nGTMty4D4@public.gmane.org>]

On Wednesday 18 July 2007, Avi Kivity wrote:
> 
> Once we're back to using fds, we might as well use ioctls.  If anything,
> an ioctl has an explicit mention of the structure it manipulates in its
> definition.  I don't care that it came in last in the last 15 annual
> kernel beauty contests.
> 
> For me, the difference between syscalls and ioctls is whether the vcpu
> is bound to a task (and the vm bound to the mm) rather than whether the
> names are spelled in lowercase or uppercase.

Ok, good point. If we go to syscalls, it absolutely makes sense to do
the automatic mm_struct<->guest and thread<->vcpu association and
not pass any file descriptors around, while as long as we're using
ioctl, having the fds point to the objects is logical.

My point regarding the move to system calls instead of ioctl was
more about emphasizing kvm as a core kernel functionality, not a device
driver, the same way that we don't have /dev/fork or /dev/epoll but
use system calls for those.

	Arnd <><

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/