--TybLhxa8M7aNoW+V
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Tue, Feb 14, 2006 at 01:06:53PM +0700, Robert Elz wrote:
> I see that the kernel has had a pair of functions
> 	ioctl_copyin()
> 	ioctl_copyout()
> since last August.   That's fine, and from what I can see the
> implementation of them looks fine.
>=20
> What I can't find is a single user of those functions anywhere
> in the kernel.   That's a pity, as I was kind of hoping to copy
> someone else's code rather than having to work it out myself.

I think the users are about to appear.  These were added for code
calling ioctl routines from elsewhere in the kernel, rather than the
usual case which is obviously that they're called from userland
devices.=20

The need to do this apparently arises in calling some new ioctl's to
be added to the CD driver for scsi-mmc use with burning from UDF.  UDF
has now been added, and the writing part along with scsi support seems
to be under way.

> Is it regarded as being safe to use those functions?   That is,
> are they expected to remain, or are they an experiment that failed?

There was some discussion about them, perhaps the above helps you
search for the answer to this.

As for the wider question, more evidence is required :)

--
Dan.
--TybLhxa8M7aNoW+V
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (NetBSD)

iD8DBQFD8XcuEAVxvV4N66cRAsVYAKDuEfM/pfDACdkXtwRXndofXpxziQCeIAh/
X7y0111YKEuRMXbHTtMMxpE=
=VSp9
-----END PGP SIGNATURE-----

--TybLhxa8M7aNoW+V--