Re: compat: autofs v5 packet size ambiguity - update

[Ian Kent <raven@themaw.net>]

On Thu, 2012-02-23 at 19:20 +0800, Ian Kent wrote:
> On Wed, 2012-02-22 at 22:31 -0800, H. Peter Anvin wrote:
> > However, I liked Linus' approach.of using the version ioctl; any reason not to?
> 
> Not really, although that ioctl is used it doesn't have to be but you
> must mount the autofs file system to use it. Setting a timeout is
> another ioctl that is almost mandatory although not using it amounts to
> saying you don't want anything to expire which is valid.

The open, to get an ioctl is another that pretty much must be done to
use the mount.

> 
> > 
> > Ian Kent <raven@themaw.net> wrote:
> > 
> > >On Thu, 2012-02-23 at 10:21 +0800, Ian Kent wrote:
> > >> On Thu, 2012-02-23 at 09:54 +0800, Ian Kent wrote:
> > >> > On Thu, 2012-02-23 at 09:48 +0800, Ian Kent wrote:
> > >> > > On Wed, 2012-02-22 at 08:12 -0800, Linus Torvalds wrote:
> > >> > > > On Wed, Feb 22, 2012 at 1:32 AM, Ian Kent <raven@themaw.net>
> > >wrote:
> > >> > > > >
> > >> > > > > Oh, DOH ...
> > >> > > > >
> > >> > > > > The process mounting the autofs mount is passing the pipe it
> > >will use
> > >> > > > > for communication via a mount option so we can do this right
> > >in
> > >> > > > > fill_super.
> > >> > > > 
> > >> > > > No, we really cannot.
> > >> > > 
> > >> > > Sorry, I think your wrong this time.
> > >> > 
> > >> > On second thought, I suppose other user space users don't "have" to
> > >use
> > >> > mount(2) ....
> > >> 
> > >> Thomas, what does systemd use to mount the autofs mounts that it
> > >uses?
> > >
> > >Mmm ... AFAICS systemd uses mount(2) so that looks OK too.
> > >
> > >Ian
> > 
>