On Thu, 9 Mar 2006 08:13:46 -0600
"David Berg" <drberg1000@gmail.com> wrote:
> On 3/9/06, Florian Kulzer <florian@molphys.leidenuniv.nl> wrote:
> > David Berg wrote:
> >
> > [...]
> >
> > > Correct me if I'm mistaken, being that PROGRAM is a _key_ it is run
> > > before the device is created so that it can be used to determine which
> > > name should be used. Since it is run before the device is created, it
> > > can't be used to mount a filesystem on that device.
> >
> > I think the main problem is that I did not realize that you are running
> > Sarge. In Sid hotplug is completely replaced by udev, and I think that
> > this is also what will happen when Etch is released. My information
> > about the PROGRAM key is from the manual of udev 0.086. Maybe this is
> > not implemented yet for the udev of Sarge. Can you find any information
> > about the intended interaction between hotplug and the older udev in
> > /usr/share/doc/udev/README.Debian.gz?
>
> dberg@chip:/etc/hotplug.d/default$ ls -l 10*
> lrwxrwxrwx 1 root root 14 2006-02-17 08:24 10-udev.hotplug -> /sbin/udevsend
> lrwxrwxrwx 1 root root 24 2006-02-17 08:24 20-hal.hotplug ->
> /usr/lib/hal/hal.hotplug
> -rwxr-xr-x 1 root root 2920 2006-03-08 19:37 default.hotplug
>
> From udevd/udevsend manpage:
> "udevd receives the events from udevsend"
>
> So udev is called before the rest of the hotplug scripts are executed
> but apparently it doesn't finish fast enough. udevsend must return
> before the device is created because /sbin/hotplug doesn't put the
> scripts it calls in the background:
>
> #/sbin/hotplug
> DIR="/etc/hotplug.d"
>
> for I in "${DIR}/$1/"*.hotplug "${DIR}/"default/*.hotplug ; do
> test -x "$I" && "$I" "$1"
>
> But again none of my mesg commands are making it to the logs while
> hotplug's usbagent reports that "photodl: loaded successfully". To me
> this seems like the primary problem. If my sleep command were
> actually executed, I should be able to make it wait a sufficient
> amount of time for the devices to come up. Perhaps not as elegant as
> it could be but I'm quite ready to be done with this little task and
> move on to something else.
just a quick thought. is it possible that udev doesn't actually process everything on the one line as suggested by Florian? Maybe it only processes one item at a time until it satisfies the rule and thus, your PROGRAM at the end of the line doesn't ever actually run as the other keys satisfy and generate the device. Also, its my understanding that the PROGRAM is just like any other key and udev is expecting it to return a value to help udev determine what devices to create.
A
>
> --Dave
Attachment:
pgpD1CqtrNTEV.pgp
Description: PGP signature