[libhid-discuss] hid_force_open failed with return code 12

[Marius Scurtescu]

On 3/21/07, Charles Lepple <clepple at ghz.cc> wrote:
>
> Marius Scurtescu wrote:
> > On 3/21/07, Charles Lepple <clepple at ghz.cc> wrote:
> >> Marius Scurtescu wrote:
> >> > On 3/21/07, Charles Lepple <clepple at ghz.cc> wrote:
> >> >> On Mar 20, 2007, at 10:34 PM, Marius Scurtescu wrote:
> >> >>
> >> >> > Hi,
> >> >> >
> >> >> > I am using libhid under Ubuntu Edgy, tried both i386 and amd64
> >> >> > (0.2.15+20060325-2ubuntu1).
> >> >> >
> >> >> > Trying to access a Radio Shark 2 device. The problem is that libhid
> >> >> > cannot detach the kernel driver from the USB device.
> >> >> >
> >> >> > Using libhid-detach-device, same thing.
> >> >> >
> >> >> > Any suggestions?
> >> >>
> >> >> The most common problem is not having write permission to the device
> >> >> node in /dev/bus/usb/*/*.
> >> >
> >> > That must be it. If I am running as root then it detaches fine. Only
> >> > root has write access to these devices. How can I change this? Should
> >> > I change the group or force write for all?
> >> >
> >> > Also, the device gets created when you insert the USB connector, can
> >> > you control permissions on that?
> >>
> >> You can do something like what we have in Network UPS Tools for USB
> >> UPSes:
> >>
> >> http://boxster.ghz.cc/projects/nut/browser/trunk/scripts/udev/nut-usbups.rules.in
> >>
> >> I think the README file one level up has some info on installing the
> >> udev
> >> rules file. (Note that the autoconf script would take the
> >> nut-usbups.rules.in file and replace @RUN_AS_GROUP@ with the actual
> >> configured group, writing the output to nut-usbups.rules. Your udev
> >> script
> >> could do something similar with your userid instead of the group.)
> >
> > Thanks for the suggestion. I will look into this.
>
> OK. Don't let the apparent complexity of the file scare you - it's only
> that big because we are trying to support a large number of USB UPS
> models.

It all worked fine. I attached the rules file I used.

Thanks again,
Marius