| 1 |
On Thu, 13 Sep 2012 02:50:27 -0400, Walter Dnes wrote: |
| 2 |
|
| 3 |
> > I don't understand, why are you using sudo to run pmount when its core |
| 4 |
> > purpose is to be run by normal users? |
| 5 |
> > |
| 6 |
> > % whatis pmount |
| 7 |
> > pmount (1) - mount arbitrary hotpluggable devices as normal user |
| 8 |
> |
| 9 |
> A normal user can pumount *WHAT THAT SAME USER* has pmounted. |
| 10 |
|
| 11 |
Ah, sorry. I didn't read your post carefully enough and saw pmount when |
| 12 |
you were using pumount. Using sudo for that makes sense. |
| 13 |
|
| 14 |
> Now try |
| 15 |
> for a general solution. If you're the only user on the system, it's |
| 16 |
> probably safe to keep an open xterm logged in to root. The problem is |
| 17 |
> that inserting a USB device sets off a kernel event, that is passed to |
| 18 |
> mdev, which looks for a script name in /etc/mdev.conf. If a script is |
| 19 |
> found that matches the device spec (i.e. sd[a-z].*), e.g. my automount |
| 20 |
> script, then the script is launched *AS ROOT*. Given that root has |
| 21 |
> mounted the device, only root can unmount it. |
| 22 |
|
| 23 |
It's exactly the same problem when udev is used to mount a device, which |
| 24 |
is why I prefer to have a process running as the logged in user doing |
| 25 |
that, whether it be something incorporated into the DE or a separate |
| 26 |
daemon started from ~/.xinitrc. |
| 27 |
|
| 28 |
It gets messy if you are running a multi-seat system, but if only one |
| 29 |
user is running X at a time, it is the cleanest way. Of course, as you |
| 30 |
are having mdev run a script, you could get that script to check which |
| 31 |
user is logged into X and use su to mount it for them, avoiding the |
| 32 |
umount issues. |
| 33 |
|
| 34 |
|
| 35 |
-- |
| 36 |
Neil Bothwick |
| 37 |
|
| 38 |
Like an atheist in a grave: all dressed up and no place to go. |
Archives