Devlabel and Fstab with usb pendrive

[Darragh Bailey]

> Got the pendrive working a while ago but I noticed some minor
> irratating issues that I can't seem to solve.
> 
> I created a new file /etc/updfstab.conf.local and included into
> /etc/updfstab and now my TinyDisk pendrive mounts on /mnt/flash (which
> is created as required) when you plug it in. Only irk with that is I
> can't control the mount options very well, i.e. I can't set it up so
> that the options noatime and users so that
> 
> 1) access times are not written which use up the finite number of
> writes that flash has (no point in wasting them)
> 2) so that if one user mounts this on the selected machine when
> another user comes along they can umount it.
> 
> By default it can only be umounted by the user who mounted it or root
> and I don't like su'ing just because someone forgot to remove the
> pendrive when they were finished with it.
> 
> Anyone know how to either use updfstab to get it to set these options
> or at least some other way, perhaps auto remounting?
> 
> The other problem I've noticed is that if I have the default file
> browser open in FC2 it appears that fam is running in the background
> and prevents the pendrive from being unmounted until all the file
> browser windows are closed at which point fam exits and you can then
> unmount the pendrive. Problem is this occurs even if the file browser
> was not being used to access the pendrive. I could just be viewing my
> own home directory and it would still block the umounting of the
> pendrive.
> 
> I think I ran into this before but could never work out what the
> problem was when I had a second harddrive of shared partitions that I
> would mount and umount as required. Same issue appeared that sometimes
> even though nothing appeared to be accessing the harddrive it was not
> unmounted.
> 
> Any ideas on how to solve this or is it currently just the way the
> system is supposed to behave until either the file browser or fam is
> modified? I'm not certain but could the use of noatime have an effect
> on this problem?


Never mind, I found out that there's a feature request for this for
about the last year, plus an additional patch. Looks like it hasn't
been included though.


-- 
Darragh Bailey
"Nothing is foolproof to a sufficiently talented fool"