Bug 873277

Just discovered that using the GNOME "Disks" utility, it's possible to add the mount options "acl,user_xattr" in /etc/fstab - but when the FS is mounted, these options won't be used (silently - there's no error message about forbidden mount options in this case).

> It's hard to believe that a tool that's intended to become a standard for
> the Linux desktop doesn't support any native Linux file system

GNOME Disks / udisks2 does support any random filesystem you throw at it. For security-reasons, it just won't accept any random mount option.

As you found out in comment 1, the way it's supposed to work is that if you want non-standard mount options for a device, you simply just add an entry to /etc/fstab for the device in question. That's really all there is to it.

If that's not working then it's not really a udisks2 bug - all that udisksd does is run "mount /dev/sdc1".

> but when the FS is mounted, these options won't be used (silently -
> there's no error message about forbidden mount options in this case).

Care to post the output of

 # mount /dev/sdc1
 # cat /proc/self/mountinfo |grep sdc1

and then (after unmounting the device)

 # udisksctl mount -b /dev/sdc1
 # cat /proc/self/mountinfo |grep sdc1

Also, obviously, please also include the relevant line of /etc/fstab. Thanks!

Sorry, I was confused by the fact that "acl,user_xattr" is apparently now the default and thus not explicitly printed (any more?) by the kernel in /proc/self/mountinfo.

I'm closing this bug. 

Nevertheless it's remarkable that you refer to ext4 as "any random file system you throw at it".

There are lots of ext2/3/4 specific mount options listed in mount(8), and none of them is supported by udisks2. Sure, some of them should be avoided "for security reasons" or in light of the recent "ext4 kernel bug" discussions. But not all.

(In reply to comment #4)
> Nevertheless it's remarkable that you refer to ext4 as "any random file
> system you throw at it".

There is absolutely no need for such sensational language - especially if you do not understand why udisks2 works the way it does and haven't even bothered to read the documentation. Good luck.

(In reply to comment #5)
> (In reply to comment #4)
> > Nevertheless it's remarkable that you refer to ext4 as "any random file
> > system you throw at it".
> There is absolutely no need for such sensational language - especially if

David, that was your own language. I only added the words "Nevertheless it's remarkable that you refer to ext4 as". Not being a native English speaker, that sounded like pretty plain language to me. I apologize if it sounded offensive to you.

> you do not understand why udisks2 works the way it does and haven't even
> bothered to read the documentation. Good luck.

I did bother. I read the available man pages and searched the web. I even read parts of the udisks2 source code. I didn't find any documentation "how udisk2 works" - there were the pretty terse man pages, and there was the API documentation and the source code. Nothing in between. Your Blog post explaining the ideas behind udisks2 (http://davidz25.blogspot.de/2012/03/simpler-faster-better.html) is unfortunately not among the top Google results for "udisks2". I found it only after I had already written this BZ, ironically from a link on a web page which is critical about udisks2.

I usually try to understand exactly what's going wrong before I file a bug. Feel free to search bugzilla for my name if you want to verify that. This time I failed, and I apologize for having wasted your time.