Solve the problem that umount will fail when an opened file isn't closed

Solve the problem that umount will fail when an opened file isn't closed

[colin]

Hi all,
I have implemented an auto-mount & auto-umount facility based on USB
hotplug.
An annoying problem will occur when some process doesn't close its open file
and auto-umount is trying to umount that mount point after usb disk has been
unplugged.
Is there any way to force it to be umounted in this situation?

Regards,
Colin

Re: Solve the problem that umount will fail when an opened file isn't closed

[Jan Engelhardt]

>
>Hi all,
>I have implemented an auto-mount & auto-umount facility based on USB
>hotplug.
>An annoying problem will occur when some process doesn't close its open file
>and auto-umount is trying to umount that mount point after usb disk has been
>unplugged.
>Is there any way to force it to be umounted in this situation?
>

fs/super.c:

    /* Forget any remaining inodes */
    if (invalidate_inodes(sb)) {
        printk("VFS: Busy inodes after unmount of %s. "
           "Self-destruct in 5 seconds.  Have a nice day...\n",
           sb->s_id);
    }

That's what happens if you eject a CD. The box won't explode though.


Jan Engelhardt
-- 

Re: Solve the problem that umount will fail when an opened file isn't closed

[colin]

Hi Jan,
I found that if the process doesn't close the file after usb disk is
unplugged, generic_shutdown_super won't be invoked.
generic_shutdown_super will be invoked right after all opened files are
closed.

Regards,
Colin


----- Original Message ----- 
From: "Jan Engelhardt" <jengelh@linux01.gwdg.de>
To: "colin" <colin@realtek.com.tw>
Cc: <linux-kernel@vger.kernel.org>
Sent: Friday, June 16, 2006 11:34 PM
Subject: Re: Solve the problem that umount will fail when an opened file
isn't closed


> >
> >Hi all,
> >I have implemented an auto-mount & auto-umount facility based on USB
> >hotplug.
> >An annoying problem will occur when some process doesn't close its open
file
> >and auto-umount is trying to umount that mount point after usb disk has
been
> >unplugged.
> >Is there any way to force it to be umounted in this situation?
> >
>
> fs/super.c:
>
>     /* Forget any remaining inodes */
>     if (invalidate_inodes(sb)) {
>         printk("VFS: Busy inodes after unmount of %s. "
>            "Self-destruct in 5 seconds.  Have a nice day...\n",
>            sb->s_id);
>     }
>
> That's what happens if you eject a CD. The box won't explode though.
>
>
> Jan Engelhardt
> --

Re: Solve the problem that umount will fail when an opened file isn't closed

[colin]

Hi Mike,
Do you want to see the hotplug agent code?
I can mail it to you and I would be very grateful if you can give me any
suggestion to improve it.
It will be used on a consumer product and would be better to be very stable.

Regards,
Colin


----- Original Message ----- 
From: "Mike Smullin" <mikesmullin@s161200816.onlinehome.us>
To: "Jan Engelhardt" <jengelh@linux01.gwdg.de>; "colin"
<colin@realtek.com.tw>
Cc: <linux-kernel@vger.kernel.org>
Sent: Saturday, June 17, 2006 12:39 AM
Subject: Re: Solve the problem that umount will fail when an opened file
isn't closed


> Thanks for that, Jan. What a great idea!
>
> > >I have implemented an auto-mount & auto-umount
> > facility based on USB
> > >hotplug.
>
> Colin can you explain how you implemented this? I would like to try it,
too.
>
> Thanks,
> Mike
>
> --
> Mike Smullin
> "The day I come in front of the Gartner audience and say we have a
> better Unix than Linux, that'll be a good day" -- Steve Ballmer
> http://www.mikesmullin.com
>
> --- Jan Engelhardt <jengelh@linux01.gwdg.de> wrote:
>
> > >
> > >Hi all,
> > >I have implemented an auto-mount & auto-umount
> > facility based on USB
> > >hotplug.
> > >An annoying problem will occur when some process
> > doesn't close its open file
> > >and auto-umount is trying to umount that mount
> > point after usb disk has been
> > >unplugged.
> > >Is there any way to force it to be umounted in this
> > situation?
> > >
> >
> > fs/super.c:
> >
> >     /* Forget any remaining inodes */
> >     if (invalidate_inodes(sb)) {
> >         printk("VFS: Busy inodes after unmount of
> > %s. "
> >            "Self-destruct in 5 seconds.  Have a nice
> > day...\n",
> >            sb->s_id);
> >     }
> >
> > That's what happens if you eject a CD. The box won't
> > explode though.
> >
> >
> > Jan Engelhardt
> > --
> > -
> > To unsubscribe from this list: send the line
> > "unsubscribe linux-kernel" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at
> > http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at  http://www.tux.org/lkml/
> >

Re: Solve the problem that umount will fail when an opened file isn't closed

[Mike Smullin]

Thanks for that, Jan. What a great idea!

> >I have implemented an auto-mount & auto-umount
> facility based on USB
> >hotplug.

Colin can you explain how you implemented this? I would like to try it, too.

Thanks,
Mike

--
Mike Smullin
"The day I come in front of the Gartner audience and say we have a
better Unix than Linux, that'll be a good day" -- Steve Ballmer
http://www.mikesmullin.com

--- Jan Engelhardt <jengelh@linux01.gwdg.de> wrote:

> >
> >Hi all,
> >I have implemented an auto-mount & auto-umount
> facility based on USB
> >hotplug.
> >An annoying problem will occur when some process
> doesn't close its open file
> >and auto-umount is trying to umount that mount
> point after usb disk has been
> >unplugged.
> >Is there any way to force it to be umounted in this
> situation?
> >
>
> fs/super.c:
>
>     /* Forget any remaining inodes */
>     if (invalidate_inodes(sb)) {
>         printk("VFS: Busy inodes after unmount of
> %s. "
>            "Self-destruct in 5 seconds.  Have a nice
> day...\n",
>            sb->s_id);
>     }
>
> That's what happens if you eject a CD. The box won't
> explode though.
>
>
> Jan Engelhardt
> --
> -
> To unsubscribe from this list: send the line
> "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at
> http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>