Plasma5 + Kernel 4.14 + mtp

DarkVision · 01-07-2018, 03:34 PM

Is anyone here having the same problems? I just updated my desktop to -current with self-compiled packages of KDE/plasma and i can't mount my mobile phone via mtp anymore.

I have tested latest LTS-4.4.x kernel and everything is fine.

I found a bug report here but the affected kernel versions are different to -current/4.14.11.

4.4 fixed it for me, i'll try some older 4.14 releases soon... just interested if other do have similar issues.

gmgf · 01-07-2018, 11:40 PM

Here, work with kdeconnect, work as camera (with mtp, and ptp protocole) not work, as mtp device (kernel-4.14.12, but same with all 4.14)

rworkman · 01-08-2018, 12:11 AM

See if upgrading with this helps: http://slackware.com/~rworkman/gvfs/

EDIT: Probably disregard; that fixes a different problem (I should have read more carefully). However, you are likely experiencing that problem too *if* you're able to mount the phone with mtp...

gmgf · 01-08-2018, 12:42 AM

Quote:

Originally Posted by rworkman

See if upgrading with this helps: http://slackware.com/~rworkman/gvfs/

EDIT: Probably disregard; that fixes a different problem (I should have read more carefully). However, you are likely experiencing that problem too *if* you're able to mount the phone with mtp...

Thanks, Robby , but it seem that not fixe the problem, I've been looking for the solution for a while, but I can not find it, I just tested, on a OpenSuse tumbleweed , and it does not have this problem.

Nille_kungen · 01-12-2018, 12:46 PM

I got the same with KDE4
https://www.linuxquestions.org/quest...nt-4175618627/
I started to bisect but since this is an laptop it will take some time.

DarkVision · 01-12-2018, 03:06 PM

Meanwhile i moved to LTS kernel 4.9.75 and everything is fine.

Nille_kungen · 01-28-2018, 04:12 PM

My bisect showed commit 1455cf8dbfd06aa7651dcfccbadb7a093944ca65 as the first bad.

Code:

# git bisect bad | tee -a /root/bisect.log                 1455cf8dbfd06aa7651dcfccbadb7a093944ca65 is the first bad commit
commit 1455cf8dbfd06aa7651dcfccbadb7a093944ca65
Author: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Date:   Wed Jul 19 17:24:30 2017 -0700

    driver core: emit uevents when device is bound to a driver
    
    There are certain touch controllers that may come up in either normal
    (application) or boot mode, depending on whether firmware/configuration is
    corrupted when they are powered on. In boot mode the kernel does not create
    input device instance (because it does not necessarily know the
    characteristics of the input device in question).
    
    Another number of controllers does not store firmware in a non-volatile
    memory, and they similarly need to have firmware loaded before input device
    instance is created. There are also other types of devices with similar
    behavior.
    
    There is a desire to be able to trigger firmware loading via udev, but it
    has to happen only when driver is bound to a physical device (i2c or spi).
    These udev actions can not use ADD events, as those happen too early, so we
    are introducing BIND and UNBIND events that are emitted at the right
    moment.
    
    Also, many drivers create additional driver-specific device attributes
    when binding to the device, to provide userspace with additional controls.
    The new events allow userspace to adjust these driver-specific attributes
    without worrying that they are not there yet.
    
    Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
    Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

:040000 040000 fef96678b0a5ea8f89c41f006a3ccd013b4a36f4 34b72074bd6fab60bea1ec8066cf5ae9c600b37c M    drivers
:040000 040000 5657f6e76558ed4e14f17edabc6c602b93011d6d 1dbac2e41568c1638ea059fc8bc1afe8878353b8 M    include
:040000 040000 88854205d5255fbec438594cb3539e01c3610f82 54ad1fcb19ca64af9660e170ae0d7af186bd51ad M    lib

Quote:

Originally Posted by rworkman

See if upgrading with this helps: http://slackware.com/~rworkman/gvfs/

EDIT: Probably disregard; that fixes a different problem (I should have read more carefully). However, you are likely experiencing that problem too *if* you're able to mount the phone with mtp...

I think that only helps udev users but not eudev users like slackware.
My understanding is that things was changed in the 4.14 kernel that needed patches in udev and systemd, i saw it discussed on the kernel mailing list when i searched for commit 1455cf8dbfd06aa7651dcfccbadb7a093944ca65 maybe eudev hasn't got all that love yet or the gvfs patches need some tweaking for eudev.
This isn't my area of expertise so please view this as an fools answer since i don't understand the inner working of this stuff.

But the fact remains that it does work with thunar or in xfce.

gmgf · 01-29-2018, 02:59 AM

The problem is already reported on manjaro, and it seem that manjaro use only systemd now, but....

Nille_kungen · 01-29-2018, 04:19 AM

Quote:

Originally Posted by gmgf

The problem is already reported on manjaro, and it seem that manjaro use only systemd now, but....

Yes but that problems was also solved by an update, i think it was the gvfs patches (and maybe some others) that solved it for udev/systemd users, for eudev users it's still not solved.
But there is another Manjaro report of it not working....

gmgf · 01-29-2018, 06:50 AM

I think, i have found the problem, it's a libgphoto2 build broblem, I will post the solution in 'Request for -current' thread, during the day

gmgf · 01-29-2018, 08:41 AM

If you want to test my solution:

https://www.linuxquestions.org/quest...ml#post5812835

Nille_kungen · 01-29-2018, 08:46 AM

Quote:

Originally Posted by gmgf

I think, i have found the problem, it's a libgphoto2 build broblem, I will post the solution in 'Request for -current' thread, during the day

I tried the patch and it sort of works.

Code:

diff -ur a/libgphoto2.SlackBuild b/libgphoto2.SlackBuild
--- a/libgphoto2.SlackBuild     2017-01-05 02:59:57.000000000 +0100
+++ b/libgphoto2.SlackBuild     2018-01-29 14:22:43.000000000 +0100
@@ -111,13 +111,15 @@
 export LD_LIBRARY_PATH CAMLIBS LIBDIR
 # Generate udev rules -- for udev <= 200
 mkdir -p $PKG/lib/udev/rules.d
-$PKG/usr/bin/print-camera-list udev-rules version 175 mode 0660 owner root group plugdev \
-  > $PKG/lib/udev/rules.d/40-libgphoto2.rules
+$PKG/usr/bin/print-camera-list udev-rules version 201 mode 0660 owner root group plugdev \
+  > $PKG/lib/udev/rules.d/40-gphoto.rules
 # For udev >= 201
 mkdir -p $PKG/lib/udev/hwdb.d/
-$PKG/usr/bin/print-camera-list hwdb > $PKG/lib/udev/hwdb.d/20-gphoto.conf
+$PKG/usr/bin/print-camera-list hwdb > $PKG/lib/udev/hwdb.d/20-gphoto.hwdb
 unset LD_LIBRARY_PATH CAMLIBS LIBDIR   # Unset these just in case
 
+
+
 # Compress and link manpages, if any:
 if [ -d $PKG/usr/man ]; then
   ( cd $PKG/usr/man

I can "mount" the mtp device in dolphin so it works that much.
I can't open the device with kde pop up the icon that looks like an usb logo (localized kde so i can't see what it's called in english)
MTP keeps dying on me and often have write problems like it has no permission to write like if i try to edit an text file..
With the patch i see my device but also an no named ghost device that shouldn't be there.

Nille_kungen · 01-29-2018, 09:23 AM

MTP keeps dying for me what ever i do, but sometimes it works, it's very much a hit or miss.

gmgf · 01-29-2018, 09:37 AM

I agree, it seem my solution solve a part of the problem, but not entirely

gmgf · 01-30-2018, 08:00 AM

Quote:

Originally Posted by Nille_kungen

MTP keeps dying for me what ever i do, but sometimes it works, it's very much a hit or miss.

With my Moto X 2:

I have tested here, with one 'openSUSE Tumbleweed', and with one 'Manjaro 17.1.3 KDE Edition', (these two have systemd and the same kernel-4.14.15) , the MTP device, have exactly the same result ,and the same comportment, that i have, with it, on slack-current with dolphin and plasma5 and with my libmtp rebuild package.