Plasma-mobile on cubox-i

[Rainer Dorsch]

Hi Bhushan and Matthias,

I closely followed https://puri.sm/posts/running-plasma-mobile-on-an-imx6-test-board/

but after starting

kwin_wayland --drm --xwayland plasma-phone

all I got was a segmentation fault. I also tried

kwin_wayland --drm --xwayland plasma-phone

but same result.

Do you have any idea what I might have done wrong? Are there any logs
available (first time I am running wayland and not an xserver)?

Kind regards and many thanks
Rainer

[Paul Wise]

Post by Rainer Dorsch
Do you have any idea what I might have done wrong? Are there any logs
available (first time I am running wayland and not an xserver)?

Sounds like a bug in kwin, please install debug symbols (run
find-dbgsym-packages from debian-goodies on the core dump) and report
a bug.

[Rainer Dorsch]

Hi Paul,

many thanks for your reply.

I installed the dbgsym packages, but when running inside gdb

***@xbian:~$ gdb kwin_wayland
GNU gdb (Debian 7.12-6+b1) 7.12.0.20161007-git
Copyright (C) 2016 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law. Type "show copying"
and "show warranty" for details.
This GDB was configured as "arm-linux-gnueabihf".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>.
Find the GDB manual and other documentation resources online at:
<http://www.gnu.org/software/gdb/documentation/>.
For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from kwin_wayland...Reading symbols from /usr/lib/
debug/.build-id/91/494219560dcdb6f7874172e03bcac638cca773.debug...done.
done.
(gdb) set pagination 0
(gdb) run --drm --xwayland plasma-phone
Starting program: /usr/bin/kwin_wayland --drm --xwayland plasma-phone
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/arm-linux-gnueabihf/libthread_db.so.1".

it consumes a large amout of memory and I do not reach the segfault but I also
do not see any UI coming up :-/

Is there something I did wrong or could do more efficiently? Would running
standalone and then loading the core be more memory efficient? If yes, how do I
do run standalone with the dbgsym packages? Installing them seems not
sufficient....

Thanks
Rainer

Post by Paul Wise

Post by Rainer Dorsch
Do you have any idea what I might have done wrong? Are there any logs
available (first time I am running wayland and not an xserver)?

Sounds like a bug in kwin, please install debug symbols (run
find-dbgsym-packages from debian-goodies on the core dump) and report
a bug.

[Paul Wise]

Post by Rainer Dorsch
Is there something I did wrong or could do more efficiently? Would running
standalone and then loading the core be more memory efficient? If yes, how do I
do run standalone with the dbgsym packages? Installing them seems not
sufficient....

I think it would be indeed better to let kwin crash and then load the core dump:

gdb -batch -n -ex 'set pagination off' -ex bt -ex 'thread apply all bt
full' --core $core /usr/bin/kwin_wayland

BTW, you might want to consider using one of these to get auto-core-dumps:

$ aptitude show core-dump-handler
No candidate version found for core-dump-handler
Package: core-dump-handler
State: not a real package
Provided by: apport (2.20.4-3), corekeeper (1.6), minicoredumper
(2.0.0-3), systemd-coredump (236-1)

[Rainer Dorsch]

Hi Paul,

thanks again for your answer.

Post by Paul Wise

Post by Rainer Dorsch
Is there something I did wrong or could do more efficiently? Would running
standalone and then loading the core be more memory efficient? If yes, how
do I do run standalone with the dbgsym packages? Installing them seems
not sufficient....

gdb -batch -n -ex 'set pagination off' -ex bt -ex 'thread apply all bt
full' --core $core /usr/bin/kwin_wayland

This also runs out of memory (1G+1G swap) even though the core file is not that
big:

***@xbian:~$ ls -l core
-rw------- 1 rd rd 29470720 Dec 29 12:49 core
***@xbian:~$

:-/

Post by Paul Wise
$ aptitude show core-dump-handler
No candidate version found for core-dump-handler
Package: core-dump-handler
State: not a real package
Provided by: apport (2.20.4-3), corekeeper (1.6), minicoredumper
(2.0.0-3), systemd-coredump (236-1)

I need to investigate this later, ....do you know if these help to reduce the
memory consumption for gdb?

Thanks
Rainer

[Paul Wise]

Post by Rainer Dorsch
This also runs out of memory (1G+1G swap)

Maybe just bt the current thread instead of all of them.

Post by Rainer Dorsch
I need to investigate this later, ....do you know if these help to reduce the
memory consumption for gdb?

No idea, but this may help:

maint set dwarf max-cache-age 0

https://sourceware.org/gdb/onlinedocs/gdb/Maintenance-Commands.html

[Rainer Dorsch]

Post by Paul Wise

Post by Rainer Dorsch
This also runs out of memory (1G+1G swap)

Maybe just bt the current thread instead of all of them.

Post by Rainer Dorsch
I need to investigate this later, ....do you know if these help to reduce
the memory consumption for gdb?

maint set dwarf max-cache-age 0
https://sourceware.org/gdb/onlinedocs/gdb/Maintenance-Commands.html

I found that the kernel parameter cma=256M makes the segfault going away.

Nevertheless plasma did not come up properly. I tried gnome with gdm, that
came up flawless. It was slow, but that was probably due to my old SD card.

Rainer