Need help debugging

[Tim]

Hi,

Since a couple of days, I'm a proud owner of a ColorHug2, but I haven't
been able to calibrate my screen so far. I'm not sure whether I'm
hitting a software bug or a hardware one. I hope that you could help me
with some pointers.

I have now tried to calibrate two different monitors on two different
computers. On my desktop, I run into the following problems, during
calibration using the GNOME control center. At about 60–70% into the
calibration process I reliably get an error message that calibration
failed. Running the command from the command line gave me the following
hints.

[liveuser@localhost ~]$ gnome-control-center --verbose color
[...snip...]
(gnome-control-center:2740): color-cc-panel-DEBUG: selected device is: xrandr-Samsung Electric Company-SA300/SA350-H4XC200641
[...snip...]
(gnome-control-center:2740): color-cc-panel-WARNING **: calibration failed with code 1: Failed to create file '/tmp/colord-session-ID6PUY/Samsung SA300/SA350 (medium) 2017-01-26 22-11-14 colorhug2.ti3.YL2PUY': No such file or directory
(gnome-control-center:2740): color-cc-panel-WARNING **: failed to start calibrate: failed to calibrate

This error message made me believe that it might be a problem that a
slash occurs in the monitor name. However, looking at `journalctl` gave
me a completely different error message.

Jan 26 17:11:58 localhost org.freedesktop.ColorHelper[1095]: steps were set as [ 3, 97, -1 ] but should have been: [ 2, 98, -1 ] at cd-main.c:475
Jan 26 17:12:19 localhost org.freedesktop.ColorHelper[1095]: steps were set as [ 3, 97, -1 ] but should have been: [ 4, 96, -1 ] at cd-main.c:475
Jan 26 17:12:43 localhost org.freedesktop.ColorHelper[1095]: steps were set as [ 3, 97, -1 ] but should have been: [ 3, 97, -1 ] at cd-main.c:475
Jan 26 17:13:07 localhost org.freedesktop.ColorHelper[1095]: steps were set as [ 3, 97, -1 ] but should have been: [ 2, 98, -1 ] at cd-main.c:475
Jan 26 17:13:25 localhost org.freedesktop.ColorHelper[1095]: steps were set as [ 3, 97, -1 ] but should have been: [ 3, 97, -1 ] at cd-main.c:475
Jan 26 17:13:42 localhost org.freedesktop.ColorHelper[1095]: steps were set as [ 3, 97, -1 ] but should have been: [ 3, 97, -1 ] at cd-main.c:475
Jan 26 17:13:59 localhost org.freedesktop.ColorHelper[1095]: steps were set as [ 3, 97, -1 ] but should have been: [ 3, 97, -1 ] at cd-main.c:475
Jan 26 17:14:28 localhost org.freedesktop.ColorHelper[1095]: steps were set as [ 3, 97, -1 ] but should have been: [ 3, 97, -1 ] at cd-main.c:475
Jan 26 17:15:30 localhost org.freedesktop.ColorHelper[1095]: steps were set as [ 3, 97, -1 ] but should have been: [ 3, 97, -1 ] at cd-main.c:475
Jan 26 17:16:22 localhost org.freedesktop.ColorHelper[1095]: steps were set as [ 3, 97, -1 ] but should have been: [ 4, 96, -1 ] at cd-main.c:475
Jan 26 17:16:22 localhost org.freedesktop.ColorHelper[1095]: child is at 9/10 steps and parent done [cd-main.c:788]
Jan 26 17:16:22 localhost org.freedesktop.ColorHelper[1095]: 2) cd-main.c:1341 (0/3)
Jan 26 17:16:22 localhost org.freedesktop.ColorHelper[1095]: 1) cd-main.c:682 (2/5)
Jan 26 17:16:22 localhost org.freedesktop.ColorHelper[1095]: 0) cd-main.c:766 (9/10)
Jan 26 17:16:23 localhost org.freedesktop.ColorHelper[1095]: child is at 9/10 steps and parent done [cd-main.c:800]
Jan 26 17:16:23 localhost org.freedesktop.ColorHelper[1095]: 2) cd-main.c:1341 (0/3)
Jan 26 17:16:23 localhost org.freedesktop.ColorHelper[1095]: 1) cd-main.c:682 (2/5)
Jan 26 17:16:23 localhost org.freedesktop.ColorHelper[1095]: 0) cd-main.c:766 (9/10)
Jan 26 17:16:23 localhost org.freedesktop.ColorHelper[1095]: child is at 9/10 steps and parent done [cd-main.c:841]
Jan 26 17:16:23 localhost org.freedesktop.ColorHelper[1095]: 2) cd-main.c:1341 (0/3)
Jan 26 17:16:23 localhost org.freedesktop.ColorHelper[1095]: 1) cd-main.c:682 (2/5)
Jan 26 17:16:23 localhost org.freedesktop.ColorHelper[1095]: 0) cd-main.c:766 (9/10)
Jan 26 17:16:23 localhost org.freedesktop.ColorHelper[1095]: child is at 2/5 steps and parent done [cd-main.c:1369]
Jan 26 17:16:23 localhost org.freedesktop.ColorHelper[1095]: 1) cd-main.c:1341 (0/3)
Jan 26 17:16:23 localhost org.freedesktop.ColorHelper[1095]: 0) cd-main.c:682 (2/5)
Jan 26 17:16:23 localhost org.freedesktop.ColorHelper[1095]: 0) cd-main.c:1341 (0/3)
Jan 26 17:16:23 localhost org.freedesktop.ColorHelper[1095]: (colord-session:2747): Cd-CRITICAL **: percentage should not go down from 65 to 0 on 0x803ee930!

On a notebook I got no error when I tried to calibrate with the “low”
option, but I couldn't get it working with “medium” or “high.”

(gnome-control-center:2361): color-cc-panel-DEBUG: selected device is: xrandr-eDP1
(gnome-control-center:2361): color-cc-panel-WARNING **: calibration failed with code 5: Gamma correction table was non-monotonic: 0.000000,0.000000,0.000000 0.203125,0.138246,0.019933 0.209375,0.145243,0.021115 0.250000,0.180364,0.044173 0.234375,0.165486,0.035980 0.318750,0.238107,0.084038 0.334375,0.254478,0.088345 0.375000,0.289600,0.117653 0.431250,0.330971,0.153211 0.484375,0.372343,0.188768 0.540625,0.413714,0.221201 0.590625,0.461336,0.253633 0.646875,0.512082,0.295441 0.671875,0.537828,0.315373 0.721875,0.579200,0.347806 0.778125,0.620571,0.383364 0.821875,0.661943,0.428296 0.837500,0.678314,0.438854 0.906250,0.744685,0.502536 0.953125,0.786057,0.550594 1.000000,0.827428,0.586151

Trying again results in a reliably reproducible segfault.

(gnome-control-center:2361): color-cc-panel-CRITICAL **: cc_color_calibrate_setup: assertion 'calibrate->priv->device_kind != CD_SENSOR_CAP_UNKNOWN' failed
Segmentation fault (core dumped)

Does anyone know what might be going on or how to debug this further and
to which packages file the bugs against?


Cheers,

Tim

PS: For some reason my first email didn't get through to the mailing
list. I'm not sure why.

[berend.de...@gmail.com]

I have an identical problem.  Calibration on medium or high quality on two different monitors results in "Gamma correction table was non-monotonic" after about 65%.  Calibration on low quality works.  It is assumed that the non-monotonic message is the root cause.

Running again also produces a segfault; however I believe that's a bug in gnome-control-center.  This now has BZ https://bugzilla.redhat.com/show_bug.cgi?id=1567496