Discussion:
Heavily disappointed about openCL
(too old to reply)
Bernhard
2018-02-20 19:01:33 UTC
Permalink
Hi,

I ran into another problem with my new GeForce GTX 1050 Ti in my system
regarding openCL.
Every now and then I see a message about darktable finding problems with
openCL and disabling it "for this session".
I then have to reboot the complete machine to get this working again -
but not for a long time and I have the same message again.


System: Host: benutzer Kernel: 4.13.0-32-generic x86_64 (64 bit gcc:
5.4.0)
Desktop: Cinnamon 3.4.6 (Gtk 3.18.9-1ubuntu3.3)
Distro: Linux Mint 18.2 Sonya
Machine: Mobo: ASUSTeK model: P8Z77-M v: Rev 1.xx
Bios: American Megatrends v: 2003 date: 05/09/2013
CPU: Quad core Intel Core i5-3570 (-MCP-) cache: 6144 KB
flags: (lm nx sse sse2 sse3 sse4_1 sse4_2 ssse3 vmx) bmips:
27280
clock speeds: max: 3800 MHz 1: 3410 MHz 2: 3410 MHz 3: 3410 MHz
4: 3410 MHz
Graphics: Card: Intel Xeon E3-1200 v2/3rd Gen Core processor Graphics
Controller
bus-ID: 00:02.0
Display Server: X.Org 1.18.4 driver: intel
Resolution: ***@59.95hz
GLX Renderer: Mesa DRI Intel Ivybridge Desktop
GLX Version: 3.0 Mesa 17.2.4 Direct Rendering: Yes

~ $ darktable -d opencl

reports the following while opening some pictures in darkroom mode:

(...)
[pixelpipe_process] [thumbnail] using device 0
[pixelpipe_process] [full] using device -1
[opencl_pixelpipe] couldn't copy image to opencl device for module
rawprepare
[opencl_pixelpipe] could not run module 'rawprepare' on gpu. falling
back to cpu path
[opencl_pixelpipe (b)] late opencl error detected while copying back to
cpu buffer: -5
[pixelpipe_process] [thumbnail] falling back to cpu path
[pixelpipe_process] [full] using device 0
[pixelpipe_process] [preview] using device -1
[pixelpipe_process] [full] using device 0
[pixelpipe_process] [preview] using device -1
[pixelpipe_process] [thumbnail] using device 0
[opencl_pixelpipe] couldn't copy image to opencl device for module
rawprepare
[opencl_pixelpipe] could not run module 'rawprepare' on gpu. falling
back to cpu path
[opencl_pixelpipe (b)] late opencl error detected while copying back to
cpu buffer: -5
[opencl] frequent opencl errors encountered; disabling opencl for this
session!
[pixelpipe_process] [thumbnail] falling back to cpu path
[pixelpipe_process] [full] using device -1
[pixelpipe_process] [full] using device -1
[pixelpipe_process] [preview] using device -1

Does anyone have an idea what I could look for?
--
regards
Bernhard

https://www.bilddateien.de


____________________________________________________________________________
darktable user mailing list
to unsubscribe send a mail to darktable-user+***@lists.darktable.org
Ulrich Pegelow
2018-02-20 19:09:45 UTC
Permalink
That's an out-of-resources problem on your graphics card. Try to
increase darktable's config variable opencl_memory_headroom (in file
darktablerc) to something like 400.

Please also make sure that no other application uses substantial amounts
of GPU memory. You can use program nvidia-smi to find out. Here on a
GTX1060 with 6GB it looks like below, indicating that only about 200MB
of GPU memory are in use by the system or any other apps.

Best wishes

Ulrich


Tue Feb 20 20:07:11 2018
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 384.111 Driver Version: 384.111
|
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile
Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util
Compute M. |
|===============================+======================+======================|
| 0 GeForce GTX 106... Off | 00000000:01:00.0 On |
N/A |
| 4% 50C P8 10W / 120W | 182MiB / 6065MiB | 0%
Default |
+-------------------------------+----------------------+----------------------+


+-----------------------------------------------------------------------------+
| Processes: GPU
Memory |
| GPU PID Type Process name Usage
|
|=============================================================================|
| 0 3534 G /usr/bin/X
107MiB |
| 0 4201 G kwin_x11
16MiB |
| 0 4216 G /usr/bin/krunner
1MiB |
| 0 4217 G /usr/bin/plasmashell
51MiB |
| 0 4261 G /usr/bin/kgpg
2MiB |
+-----------------------------------------------------------------------------+
Post by Bernhard
Hi,
I ran into another problem with my new GeForce GTX 1050 Ti in my system
regarding openCL.
Every now and then I see a message about darktable finding problems with
openCL and disabling it "for this session".
I then have to reboot the complete machine to get this working again -
but not for a long time and I have the same message again.
5.4.0)
           Desktop: Cinnamon 3.4.6 (Gtk 3.18.9-1ubuntu3.3)
           Distro: Linux Mint 18.2 Sonya
Machine:   Mobo: ASUSTeK model: P8Z77-M v: Rev 1.xx
           Bios: American Megatrends v: 2003 date: 05/09/2013
CPU:       Quad core Intel Core i5-3570 (-MCP-) cache: 6144 KB
27280
           clock speeds: max: 3800 MHz 1: 3410 MHz 2: 3410 MHz 3: 3410 MHz
           4: 3410 MHz
Graphics:  Card: Intel Xeon E3-1200 v2/3rd Gen Core processor Graphics
Controller
           bus-ID: 00:02.0
           Display Server: X.Org 1.18.4 driver: intel
           GLX Renderer: Mesa DRI Intel Ivybridge Desktop
           GLX Version: 3.0 Mesa 17.2.4 Direct Rendering: Yes
~ $ darktable -d opencl
(...)
[pixelpipe_process] [thumbnail] using device 0
[pixelpipe_process] [full] using device -1
[opencl_pixelpipe] couldn't copy image to opencl device for module
rawprepare
[opencl_pixelpipe] could not run module 'rawprepare' on gpu. falling
back to cpu path
[opencl_pixelpipe (b)] late opencl error detected while copying back to
cpu buffer: -5
[pixelpipe_process] [thumbnail] falling back to cpu path
[pixelpipe_process] [full] using device 0
[pixelpipe_process] [preview] using device -1
[pixelpipe_process] [full] using device 0
[pixelpipe_process] [preview] using device -1
[pixelpipe_process] [thumbnail] using device 0
[opencl_pixelpipe] couldn't copy image to opencl device for module
rawprepare
[opencl_pixelpipe] could not run module 'rawprepare' on gpu. falling
back to cpu path
[opencl_pixelpipe (b)] late opencl error detected while copying back to
cpu buffer: -5
[opencl] frequent opencl errors encountered; disabling opencl for this
session!
[pixelpipe_process] [thumbnail] falling back to cpu path
[pixelpipe_process] [full] using device -1
[pixelpipe_process] [full] using device -1
[pixelpipe_process] [preview] using device -1
Does anyone have an idea what I could look for?
Bernhard
2018-02-20 19:45:30 UTC
Permalink
Hi Ulrich,

thanks for quick response. I found this
http://darktable-devel.narkive.com/K9FwaE0y/opencl-problem in the
meantime and already tried a value of 500 (does this have side affects?)
and could not reproduce the problem til now.
darktable reports

[pixelpipe_process] [thumbnail] using device 0
[pixelpipe_process] [full] using device -1
[pixelpipe_process] [full] using device 0
[pixelpipe_process] [preview] using device -1
[pixelpipe_process] [full] using device 0
[pixelpipe_process] [full] using device 0
[pixelpipe_process] [preview] using device -1
[pixelpipe_process] [thumbnail] using device 0
wait time 0,144355s
try- wait time 0,143256s
wait time 0,165599s
try- wait time 0,164212s
[pixelpipe_process] [full] using device 0
[pixelpipe_process] [preview] using device -1
[pixelpipe_process] [thumbnail] using device 0
[pixelpipe_process] [full] using device -1
[pixelpipe_process] [full] using device 0
[pixelpipe_process] [preview] using device -1

when doing the same thing (switching between images in darkroom mode).
Does that _wait time_ indicate another problem?

And thanks for this

~ $ nvidia-smi
Tue Feb 20 20:29:46 2018
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 384.111                Driver Version:
384.111                   |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile
Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util 
Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX 105...  Off  | 00000000:01:00.0  On
|                  N/A |
| 20%   42C    P8   ERR! /  75W |    269MiB /  4035MiB |      0%     
Default |
+-------------------------------+----------------------+----------------------+

+-----------------------------------------------------------------------------+
| Processes: GPU Memory |
|  GPU       PID   Type   Process name                            
Usage      |
|=============================================================================|
|    0      1363      G /usr/lib/xorg/Xorg                          
148MiB |
|    0      2005      G cinnamon                                     
74MiB |
|    0      3425      C darktable                                    
43MiB |
+-----------------------------------------------------------------------------+

I didn't know about it.

Does this still apply?
Ulrich Pegelow
3 years agoPermalink <http://narkive.com/K9FwaE0y.7>
Forgot to mention. If you have other applications which consume
significant amounts of GPU memory this could also cause OpenCL in
darktable to fail. Unfortunately there is no way to find out at any time
which amount of GPU memory is still available. Therefore darktable
assumes it can have all minus the amount given in opencl_memory_headroom.
So - if darktable fails => I should see high volumes of GPU memory
usage? And also the apps that cause the problem, I assume?
I'll watch this. Thank you.
--
regards
Bernhard

https://www.bilddateien.de
That's an out-of-resources problem on your graphics card. Try to
increase darktable's config variable opencl_memory_headroom (in file
darktablerc) to something like 400.
Please also make sure that no other application uses substantial
amounts of GPU memory. You can use program nvidia-smi to find out.
Here on a GTX1060 with 6GB it looks like below, indicating that only
about 200MB of GPU memory are in use by the system or any other apps.
Best wishes
Ulrich
Tue Feb 20 20:07:11 2018
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 384.111                Driver Version: 384.111      |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile
Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util
Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX 106...  Off  | 00000000:01:00.0  On |  N/A |
|  4%   50C    P8    10W / 120W |    182MiB /  6065MiB |      0%
Default |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: GPU Memory |
|  GPU       PID   Type   Process name                            
Usage      |
|=============================================================================|
|    0      3534      G   /usr/bin/X 107MiB |
|    0      4201      G   kwin_x11 16MiB |
|    0      4216      G   /usr/bin/krunner 1MiB |
|    0      4217      G   /usr/bin/plasmashell 51MiB |
|    0      4261      G   /usr/bin/kgpg 2MiB |
+-----------------------------------------------------------------------------+
Post by Bernhard
Hi,
I ran into another problem with my new GeForce GTX 1050 Ti in my
system regarding openCL.
Every now and then I see a message about darktable finding problems
with openCL and disabling it "for this session".
I then have to reboot the complete machine to get this working again
- but not for a long time and I have the same message again.
System:    Host: benutzer Kernel: 4.13.0-32-generic x86_64 (64 bit
gcc: 5.4.0)
           Desktop: Cinnamon 3.4.6 (Gtk 3.18.9-1ubuntu3.3)
           Distro: Linux Mint 18.2 Sonya
Machine:   Mobo: ASUSTeK model: P8Z77-M v: Rev 1.xx
           Bios: American Megatrends v: 2003 date: 05/09/2013
CPU:       Quad core Intel Core i5-3570 (-MCP-) cache: 6144 KB
           flags: (lm nx sse sse2 sse3 sse4_1 sse4_2 ssse3 vmx)
bmips: 27280
3410 MHz
           4: 3410 MHz
Graphics:  Card: NVIDIA Device 1c82 bus-ID: 01:00.0
fbdev,vesa,nouveau)
           GLX Renderer: GeForce GTX 1050 Ti/PCIe/SSE2
           GLX Version: 4.5.0 NVIDIA 384.111 Direct Rendering: Yes
~ $ darktable -d opencl
(...)
[pixelpipe_process] [thumbnail] using device 0
[pixelpipe_process] [full] using device -1
[opencl_pixelpipe] couldn't copy image to opencl device for module
rawprepare
[opencl_pixelpipe] could not run module 'rawprepare' on gpu. falling
back to cpu path
[opencl_pixelpipe (b)] late opencl error detected while copying back
to cpu buffer: -5
[pixelpipe_process] [thumbnail] falling back to cpu path
[pixelpipe_process] [full] using device 0
[pixelpipe_process] [preview] using device -1
[pixelpipe_process] [full] using device 0
[pixelpipe_process] [preview] using device -1
[pixelpipe_process] [thumbnail] using device 0
[opencl_pixelpipe] couldn't copy image to opencl device for module
rawprepare
[opencl_pixelpipe] could not run module 'rawprepare' on gpu. falling
back to cpu path
[opencl_pixelpipe (b)] late opencl error detected while copying back
to cpu buffer: -5
[opencl] frequent opencl errors encountered; disabling opencl for
this session!
[pixelpipe_process] [thumbnail] falling back to cpu path
[pixelpipe_process] [full] using device -1
[pixelpipe_process] [full] using device -1
[pixelpipe_process] [preview] using device -1
Does anyone have an idea what I could look for?
____________________________________________________________________________
darktable user mailing list
to unsubscribe send a mail to
Peter Mc Donough
2018-02-20 19:46:53 UTC
Permalink
Post by Ulrich Pegelow
That's an out-of-resources problem on your graphics card. Try to
increase darktable's config variable opencl_memory_headroom (in file
darktablerc) to something like 400.
Shouldn't that be configured in darktable GUI settings options?

e.g. With graphics cards with less than "whatever" memory set headroom
to "whatever" and avoid running other graphics card loads.

cu
Peter
Ulrich Pegelow
2018-02-20 20:57:31 UTC
Permalink
With today's typical amount of graphics cards memory we should probably
increase the default setting of that parameter to maybe 400 or 450.

In "the old days" when we only had like 1GB a too high value would have
forced darktable to go into useless tiling, but with more GPU memory
that's really not an issue any longer.

Ulrich
Post by Peter Mc Donough
Post by Ulrich Pegelow
That's an out-of-resources problem on your graphics card. Try to
increase darktable's config variable opencl_memory_headroom (in file
darktablerc) to something like 400.
Shouldn't that be configured in darktable GUI settings options?
e.g. With graphics cards with less than "whatever" memory set headroom
to "whatever" and avoid running other graphics card loads.
cu
Peter
Michael Rasmussen
2018-02-20 21:57:24 UTC
Permalink
On Tue, 20 Feb 2018 21:57:31 +0100
With today's typical amount of graphics cards memory we should probably increase the default setting of that parameter to maybe 400 or 450.
With 2GB of GPU RAM I have found the sweet spot to be
opencl_memory_headroom=400
opencl_memory_requirement=768
--
Hilsen/Regards
Michael Rasmussen

Get my public GnuPG keys:
michael <at> rasmussen <dot> cc
http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E
mir <at> datanom <dot> net
http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C
mir <at> miras <dot> org
http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917
--------------------------------------------------------------
/usr/games/fortune -es says:
The secret of healthy hitchhiking is to eat junk food.
Bernhard
2018-02-21 06:35:32 UTC
Permalink
Post by Michael Rasmussen
On Tue, 20 Feb 2018 21:57:31 +0100
With today's typical amount of graphics cards memory we should probably increase the default setting of that parameter to maybe 400 or 450.
With 2GB of GPU RAM I have found the sweet spot to be
opencl_memory_headroom=400
opencl_memory_requirement=768
Just had the issue again with

~ $ nvidia-smi
Wed Feb 21 07:17:20 2018
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 384.111                Driver Version:
384.111                   |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile
Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util
Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX 105...  Off  | 00000000:01:00.0  On
|                  N/A |
| 20%   30C    P8   ERR! /  75W |    328MiB /  4035MiB | 5%      Default |
+-------------------------------+----------------------+----------------------+

+-----------------------------------------------------------------------------+
| Processes: GPU Memory |
|  GPU       PID   Type   Process name Usage      |
|=============================================================================|
|    0      1363      G /usr/lib/xorg/Xorg                          
207MiB |
|    0      2005      G cinnamon                                     
57MiB |
|    0      3425      C darktable                                    
43MiB |
+-----------------------------------------------------------------------------+

I exported several pictures and toggled nvidia-smi in another terminal
to see if GPU-Memory usage will increase => nothing the like.

darktable -d reported

[pixelpipe_process] [thumbnail] using device 0
[pixelpipe_process] [thumbnail] using device 0
wait time 0,109648s
wait time 0,110881s
[pixelpipe_process] [thumbnail] using device 0
[pixelpipe_process] [thumbnail] using device 0
[pixelpipe_process] [export] using device 0
[opencl_pixelpipe] couldn't copy image to opencl device for module
rawprepare
[opencl_pixelpipe] could not run module 'rawprepare' on gpu. falling
back to cpu path
[opencl_pixelpipe (b)] late opencl error detected while copying back to
cpu buffer: -5
[pixelpipe_process] [export] falling back to cpu path
[export_job] exported to
`/home/username/path-to-export/2018-02-21_07-16-45/20180218_n5x0268.jpg'
[pixelpipe_process] [export] using device 0
[opencl_pixelpipe] couldn't copy image to opencl device for module
rawprepare
[opencl_pixelpipe] could not run module 'rawprepare' on gpu. falling
back to cpu path
[opencl_pixelpipe (b)] late opencl error detected while copying back to
cpu buffer: -5
[pixelpipe_process] [export] falling back to cpu path
[export_job] exported to
`/home/username/path-to-export/2018-02-21_07-16-45/20180218_n5x0266.jpg'
[pixelpipe_process] [export] using device 0
[opencl_pixelpipe] couldn't copy image to opencl device for module
rawprepare
[opencl_pixelpipe] could not run module 'rawprepare' on gpu. falling
back to cpu path
[opencl_pixelpipe (b)] late opencl error detected while copying back to
cpu buffer: -5
[pixelpipe_process] [export] falling back to cpu path
[export_job] exported to
`/home/username/path-to-export/2018-02-21_07-16-45/20180218_n5x0252.jpg'
[pixelpipe_process] [export] using device 0
[opencl_pixelpipe] couldn't copy image to opencl device for module
rawprepare
[opencl_pixelpipe] could not run module 'rawprepare' on gpu. falling
back to cpu path
[opencl_pixelpipe (b)] late opencl error detected while copying back to
cpu buffer: -5
[pixelpipe_process] [export] falling back to cpu path
[export_job] exported to
`/home/username/path-to-export/2018-02-21_07-16-45/20180218_n5x0246.jpg'
[pixelpipe_process] [export] using device 0
[opencl_pixelpipe] couldn't copy image to opencl device for module
rawprepare
[opencl_pixelpipe] could not run module 'rawprepare' on gpu. falling
back to cpu path
[opencl_pixelpipe (b)] late opencl error detected while copying back to
cpu buffer: -5
[opencl] frequent opencl errors encountered; disabling opencl for this
session!
[pixelpipe_process] [export] falling back to cpu path
[export_job] exported to
`/home/username/path-to-export/2018-02-21_07-16-45/20180218_n5x0245.jpg'
[pixelpipe_process] [export] using device -1
[export_job] exported to
`/home/username/path-to-export/2018-02-21_07-16-45/20180218_n5x0242.jpg'
[pixelpipe_process] [export] using device -1
[export_job] exported to
`/home/username/path-to-export/2018-02-21_07-16-45/20180218_n5x0240.jpg'

opencl_memory_headroom=500
opencl_memory_requirement=768
--
regards
Bernhard

https://www.bilddateien.de
Michael Rasmussen
2018-02-21 06:54:43 UTC
Permalink
On Wed, 21 Feb 2018 07:35:32 +0100
Post by Bernhard
~ $ nvidia-smi
Wed Feb 21 07:17:20 2018
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 384.111                Driver Version: 384.111                   |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX 105...  Off  | 00000000:01:00.0  On |                  N/A |
| 20%   30C    P8   ERR! /  75W |    328MiB /  4035MiB | 5%      Default |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: GPU Memory |
|  GPU       PID   Type   Process name Usage      |
|=============================================================================|
|    0      1363      G /usr/lib/xorg/Xorg                           207MiB |
|    0      2005      G cinnamon                                      57MiB |
|    0      3425      C darktable                                     43MiB |
+-----------------------------------------------------------------------------+
Try running this command in another terminal while running the
darktable command: watch -n 1 nvidia-smi (this will call nvidia-smi
every second)
--
Hilsen/Regards
Michael Rasmussen

Get my public GnuPG keys:
michael <at> rasmussen <dot> cc
http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E
mir <at> datanom <dot> net
http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C
mir <at> miras <dot> org
http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917
--------------------------------------------------------------
/usr/games/fortune -es says:
It's the thought, if any, that counts!
Christian Kanzian
2018-02-21 07:31:47 UTC
Permalink
Hi,
Post by Bernhard
Post by Michael Rasmussen
On Tue, 20 Feb 2018 21:57:31 +0100
Post by Ulrich Pegelow
With today's typical amount of graphics cards memory we should
probably increase the default setting of that parameter to maybe 400
or 450.
With 2GB of GPU RAM I have found the sweet spot to be
opencl_memory_headroom=400
opencl_memory_requirement=768
Just had the issue again with
Your's
Post by Bernhard
~ $ nvidia-smi
Wed Feb 21 07:17:20 2018
+-----------------------------------------------------------------------------+
384.111                   |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile
Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util
Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX 105...  Off  | 00000000:01:00.0  On
|                  N/A |
| 20%   30C    P8   ERR! /  75W |    328MiB /  4035MiB | 5%      Default |
+-------------------------------+----------------------+----------------------+
Ulrich's

+-----------------------------------------------------------------------------+
| NVIDIA-SMI 384.111 Driver Version: 384.111 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile
Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util
Compute M. |
|===============================+======================+======================|
| 0 GeForce GTX 106... Off | 00000000:01:00.0 On | N/A |
| 4% 50C P8 10W / 120W | 182MiB / 6065MiB | 0% Default
|
+-------------------------------+----------------------+----------------------+

What about this "ERR!". Are you sure that there is no hardware issue?
Power Supply? darktable puts a have load on the GPU.

Ciao,
Christian
Bernhard
2018-02-21 19:43:45 UTC
Permalink
Post by Christian Kanzian
Hi,
Post by Bernhard
Post by Michael Rasmussen
On Tue, 20 Feb 2018 21:57:31 +0100
Post by Ulrich Pegelow
With today's typical amount of graphics cards memory we should
probably increase the default setting of that parameter to maybe
400 or 450.
With 2GB of GPU RAM I have found the sweet spot to be
opencl_memory_headroom=400
opencl_memory_requirement=768
Just had the issue again with
Your's
Post by Bernhard
~ $ nvidia-smi
Wed Feb 21 07:17:20 2018
+-----------------------------------------------------------------------------+
384.111                   |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile
Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util
Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX 105...  Off  | 00000000:01:00.0  On
|                  N/A |
| 20%   30C    P8   ERR! /  75W |    328MiB /  4035MiB | 5% Default |
+-------------------------------+----------------------+----------------------+
Ulrich's
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 384.111                Driver Version: 384.111      |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile
Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util
Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX 106...  Off  | 00000000:01:00.0  On |  N/A |
|  4%   50C    P8    10W / 120W |    182MiB /  6065MiB |      0%
Default |
+-------------------------------+----------------------+----------------------+
What about this "ERR!".
I really don't know. I found lots of reports though in the internet
about people reporting on this issue on the GTX1050 - seems to be some
dependency on the driver version.
Post by Christian Kanzian
Are you sure that there is no hardware issue?
At least the graphics card is completely new. Everything else worked
without issues so far.
Post by Christian Kanzian
Power Supply?
I tried 'sensors' but it does report temperatures only - no Voltages on
my board.
--
regards
Bernhard

https://www.bilddateien.de
Patrick Shanahan
2018-02-21 19:54:56 UTC
Permalink
Post by Christian Kanzian
Hi,
Post by Bernhard
Post by Michael Rasmussen
On Tue, 20 Feb 2018 21:57:31 +0100
Post by Ulrich Pegelow
With today's typical amount of graphics cards memory we should
probably increase the default setting of that parameter to maybe 400
or 450.
With 2GB of GPU RAM I have found the sweet spot to be
opencl_memory_headroom=400
opencl_memory_requirement=768
Just had the issue again with
Your's
Post by Bernhard
~ $ nvidia-smi
Wed Feb 21 07:17:20 2018
+-----------------------------------------------------------------------------+
384.111                   |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile
Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util
Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX 105...  Off  | 00000000:01:00.0  On
|                  N/A |
| 20%   30C    P8   ERR! /  75W |    328MiB /  4035MiB | 5% Default |
+-------------------------------+----------------------+----------------------+
Ulrich's
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 384.111                Driver Version: 384.111      |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr.
ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util Compute
M. |
|===============================+======================+======================|
|   0  GeForce GTX 106...  Off  | 00000000:01:00.0  On |  N/A |
|  4%   50C    P8    10W / 120W |    182MiB /  6065MiB |      0% Default |
+-------------------------------+----------------------+----------------------+
What about this "ERR!".
I really don't know. I found lots of reports though in the internet about
people reporting on this issue on the GTX1050 - seems to be some dependency
on the driver version.
Post by Christian Kanzian
Are you sure that there is no hardware issue?
At least the graphics card is completely new. Everything else worked without
issues so far.
Post by Christian Kanzian
Power Supply?
I tried 'sensors' but it does report temperatures only - no Voltages on my
board.
most "sensors" that I have used required setting up. my chosen distro,
openSUSE Tumbleweed uses /usr/sbin/sensors-detect
--
(paka)Patrick Shanahan Plainfield, Indiana, USA @ptilopteri
http://en.opensuse.org openSUSE Community Member facebook/ptilopteri
Registered Linux User #207535 @ http://linuxcounter.net
Photos: http://wahoo.no-ip.org/piwigo paka @ IRCnet freenode
Peter Cripps
2018-02-25 18:42:53 UTC
Permalink
I have darktable installed on two machines: 2.4.0rc0 on Windows, and
2.4.1 on Linux. If I make edits on the Windows machine, close darktable,
and then open on Linux, the edits don't carry over. Same in the reverse
direction, Linux to Windows.

This used to work at one point. If I open earlier images, the same edits
appear on both systems. I'm wondering if this is caused by having
different versions of darktable on each machine? (It doesn't look as if
2.4.1 is available in a Windows version right now.)
Pascal Obry
2018-02-25 18:59:21 UTC
Permalink
Peter,
Post by Peter Cripps
This used to work at one point. If I open earlier images, the same
edits appear on both systems. I'm wondering if this is caused by
having different versions of darktable on each machine? (It doesn't
look as if 2.4.1 is available in a Windows version right now.)
2.4.1 is available on Windows. But I don't think this is the issue. It
should be possible to use 2.4.0 and 2.4.1 together as the later is only
a corrective release fully compatible with 2.4.0.
--
Pascal Obry / Magny Les Hameaux (78)

The best way to travel is by means of imagination

http://www.obry.net

gpg --keyserver keys.gnupg.net --recv-key F949BD3B
Peter Cripps
2018-02-25 19:11:31 UTC
Permalink
Thanks Pascal.

After seeing your comment, I found 2.4.1, and installed it on Windows.
(Previously I was getting it from the link on the pixls.us site.)
Anyway, as you say, it doesn't fix the problem. I'll try reinstalling
the Linux version now, although I can't imagine that has anything to do
with it.
Post by Pascal Obry
Peter,
Post by Peter Cripps
This used to work at one point. If I open earlier images, the same
edits appear on both systems. I'm wondering if this is caused by
having different versions of darktable on each machine? (It doesn't
look as if 2.4.1 is available in a Windows version right now.)
2.4.1 is available on Windows. But I don't think this is the issue. It
should be possible to use 2.4.0 and 2.4.1 together as the later is only
a corrective release fully compatible with 2.4.0.
Patrick Shanahan
2018-02-25 19:13:18 UTC
Permalink
I have darktable installed on two machines: 2.4.0rc0 on Windows, and 2.4.1
on Linux. If I make edits on the Windows machine, close darktable, and then
open on Linux, the edits don't carry over. Same in the reverse direction,
Linux to Windows.
This used to work at one point. If I open earlier images, the same edits
appear on both systems. I'm wondering if this is caused by having different
versions of darktable on each machine? (It doesn't look as if 2.4.1 is
available in a Windows version right now.)
you have two different databases. you must import the images into the
*2nd* system in order for the edits to be read into the db. If you have
already imported them in the 2nd system, then you need to set your system
to check for new xmp (sidecar) files. see preferences -> core options ->
"look for updated mp files on startup".
--
(paka)Patrick Shanahan Plainfield, Indiana, USA @ptilopteri
http://en.opensuse.org openSUSE Community Member facebook/ptilopteri
Registered Linux User #207535 @ http://linuxcounter.net
Photos: http://wahoo.no-ip.org/piwigo paka @ IRCnet freenode
Peter Cripps
2018-02-25 19:49:22 UTC
Permalink
Thanks Patrick.

As it happens, a couple of minutes ago I found the "look for updated xmp
files on startup" option. Problem fixed!

Then I saw your reply. Classic case of 'crossing in the mail'. Thanks
again for taking the time to respond.
Post by Patrick Shanahan
I have darktable installed on two machines: 2.4.0rc0 on Windows, and 2.4.1
on Linux. If I make edits on the Windows machine, close darktable, and then
open on Linux, the edits don't carry over. Same in the reverse direction,
Linux to Windows.
This used to work at one point. If I open earlier images, the same edits
appear on both systems. I'm wondering if this is caused by having different
versions of darktable on each machine? (It doesn't look as if 2.4.1 is
available in a Windows version right now.)
you have two different databases. you must import the images into the
*2nd* system in order for the edits to be read into the db. If you have
already imported them in the 2nd system, then you need to set your system
to check for new xmp (sidecar) files. see preferences -> core options ->
"look for updated mp files on startup".
Bernhard
2018-02-26 22:24:16 UTC
Permalink
Post by Christian Kanzian
Your's
Post by Bernhard
~ $ nvidia-smi
Wed Feb 21 07:17:20 2018
+-----------------------------------------------------------------------------+
384.111                   |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile
Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util
Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX 105...  Off  | 00000000:01:00.0  On
|                  N/A |
| 20%   30C    P8   ERR! /  75W |    328MiB /  4035MiB | 5% Default |
+-------------------------------+----------------------+----------------------+
Ulrich's
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 384.111                Driver Version: 384.111      |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile
Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util
Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX 106...  Off  | 00000000:01:00.0  On |  N/A |
|  4%   50C    P8    10W / 120W |    182MiB /  6065MiB |      0%
Default |
+-------------------------------+----------------------+----------------------+
What about this "ERR!". Are you sure that there is no hardware issue?
Power Supply? darktable puts a have load on the GPU.
I was told yesterday that this card doesn't have that sensor.
Updating to newer driver shows N/A instead:

|~ $ nvidia-smi Sun Feb 25 10:57:52 2018
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 390.25                 Driver Version: 390.25             
      |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile
Uncorr. ECC | | Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage |
GPU-Util  Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX 105...  Off  | 00000000:01:00.0  On |               
  N/A | | 20%   39C    P0    N/A /  75W |    166MiB /  4036MiB |     
2%      Default |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes:                                                       GPU
Memory | |  GPU       PID   Type   Process name                         
   Usage      |
|=============================================================================|
|    0      1428      G   /usr/lib/xorg/Xorg                         
 129MiB | |    0      2032      G   cinnamon                           
          34MiB |
+-----------------------------------------------------------------------------+
|
--
regards
Bernhard

https://www.bilddateien.de
Bernhard
2018-02-26 22:28:31 UTC
Permalink
Post by Christian Kanzian
Your's
Post by Bernhard
~ $ nvidia-smi
Wed Feb 21 07:17:20 2018
+-----------------------------------------------------------------------------+
384.111                   |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile
Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util
Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX 105...  Off  | 00000000:01:00.0  On
|                  N/A |
| 20%   30C    P8   ERR! /  75W |    328MiB /  4035MiB | 5%     
Default |
+-------------------------------+----------------------+----------------------+
Ulrich's
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 384.111                Driver Version: 384.111      |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile
Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util
Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX 106...  Off  | 00000000:01:00.0  On | N/A |
|  4%   50C    P8    10W / 120W |    182MiB /  6065MiB |      0%
Default |
+-------------------------------+----------------------+----------------------+
What about this "ERR!". Are you sure that there is no hardware issue?
Power Supply? darktable puts a have load on the GPU.
I was told yesterday that this card doesn't have that sensor.
Updating to newer driver shows N/A instead:

|~ $ nvidia-smi Sun Feb 25 10:57:52 2018
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 390.25                 Driver Version: 390.25             
      |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile
Uncorr. ECC | | Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage |
GPU-Util  Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX 105...  Off  | 00000000:01:00.0  On |               
  N/A | | 20%   39C    P0    N/A /  75W |    166MiB /  4036MiB |     
2%      Default |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes:                                                       GPU
Memory | |  GPU       PID   Type   Process name                         
   Usage      |
|=============================================================================|
|    0      1428      G   /usr/lib/xorg/Xorg                         
 129MiB | |    0      2032      G   cinnamon                           
          34MiB |
+-----------------------------------------------------------------------------+
|
--
regards
Bernhard

https://www.bilddateien.de
Ulrich Pegelow
2018-02-27 06:51:26 UTC
Permalink
What are your opencl related settings in darktablerc (i.e. all config
parameters of the form opencl_*) ?

Ulrich
Post by Bernhard
I was told yesterday that this card doesn't have that sensor.
|~ $ nvidia-smi Sun Feb 25 10:57:52 2018
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 390.25                 Driver Version: 390.25
      |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile
Uncorr. ECC | | Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage |
GPU-Util  Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX 105...  Off  | 00000000:01:00.0  On |
  N/A | | 20%   39C    P0    N/A /  75W |    166MiB /  4036MiB |
2%      Default |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes:                                                       GPU
Memory | |  GPU       PID   Type   Process name
   Usage      |
|=============================================================================|
|    0      1428      G   /usr/lib/xorg/Xorg
 129MiB | |    0      2032      G   cinnamon
          34MiB |
+-----------------------------------------------------------------------------+
|
--
Bernhard
2018-02-27 22:48:05 UTC
Permalink
First of all: I just wanted to tell people about the (possible) solution
of that ERR! message.
Second: After your advice to the out-of-resource I changed
opencl_memory_headroom to 500, later I found this article
https://www.darktable.org/2012/03/darktable-and-opencl/ by using the
search for that parameter on your website - now I understand the background.
Third: For some other reasons I reinstalled my system and now the base
is a brand new LinuxMint 18.3 - so I am back with the standard nvidia
384.111
and
System:    Host: Rechner Kernel: 4.10.0-38-generic x86_64 (64 bit gcc:
5.4.0)
           Desktop: Cinnamon 3.6.7 (Gtk 3.18.9-1ubuntu3.3)
           Distro: Linux Mint 18.3 Sylvia
Machine:   Mobo: ASUSTeK model: P8Z77-M v: Rev 1.xx
           Bios: American Megatrends v: 2203 date: 12/18/2015
CPU:       Quad core Intel Core i5-3570 (-MCP-) cache: 6144 KB
           flags: (lm nx sse sse2 sse3 sse4_1 sse4_2 ssse3 vmx) bmips:
27278
           clock speeds: max: 3800 MHz 1: 2190 MHz 2: 1858 MHz 3: 1964 MHz
           4: 2103 MHz
Graphics:  Card: NVIDIA Device 1c82 bus-ID: 01:00.0
           Display Server: X.Org 1.18.4 drivers: nvidia (unloaded:
fbdev,vesa,nouveau)
           Resolution: ***@59.95hz
           GLX Renderer: GeForce GTX 1050 Ti/PCIe/SSE2
           GLX Version: 4.5.0 NVIDIA 384.111 Direct Rendering: Yes

openCL settings in darktable are default except the one mentioned above
in darktablerc

opencl=TRUE
opencl_async_pixelpipe=false
opencl_avoid_atomics=false
opencl_checksum=1868237003
opencl_device_priority=*/!0,*/*/*
opencl_disable_drivers_blacklist=false
opencl_enable_markesteijn=true
opencl_library=
opencl_mandatory_timeout=200
opencl_memory_headroom=500
opencl_memory_requirement=768
opencl_micro_nap=1000
opencl_number_event_handles=25
opencl_omit_whitebalance=
opencl_scheduling_profile=default
opencl_size_roundup=16
opencl_synch_cache=false
opencl_use_cpu_devices=false
opencl_use_pinned_memory=false

Yesterday I exported some images and it seemed to work well - machine
was in STR since then - but with my test some minutes ago I got:

[opencl_init] opencl related configuration options:
[opencl_init]
[opencl_init] opencl: 1
[opencl_init] opencl_library: ''
[opencl_init] opencl_memory_requirement: 768
[opencl_init] opencl_memory_headroom: 500
[opencl_init] opencl_device_priority: '*/!0,*/*/*'
[opencl_init] opencl_mandatory_timeout: 200
[opencl_init] opencl_size_roundup: 16
[opencl_init] opencl_async_pixelpipe: 0
[opencl_init] opencl_synch_cache: 0
[opencl_init] opencl_number_event_handles: 25
[opencl_init] opencl_micro_nap: 1000
[opencl_init] opencl_use_pinned_memory: 0
[opencl_init] opencl_use_cpu_devices: 0
[opencl_init] opencl_avoid_atomics: 0
[opencl_init]
[opencl_init] found opencl runtime library 'libOpenCL'
[opencl_init] opencl library 'libOpenCL' found on your system and loaded
*[opencl_init] could not get platforms: -1001*
*[opencl_init] FINALLY: opencl is NOT AVAILABLE on this system.*
[opencl_init] initial status of opencl enabled flag is OFF.

After restarting I got back my opencl and successfully exported images
up to 100MPx - the latter from TIF.
I attach the log.

-Bernhard-
Post by Ulrich Pegelow
What are your opencl related settings in darktablerc (i.e. all config
parameters of the form opencl_*) ?
Ulrich
Post by Bernhard
|~ $ nvidia-smi Sun Feb 25 10:57:52 2018
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 390.25                 Driver Version: 390.25           
        |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile
Uncorr. ECC | | Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage
| GPU-Util  Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX 105...  Off  | 00000000:01:00.0  On |             
    N/A | | 20%   39C    P0    N/A /  75W |    166MiB /  4036MiB |   
  2%      Default |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes:                                                     
 GPU Memory | |  GPU       PID   Type   Process name                 
           Usage      |
|=============================================================================|
|    0      1428      G   /usr/lib/xorg/Xorg                         
 129MiB | |    0      2032      G   cinnamon                         
            34MiB |
+-----------------------------------------------------------------------------+|
--
regards
Bernhard

https://www.bilddateien.de
Bernhard
2018-03-04 15:02:27 UTC
Permalink
Hi,

I'm still struggling with "part-time openCL" on my new Nvidia card.

General system data:

System:    Host: lorem_ipsum Kernel: 4.10.0-38-generic x86_64 (64 bit
gcc: 5.4.0)
           Desktop: Cinnamon 3.6.7 (Gtk 3.18.9-1ubuntu3.3)
           Distro: Linux Mint 18.3 Sylvia
Machine:   Mobo: ASUSTeK model: P8Z77-M v: Rev 1.xx
           Bios: American Megatrends v: 2203 date: 12/18/2015
CPU:       Quad core Intel Core i5-3570 (-MCP-) cache: 6144 KB
           flags: (lm nx sse sse2 sse3 sse4_1 sse4_2 ssse3 vmx) bmips:
27278
           clock speeds: max: 3800 MHz 1: 2190 MHz 2: 1858 MHz 3: 1964 MHz
           4: 2103 MHz
Graphics:  Card: NVIDIA Device 1c82 bus-ID: 01:00.0
           Display Server: X.Org 1.18.4 drivers: nvidia (unloaded:
fbdev,vesa,nouveau)
           Resolution: ***@59.95hz
           GLX Renderer: GeForce GTX 1050 Ti/PCIe/SSE2
           GLX Version: 4.5.0 NVIDIA 384.111 Direct Rendering: Yes

openCL settings in darktable are default except the one mentioned by
Ulrich days ago in darktablerc

opencl=TRUE
opencl_async_pixelpipe=false
opencl_avoid_atomics=false
opencl_checksum=1868237003
opencl_device_priority=*/!0,*/*/*
opencl_disable_drivers_blacklist=false
opencl_enable_markesteijn=true
opencl_library=
opencl_mandatory_timeout=200
*opencl_memory_headroom=500*
opencl_memory_requirement=768
opencl_micro_nap=1000
opencl_number_event_handles=25
opencl_omit_whitebalance=
opencl_scheduling_profile=default
opencl_size_roundup=16
opencl_synch_cache=false
opencl_use_cpu_devices=false
opencl_use_pinned_memory=false

I never saw too high memory consumption in nvidia-smi since then.

I compared my installation to this list
https://www.darktable.org/usermanual/en/darktable_and_opencl_system.html

$ dpkg -l | grep nvidia
ii  nvidia-384 384.111-0ubuntu0.16.04.1 amd64        NVIDIA binary
driver - version 384.111
ii  nvidia-cuda-dev 7.5.18-0ubuntu1 amd64        NVIDIA CUDA development
files
ii  nvidia-cuda-toolkit 7.5.18-0ubuntu1 amd64        NVIDIA CUDA
development toolkit
ii  nvidia-modprobe 361.28-1 amd64        utility to load NVIDIA kernel
modules and create device nodes
ii  nvidia-opencl-dev:amd64 7.5.18-0ubuntu1 amd64        NVIDIA OpenCL
development files
ii  nvidia-opencl-icd-384 384.111-0ubuntu0.16.04.1 amd64        NVIDIA
OpenCL ICD
ii  nvidia-prime 0.8.2linuxmint1 amd64        Tools to enable NVIDIA's Prime
ii  nvidia-prime-applet 1.0.6 all          An applet for NVIDIA Prime
ii  nvidia-profiler 7.5.18-0ubuntu1 amd64        NVIDIA Profiler for
CUDA and OpenCL
ii  nvidia-settings 361.42-0ubuntu1 amd64        Tool for configuring
the NVIDIA graphics driver

I find some additional CUDA packages but some other files are missing.
So e. g. nvidia-libopencl1-384 is missing but synaptic says:

This package contains NVIDIA's OpenCL implementation and the ICD Loader
library. This library supports only OpenCL 1.1.

So I find ocl-icd-libopencl1 on my system instead (installed automatically):

$ dpkg -l | grep opencl
ii  nvidia-opencl-dev:amd64 7.5.18-0ubuntu1 amd64        NVIDIA OpenCL
development files
ii  nvidia-opencl-icd-384 384.111-0ubuntu0.16.04.1 amd64        NVIDIA
OpenCL ICD
ii  ocl-icd-libopencl1:amd64 2.2.8-1 amd64        Generic OpenCL ICD Loader
ii  opencl-headers 2.0~svn32091-2 all          OpenCL (Open Computing
Language) header files

$ ls -l /dev/nvidia*
crw-rw-rw- 1 root root 195,   0 MÀr  4 09:44 /dev/nvidia0
crw-rw-rw- 1 root root 195, 255 MÀr  4 09:44 /dev/nvidiactl
crw-rw-rw- 1 root root 195, 254 MÀr  4 09:44 /dev/nvidia-modeset
crw-rw-rw- 1 root root 242,   0 MÀr  4 09:44 /dev/nvidia-uvm

$ lsmod | grep nvidia
nvidia_uvm            671744  2
nvidia_drm             45056  1
nvidia_modeset        860160  4 nvidia_drm
nvidia              13139968  429 nvidia_modeset,nvidia_uvm
drm_kms_helper        151552  1 nvidia_drm
drm                   352256  4 nvidia_drm,drm_kms_helper


I also checked /etc/OpenCL as suggested in the docs and found only one
file: /etc/OpenCL/vendors/nvidia.icd with a one-line-entry:
libnvidia-opencl.so.1
I find this in /usr/lib/x86_64-linux-gnu and also in
/usr/lib/i386-linux-gnu which seems to link to libnvidia-opencl.so.384.111


*_ERRORS:_*

I sometimes find something like this in the log:

[dev_process_thumbnail] pixel pipeline processing took 0,623 secs (1,912
CPU)
[dev] took 0,468 secs (0,172 CPU) to load the image.
[dev] took 0,000 secs (0,000 CPU) to load the image.
[pixelpipe_process] [full] using device 0
[dev_pixelpipe] took 0,010 secs (0,028 CPU) initing base buffer [full]
[opencl_pixelpipe] couldn't copy image to opencl device for module
rawprepare
[opencl_pixelpipe] could not run module 'rawprepare' on gpu. falling
back to cpu path
*[opencl_pixelpipe (b)] late opencl error detected while copying back to
cpu buffer: -5**
**[pixelpipe_process] [full] falling back to cpu path*

*/I don't find '-5' error in the docs - what does this mean?/*
This does _not_ deactivate openCL immediately, but after 5 such errors I
get:

[pixelpipe_process] [thumbnail] using device 0
[dev_pixelpipe] took 0,006 secs (0,016 CPU) initing base buffer [thumbnail]
[opencl_pixelpipe] couldn't copy image to opencl device for module
rawprepare
[opencl_pixelpipe] could not run module 'rawprepare' on gpu. falling
back to cpu path
*[opencl_pixelpipe (b)] late opencl error detected while copying back to
cpu buffer: -5**
**[opencl] frequent opencl errors encountered; disabling opencl for this
session!*

Shutting down darktable and starting again I get the following:

[opencl_init] opencl related configuration options:
[opencl_init]
[opencl_init] opencl: 1
[opencl_init] opencl_library: ''
[opencl_init] opencl_memory_requirement: 768
[opencl_init] opencl_memory_headroom: 500
[opencl_init] opencl_device_priority: '*/!0,*/*/*'
[opencl_init] opencl_mandatory_timeout: 200
[opencl_init] opencl_size_roundup: 16
[opencl_init] opencl_async_pixelpipe: 0
[opencl_init] opencl_synch_cache: 0
[opencl_init] opencl_number_event_handles: 25
[opencl_init] opencl_micro_nap: 1000
[opencl_init] opencl_use_pinned_memory: 0
[opencl_init] opencl_use_cpu_devices: 0
[opencl_init] opencl_avoid_atomics: 0
[opencl_init]
[opencl_init] found opencl runtime library 'libOpenCL'
[opencl_init] opencl library 'libOpenCL' found on your system and loaded
*[opencl_init] could not get platforms: -1001*
*[opencl_init] FINALLY: opencl is NOT AVAILABLE on this system.*
[opencl_init] initial status of opencl enabled flag is OFF.

After this I have to restart my computer to get openCL back.

Any suggestions what to do next?
--
regards
Bernhard

https://www.bilddateien.de
KOVÁCS István
2018-03-05 10:37:20 UTC
Permalink
The only memory-transfer related setting that I know of
is opencl_use_pinned_memory. The manual (
https://www.darktable.org/usermanual/en/darktable_and_opencl_optimization.html)
says for NVidia it should be set to false. You may want to check - for me,
it was set to true by default, although it did not cause problems apart
from being slow.

Kofa
Post by Bernhard
Hi,
I'm still struggling with "part-time openCL" on my new Nvidia card.
5.4.0)
Desktop: Cinnamon 3.6.7 (Gtk 3.18.9-1ubuntu3.3)
Distro: Linux Mint 18.3 Sylvia
Machine: Mobo: ASUSTeK model: P8Z77-M v: Rev 1.xx
Bios: American Megatrends v: 2203 date: 12/18/2015
CPU: Quad core Intel Core i5-3570 (-MCP-) cache: 6144 KB
27278
clock speeds: max: 3800 MHz 1: 2190 MHz 2: 1858 MHz 3: 1964 MHz
4: 2103 MHz
Graphics: Card: NVIDIA Device 1c82 bus-ID: 01:00.0
fbdev,vesa,nouveau)
GLX Renderer: GeForce GTX 1050 Ti/PCIe/SSE2
GLX Version: 4.5.0 NVIDIA 384.111 Direct Rendering: Yes
openCL settings in darktable are default except the one mentioned by
Ulrich days ago in darktablerc
opencl=TRUE
opencl_async_pixelpipe=false
opencl_avoid_atomics=false
opencl_checksum=1868237003
opencl_device_priority=*/!0,*/*/*
opencl_disable_drivers_blacklist=false
opencl_enable_markesteijn=true
opencl_library=
opencl_mandatory_timeout=200
*opencl_memory_headroom=500*
opencl_memory_requirement=768
opencl_micro_nap=1000
opencl_number_event_handles=25
opencl_omit_whitebalance=
opencl_scheduling_profile=default
opencl_size_roundup=16
opencl_synch_cache=false
opencl_use_cpu_devices=false
opencl_use_pinned_memory=false
I never saw too high memory consumption in nvidia-smi since then.
I compared my installation to this list https://www.darktable.org/
usermanual/en/darktable_and_opencl_system.html
$ dpkg -l | grep nvidia
ii nvidia-384
384.111-0ubuntu0.16.04.1 amd64
NVIDIA binary driver - version 384.111
ii nvidia-cuda-dev
7.5.18-0ubuntu1 amd64
NVIDIA CUDA development files
ii nvidia-cuda-toolkit
7.5.18-0ubuntu1 amd64
NVIDIA CUDA development toolkit
ii nvidia-modprobe
361.28-1 amd64
utility to load NVIDIA kernel modules and create device nodes
ii nvidia-opencl-dev:amd64
7.5.18-0ubuntu1 amd64
NVIDIA OpenCL development files
ii nvidia-opencl-icd-384
384.111-0ubuntu0.16.04.1 amd64
NVIDIA OpenCL ICD
ii nvidia-prime
0.8.2linuxmint1 amd64
Tools to enable NVIDIA's Prime
ii nvidia-prime-applet
1.0.6 all An
applet for NVIDIA Prime
ii nvidia-profiler
7.5.18-0ubuntu1 amd64
NVIDIA Profiler for CUDA and OpenCL
ii nvidia-settings
361.42-0ubuntu1 amd64 Tool
for configuring the NVIDIA graphics driver
I find some additional CUDA packages but some other files are missing.
This package contains NVIDIA's OpenCL implementation and the ICD Loader
library. This library supports only OpenCL 1.1.
So I find ocl-icd-libopencl1 on my system instead (installed automatically)
$ dpkg -l | grep opencl
ii nvidia-opencl-dev:amd64
7.5.18-0ubuntu1 amd64
NVIDIA OpenCL development files
ii nvidia-opencl-icd-384
384.111-0ubuntu0.16.04.1 amd64
NVIDIA OpenCL ICD
ii ocl-icd-libopencl1:amd64
2.2.8-1 amd64
Generic OpenCL ICD Loader
ii opencl-headers
2.0~svn32091-2 all
OpenCL (Open Computing Language) header files
$ ls -l /dev/nvidia*
crw-rw-rw- 1 root root 195, 0 MÀr 4 09:44 /dev/nvidia0
crw-rw-rw- 1 root root 195, 255 MÀr 4 09:44 /dev/nvidiactl
crw-rw-rw- 1 root root 195, 254 MÀr 4 09:44 /dev/nvidia-modeset
crw-rw-rw- 1 root root 242, 0 MÀr 4 09:44 /dev/nvidia-uvm
$ lsmod | grep nvidia
nvidia_uvm 671744 2
nvidia_drm 45056 1
nvidia_modeset 860160 4 nvidia_drm
nvidia 13139968 <(1)%20313%209968> 429
nvidia_modeset,nvidia_uvm
drm_kms_helper 151552 1 nvidia_drm
drm 352256 4 nvidia_drm,drm_kms_helper
I also checked /etc/OpenCL as suggested in the docs and found only one
libnvidia-opencl.so.1
I find this in /usr/lib/x86_64-linux-gnu and also in
/usr/lib/i386-linux-gnu which seems to link to libnvidia-opencl.so.384.111
*ERRORS:*
[dev_process_thumbnail] pixel pipeline processing took 0,623 secs (1,912
CPU)
[dev] took 0,468 secs (0,172 CPU) to load the image.
[dev] took 0,000 secs (0,000 CPU) to load the image.
[pixelpipe_process] [full] using device 0
[dev_pixelpipe] took 0,010 secs (0,028 CPU) initing base buffer [full]
[opencl_pixelpipe] couldn't copy image to opencl device for module
rawprepare
[opencl_pixelpipe] could not run module 'rawprepare' on gpu. falling back
to cpu path
*[opencl_pixelpipe (b)] late opencl error detected while copying back to
cpu buffer: -5*
*[pixelpipe_process] [full] falling back to cpu path*
*I don't find '-5' error in the docs - what does this mean?*
This does *not* deactivate openCL immediately, but after 5 such errors I
[pixelpipe_process] [thumbnail] using device 0
[dev_pixelpipe] took 0,006 secs (0,016 CPU) initing base buffer [thumbnail]
[opencl_pixelpipe] couldn't copy image to opencl device for module
rawprepare
[opencl_pixelpipe] could not run module 'rawprepare' on gpu. falling back
to cpu path
*[opencl_pixelpipe (b)] late opencl error detected while copying back to
cpu buffer: -5*
*[opencl] frequent opencl errors encountered; disabling opencl for this
session!*
[opencl_init]
[opencl_init] opencl: 1
[opencl_init] opencl_library: ''
[opencl_init] opencl_memory_requirement: 768
[opencl_init] opencl_memory_headroom: 500
[opencl_init] opencl_device_priority: '*/!0,*/*/*'
[opencl_init] opencl_mandatory_timeout: 200
[opencl_init] opencl_size_roundup: 16
[opencl_init] opencl_async_pixelpipe: 0
[opencl_init] opencl_synch_cache: 0
[opencl_init] opencl_number_event_handles: 25
[opencl_init] opencl_micro_nap: 1000
[opencl_init] opencl_use_pinned_memory: 0
[opencl_init] opencl_use_cpu_devices: 0
[opencl_init] opencl_avoid_atomics: 0
[opencl_init]
[opencl_init] found opencl runtime library 'libOpenCL'
[opencl_init] opencl library 'libOpenCL' found on your system and loaded
*[opencl_init] could not get platforms: -1001*
*[opencl_init] FINALLY: opencl is NOT AVAILABLE on this system.*
[opencl_init] initial status of opencl enabled flag is OFF.
After this I have to restart my computer to get openCL back.
Any suggestions what to do next?
--
regards
Bernhard
https://www.bilddateien.de
____________________________________________________________________________
darktable user mailing list to unsubscribe send a mail to
Ulrich Pegelow
2018-03-05 16:54:28 UTC
Permalink
Additionally you may want to play with opencl_number_event_handles.
Start with an extreme setting of zero. In case this would solve your
issues you can then try different values in-between.

Ulrich
Post by KOVÁCS István
The only memory-transfer related setting that I know of
is opencl_use_pinned_memory. The manual
(https://www.darktable.org/usermanual/en/darktable_and_opencl_optimization.html)
says for NVidia it should be set to false. You may want to check - for
me, it was set to true by default, although it did not cause problems
apart from being slow.
Kofa
Bernhard
2018-03-07 21:30:26 UTC
Permalink
Post by Ulrich Pegelow
Additionally you may want to play with opencl_number_event_handles.
Start with an extreme setting of zero. In case this would solve your
issues you can then try different values in-between.
Ulrich
I changed this value with no result.

Then I learned that there is a little tool called clinfo.
I fear that it's a hardware or system issue now.

I get

$ clinfo
Number of platforms                               1
  Platform Name                                   NVIDIA CUDA
  Platform Vendor                                 NVIDIA Corporation
  Platform Version                                OpenCL 1.2 CUDA 9.0.282
  Platform Profile FULL_PROFILE
  Platform Extensions cl_khr_global_int32_base_atomics
cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics
cl_khr_local_int32_extended_atomics cl_khr_fp64
cl_khr_byte_addressable_store cl_khr_icd cl_khr_gl_sharing
cl_nv_compiler_options cl_nv_device_attribute_query cl_nv_pragma_unroll
cl_nv_copy_opts cl_nv_create_buffer
  Platform Extensions function suffix             NV

  Platform Name                                   NVIDIA CUDA
Number of devices                                 1
  Device Name                                     GeForce GTX 1050 Ti
  Device Vendor                                   NVIDIA Corporation
  Device Vendor ID                                0x10de
  Device Version                                  OpenCL 1.2 CUDA
  Driver Version                                  384.111
  Device OpenCL C Version                         OpenCL C 1.2
  Device Type                                     GPU
  Device Profile FULL_PROFILE
  Device Topology (NV)                            PCI-E, 01:00.0
  Max compute units                               6
  Max clock frequency                             1392MHz
  Compute Capability (NV)                         6.1
  Device Partition                                (core)
    Max number of sub-devices                     1
    Supported partition types                     None
  Max work item dimensions                        3
  Max work item sizes 1024x1024x64
  Max work group size                             1024
  Preferred work group size multiple              32
  Warp size (NV)                                  32
  Preferred / native vector sizes
    char                                                 1 / 1
    short                                                1 / 1
    int                                                  1 / 1
    long                                                 1 / 1
    half                                                 0 / 0        (n/a)
    float                                                1 / 1
    double                                               1 / 1       
(cl_khr_fp64)
  Half-precision Floating-point support           (n/a)
  Single-precision Floating-point support         (core)
    Denormals                                     Yes
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 Yes
    Round to infinity                             Yes
    IEEE754-2008 fused multiply-add               Yes
    Support is emulated in software               No
    Correctly-rounded divide and sqrt operations  Yes
  Double-precision Floating-point support (cl_khr_fp64)
    Denormals                                     Yes
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 Yes
    Round to infinity                             Yes
    IEEE754-2008 fused multiply-add               Yes
    Support is emulated in software               No
    Correctly-rounded divide and sqrt operations  No
  Address bits                                    64, Little-Endian
  Global memory size 4231725056 (3.941GiB)
  Error Correction support                        No
  Max memory allocation 1057931264 (1009MiB)
  Unified memory for Host and Device              No
  Integrated memory (NV)                          No
  Minimum alignment for any data type             128 bytes
  Alignment of base address                       4096 bits (512 bytes)
  Global Memory cache type Read/Write
  Global Memory cache size                        98304
  Global Memory cache line                        128 bytes
  Image support                                   Yes
    Max number of samplers per kernel             32
    Max size for 1D images from buffer            134217728 pixels
    Max 1D or 2D image array size                 2048 images
    Max 2D image size 16384x32768 pixels
    Max 3D image size 16384x16384x16384 pixels
    Max number of read image args                 256
    Max number of write image args                16
  Local memory type                               Local
  Local memory size                               49152 (48KiB)
  Registers per block (NV)                        65536
  Max constant buffer size                        65536 (64KiB)
  Max number of constant args                     9
  Max size of kernel argument                     4352 (4.25KiB)
  Queue properties
    Out-of-order execution                        Yes
    Profiling                                     Yes
  Prefer user sync for interop                    No
  Profiling timer resolution                      1000ns
  Execution capabilities
    Run OpenCL kernels                            Yes
    Run native kernels                            No
    Kernel execution timeout (NV)                 Yes
  Concurrent copy and kernel execution (NV)       Yes
    Number of async copy engines                  2
  printf() buffer size                            1048576 (1024KiB)
  Built-in kernels
  Device Available                                Yes
  Compiler Available                              Yes
  Linker Available                                Yes
  Device Extensions cl_khr_global_int32_base_atomics
cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics
cl_khr_local_int32_extended_atomics cl_khr_fp64
cl_khr_byte_addressable_store cl_khr_icd cl_khr_gl_sharing
cl_nv_compiler_options cl_nv_device_attribute_query cl_nv_pragma_unroll
cl_nv_copy_opts cl_nv_create_buffer

NULL platform behavior
  clGetPlatformInfo(NULL, CL_PLATFORM_NAME, ...)  NVIDIA CUDA
  clGetDeviceIDs(NULL, CL_DEVICE_TYPE_ALL, ...)   Success [NV]
  clCreateContext(NULL, ...) [default]            Success [NV]
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_CPU)  No devices found
in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_GPU)  No platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_ACCELERATOR)  No devices
found in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_CUSTOM)  No devices
found in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_ALL)  No platform

ICD loader properties
  ICD loader Name                                 OpenCL ICD Loader
  ICD loader Vendor                               OCL Icd free software
  ICD loader Version                              2.2.8
  ICD loader Profile                              OpenCL 1.2
    NOTE:    your OpenCL library declares to support OpenCL 1.2,
        but it seems to support up to OpenCL 2.1 too.

and all is well.
BUT: after a STR condition I get

 $ clinfo
Number of platforms                               0

instead.
Something seems to be very wrong with my fresh-installed LinuxMint 18.3 :(

Now I have to reboot again ...

Thanks so far.
--
regards
Bernhard

https://www.bilddateien.de
Šarūnas
2018-03-07 22:30:42 UTC
Permalink
Post by Bernhard
Post by Ulrich Pegelow
Additionally you may want to play with opencl_number_event_handles.
Start with an extreme setting of zero. In case this would solve your
issues you can then try different values in-between.
Ulrich
I changed this value with no result.
Then I learned that there is a little tool called clinfo.
I fear that it's a hardware or system issue now.
I get
$ clinfo
Number of platforms                               1
  Platform Name                                   NVIDIA CUDA
  Platform Vendor                                 NVIDIA Corporation
  Platform Version                                OpenCL 1.2 CUDA 9.0.282
  Platform Profile FULL_PROFILE
  Platform Extensions cl_khr_global_int32_base_atomics
cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics
cl_khr_local_int32_extended_atomics cl_khr_fp64
cl_khr_byte_addressable_store cl_khr_icd cl_khr_gl_sharing
cl_nv_compiler_options cl_nv_device_attribute_query cl_nv_pragma_unroll
cl_nv_copy_opts cl_nv_create_buffer
  Platform Extensions function suffix             NV
  Platform Name                                   NVIDIA CUDA
Number of devices                                 1
  Device Name                                     GeForce GTX 1050 Ti
  Device Vendor                                   NVIDIA Corporation
  Device Vendor ID                                0x10de
  Device Version                                  OpenCL 1.2 CUDA
  Driver Version                                  384.111
  [...]
Perhaps it is possible to try different versions of nvidia driver and
CUDA, older/newer?

Nvidia has a network repository at
http://developer.download.nvidia.com/compute/cuda/repos/
--
Šarūnas Burdulis
math.dartmouth.edu/~sarunas
Bernhard
2018-03-10 11:04:26 UTC
Permalink
Post by Šarūnas
Perhaps it is possible to try different versions of nvidia driver and
CUDA, older/newer?
Nvidia has a network repository at
http://developer.download.nvidia.com/compute/cuda/repos/
-- Šarūnas Burdulis math.dartmouth.edu/~sarunas
Thanks - I'm on my way to try this:
https://launchpad.net/~graphics-drivers/+archive/ubuntu/ppa/ and then
compare with your link
--
regards
Bernhard

https://www.bilddateien.de
Loading...