Hi,

Since 0.xx version (1.2 now) I've problems with DT and OpenCL activated.
Everything runs smoothly on CPU (changing parameteres in modules,
preview and so on), and after I turn on OpenCL I can't smoothly operate
with sliders (changing params), preview is freezing for some seconds.
Saying straight  - DT slows down. Investigate with '-d perf -d opencl'
shows, that there's something wrong with the 'splat' function/module
(??). Each plugin/function does job in few ms, except 'splat', which
tooks ~ 1.2 second - no matter what I set in preferences (bilinear,
ppg, dither, etc.).

In log you can see what I'm talking about - the same plugins, the same
photo, first section WITHOUT OpenCL, second WITH OpenCL.

I've no idea, what cause this... 

The problem remains the same, with different kernels, different NVIDIA
drivers (I've 313.30 just now).

-- 
Dariusz

[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: 300
[opencl_init] opencl_device_priority: '*/!0,*/*/*'
[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_avoid_atomics: 0
[opencl_init] opencl_omit_whitebalance: 0
[opencl_init] 
[opencl_init] trying to load opencl library: '<system default>'
[opencl_init] opencl library 'libOpenCL' found on your system and loaded
[opencl_init] found 1 platform
[opencl_init] found 1 device
[opencl_init] device 0 `GeForce GT 240' doesn't have sm_20 support.
[opencl_init] device 0 `GeForce GT 240' supports image sizes of 4096 x 16383
[opencl_init] device 0 `GeForce GT 240' allows GPU memory allocations of up to 
255MB
[opencl_init] device 0: GeForce GT 240 
     GLOBAL_MEM_SIZE:          1023MB
     MAX_WORK_GROUP_SIZE:      512
     MAX_WORK_ITEM_DIMENSIONS: 3
     MAX_WORK_ITEM_SIZES:      [ 512 512 64 ]
     DRIVER_VERSION:           313.30
     DEVICE_VERSION:           OpenCL 1.0 CUDA
[opencl_init] compiling program `demosaic_ppg.cl' ..
[opencl_load_program] loaded cached binary program from file 
`/home/dhor/.cache/darktable/cached_kernels_for_GeForceGT240/demosaic_ppg.cl.bin'
[opencl_load_program] successfully loaded program from 
`/usr/share/darktable/kernels/demosaic_ppg.cl'
[opencl_build_program] successfully built program
[opencl_build_program] BUILD STATUS: 0

# WITHOUT OPENCL ##################
[pixelpipe_process] [full] using device -1
[dev_pixelpipe] took 0.000 secs (0.000 CPU) initing base buffer [full]
[dev_pixelpipe] took 0.006 secs (0.020 CPU) processing `balans bieli' [full]
[dev_pixelpipe] took 0.007 secs (0.020 CPU) processing `ratowanie 
prześwietleń' [full]
[dev_pixelpipe] took 0.033 secs (0.072 CPU) processing `demozaikowanie' [full]
[dev_pixelpipe] took 0.005 secs (0.012 CPU) processing `krzywa bazowa' [full]
[dev_pixelpipe] took 0.012 secs (0.024 CPU) processing `wejściowy profil 
koloru' [full]
[dev_pixelpipe] took 0.066 secs (0.152 CPU) processing `cienie i światła' 
[full]
[dev_pixelpipe] took 0.052 secs (0.164 CPU) processing `kontrast miejscowy' 
[full]
[dev_pixelpipe] took 0.010 secs (0.024 CPU) processing `poziomy' [full]
[dev_pixelpipe] took 0.015 secs (0.032 CPU) processing `wyostrzenie' [full]
[dev_pixelpipe] took 0.020 secs (0.040 CPU) processing `wyjściowy profil 
koloru' [full]
[dev_pixelpipe] took 0.011 secs (0.024 CPU) processing `velvia' [full]
[dev_pixelpipe] took 0.003 secs (0.000 CPU) processing `prześwietlenie' [full]
[dev_pixelpipe] took 0.005 secs (0.012 CPU) processing `gamma' [full]
[dev_process_image] pixel pipeline processing took 0.246 secs (0.596 CPU)
[pixelpipe_process] [preview] using device -1
[dev_pixelpipe] took 0.000 secs (0.000 CPU) initing base buffer [preview]
[dev_pixelpipe] took 0.004 secs (0.012 CPU) processing `balans bieli' [preview]
[dev_pixelpipe] took 0.003 secs (0.008 CPU) processing `ratowanie 
prześwietleń' [preview]
[dev_pixelpipe] took 0.008 secs (0.020 CPU) processing `krzywa bazowa' [preview]
[dev_pixelpipe] took 0.011 secs (0.024 CPU) processing `wejściowy profil 
koloru' [preview]
[dev_pixelpipe] took 0.058 secs (0.172 CPU) processing `cienie i światła' 
[preview]
[dev_pixelpipe] took 0.055 secs (0.172 CPU) processing `kontrast miejscowy' 
[preview]
[dev_pixelpipe] took 0.013 secs (0.024 CPU) processing `poziomy' [preview]
[dev_pixelpipe] took 0.012 secs (0.044 CPU) processing `wyostrzenie' [preview]
[dev_pixelpipe] took 0.019 secs (0.040 CPU) processing `wyjściowy profil 
koloru' [preview]
[dev_pixelpipe] took 0.012 secs (0.024 CPU) processing `velvia' [preview]
[dev_pixelpipe] took 0.003 secs (0.004 CPU) processing `prześwietlenie' 
[preview]
[dev_pixelpipe] took 0.004 secs (0.012 CPU) processing `gamma' [preview]
[dev_process_preview] pixel pipeline processing took 0.225 secs (0.580 CPU)

# WITH OPENCL ##################
[pixelpipe_process] [full] using device 0
[dev_pixelpipe] took 0.000 secs (0.000 CPU) initing base buffer [full]
[default_process_tiling_cl_ptp] use tiling on module 'temperature' for image 
with full size 4310 x 2868
[default_process_tiling_cl_ptp] (2 x 1) tiles with max dimensions 4096 x 2868 
and overlap 0
[default_process_tiling_cl_ptp] tile (0, 0) with 4096 x 2868 at origin [0, 0]
[default_process_tiling_cl_ptp] tile (1, 0) with 214 x 2868 at origin [4096, 0]
[dev_pixelpipe] took 0.029 secs (0.020 CPU) processing `balans bieli' [full]
[default_process_tiling_cl_ptp] use tiling on module 'highlights' for image 
with full size 4310 x 2868
[default_process_tiling_cl_ptp] (2 x 1) tiles with max dimensions 4096 x 2868 
and overlap 0
[default_process_tiling_cl_ptp] tile (0, 0) with 4096 x 2868 at origin [0, 0]
[default_process_tiling_cl_ptp] tile (1, 0) with 214 x 2868 at origin [4096, 0]
[dev_pixelpipe] took 0.051 secs (0.036 CPU) processing `ratowanie 
prześwietleń' [full]
[default_process_tiling_cl_roi] use tiling on module 'demosaic' for image with 
full input size 4310 x 2868
[default_process_tiling_cl_roi] (2 x 1) tiles with max input dimensions 4096 x 
2868
[default_process_tiling_cl_roi] tile (0, 0) with 2175 x 2868 at origin [0, 0]
[default_process_tiling_cl_roi] tile (1, 0) with 2161 x 2868 at origin [2148, 0]
[dev_pixelpipe] took 0.041 secs (0.016 CPU) processing `demozaikowanie' [full]
[dev_pixelpipe] took 0.013 secs (0.012 CPU) processing `krzywa bazowa' [full]
[dev_pixelpipe] took 0.008 secs (0.000 CPU) processing `wejściowy profil 
koloru' [full]
[dev_pixelpipe] took 0.029 secs (0.004 CPU) processing `cienie i światła' 
[full]
[dev_pixelpipe] took 1.058 secs (0.400 CPU) processing `kontrast miejscowy' 
[full]
[dev_pixelpipe] took 0.009 secs (0.000 CPU) processing `poziomy' [full]
[dev_pixelpipe] took 0.011 secs (0.000 CPU) processing `wyostrzenie' [full]
[dev_pixelpipe] took 0.012 secs (0.004 CPU) processing `wyjściowy profil 
koloru' [full]
[dev_pixelpipe] took 0.006 secs (0.004 CPU) processing `velvia' [full]
[dev_pixelpipe] took 0.003 secs (0.000 CPU) processing `prześwietlenie' [full]
[dev_pixelpipe] took 0.010 secs (0.020 CPU) processing `gamma' [full]
[opencl_profiling] spent  0.0340 seconds in [Write Image (from host to device)]
[opencl_profiling] spent  0.0077 seconds in whitebalance_1ui
[opencl_profiling] spent  0.0373 seconds in [Read Image (from device to host)]
[opencl_profiling] spent  0.0227 seconds in highlights_1f
[opencl_profiling] spent  0.0230 seconds in ppg_demosaic_green
[opencl_profiling] spent  0.0424 seconds in ppg_demosaic_redblue
[opencl_profiling] spent  0.0214 seconds in border_interpolate
[opencl_profiling] spent  0.0232 seconds in clip_and_zoom
[opencl_profiling] spent  0.0030 seconds in basecurve
[opencl_profiling] spent  0.0041 seconds in colorin
[opencl_profiling] spent  0.0015 seconds in [Copy Image to Buffer (on device)]
[opencl_profiling] spent  0.0112 seconds in gaussian_column_4c
[opencl_profiling] spent  0.0025 seconds in gaussian_transpose_4c
[opencl_profiling] spent  0.0012 seconds in [Copy Buffer to Image (on device)]
[opencl_profiling] spent  0.0027 seconds in shadows_highlights_mix
[opencl_profiling] spent  0.0000 seconds in zero
[opencl_profiling] spent  1.1477 seconds in splat
[opencl_profiling] spent  0.0000 seconds in [Copy Buffer to Buffer (on device)]
[opencl_profiling] spent  0.0003 seconds in blur_line
[opencl_profiling] spent  0.0010 seconds in blur_line_z
[opencl_profiling] spent  0.0036 seconds in slice
[opencl_profiling] spent  0.0028 seconds in levels
[opencl_profiling] spent  0.0032 seconds in sharpen_hblur
[opencl_profiling] spent  0.0028 seconds in sharpen_vblur
[opencl_profiling] spent  0.0026 seconds in sharpen_mix
[opencl_profiling] spent  0.0064 seconds in colorout
[opencl_profiling] spent  0.0021 seconds in velvia
[opencl_profiling] spent  0.0012 seconds in [Copy Image (on device)]
[opencl_profiling] spent  1.4115 seconds totally in command queue (with 0 
events missing)
[dev_process_image] pixel pipeline processing took 1.464 secs (0.648 CPU)
[pixelpipe_process] [preview] using device -1
[dev_pixelpipe] took 0.000 secs (0.000 CPU) initing base buffer [preview]
[dev_pixelpipe] took 0.004 secs (0.008 CPU) processing `balans bieli' [preview]
[dev_pixelpipe] took 0.003 secs (0.008 CPU) processing `ratowanie 
prześwietleń' [preview]
[dev_pixelpipe] took 0.007 secs (0.016 CPU) processing `krzywa bazowa' [preview]
[dev_pixelpipe] took 0.007 secs (0.024 CPU) processing `wejściowy profil 
koloru' [preview]
[dev_pixelpipe] took 0.057 secs (0.148 CPU) processing `cienie i światła' 
[preview]
[dev_pixelpipe] took 0.050 secs (0.160 CPU) processing `kontrast miejscowy' 
[preview]
[dev_pixelpipe] took 0.009 secs (0.028 CPU) processing `poziomy' [preview]
[dev_pixelpipe] took 0.015 secs (0.036 CPU) processing `wyostrzenie' [preview]
[dev_pixelpipe] took 0.018 secs (0.040 CPU) processing `wyjściowy profil 
koloru' [preview]
[dev_pixelpipe] took 0.011 secs (0.020 CPU) processing `velvia' [preview]
[dev_pixelpipe] took 0.003 secs (0.004 CPU) processing `prześwietlenie' 
[preview]
[dev_pixelpipe] took 0.003 secs (0.012 CPU) processing `gamma' [preview]
[dev_process_preview] pixel pipeline processing took 0.211 secs (0.528 CPU)


------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
darktable-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/darktable-devel

Reply via email to