- User Since
- Sep 4 2007, 4:11 PM (649 w, 6 d)
Nov 6 2019
Yes. Using the the "spring" scene as test again, the same error occures with only the screen driving gpu should render.
So the issue seems not to be the balancing between differently VRAM equipped gpu's, your fix looks okay here.
I now follow your assumption that its another issue.
I can get the scene to render by either reducing the haircount or other simplifications that reduce memory usage.
I think now that the memory used for the attched screens is not taken into account. ( ? )
Unfortunately this does not fix the Optix render if one of the gpu's is driving the screen.
I get with 2 exact same RTX 2080 and the spring test scene:
Aug 3 2019
Mar 15 2019
Nov 24 2018
Jul 26 2018
May 2 2018
Feb 25 2018
Feb 24 2018
Feb 4 2018
As we still are default c99, a patch is needed:
Jan 14 2018
In Linux or macOS i would make those symbols local in the exports map.
I suggest using a windows technique to archieve here those symbols are local.
This would avoid clashes.
Nov 6 2017
Nov 4 2017
Nov 3 2017
Oct 13 2017
I use a quickfix for now cause it seems rle is always handled/expected with alpha in blender:
May 9 2017
Apr 28 2017
Apr 26 2017
Apr 25 2017
Apr 22 2017
Apr 21 2017
Apr 15 2017
Mar 23 2017
Feb 25 2017
Feb 22 2017
Feb 12 2017
Okay, the bell rang loud here so i checkout one commit before "fd7e9f7 Cycles: Fix pointiness attribute giving wrong results with autosplit" and this worked again. So the culprit is found and the task attached to the right person :P
Did some more investigation and extracted a simplified testcase from victor. It consist only one tree + material. The hang goes away if you either disable multires or cut both pointiness routes !
Feb 11 2017
To be more precise:
I disabled renderability of multires/subsurf for some "critical"
objects ( GEO_tree_sculpt and GEO_rocked_detailed ) and the sychronize is way faster an i came to render again.
Also: Exactly whats stallling is only the "synchronize" part.
Feb 9 2017
I tested the latest patch and found it working nicely.
It behaves a bit different than before where i could just add a bit tranparency to the shader to preserve
This can now be obtained other way ( see image ) so i would say all works as expected.
Also the hdri MIS problem when mixing various lamp types seems to be gone.
Made a lot of A/B comparison to find odd effects from SC to no avail.
Feb 2 2017
Please give the full log from the crash here, the simple bugreport always shifts the true culprit.
10.7 should work.
Blindly guessing such crashes on start are most likely from fs_menu.
You could try disable in finder sidebar such as iCloud and "find all my files"
Jan 27 2017
Could not reproduce on linux, worked as expected with any build.
Dec 15 2016
Oct 4 2016
Oct 2 2016
Aug 15 2016
Jul 22 2016
Jul 21 2016
Jul 20 2016
Jul 19 2016
Jul 18 2016
Jul 14 2016
Jul 7 2016
Jul 3 2016
Jul 2 2016
Jul 1 2016
I reconfigured my buildbot for the new alembic-1.6/hdf5-less method now, first of those type:
Jun 25 2016
Jun 17 2016
Tested your files in last linux build:
Flat geo reads at 60+ fps whereas packed object reach ~24fps.
So packed IS slower but somewhat acceptable on an actual highend cpu ?
Jun 11 2016
Until i abandoned OSX last years may, i always used to work in blender exclusively
with a wacom graphire4. I never experienced such problem, had used both pen buttons
always. ( but not flipped )
So this may be a new issue ?
Jun 8 2016
OSX build also available and so far tested to be on par with linux:
Jun 3 2016
Sep 8 2015
Jun 15 2015
ideasman proposal to change line 1096 in colortools.c to:
- thread #1: tid = 0x4e8b5, 0x0000000100f09f81 blender`save_sample_line(scopes=0x000000010e121668, idx=93182, fx=0.532291651, rgb=0x00007fff5fbadd90, ycc=0x00007fff5fbadd84) + 97 at colortools.c:943, queue = 'com.apple.main-thread', stop reason = EXC_BAD_ACCESS (code=1, address=0x12c005000)
- frame #0: 0x0000000100f09f81 blender`save_sample_line(scopes=0x000000010e121668, idx=93182, fx=0.532291651, rgb=0x00007fff5fbadd90, ycc=0x00007fff5fbadd84) + 97 at colortools.c:943 frame #1: 0x0000000100f09783 blender`scopes_update(scopes=0x000000010e121668, ibuf=0x000000010fc86608, view_settings=0x000000010e1362a8, display_settings=0x000000010e136348) + 3635 at colortools.c:1187 frame #2: 0x00000001001ae23f blender`ED_space_image_scopes_update(C=0x00006080000c0078, sima=0x000000010e121608, ibuf=0x000000010fc86608, use_view_settings=true) + 223 at image_edit.c:305 frame #3: 0x00000001001bba49 blender`image_tools_area_draw(C=0x00006080000c0078, ar=0x000000010e84d078) + 233 at space_image.c:904 frame #4: 0x000000010073e725 blender`ED_region_do_draw(C=0x00006080000c0078, ar=0x000000010e84d078) + 549 at area.c:517 frame #5: 0x00000001001339c4 blender`wm_method_draw_triple(C=0x00006080000c0078, win=0x000000010e837b68) + 612 at wm_draw.c:640 frame #6: 0x0000000100132bcf blender`wm_draw_update(C=0x00006080000c0078) + 399 at wm_draw.c:1047 frame #7: 0x00000001001308c0 blender`WM_main(C=0x00006080000c0078) + 48 at wm.c:493 frame #8: 0x000000010012be09 blender`main(argc=1, argv=0x00007fff5fbff490) + 1497 at creator.c:1917 frame #9: 0x000000010012b81c blender`start + 52
This is how notifications work normally: they inform you about bg tasks.
I tried different instancing too: 2 times started blender from binary in dock - strted blender 2.74 and 2.75rc -> always correct behaviour
Perhaps check your notification center for settings, but 'am not aware you could affect this.
All is like it should be !
Jun 7 2015
@ sebastian: use "thread apply all bt"
May 17 2015
Tested on 10.10 with bluetooth trackpad.
Works as expected.
May 16 2015
Cannot reproduce, MacPro 5.1, AMD 7970
May 13 2015
May 9 2015
@ Dalai: the python binary i added to darwin libs is fully static against python and has nothing todo with python_framework.
Place it in the bundle in an apropiate way ( ....2.47/python/bin ) and you are done.
Now you must just call it instead of new blender py instances for subprocesses ....
May 1 2015
Apr 30 2015
Filip, you described it good enough.
I was just too tired to get it first time just reading ( late night pebkac )
The screenshot opened my eyes :-)
Apr 29 2015
Closed by commit rBce32aae80c5c.
Confirm dir named "öäü" also not showing up !
Trying to fix now....
Now i first time understand what you mean :-)
Have to find a way to reproduce this.
Apr 28 2015
I checked here with german umlauts öäü and åœ etc.
Could it be it is simply forgotten in your translation set ?
Apr 24 2015
2.74 release is working okay with testconditions
Crashpoint is here:
I confirm crash with same report on my MacPro/AMD
Apr 6 2015
Mar 28 2015
Mar 26 2015
Afaik the accessible buttonfunctions in ndof are hardcoded. Mike Erwin should know more here.
Mar 25 2015
Okay, i started testing on OSX.
Made the needed g++ adaption ( foreach ... ) and fixed doublespaces between definitions.
The kernels seem to compile fine, although i run into a sigbart.
I know OSX 10.10 drivers are a bit behind ( sigh ), but in luxrays ( 'am one of the lux dev's ) this works
fine so far ( we have long time microkernels ;-) ), except on extra huge scenes ( also OSX prob only ).