User Details
- User Since
- Oct 2 2016, 4:45 PM (149 w, 6 d)
Feb 10 2019
@roberto (robertora)
There are newer drivers from 2015 (Catalyst) and 2016 (Crimson) - link. YMMV, sometimes newer drivers make things worse.
May 31 2018
Sep 16 2017
Jun 13 2017
Jun 11 2017
May 17 2017
May 11 2017
May 5 2017
May 3 2017
Rebased on master. No shader sorting for CUDA split kernel.
May 2 2017
Apr 25 2017
Apr 21 2017
Updated all minor stuff.
Apr 14 2017
shader_bsdf_blur is ok, my mistake.
- Kernels fail to compile when using PT with volume, and when using BPT without subsurface.
- The body of kernel_lamp_emission should be skipped during regular BPT iteration.
- In kernel_holdout_... houldout and kernel_write_passes should be skipped during indirect BPT iteration.
- In same kernel shader_bsdf_blur should be skipped during regular BPT iteration.
Apr 12 2017
- Renamed KSES_* defines to more meaningful ones
- Moved SHADER_EVAL_SORT_SIZE to kernel_types.h
- Moved local variables to __kernel function
- Fixed style issues
Apr 4 2017
Apr 1 2017
Mar 29 2017
Mar 24 2017
Cause is unsafe compiler options. Changing it to a safe one does performance penalty 1%.
Mar 23 2017
Already in master rB57e26627c4.
This was due to a combination of division by zero and unsafe OpenCL optimization options.
Mar 13 2017
Mar 9 2017
@James W E Bird (3dLuver) Please also share the git hash (first 8 letters) of the master which you use.
Mar 8 2017
Feb 3 2017
@Mai Lavelle (maiself)
Volume scattering doesn't work due to NO_TRANSPARENT_SHADOWS.
Subsurface scattering works ok when using a single program - all kernels in one file.
Dec 10 2016
Update against latest master.
Update against latest master.
Nov 14 2016
Changing the behavior of atomic_inc is misleading, both CUDA & OpenCL return the old value. Further, using atomic_inc_uint32 and atomic_add_uint32 produces less readable code, since we have to subtract what was just added. Maybe there should be indication in atomic_ops.h, i.e. atomic_add_and_get_uint32.
Nov 10 2016
The patch had wrong newline characters.
The patch had wrong newline characters.
Update against latest master. Fixed queue bug. Fixed indentation.
Nov 8 2016
Fixed missing ccl_fetch.
Updated to work with current master.
Oct 25 2016
Fixed more indentation.
Fixed queuing and indentation. SOA is not supported.
Oct 24 2016
In kernel_shadow_blocked.h must use SD_THREAD instead of ray_index .
Oct 23 2016
Fixed indentation, a typo, removed shadow_blocked_addrspace.
Oct 21 2016
Switched back to uint4. Also fixed a bug - missing offset.
Oct 20 2016
More missing ccl_global.
Added missing ccl_global definitions.
Minor changes: added missing rays state in queue_enqueue.
Oct 18 2016
Currently it's buggy (default cube with quick smoke+fire at frame 30):
Fixed kernel_image_3d_interp calls for OpenCL.
Oct 14 2016
Tested only Split kernel, Mega should work, kernel_image_opencl.h is included. There might be problems with CUDA though.