Age | Commit message (Collapse) | Author |
|
|
|
|
|
|
|
|
|
This is not a real solution and colored dots still remain, just rejecting some
pixels that seem clearly wrong.
|
|
|
|
|
|
Always use native function since this was already the case due to
__CL_USE_NATIVE__ not being defined in time, and seems to have caused no
known issues.
|
|
|
|
Skip shader evaluation then, as we already do for lights. Less than
1% faster in my tests, but might as well be consistent for both.
|
|
No functional changes, logic here got too complex after many changes over
the years.
|
|
It's effectively always enabled, only not on some unsupported OpenCL devices.
For testing those it's not useful to disable these features. This is replaced
by the more fine grained feature toggles that we have now.
|
|
|
|
This version fixes various bugs, and there is no need anymore to use both
9.1 and 10.0 for different cards.
There is a bug related to WITH_CYCLES_CUBIN_COMPILER and bump mapping in the
regression tests, so that remains disabled same as it was for CUDA 10.0.
Fix T59286: CUDA bake failing on some cards.
Fix T56858: CUDA 9.2 and 10 issues.
|
|
|
|
The main goals of this change is faster starting when using foreground
rendering.
This patch will build kernels in parallel to the update process of
the scene. When these optimized kernels are not available (yet) an AO
kernel will be used.
These AO kernels are fast to compile (3-7 seconds) and can be
reused by all scenes. When the final kernels become available we
will switch to these kernels.
In background mode the AO kernels will not be used.
Some kernels are being used during Scene update (displace, background
light). When these kernels are being used the process can halt until
these become available.
Reviewed By: brecht, #cycles
Maniphest Tasks: T61752
Differential Revision: https://developer.blender.org/D4428
|
|
|
|
This patch puts threads that render the same pixel closer together,
as opposed to threads that render the same sample. Thus threads
within a warp are more coherent in memory access and control flow,
leading to performance improvements.
Example benchmarks on a Quadro RTX4000 (WDDM) on Windows 10:
Koro: 4:23 -> 3:46
BMW: 1:18 -> 1:25
Barbershop Interior: 17:52 -> 14:55
Classroom: 4:37 -> 3:45
Performance differences on OpenCL/AMD were hit and miss, some scenes
became faster, others lost significantly. Therefore, this is kept as
CUDA only change for now.
|
|
|
|
|
|
|
|
|
|
Conflicts:
intern/cycles/blender/addon/properties.py
intern/cycles/device/opencl/opencl_split.cpp
|
|
Part of the cleanup of the OpenCL codebase.
Single program is not effective when using OpenCL, it is slower
to compile and slower during rendering (when used in for example
`barbershop` or `victor`).
Reviewers: brecht, #cycles
Maniphest Tasks: T62267
Differential Revision: https://developer.blender.org/D4481
|
|
|
|
Float2 are now a new type for attributes in Cycles. Before, the choices
for attribute storage were float and float3, the latter padded to
float4. This meant that UV maps were inflated to twice the size
necessary.
Reviewers: brecht, sergey
Reviewed By: brecht
Subscribers: #cycles
Tags: #cycles
Differential Revision: https://developer.blender.org/D4409
|
|
|
|
The Lamp data was not always set. When using CUDA or CPU it was, but when using OpenCL
without `OBJECT_MOTION` `sd->lamp` not updated to the actual lamp. This made the
TextureCoordinate output the wrong normal when used in a light shader.
As the normal was incorrect it made the IES node render incorrectly.
(what is the default for the IES node).
By setting the lamp data when no `__OBJECT_MOTION__` compile directive is present makes
sure that the normal is correctly calculated.
Fix D4450
Reviewed By: Brecht van Lommel
|
|
|
|
The goal of this patch is to have limit the number of times
kernels needs to be compiled and are reused as kernels with
different compile directives can lead to identical same
binaries.
The implementation does this by stripping the compile directives.
and reshuffling kernels so the output is more likely to be the
same.
We focussed on the kernels where it was easy to detect and maintain
(bundle, bake, displace, do_volume and background). More optimizations
could be done but they are probably less obvious.
Merged the data_init and state_buffer_size kernels to split_bundle.
This patch will also remove empty kernels for do_volume and bake
when their features are not enabled.
When using the benchmark files there are less background, bake and
do_volume kernels compiled.
Fix: T61576, T61501, T61466
Reviewed By: brecht, #cycles
Differential Revision: https://developer.blender.org/D4390
|
|
|
|
|
|
The bake kernels are also used during mesh displacement and light
importance sampling. We disabled the implementation of these kernels
when baking was not enabled.
|
|
|
|
Was happening when looking for all intersections for transparent shadow rays
in the case the ray is degenerate.
Still quesitonable whether we should consider this a transparent or opaque
configuraiton. Ideally, we should prevent such rays from happening, but that
is another vector of debugging.
|
|
added missing quote, introduced in rB15edda3a8e07003bef695cca939744bbea80ad18
|
|
|
|
Using OpenCL MegaKernel has been slow and therefore not usefull.
This patch will remove the mega kernel from the OpenCL codebase
and the OpenCLDeviceBase class.
T61736: removal of mega kernel
T61703: baking does not work with mega kernel
Tags: #cycles
Differential Revision: https://developer.blender.org/D4383
|
|
Cycles OpenCL: Split baking kernels in own program
Fix T61463. Before this patch baking was part of the base kernels. There
are 3 baking kernels that and all 3 uses shader evaluation. Only for one
of these kernels the functionality was wrapped in the __NO_BAKING__
compile directive.
When you start baking this leads to long compile times. By separating
in individual programs will reduce the compile times.
Also wrapped all baking kernels with __NO_BAKING__ to reduce the
compilation times.
Impact on compilation time
job | scene_name | previous | new | percentage
--------+-----------------+----------+-------+------------
T61463 | empty | 10.63 | 7.27 | 32%
T61463 | bmw | 17.91 | 14.24 | 20%
T61463 | fishycat | 19.57 | 15.08 | 23%
T61463 | barbershop | 54.10 | 48.18 | 11%
T61463 | classroom | 17.55 | 14.42 | 18%
T61463 | koro | 18.92 | 17.15 | 9%
T61463 | pavillion | 17.43 | 14.23 | 18%
T61463 | splash279 | 16.48 | 15.33 | 7%
T61463 | volume_emission | 36.22 | 34.19 | 6%
Impact on render time
job | scene_name | previous | new | percentage
--------+-----------------+----------+---------+------------
T61463 | empty | 21.06 | 20.54 | 2%
T61463 | bmw | 198.44 | 189.59 | 4%
T61463 | fishycat | 394.20 | 388.50 | 1%
T61463 | barbershop | 1188.16 | 1185.49 | 0%
T61463 | classroom | 341.08 | 339.27 | 1%
T61463 | koro | 472.43 | 360.70 | 24%
T61463 | pavillion | 905.77 | 902.14 | 0%
T61463 | splash279 | 55.26 | 54.92 | 1%
T61463 | volume_emission | 62.59 | 39.09 | 38%
I don't have a grounded explanation why koro and volume_emission is this much
faster; I have done several tests though...
Maniphest Tasks: T61463
Differential Revision: https://developer.blender.org/D4376
|
|
Cycles OpenCL: Split baking kernels in own program
Fix T61463. Before this patch baking was part of the base kernels. There
are 3 baking kernels that and all 3 uses shader evaluation. Only for one
of these kernels the functionality was wrapped in the __NO_BAKING__
compile directive.
When you start baking this leads to long compile times. By separating
in individual programs will reduce the compile times.
Also wrapped all baking kernels with __NO_BAKING__ to reduce the
compilation times.
Impact on compilation time
job | scene_name | previous | new | percentage
--------+-----------------+----------+-------+------------
T61463 | empty | 10.63 | 7.27 | 32%
T61463 | bmw | 17.91 | 14.24 | 20%
T61463 | fishycat | 19.57 | 15.08 | 23%
T61463 | barbershop | 54.10 | 48.18 | 11%
T61463 | classroom | 17.55 | 14.42 | 18%
T61463 | koro | 18.92 | 17.15 | 9%
T61463 | pavillion | 17.43 | 14.23 | 18%
T61463 | splash279 | 16.48 | 15.33 | 7%
T61463 | volume_emission | 36.22 | 34.19 | 6%
Impact on render time
job | scene_name | previous | new | percentage
--------+-----------------+----------+---------+------------
T61463 | empty | 21.06 | 20.54 | 2%
T61463 | bmw | 198.44 | 189.59 | 4%
T61463 | fishycat | 394.20 | 388.50 | 1%
T61463 | barbershop | 1188.16 | 1185.49 | 0%
T61463 | classroom | 341.08 | 339.27 | 1%
T61463 | koro | 472.43 | 360.70 | 24%
T61463 | pavillion | 905.77 | 902.14 | 0%
T61463 | splash279 | 55.26 | 54.92 | 1%
T61463 | volume_emission | 62.59 | 39.09 | 38%
I don't have a grounded explanation why koro and volume_emission is this much
faster; I have done several tests though...
Maniphest Tasks: T61463
Differential Revision: https://developer.blender.org/D4376
|
|
There is a generic function to retrieve float and float3 attributes
`primitive_attribute_float` and primitive_attribute_float3`. Inside
these functions an prioritised if-else construction checked where
the attribute is stored and then retrieved from that location.
Actually the calling function most of the time already knows where
the data is stored. So we could simplify this by splitting these
functions and remove the check logic.
This patch splits the `primitive_attribute_float?` functions into
`primitive_surface_attribute_float?` and `primitive_volume_attribute_float?`.
What leads to less branching and more optimum kernels.
The original function is still being used by OSL and `svm_node_attr`.
This will reduce the compilation time and render time for kernels.
Especially in production scenes there is a lot of benefit.
Impact in compilation times
job | scene_name | previous | new | percentage
-------+-----------------+----------+-------+------------
t61513 | empty | 10.63 | 10.66 | 0%
t61513 | bmw | 17.91 | 17.65 | 1%
t61513 | fishycat | 19.57 | 17.68 | 10%
t61513 | barbershop | 54.10 | 24.41 | 55%
t61513 | classroom | 17.55 | 16.29 | 7%
t61513 | koro | 18.92 | 18.05 | 5%
t61513 | pavillion | 17.43 | 16.52 | 5%
t61513 | splash279 | 16.48 | 14.91 | 10%
t61513 | volume_emission | 36.22 | 21.60 | 40%
Impact in render times
job | scene_name | previous | new | percentage
-------+-----------------+----------+--------+------------
61513 | empty | 21.06 | 20.35 | 3%
61513 | bmw | 198.44 | 190.05 | 4%
61513 | fishycat | 394.20 | 401.25 | -2%
61513 | barbershop | 1188.16 | 912.39 | 23%
61513 | classroom | 341.08 | 340.38 | 0%
61513 | koro | 472.43 | 471.80 | 0%
61513 | pavillion | 905.77 | 899.80 | 1%
61513 | splash279 | 55.26 | 54.86 | 1%
61513 | volume_emission | 62.59 | 61.70 | 1%
There is also a possitive impact when using CPU and CUDA, but they are small.
I didn't split the hair logic from the surface logic due to:
* Hair and surface use same attribute types. It was not clear if it could be
splitted when looking at the code only.
* Hair and surface are quick to compile and to read. So the benefit is quite
small.
Differential Revision: https://developer.blender.org/D4375
|
|
There is a generic function to retrieve float and float3 attributes
`primitive_attribute_float` and primitive_attribute_float3`. Inside
these functions an prioritised if-else construction checked where
the attribute is stored and then retrieved from that location.
Actually the calling function most of the time already knows where
the data is stored. So we could simplify this by splitting these
functions and remove the check logic.
This patch splits the `primitive_attribute_float?` functions into
`primitive_surface_attribute_float?` and `primitive_volume_attribute_float?`.
What leads to less branching and more optimum kernels.
The original function is still being used by OSL and `svm_node_attr`.
This will reduce the compilation time and render time for kernels.
Especially in production scenes there is a lot of benefit.
Impact in compilation times
job | scene_name | previous | new | percentage
-------+-----------------+----------+-------+------------
t61513 | empty | 10.63 | 10.66 | 0%
t61513 | bmw | 17.91 | 17.65 | 1%
t61513 | fishycat | 19.57 | 17.68 | 10%
t61513 | barbershop | 54.10 | 24.41 | 55%
t61513 | classroom | 17.55 | 16.29 | 7%
t61513 | koro | 18.92 | 18.05 | 5%
t61513 | pavillion | 17.43 | 16.52 | 5%
t61513 | splash279 | 16.48 | 14.91 | 10%
t61513 | volume_emission | 36.22 | 21.60 | 40%
Impact in render times
job | scene_name | previous | new | percentage
-------+-----------------+----------+--------+------------
61513 | empty | 21.06 | 20.35 | 3%
61513 | bmw | 198.44 | 190.05 | 4%
61513 | fishycat | 394.20 | 401.25 | -2%
61513 | barbershop | 1188.16 | 912.39 | 23%
61513 | classroom | 341.08 | 340.38 | 0%
61513 | koro | 472.43 | 471.80 | 0%
61513 | pavillion | 905.77 | 899.80 | 1%
61513 | splash279 | 55.26 | 54.86 | 1%
61513 | volume_emission | 62.59 | 61.70 | 1%
There is also a possitive impact when using CPU and CUDA, but they are small.
I didn't split the hair logic from the surface logic due to:
* Hair and surface use same attribute types. It was not clear if it could be
splitted when looking at the code only.
* Hair and surface are quick to compile and to read. So the benefit is quite
small.
Differential Revision: https://developer.blender.org/D4375
|
|
This patch implements a workaround to get the multithreaded compilation from D2231 working.
So far, it only works for Blender, not for Cycles Standalone. Also, I have only tested the Linux codepath in the helper function.
Depends on D2231.
Patch by lukasstockner97, jbakker, brecht
job | scene_name | compilation_time
----------+-----------------+------------------
Baseline | empty | 22.73
D2264 | empty | 13.94
Baseline | bmw | 56.44
D2264 | bmw | 41.32
Baseline | fishycat | 59.50
D2264 | fishycat | 45.19
Baseline | barbershop | 212.28
D2264 | barbershop | 169.81
Baseline | victor | 67.51
D2264 | victor | 53.60
Baseline | classroom | 51.46
D2264 | classroom | 39.02
Baseline | koro | 62.48
D2264 | koro | 49.03
Baseline | pavillion | 54.37
D2264 | pavillion | 38.82
Baseline | splash279 | 47.43
D2264 | splash279 | 37.94
Baseline | volume_emission | 145.22
D2264 | volume_emission | 121.10
This patch reduced compilation time as the split kernels and base
kernels are compiled in parallel. In cycles debug mode (256) you can set
unmark the opencl single program file, what reduces the compilation time
even further (bmw 17 seconds, barbershop 53 seconds).
Reviewers: brecht, dingto, sergey, juicyfruit, lukasstockner97
Reviewed By: brecht
Subscribers: Loner, jbakker, candreacchio, 3dLuver, LazyDodo, bliblubli
Differential Revision: https://developer.blender.org/D2264
|
|
This patch implements a workaround to get the multithreaded compilation from D2231 working.
So far, it only works for Blender, not for Cycles Standalone. Also, I have only tested the Linux codepath in the helper function.
Depends on D2231.
Reviewers: brecht, dingto, sergey, juicyfruit, lukasstockner97
Reviewed By: brecht
Subscribers: Loner, jbakker, candreacchio, 3dLuver, LazyDodo, bliblubli
Differential Revision: https://developer.blender.org/D2264
|
|
|
|
|
|
We should clamp the result after multiplication.
|
|
|
|
Values outside the 0..1 range produce negative colors, so now clamp to that
range everywhere. Also fixes improper handling of hue > 2.0 in some places.
|
|
|