Blender Git Loki

Kaikki Blender Git kommitit.

Page: 4249 / 8462

August 9, 2016, 09:33 (GMT)
Half Floats: Initial support for CUDA.

This is not working yet, due to an error on rendertime, I assume it's an error on reading the texture. If I render a scene, with the half float texture outside of the camera, it works.

* Enable half floats for Kepler cards (1024 for half and 1024 for half4)
* Use proper half data type for CUDA, don't declare our own.
* Change __float2half_rn to __float2half, which should be the same according to http://stackoverflow.com/questions/35198856/half-precision-difference-between-float2half-vs-float2half-rn
August 9, 2016, 09:16 (GMT)
Merge remote-tracking branch 'origin/master' into soc-2016-cycles_images
Revision f3d65ad by Mike Erwin (master)
August 9, 2016, 07:29 (GMT)
OpenGL: debug context on Windows

Enable based on --debug-gpu at the command line. Linux already works
this way.
Revision b7f3fb0 by Mike Erwin (master)
August 9, 2016, 06:27 (GMT)
Gawain: fix Windows build error

MSVC is more strict than gcc or clang about pointer arithmetic. Also
fixed pointer cast warnings.
Revision a77e775 by Mike Erwin (master)
August 9, 2016, 05:29 (GMT)
OpenGL: enhance debug output

When running blender --debug-gpu

Display which debug facilities are available. One of these, in order of preference:
- OpenGL 4.3
- KHR_debug
- ARB_debug_output
- AMD_debug_output

All messages are logged now, not just errors. Will probably turn some of these off later.

GL_DEBUG_OUTPUT_SYNCHRONOUS lets us break on errors and backtrace to the exact trouble spot.

Callers of GPU_string_marker no longer pass in a message length, just the message itself (null terminated).

Apple provides no GL debug logging features.
August 9, 2016, 02:38 (GMT)
Merge branch 'master' into temp_localview_split
August 9, 2016, 02:32 (GMT)
Cycles: Fix crash in empty scenes

To recreate, just delete default cube and render.

Simple fix, but @maiself or @sergey may want to double-check ;)
August 9, 2016, 01:50 (GMT)
Cycles: Implement SSE3-optimized NLM prefiltering kernel
August 9, 2016, 01:50 (GMT)
Cycles: Implement SSE3-optimized denoising kernel
August 9, 2016, 01:50 (GMT)
Cycles: Add a few SSE utilities
August 9, 2016, 01:50 (GMT)
Cycles: Move denoising utility functions into a separate file
August 9, 2016, 01:50 (GMT)
Cycles: Move prefiltering functions into a separate file
August 9, 2016, 01:50 (GMT)
Cycles: Fix preprocessor directives around SSE3 replacement functions

The code is supposed to implement replacements for a few SSE4.1-specific functions so that they can be used with SSE3 as well.
Therefore, it was enabled when __KERNEL_SSE3__ was set, but __KERNEL_SSE4__ wasn't.

However, __KERNEL_SSE4__ is never set anywhere - the correct one is __KERNEL_SSE41__.
Because of that, the replacements were enabled for SSE4.1 and better (AVX) as well, where they're not needed, but only slow things down.
August 9, 2016, 01:50 (GMT)
Cycles: Fix denoising feature matrix norm calculation
August 9, 2016, 01:50 (GMT)
Cycles: Remove a few useless temporary files
August 9, 2016, 01:50 (GMT)
Cycles: Fix GGX shader after recent changes to numerical stability
August 9, 2016, 01:50 (GMT)
Cycles: Fix building after the debug_fpe commit
August 8, 2016, 23:39 (GMT)
Added initial "Use concave hulls" option to operator
August 8, 2016, 23:31 (GMT)
Fix Cycles CUDA adaptive kernel not working correctly after recent closure changes.
August 8, 2016, 22:57 (GMT)
Alembic: fix finding boost headers for win32
Tehnyt: Miika HämäläinenViimeksi päivitetty: 07.11.2014 14:18MiikaH:n Sivut a.k.a. MiikaHweb | 2003-2021