Subversion Repository Public Repository

Divide-Framework

This repository has no backups
This repository's network speed is throttled to 100KB/sec

Entry Size Revision Updated
Buffers 765 Mon 22 Aug, 2016 15:48:12 +0000
Direct3D 762 Thu 11 Aug, 2016 16:18:35 +0000
Headers 770 Tue 20 Sep, 2016 16:12:27 +0000
Null 334 Wed 10 Dec, 2014 17:21:22 +0000
OpenGL 770 Tue 20 Sep, 2016 16:12:27 +0000
Shaders 770 Tue 20 Sep, 2016 16:12:27 +0000
Textures 759 Sun 07 Aug, 2016 18:34:32 +0000
GenericCommandPool.cpp 382 bytes 726 Fri 10 Jun, 2016 16:13:52 +0000
GenericDrawCommand.cpp 4.1K 767 Mon 22 Aug, 2016 16:25:37 +0000
GFXDevice.cpp 33K 770 Tue 20 Sep, 2016 16:12:27 +0000
GFXDeviceDebug.cpp 6.8K 770 Tue 20 Sep, 2016 16:12:27 +0000
GFXDeviceDraw.cpp 16.9K 767 Mon 22 Aug, 2016 16:25:37 +0000
GFXDeviceObjects.cpp 8.1K 734 Tue 21 Jun, 2016 16:28:11 +0000
GFXDeviceState.cpp 18.9K 770 Tue 20 Sep, 2016 16:12:27 +0000
GFXRTPool.cpp 2.3K 770 Tue 20 Sep, 2016 16:12:27 +0000
GFXState.cpp 1.2K 714 Thu 26 May, 2016 16:26:23 +0000
IMPrimitive.cpp 6.3K 721 Wed 08 Jun, 2016 15:47:49 +0000
RenderStateBlock.cpp 5.6K 750 Thu 07 Jul, 2016 16:02:03 +0000
ScopedStates.cpp 586 bytes 762 Thu 11 Aug, 2016 16:18:35 +0000

Commits for Divide-Framework/trunk/Source Code/Platform/Video

Revision Author Commited Message
770 IonutCava picture IonutCava Tue 20 Sep, 2016 16:12:27 +0000

[IonutCava]
- Prepare velocity map calculation (later used for temporal AA and motion blur):
— Add a previous depth buffer render target that blits the main depth buffer at the end of the frame
— Add a new render target to the screen FBO: velocity map
- Store normals as RG instead of RGB and pack/unpack in shaders. (compute >>>> bandwith)
- Rename Forward+ renderer to something more descriptive

(Requires deletion of all parsed & cached data)

769 IonutCava picture IonutCava Fri 16 Sep, 2016 15:18:09 +0000

[IonutCava]
- Clean up camera class:
— Remove “reflection” specific code and add a generic “reflect” method to allow the calling code to handle camera placements and reflections. This got rid of the “reflectionPlane” member in the RenderParams struct
— Remove some member variables and replace them with on-demand computed variables on the stack.
— Add a “destroyCamera” call to the CameraManager to allow better resource cleanup when needed
- Prepare a rendering system overhaul:
— Add dual screen render targets: current frame and previous frame. Used for future features such as TXAA and Motion Blur.
— Prepare code for an implementation similar to the Doom (2016) system: http://www.adriancourreges.com/blog/2016/09/09/doom-2016-graphics-study/ (the code was already going in that direction)

768 IonutCava picture IonutCava Fri 09 Sep, 2016 14:38:51 +0000

[IonutCava]
- Refactored framework entry points (init, step, shutdown) to allow the addition of hot code reloading in the future

767 IonutCava picture IonutCava Mon 22 Aug, 2016 16:25:37 +0000

[IonutCava]
- Descriptor/State cleanup part 3/3: GenericDrawCommand cleanup

765 IonutCava picture IonutCava Mon 22 Aug, 2016 15:48:12 +0000

[IonutCava]
- Descriptor/State cleanup part 1/3: RTDrawDescriptor refactoring

764 IonutCava picture IonutCava Mon 22 Aug, 2016 15:38:11 +0000

[IonutCava]
- Add some explicit include for RenderStateBlock as it will be removed from the RenderAPIWrapper.h file

763 IonutCava picture IonutCava Mon 22 Aug, 2016 15:34:32 +0000

[IonutCava]
- Small fix to sampler binding

762 IonutCava picture IonutCava Thu 11 Aug, 2016 16:18:35 +0000

[IonutCava]
- Removed a lot of high level GFX state and merged it back on a lower level:
— depth mask is now a framebuffer specific toggle controlled by the RenderPolicy
— rasterization is now a per-buffer draw switch toggled by a rendering flag in GenericDrawCommand
- Replaced old style GL texture binding code with DSA style glBindTextures and glBindSamplers even for single textures
— completely removed the concept of a active texture unit in the GL code
- Fixed some draw policy settings that were clearing the depth buffer in the PostFX passes
- More build type flag replacement of macros
- Render pass system bug fixing
- Add option to skip occlusion culling for certain nodes

761 IonutCava picture IonutCava Wed 10 Aug, 2016 16:24:04 +0000

[IonutCava]
- More work on the new render pass system
- Added new HiZ culling code from nvidia samples: https://github.com/nvpro-samples/gl_occlusion_culling

760 IonutCava picture IonutCava Tue 09 Aug, 2016 16:30:12 +0000

[IonutCava]
- Complete rewrite of RenderPass system:
— Allow every rendering pass (reflection, environment, etc) to use the same rendering system: occlusion culling, light culling, etc
— Define render passes by a param based token similar to modern API systems (DX12/Vulkan)
— W.I.P.: Depth rendering slightly wrong; Environment Mapping ,reflection and refraction rendering currently disabled
- Add compile time variables for DEBUG/PROFILE/RELEASE builds to validate all code for every build