This was more of a general nvidia driver bug. Was caused by a
drawcall that had 0 vertex count. This worked in normal drawcalls
but not in indirect drawcalls.
For clarity sake, the batch cache now uses exclusively per Loop attributes.
While this is a bit of a waste of VRAM (for the few case where per vert
attribs are enough) it reduces the complexity and amount of overall VBO
to update in general situations.
This patch also makes the VertexBuffers filling multithreaded. This make
the update of dense meshes a bit faster. The main bottleneck is the
IndexBuffers update which cannot be multithreaded efficiently (have to
increment a counter and/or do a final sorting pass).
We introduce the concept of "extract" functions/step.
All extract functions are executed in one thread each and if possible,
using multiple thread for looping over all elements.
Reviewed By: brecht
Differential Revision: http://developer.blender.org/D5424
The problem is that the `glDrawArraysInstancedBaseInstance` is ignoring the last parameter.
The solution is to indicate that `GLEW_ARB_base_instance` is not supported in these cases.
Reviewers: fclem, brecht, jbakker
Reviewed By: fclem, brecht
Differential Revision: https://developer.blender.org/D5383
Restart index can have been changed in another context and the static
var can get out of sync. A better solution is to set the restart index
when binding the VAO. It also have less perf impact.
Fix T65364 Corrupted mesh display on macOS
This type of indices is not natively supported on modern GPU and
gives warning on some implementation. The memory savings it
provides is also quite minimal and unlikely to be visible on
nowadays hardware.
This remove some uneeded struct members and makes primitive
restart always enabled by default. This can be broken by addons
if they are not careful enough but many other states have this
problem.
Also leverage GL_PRIMITIVE_RESTART_FIXED_INDEX if
ARB_ES3_compatibility is supported. This removes all API calls
to change restart index depending on indices length.
Move free callback call to GPU_batch_discard to prevent the crash.
The issue was that clearing can happen after referencing to an instance
buffer and that's perfectly legal.
BF-admins agree to remove header information that isn't useful,
to reduce noise.
- BEGIN/END license blocks
Developers should add non license comments as separate comment blocks.
No need for separator text.
- Contributors
This is often invalid, outdated or misleading
especially when splitting files.
It's more useful to git-blame to find out who has developed the code.
See P901 for script to perform these edits.
This is in order to make the API more multithread friendly inside the
draw manager.
GPU_shader_get_uniform will only serve to query the shader interface and
not do any GL call, making it threadsafe.
For now it only print a warning if the uniform was not queried before.
GPUFrameBuffers were being free when no context was attached or in the
wrong gl context. This make sure this does not happen again.
You can now safely free any gl resource from any thread (well as long as
it's not used anymore!).
This does not include all the struct and type renaming. Only files were
renamed.
gwn_batch.c/h was fusioned with GPU_batch.c/h
gwn_immediate.c/h was fusioned with GPU_immediate.c/h
gwn_imm_util.c/h was fusioned with GPU_immediate_util.c/h
