319ff28b7b
Motivation: When discussing with @Jeroen-Bakker and @aras_p about how to approach sorting when rendering Gaussian splats in Blender, we realised that compute shaders could help there (and have many other use cases), and that also due to Blender 4.0 being on OpenGL >= 4.3, we can now rely on compute shaders existing. This PR is an initial pass for that functionality. It comes with a Python example, which runs a compute shader and saves the output to a texture, which is then rendered in the viewport. There is no exposed support for storage buffers yet, but I expect I'll be able to work on them soon if this is accepted. The newly added parts are: 1. `gpu.compute.dispatch()` 2. a way set the compute source to `GPUShaderCreateInfo`: `shader_info.compute_source()` 3. a way to set the image store for `GPUShaderCreateInfo`: `shader_info.image()` 4. a way to set the `local_group_size` for `GPUShaderCreateInfo`: `shader_info.local_group_size(x,y,z)` 5. a way to get `max_work_group_size` from capabilities: `gpu.capabilities.max_work_group_size_get(index)` 6. a way to get `max_work_group_count` from capabilities: `gpu.capabilities.max_work_group_count_get(index)` Pull Request: https://projects.blender.org/blender/blender/pulls/114238
389 lines
12 KiB
C++
389 lines
12 KiB
C++
/* SPDX-FileCopyrightText: 2023 Blender Authors
|
|
*
|
|
* SPDX-License-Identifier: GPL-2.0-or-later */
|
|
|
|
/** \file
|
|
* \ingroup bpygpu
|
|
*
|
|
* - Use `bpygpu_` for local API.
|
|
* - Use `BPyGPU` for public API.
|
|
*/
|
|
|
|
#include <Python.h>
|
|
|
|
#include "BLI_utildefines.h"
|
|
|
|
#include "GPU_capabilities.h"
|
|
|
|
#include "gpu_py.h"
|
|
#include "gpu_py_capabilities.h" /* own include */
|
|
|
|
/* -------------------------------------------------------------------- */
|
|
/** \name Functions
|
|
* \{ */
|
|
|
|
PyDoc_STRVAR(pygpu_max_texture_size_get_doc,
|
|
".. function:: max_texture_size_get()\n"
|
|
"\n"
|
|
" Get estimated maximum texture size to be able to handle.\n"
|
|
"\n"
|
|
" :return: Texture size.\n"
|
|
" :rtype: int\n");
|
|
static PyObject *pygpu_max_texture_size_get(PyObject * /*self*/)
|
|
{
|
|
return PyLong_FromLong(GPU_max_texture_size());
|
|
}
|
|
|
|
PyDoc_STRVAR(pygpu_max_texture_layers_get_doc,
|
|
".. function:: max_texture_layers_get()\n"
|
|
"\n"
|
|
" Get maximum number of layers in texture.\n"
|
|
"\n"
|
|
" :return: Number of layers.\n"
|
|
" :rtype: int\n");
|
|
static PyObject *pygpu_max_texture_layers_get(PyObject * /*self*/)
|
|
{
|
|
return PyLong_FromLong(GPU_max_texture_layers());
|
|
}
|
|
|
|
PyDoc_STRVAR(pygpu_max_textures_get_doc,
|
|
".. function:: max_textures_get()\n"
|
|
"\n"
|
|
" Get maximum supported texture image units used for\n"
|
|
" accessing texture maps from the vertex shader and the\n"
|
|
" fragment processor.\n"
|
|
"\n"
|
|
" :return: Texture image units.\n"
|
|
" :rtype: int\n");
|
|
static PyObject *pygpu_max_textures_get(PyObject * /*self*/)
|
|
{
|
|
return PyLong_FromLong(GPU_max_textures());
|
|
}
|
|
|
|
PyDoc_STRVAR(pygpu_max_textures_vert_get_doc,
|
|
".. function:: max_textures_vert_get()\n"
|
|
"\n"
|
|
" Get maximum supported texture image units used for\n"
|
|
" accessing texture maps from the vertex shader.\n"
|
|
"\n"
|
|
" :return: Texture image units.\n"
|
|
" :rtype: int\n");
|
|
static PyObject *pygpu_max_textures_vert_get(PyObject * /*self*/)
|
|
{
|
|
return PyLong_FromLong(GPU_max_textures_vert());
|
|
}
|
|
|
|
PyDoc_STRVAR(pygpu_max_textures_geom_get_doc,
|
|
".. function:: max_textures_geom_get()\n"
|
|
"\n"
|
|
" Get maximum supported texture image units used for\n"
|
|
" accessing texture maps from the geometry shader.\n"
|
|
"\n"
|
|
" :return: Texture image units.\n"
|
|
" :rtype: int\n");
|
|
static PyObject *pygpu_max_textures_geom_get(PyObject * /*self*/)
|
|
{
|
|
return PyLong_FromLong(GPU_max_textures_geom());
|
|
}
|
|
|
|
PyDoc_STRVAR(pygpu_max_textures_frag_get_doc,
|
|
".. function:: max_textures_frag_get()\n"
|
|
"\n"
|
|
" Get maximum supported texture image units used for\n"
|
|
" accessing texture maps from the fragment shader.\n"
|
|
"\n"
|
|
" :return: Texture image units.\n"
|
|
" :rtype: int\n");
|
|
static PyObject *pygpu_max_textures_frag_get(PyObject * /*self*/)
|
|
{
|
|
return PyLong_FromLong(GPU_max_textures_frag());
|
|
}
|
|
|
|
PyDoc_STRVAR(pygpu_max_uniforms_vert_get_doc,
|
|
".. function:: max_uniforms_vert_get()\n"
|
|
"\n"
|
|
" Get maximum number of values held in uniform variable\n"
|
|
" storage for a vertex shader.\n"
|
|
"\n"
|
|
" :return: Number of values.\n"
|
|
" :rtype: int\n");
|
|
static PyObject *pygpu_max_uniforms_vert_get(PyObject * /*self*/)
|
|
{
|
|
return PyLong_FromLong(GPU_max_uniforms_vert());
|
|
}
|
|
|
|
PyDoc_STRVAR(pygpu_max_uniforms_frag_get_doc,
|
|
".. function:: max_uniforms_frag_get()\n"
|
|
"\n"
|
|
" Get maximum number of values held in uniform variable\n"
|
|
" storage for a fragment shader.\n"
|
|
"\n"
|
|
" :return: Number of values.\n"
|
|
" :rtype: int\n");
|
|
static PyObject *pygpu_max_uniforms_frag_get(PyObject * /*self*/)
|
|
{
|
|
return PyLong_FromLong(GPU_max_uniforms_frag());
|
|
}
|
|
|
|
PyDoc_STRVAR(pygpu_max_batch_indices_get_doc,
|
|
".. function:: max_batch_indices_get()\n"
|
|
"\n"
|
|
" Get maximum number of vertex array indices.\n"
|
|
"\n"
|
|
" :return: Number of indices.\n"
|
|
" :rtype: int\n");
|
|
static PyObject *pygpu_max_batch_indices_get(PyObject * /*self*/)
|
|
{
|
|
return PyLong_FromLong(GPU_max_batch_indices());
|
|
}
|
|
|
|
PyDoc_STRVAR(pygpu_max_batch_vertices_get_doc,
|
|
".. function:: max_batch_vertices_get()\n"
|
|
"\n"
|
|
" Get maximum number of vertex array vertices.\n"
|
|
"\n"
|
|
" :return: Number of vertices.\n"
|
|
" :rtype: int\n");
|
|
static PyObject *pygpu_max_batch_vertices_get(PyObject * /*self*/)
|
|
{
|
|
return PyLong_FromLong(GPU_max_batch_vertices());
|
|
}
|
|
|
|
PyDoc_STRVAR(pygpu_max_vertex_attribs_get_doc,
|
|
".. function:: max_vertex_attribs_get()\n"
|
|
"\n"
|
|
" Get maximum number of vertex attributes accessible to\n"
|
|
" a vertex shader.\n"
|
|
"\n"
|
|
" :return: Number of attributes.\n"
|
|
" :rtype: int\n");
|
|
static PyObject *pygpu_max_vertex_attribs_get(PyObject * /*self*/)
|
|
{
|
|
return PyLong_FromLong(GPU_max_vertex_attribs());
|
|
}
|
|
|
|
PyDoc_STRVAR(pygpu_max_varying_floats_get_doc,
|
|
".. function:: max_varying_floats_get()\n"
|
|
"\n"
|
|
" Get maximum number of varying variables used by\n"
|
|
" vertex and fragment shaders.\n"
|
|
"\n"
|
|
" :return: Number of variables.\n"
|
|
" :rtype: int\n");
|
|
static PyObject *pygpu_max_varying_floats_get(PyObject * /*self*/)
|
|
{
|
|
return PyLong_FromLong(GPU_max_varying_floats());
|
|
}
|
|
|
|
PyDoc_STRVAR(pygpu_extensions_get_doc,
|
|
".. function:: extensions_get()\n"
|
|
"\n"
|
|
" Get supported extensions in the current context.\n"
|
|
"\n"
|
|
" :return: Extensions.\n"
|
|
" :rtype: tuple of string\n");
|
|
static PyObject *pygpu_extensions_get(PyObject * /*self*/)
|
|
{
|
|
int extensions_len = GPU_extensions_len();
|
|
PyObject *ret = PyTuple_New(extensions_len);
|
|
PyObject **ob_items = ((PyTupleObject *)ret)->ob_item;
|
|
for (int i = 0; i < extensions_len; i++) {
|
|
ob_items[i] = PyUnicode_FromString(GPU_extension_get(i));
|
|
}
|
|
|
|
return ret;
|
|
}
|
|
|
|
PyDoc_STRVAR(pygpu_compute_shader_support_get_doc,
|
|
".. function:: compute_shader_support_get()\n"
|
|
"\n"
|
|
" Are compute shaders supported.\n"
|
|
"\n"
|
|
" :return: True when supported, False when not supported.\n"
|
|
" :rtype: bool\n");
|
|
static PyObject *pygpu_compute_shader_support_get(PyObject * /*self*/)
|
|
{
|
|
return PyBool_FromLong(GPU_compute_shader_support());
|
|
}
|
|
|
|
PyDoc_STRVAR(pygpu_shader_image_load_store_support_get_doc,
|
|
".. function:: shader_image_load_store_support_get()\n"
|
|
"\n"
|
|
" Is image load/store supported.\n"
|
|
"\n"
|
|
" :return: True when supported, False when not supported.\n"
|
|
" :rtype: bool\n");
|
|
static PyObject *pygpu_shader_image_load_store_support_get(PyObject * /*self*/)
|
|
{
|
|
return PyBool_FromLong(GPU_shader_image_load_store_support());
|
|
}
|
|
|
|
PyDoc_STRVAR(pygpu_hdr_support_get_doc,
|
|
".. function:: hdr_support_get()\n"
|
|
"\n"
|
|
" Return whether GPU backend supports High Dynamic range for viewport.\n"
|
|
"\n"
|
|
" :return: HDR support available.\n"
|
|
" :rtype: bool\n");
|
|
static PyObject *pygpu_hdr_support_get(PyObject * /*self*/)
|
|
{
|
|
return PyBool_FromLong(GPU_hdr_support());
|
|
}
|
|
|
|
PyDoc_STRVAR(pygpu_max_work_group_count_get_doc,
|
|
".. function:: max_work_group_count_get(index)\n"
|
|
"\n"
|
|
" Get maximum number of work groups that may be dispatched to a compute shader.\n"
|
|
"\n"
|
|
" :arg index: Index of the dimension.\n"
|
|
" :type index: int\n"
|
|
" :return: Maximum number of work groups for the queried dimension.\n"
|
|
" :rtype: int\n");
|
|
static PyObject *pygpu_max_work_group_count_get(PyObject * /*self*/, PyObject *args)
|
|
{
|
|
int index;
|
|
if (!PyArg_ParseTuple(args, "i", &index)) {
|
|
return nullptr;
|
|
}
|
|
|
|
const int max_work_group_count = GPU_max_work_group_count(index);
|
|
|
|
return PyLong_FromLong(max_work_group_count);
|
|
}
|
|
|
|
PyDoc_STRVAR(pygpu_max_work_group_size_get_doc,
|
|
".. function:: max_work_group_size_get(index)\n"
|
|
"\n"
|
|
" Get maximum size of a work group that may be dispatched to a compute shader.\n"
|
|
"\n"
|
|
" :arg index: Index of the dimension.\n"
|
|
" :type index: int\n"
|
|
" :return: Maximum size of a work group for the queried dimension.\n"
|
|
" :rtype: int\n");
|
|
static PyObject *pygpu_max_work_group_size_get(PyObject * /*self*/, PyObject *args)
|
|
{
|
|
int index;
|
|
if (!PyArg_ParseTuple(args, "i", &index)) {
|
|
return nullptr;
|
|
}
|
|
|
|
const int max_work_group_size = GPU_max_work_group_size(index);
|
|
|
|
return PyLong_FromLong(max_work_group_size);
|
|
}
|
|
|
|
/** \} */
|
|
|
|
/* -------------------------------------------------------------------- */
|
|
/** \name Module
|
|
* \{ */
|
|
|
|
#if (defined(__GNUC__) && !defined(__clang__))
|
|
# pragma GCC diagnostic push
|
|
# pragma GCC diagnostic ignored "-Wcast-function-type"
|
|
#endif
|
|
|
|
static PyMethodDef pygpu_capabilities__tp_methods[] = {
|
|
{"max_texture_size_get",
|
|
(PyCFunction)pygpu_max_texture_size_get,
|
|
METH_NOARGS,
|
|
pygpu_max_texture_size_get_doc},
|
|
{"max_texture_layers_get",
|
|
(PyCFunction)pygpu_max_texture_layers_get,
|
|
METH_NOARGS,
|
|
pygpu_max_texture_layers_get_doc},
|
|
{"max_textures_get",
|
|
(PyCFunction)pygpu_max_textures_get,
|
|
METH_NOARGS,
|
|
pygpu_max_textures_get_doc},
|
|
{"max_textures_vert_get",
|
|
(PyCFunction)pygpu_max_textures_vert_get,
|
|
METH_NOARGS,
|
|
pygpu_max_textures_vert_get_doc},
|
|
{"max_textures_geom_get",
|
|
(PyCFunction)pygpu_max_textures_geom_get,
|
|
METH_NOARGS,
|
|
pygpu_max_textures_geom_get_doc},
|
|
{"max_textures_frag_get",
|
|
(PyCFunction)pygpu_max_textures_frag_get,
|
|
METH_NOARGS,
|
|
pygpu_max_textures_frag_get_doc},
|
|
{"max_uniforms_vert_get",
|
|
(PyCFunction)pygpu_max_uniforms_vert_get,
|
|
METH_NOARGS,
|
|
pygpu_max_uniforms_vert_get_doc},
|
|
{"max_uniforms_frag_get",
|
|
(PyCFunction)pygpu_max_uniforms_frag_get,
|
|
METH_NOARGS,
|
|
pygpu_max_uniforms_frag_get_doc},
|
|
{"max_batch_indices_get",
|
|
(PyCFunction)pygpu_max_batch_indices_get,
|
|
METH_NOARGS,
|
|
pygpu_max_batch_indices_get_doc},
|
|
{"max_batch_vertices_get",
|
|
(PyCFunction)pygpu_max_batch_vertices_get,
|
|
METH_NOARGS,
|
|
pygpu_max_batch_vertices_get_doc},
|
|
{"max_vertex_attribs_get",
|
|
(PyCFunction)pygpu_max_vertex_attribs_get,
|
|
METH_NOARGS,
|
|
pygpu_max_vertex_attribs_get_doc},
|
|
{"max_varying_floats_get",
|
|
(PyCFunction)pygpu_max_varying_floats_get,
|
|
METH_NOARGS,
|
|
pygpu_max_varying_floats_get_doc},
|
|
{"extensions_get", (PyCFunction)pygpu_extensions_get, METH_NOARGS, pygpu_extensions_get_doc},
|
|
|
|
{"compute_shader_support_get",
|
|
(PyCFunction)pygpu_compute_shader_support_get,
|
|
METH_NOARGS,
|
|
pygpu_compute_shader_support_get_doc},
|
|
{"shader_image_load_store_support_get",
|
|
(PyCFunction)pygpu_shader_image_load_store_support_get,
|
|
METH_NOARGS,
|
|
pygpu_shader_image_load_store_support_get_doc},
|
|
{"hdr_support_get",
|
|
(PyCFunction)pygpu_hdr_support_get,
|
|
METH_NOARGS,
|
|
pygpu_hdr_support_get_doc},
|
|
{"max_work_group_count_get",
|
|
(PyCFunction)pygpu_max_work_group_count_get,
|
|
METH_VARARGS,
|
|
pygpu_max_work_group_count_get_doc,
|
|
},
|
|
{"max_work_group_size_get",
|
|
(PyCFunction)pygpu_max_work_group_size_get,
|
|
METH_VARARGS,
|
|
pygpu_max_work_group_size_get_doc,
|
|
},
|
|
{nullptr, nullptr, 0, nullptr},
|
|
};
|
|
|
|
#if (defined(__GNUC__) && !defined(__clang__))
|
|
# pragma GCC diagnostic pop
|
|
#endif
|
|
|
|
PyDoc_STRVAR(pygpu_capabilities__tp_doc, "This module provides access to the GPU capabilities.");
|
|
static PyModuleDef pygpu_capabilities_module_def = {
|
|
/*m_base*/ PyModuleDef_HEAD_INIT,
|
|
/*m_name*/ "gpu.capabilities",
|
|
/*m_doc*/ pygpu_capabilities__tp_doc,
|
|
/*m_size*/ 0,
|
|
/*m_methods*/ pygpu_capabilities__tp_methods,
|
|
/*m_slots*/ nullptr,
|
|
/*m_traverse*/ nullptr,
|
|
/*m_clear*/ nullptr,
|
|
/*m_free*/ nullptr,
|
|
};
|
|
|
|
PyObject *bpygpu_capabilities_init()
|
|
{
|
|
PyObject *submodule;
|
|
|
|
submodule = bpygpu_create_module(&pygpu_capabilities_module_def);
|
|
|
|
return submodule;
|
|
}
|
|
|
|
/** \} */
|