Brecht Van Lommel
92a99c1496
Backend initialization needs to be delayed until after the OpenGL context is created. This worked fine in foreground mode because the OpenGL context already exists for the window at the point GPU_backend_init_once was called, but not for background mode. Create the backend just in time in GPU_context_create as before, and automatically free it when the last context id discarded. But check if any GPU backend is supported before creating the OpenGL context. Ref D15463, D15465
67 lines
2.1 KiB
C++
67 lines
2.1 KiB
C++
/* SPDX-License-Identifier: GPL-2.0-or-later
|
|
* Copyright 2016 by Mike Erwin. All rights reserved. */
|
|
|
|
/** \file
|
|
* \ingroup gpu
|
|
*
|
|
* This interface allow GPU to manage VAOs for multiple context and threads.
|
|
*/
|
|
|
|
#pragma once
|
|
|
|
#include "GPU_batch.h"
|
|
#include "GPU_common.h"
|
|
#include "GPU_platform.h"
|
|
|
|
#ifdef __cplusplus
|
|
extern "C" {
|
|
#endif
|
|
|
|
/* GPU backends abstract the differences between different APIs. GPU_context_create
|
|
* automatically initializes the backend, and GPU_context_discard frees it when there
|
|
* are no more contexts. */
|
|
bool GPU_backend_supported(void);
|
|
eGPUBackendType GPU_backend_get_type(void);
|
|
|
|
/** Opaque type hiding blender::gpu::Context. */
|
|
typedef struct GPUContext GPUContext;
|
|
|
|
GPUContext *GPU_context_create(void *ghost_window);
|
|
/**
|
|
* To be called after #GPU_context_active_set(ctx_to_destroy).
|
|
*/
|
|
void GPU_context_discard(GPUContext *);
|
|
|
|
/**
|
|
* Ctx can be NULL.
|
|
*/
|
|
void GPU_context_active_set(GPUContext *);
|
|
GPUContext *GPU_context_active_get(void);
|
|
|
|
/* Begin and end frame are used to mark the singular boundary representing the lifetime of a whole
|
|
* frame. This also acts as a divisor for ensuring workload submission and flushing, especially for
|
|
* background rendering when there is no call to present.
|
|
* This is required by explicit-API's where there is no implicit workload flushing. */
|
|
void GPU_context_begin_frame(GPUContext *ctx);
|
|
void GPU_context_end_frame(GPUContext *ctx);
|
|
|
|
/* Legacy GPU (Intel HD4000 series) do not support sharing GPU objects between GPU
|
|
* contexts. EEVEE/Workbench can create different contexts for image/preview rendering, baking or
|
|
* compiling. When a legacy GPU is detected (`GPU_use_main_context_workaround()`) any worker
|
|
* threads should use the draw manager opengl context and make sure that they are the only one
|
|
* using it by locking the main context using these two functions. */
|
|
void GPU_context_main_lock(void);
|
|
void GPU_context_main_unlock(void);
|
|
|
|
/* GPU Begin/end work blocks */
|
|
void GPU_render_begin(void);
|
|
void GPU_render_end(void);
|
|
|
|
/* For operations which need to run exactly once per frame -- even if there are no render updates.
|
|
*/
|
|
void GPU_render_step(void);
|
|
|
|
#ifdef __cplusplus
|
|
}
|
|
#endif
|