4ebd7d84d1
Batches don't always provide all required resources for a shader. This can happen when the resource is defined in the shader, but not used in a specific variant. The definition is still there and needs to be filled. This PR adds a dummy buffer to fill the missing attributes. It also fixes an issue when using vertex attributes that uses multiple binding locations (mat4). Pull Request: https://projects.blender.org/blender/blender/pulls/111350
79 lines
1.6 KiB
C++
79 lines
1.6 KiB
C++
/* SPDX-FileCopyrightText: 2023 Blender Authors
|
|
*
|
|
* SPDX-License-Identifier: GPL-2.0-or-later */
|
|
|
|
/** \file
|
|
* \ingroup gpu
|
|
*/
|
|
|
|
#pragma once
|
|
|
|
#include "gpu_context_private.hh"
|
|
|
|
#include "vk_common.hh"
|
|
|
|
namespace blender::gpu {
|
|
class VKContext;
|
|
|
|
/**
|
|
* Class for handing vulkan buffers (allocation/updating/binding).
|
|
*/
|
|
class VKBuffer {
|
|
int64_t size_in_bytes_ = 0;
|
|
VkBuffer vk_buffer_ = VK_NULL_HANDLE;
|
|
VmaAllocation allocation_ = VK_NULL_HANDLE;
|
|
/* Pointer to the virtually mapped memory. */
|
|
void *mapped_memory_ = nullptr;
|
|
|
|
public:
|
|
VKBuffer() = default;
|
|
virtual ~VKBuffer();
|
|
|
|
/** Has this buffer been allocated? */
|
|
bool is_allocated() const;
|
|
|
|
bool create(int64_t size, GPUUsageType usage, VkBufferUsageFlagBits buffer_usage);
|
|
void clear(VKContext &context, uint32_t clear_value);
|
|
void update(const void *data) const;
|
|
void read(void *data) const;
|
|
bool free();
|
|
|
|
int64_t size_in_bytes() const
|
|
{
|
|
return size_in_bytes_;
|
|
}
|
|
|
|
VkBuffer vk_handle() const
|
|
{
|
|
return vk_buffer_;
|
|
}
|
|
|
|
/**
|
|
* Get the reference to the mapped memory.
|
|
*
|
|
* Can only be called when the buffer is (still) mapped.
|
|
*/
|
|
void *mapped_memory_get() const;
|
|
|
|
private:
|
|
/** Check if this buffer is mapped. */
|
|
bool is_mapped() const;
|
|
bool map();
|
|
void unmap();
|
|
};
|
|
|
|
/**
|
|
* Helper struct to enable buffers to be bound with an offset.
|
|
*
|
|
* VKImmediate mode uses a single VKBuffer with multiple vertex layouts. Those layouts are send to
|
|
* the command buffer containing an offset.
|
|
*
|
|
* VKIndexBuffer uses this when it is a subrange of another buffer.
|
|
*/
|
|
struct VKBufferWithOffset {
|
|
const VKBuffer &buffer;
|
|
VkDeviceSize offset;
|
|
};
|
|
|
|
} // namespace blender::gpu
|