write_buffer()

void MayaFlux::Buffers::DataWriteProcessor::write_buffer	(	std::shared_ptr< VKBuffer >	target,
		Kakshya::DataVariant	data,
		size_t	dst_offset = 0
	)

Write bytes into another VKBuffer, such as a plain descriptor-bound buffer.

The target is not resized. A device-local target is written through a persistent staging buffer. A buffer whose descriptor binding has a node, audio, or host-buffer source is rewritten from that source each cycle, so write to the source instead.

Parameters

target	Buffer to write; retained until the request runs.
data	Source data, converted to bytes when the request runs.
dst_offset	Byte offset into target. offset plus data size must fit its allocation.

Definition at line 53 of file DataWriteProcessor.cpp.

{
    auto payload = std::make_shared<Kakshya::DataVariant>(std::move(data));
    enqueue_write([target = std::move(target), payload, dst_offset](DataWriteProcessor& self, const std::shared_ptr<VKBuffer>&) {
        const auto bytes = Kakshya::convert_variant<uint8_t>(*payload);
        if (!target || !target->is_initialized() || bytes.size_bytes() == 0
            || dst_offset + bytes.size_bytes() > target->get_size_bytes()) {
            MF_RT_ERROR(Journal::Component::Buffers, Journal::Context::BufferProcessing,
                "DataWriteProcessor write_buffer needs an initialized target with room for the data");
            return;
        }
        if (is_device_local(target)) {
            ensure_staging(self.m_write_staging, bytes.size_bytes());
        }
        upload_to_gpu(bytes.data(), bytes.size_bytes(), target, self.m_write_staging, dst_offset);
    });
}

References MayaFlux::Journal::BufferProcessing, MayaFlux::Journal::Buffers, enqueue_write(), MayaFlux::Buffers::is_device_local(), m_write_staging, MF_RT_ERROR, payload, and MayaFlux::Buffers::upload_to_gpu().

Here is the call graph for this function: