diff options
author | Kelebek1 <eeeedddccc@hotmail.co.uk> | 2021-06-20 16:23:16 +0200 |
---|---|---|
committer | Kelebek1 <eeeedddccc@hotmail.co.uk> | 2021-06-27 16:58:07 +0200 |
commit | 0857d6a3dbcaeb30b51326419fb56d9b543601f1 (patch) | |
tree | 26221ce096533b188aa1fe8d7e82290780edee1d | |
parent | Merge pull request #6526 from bunnei/doom-update (diff) | |
download | yuzu-0857d6a3dbcaeb30b51326419fb56d9b543601f1.tar yuzu-0857d6a3dbcaeb30b51326419fb56d9b543601f1.tar.gz yuzu-0857d6a3dbcaeb30b51326419fb56d9b543601f1.tar.bz2 yuzu-0857d6a3dbcaeb30b51326419fb56d9b543601f1.tar.lz yuzu-0857d6a3dbcaeb30b51326419fb56d9b543601f1.tar.xz yuzu-0857d6a3dbcaeb30b51326419fb56d9b543601f1.tar.zst yuzu-0857d6a3dbcaeb30b51326419fb56d9b543601f1.zip |
-rw-r--r-- | src/audio_core/audio_renderer.cpp | 178 | ||||
-rw-r--r-- | src/audio_core/audio_renderer.h | 6 | ||||
-rw-r--r-- | src/core/hle/service/audio/audren_u.cpp | 10 |
3 files changed, 115 insertions, 79 deletions
diff --git a/src/audio_core/audio_renderer.cpp b/src/audio_core/audio_renderer.cpp index 80ffddb10..0757cd804 100644 --- a/src/audio_core/audio_renderer.cpp +++ b/src/audio_core/audio_renderer.cpp @@ -12,6 +12,7 @@ #include "audio_core/voice_context.h" #include "common/logging/log.h" #include "common/settings.h" +#include "core/core_timing.h" #include "core/memory.h" namespace { @@ -68,7 +69,9 @@ namespace { } // namespace namespace AudioCore { -AudioRenderer::AudioRenderer(Core::Timing::CoreTiming& core_timing, Core::Memory::Memory& memory_, +constexpr s32 NUM_BUFFERS = 2; + +AudioRenderer::AudioRenderer(Core::Timing::CoreTiming& core_timing_, Core::Memory::Memory& memory_, AudioCommon::AudioRendererParameter params, Stream::ReleaseCallback&& release_callback, std::size_t instance_number) @@ -77,7 +80,8 @@ AudioRenderer::AudioRenderer(Core::Timing::CoreTiming& core_timing, Core::Memory sink_context(params.sink_count), splitter_context(), voices(params.voice_count), memory{memory_}, command_generator(worker_params, voice_context, mix_context, splitter_context, effect_context, - memory) { + memory), + core_timing{core_timing_} { behavior_info.SetUserRevision(params.revision); splitter_context.Initialize(behavior_info, params.splitter_count, params.num_splitter_send_channels); @@ -86,16 +90,27 @@ AudioRenderer::AudioRenderer(Core::Timing::CoreTiming& core_timing, Core::Memory stream = audio_out->OpenStream( core_timing, params.sample_rate, AudioCommon::STREAM_NUM_CHANNELS, fmt::format("AudioRenderer-Instance{}", instance_number), std::move(release_callback)); - audio_out->StartStream(stream); - - QueueMixedBuffer(0); - QueueMixedBuffer(1); - QueueMixedBuffer(2); - QueueMixedBuffer(3); + process_event = Core::Timing::CreateEvent( + fmt::format("AudioRenderer-Instance{}-Process", instance_number), + [this](std::uintptr_t, std::chrono::nanoseconds) { ReleaseAndQueueBuffers(); }); + for (s32 i = 0; i < NUM_BUFFERS; ++i) { + QueueMixedBuffer(i); + } } AudioRenderer::~AudioRenderer() = default; +ResultCode AudioRenderer::Start() { + audio_out->StartStream(stream); + ReleaseAndQueueBuffers(); + return ResultSuccess; +} + +ResultCode AudioRenderer::Stop() { + audio_out->StopStream(stream); + return ResultSuccess; +} + u32 AudioRenderer::GetSampleRate() const { return worker_params.sample_rate; } @@ -114,89 +129,88 @@ Stream::State AudioRenderer::GetStreamState() const { ResultCode AudioRenderer::UpdateAudioRenderer(const std::vector<u8>& input_params, std::vector<u8>& output_params) { + { + std::scoped_lock lock{mutex}; + InfoUpdater info_updater{input_params, output_params, behavior_info}; - InfoUpdater info_updater{input_params, output_params, behavior_info}; - - if (!info_updater.UpdateBehaviorInfo(behavior_info)) { - LOG_ERROR(Audio, "Failed to update behavior info input parameters"); - return AudioCommon::Audren::ERR_INVALID_PARAMETERS; - } - - if (!info_updater.UpdateMemoryPools(memory_pool_info)) { - LOG_ERROR(Audio, "Failed to update memory pool parameters"); - return AudioCommon::Audren::ERR_INVALID_PARAMETERS; - } - - if (!info_updater.UpdateVoiceChannelResources(voice_context)) { - LOG_ERROR(Audio, "Failed to update voice channel resource parameters"); - return AudioCommon::Audren::ERR_INVALID_PARAMETERS; - } + if (!info_updater.UpdateBehaviorInfo(behavior_info)) { + LOG_ERROR(Audio, "Failed to update behavior info input parameters"); + return AudioCommon::Audren::ERR_INVALID_PARAMETERS; + } - if (!info_updater.UpdateVoices(voice_context, memory_pool_info, 0)) { - LOG_ERROR(Audio, "Failed to update voice parameters"); - return AudioCommon::Audren::ERR_INVALID_PARAMETERS; - } + if (!info_updater.UpdateMemoryPools(memory_pool_info)) { + LOG_ERROR(Audio, "Failed to update memory pool parameters"); + return AudioCommon::Audren::ERR_INVALID_PARAMETERS; + } - // TODO(ogniK): Deal with stopped audio renderer but updates still taking place - if (!info_updater.UpdateEffects(effect_context, true)) { - LOG_ERROR(Audio, "Failed to update effect parameters"); - return AudioCommon::Audren::ERR_INVALID_PARAMETERS; - } + if (!info_updater.UpdateVoiceChannelResources(voice_context)) { + LOG_ERROR(Audio, "Failed to update voice channel resource parameters"); + return AudioCommon::Audren::ERR_INVALID_PARAMETERS; + } - if (behavior_info.IsSplitterSupported()) { - if (!info_updater.UpdateSplitterInfo(splitter_context)) { - LOG_ERROR(Audio, "Failed to update splitter parameters"); + if (!info_updater.UpdateVoices(voice_context, memory_pool_info, 0)) { + LOG_ERROR(Audio, "Failed to update voice parameters"); return AudioCommon::Audren::ERR_INVALID_PARAMETERS; } - } - const auto mix_result = info_updater.UpdateMixes(mix_context, worker_params.mix_buffer_count, - splitter_context, effect_context); + // TODO(ogniK): Deal with stopped audio renderer but updates still taking place + if (!info_updater.UpdateEffects(effect_context, true)) { + LOG_ERROR(Audio, "Failed to update effect parameters"); + return AudioCommon::Audren::ERR_INVALID_PARAMETERS; + } - if (mix_result.IsError()) { - LOG_ERROR(Audio, "Failed to update mix parameters"); - return mix_result; - } + if (behavior_info.IsSplitterSupported()) { + if (!info_updater.UpdateSplitterInfo(splitter_context)) { + LOG_ERROR(Audio, "Failed to update splitter parameters"); + return AudioCommon::Audren::ERR_INVALID_PARAMETERS; + } + } - // TODO(ogniK): Sinks - if (!info_updater.UpdateSinks(sink_context)) { - LOG_ERROR(Audio, "Failed to update sink parameters"); - return AudioCommon::Audren::ERR_INVALID_PARAMETERS; - } + const auto mix_result = info_updater.UpdateMixes( + mix_context, worker_params.mix_buffer_count, splitter_context, effect_context); - // TODO(ogniK): Performance buffer - if (!info_updater.UpdatePerformanceBuffer()) { - LOG_ERROR(Audio, "Failed to update performance buffer parameters"); - return AudioCommon::Audren::ERR_INVALID_PARAMETERS; - } + if (mix_result.IsError()) { + LOG_ERROR(Audio, "Failed to update mix parameters"); + return mix_result; + } - if (!info_updater.UpdateErrorInfo(behavior_info)) { - LOG_ERROR(Audio, "Failed to update error info"); - return AudioCommon::Audren::ERR_INVALID_PARAMETERS; - } + // TODO(ogniK): Sinks + if (!info_updater.UpdateSinks(sink_context)) { + LOG_ERROR(Audio, "Failed to update sink parameters"); + return AudioCommon::Audren::ERR_INVALID_PARAMETERS; + } - if (behavior_info.IsElapsedFrameCountSupported()) { - if (!info_updater.UpdateRendererInfo(elapsed_frame_count)) { - LOG_ERROR(Audio, "Failed to update renderer info"); + // TODO(ogniK): Performance buffer + if (!info_updater.UpdatePerformanceBuffer()) { + LOG_ERROR(Audio, "Failed to update performance buffer parameters"); return AudioCommon::Audren::ERR_INVALID_PARAMETERS; } - } - // TODO(ogniK): Statistics - if (!info_updater.WriteOutputHeader()) { - LOG_ERROR(Audio, "Failed to write output header"); - return AudioCommon::Audren::ERR_INVALID_PARAMETERS; - } + if (!info_updater.UpdateErrorInfo(behavior_info)) { + LOG_ERROR(Audio, "Failed to update error info"); + return AudioCommon::Audren::ERR_INVALID_PARAMETERS; + } - // TODO(ogniK): Check when all sections are implemented + if (behavior_info.IsElapsedFrameCountSupported()) { + if (!info_updater.UpdateRendererInfo(elapsed_frame_count)) { + LOG_ERROR(Audio, "Failed to update renderer info"); + return AudioCommon::Audren::ERR_INVALID_PARAMETERS; + } + } + // TODO(ogniK): Statistics - if (!info_updater.CheckConsumedSize()) { - LOG_ERROR(Audio, "Audio buffers were not consumed!"); - return AudioCommon::Audren::ERR_INVALID_PARAMETERS; - } + if (!info_updater.WriteOutputHeader()) { + LOG_ERROR(Audio, "Failed to write output header"); + return AudioCommon::Audren::ERR_INVALID_PARAMETERS; + } - ReleaseAndQueueBuffers(); + // TODO(ogniK): Check when all sections are implemented + if (!info_updater.CheckConsumedSize()) { + LOG_ERROR(Audio, "Audio buffers were not consumed!"); + return AudioCommon::Audren::ERR_INVALID_PARAMETERS; + } + } return ResultSuccess; } @@ -315,10 +329,24 @@ void AudioRenderer::QueueMixedBuffer(Buffer::Tag tag) { } void AudioRenderer::ReleaseAndQueueBuffers() { - const auto released_buffers{audio_out->GetTagsAndReleaseBuffers(stream)}; - for (const auto& tag : released_buffers) { - QueueMixedBuffer(tag); + if (!stream->IsPlaying()) { + return; + } + + { + std::scoped_lock lock{mutex}; + const auto released_buffers{audio_out->GetTagsAndReleaseBuffers(stream)}; + for (const auto& tag : released_buffers) { + QueueMixedBuffer(tag); + } } + + const f32 sample_rate = static_cast<f32>(GetSampleRate()); + const f32 sample_count = static_cast<f32>(GetSampleCount()); + const f32 consume_rate = sample_rate / (sample_count * (sample_count / 240)); + const s32 ms = (1000 / static_cast<s32>(consume_rate)) - 1; + const std::chrono::milliseconds next_event_time(std::max(ms / NUM_BUFFERS, 1)); + core_timing.ScheduleEvent(next_event_time, process_event, {}); } } // namespace AudioCore diff --git a/src/audio_core/audio_renderer.h b/src/audio_core/audio_renderer.h index 18567f618..88fdd13dd 100644 --- a/src/audio_core/audio_renderer.h +++ b/src/audio_core/audio_renderer.h @@ -6,6 +6,7 @@ #include <array> #include <memory> +#include <mutex> #include <vector> #include "audio_core/behavior_info.h" @@ -45,6 +46,8 @@ public: [[nodiscard]] ResultCode UpdateAudioRenderer(const std::vector<u8>& input_params, std::vector<u8>& output_params); + [[nodiscard]] ResultCode Start(); + [[nodiscard]] ResultCode Stop(); void QueueMixedBuffer(Buffer::Tag tag); void ReleaseAndQueueBuffers(); [[nodiscard]] u32 GetSampleRate() const; @@ -68,6 +71,9 @@ private: Core::Memory::Memory& memory; CommandGenerator command_generator; std::size_t elapsed_frame_count{}; + Core::Timing::CoreTiming& core_timing; + std::shared_ptr<Core::Timing::EventType> process_event; + std::mutex mutex; }; } // namespace AudioCore diff --git a/src/core/hle/service/audio/audren_u.cpp b/src/core/hle/service/audio/audren_u.cpp index 800feba6e..feb5150e1 100644 --- a/src/core/hle/service/audio/audren_u.cpp +++ b/src/core/hle/service/audio/audren_u.cpp @@ -110,17 +110,19 @@ private: void Start(Kernel::HLERequestContext& ctx) { LOG_WARNING(Service_Audio, "(STUBBED) called"); - IPC::ResponseBuilder rb{ctx, 2}; + const auto result = renderer->Start(); - rb.Push(ResultSuccess); + IPC::ResponseBuilder rb{ctx, 2}; + rb.Push(result); } void Stop(Kernel::HLERequestContext& ctx) { LOG_WARNING(Service_Audio, "(STUBBED) called"); - IPC::ResponseBuilder rb{ctx, 2}; + const auto result = renderer->Stop(); - rb.Push(ResultSuccess); + IPC::ResponseBuilder rb{ctx, 2}; + rb.Push(result); } void QuerySystemEvent(Kernel::HLERequestContext& ctx) { |