Merge pull request #2224 from lioncash/opus
hwopus: Leverage multistream API for decoding regular Opus packets
This commit is contained in:
commit
723ad4512f
1 changed files with 48 additions and 34 deletions
|
@ -8,6 +8,7 @@
|
|||
#include <vector>
|
||||
|
||||
#include <opus.h>
|
||||
#include <opus_multistream.h>
|
||||
|
||||
#include "common/assert.h"
|
||||
#include "common/logging/log.h"
|
||||
|
@ -18,12 +19,12 @@
|
|||
namespace Service::Audio {
|
||||
namespace {
|
||||
struct OpusDeleter {
|
||||
void operator()(void* ptr) const {
|
||||
operator delete(ptr);
|
||||
void operator()(OpusMSDecoder* ptr) const {
|
||||
opus_multistream_decoder_destroy(ptr);
|
||||
}
|
||||
};
|
||||
|
||||
using OpusDecoderPtr = std::unique_ptr<OpusDecoder, OpusDeleter>;
|
||||
using OpusDecoderPtr = std::unique_ptr<OpusMSDecoder, OpusDeleter>;
|
||||
|
||||
struct OpusPacketHeader {
|
||||
// Packet size in bytes.
|
||||
|
@ -33,7 +34,7 @@ struct OpusPacketHeader {
|
|||
};
|
||||
static_assert(sizeof(OpusPacketHeader) == 0x8, "OpusHeader is an invalid size");
|
||||
|
||||
class OpusDecoderStateBase {
|
||||
class OpusDecoderState {
|
||||
public:
|
||||
/// Describes extra behavior that may be asked of the decoding context.
|
||||
enum class ExtraBehavior {
|
||||
|
@ -49,22 +50,13 @@ public:
|
|||
Enabled,
|
||||
};
|
||||
|
||||
virtual ~OpusDecoderStateBase() = default;
|
||||
|
||||
// Decodes interleaved Opus packets. Optionally allows reporting time taken to
|
||||
// perform the decoding, as well as any relevant extra behavior.
|
||||
virtual void DecodeInterleaved(Kernel::HLERequestContext& ctx, PerfTime perf_time,
|
||||
ExtraBehavior extra_behavior) = 0;
|
||||
};
|
||||
|
||||
// Represents the decoder state for a non-multistream decoder.
|
||||
class OpusDecoderState final : public OpusDecoderStateBase {
|
||||
public:
|
||||
explicit OpusDecoderState(OpusDecoderPtr decoder, u32 sample_rate, u32 channel_count)
|
||||
: decoder{std::move(decoder)}, sample_rate{sample_rate}, channel_count{channel_count} {}
|
||||
|
||||
// Decodes interleaved Opus packets. Optionally allows reporting time taken to
|
||||
// perform the decoding, as well as any relevant extra behavior.
|
||||
void DecodeInterleaved(Kernel::HLERequestContext& ctx, PerfTime perf_time,
|
||||
ExtraBehavior extra_behavior) override {
|
||||
ExtraBehavior extra_behavior) {
|
||||
if (perf_time == PerfTime::Disabled) {
|
||||
DecodeInterleavedHelper(ctx, nullptr, extra_behavior);
|
||||
} else {
|
||||
|
@ -135,7 +127,7 @@ private:
|
|||
|
||||
const int frame_size = (static_cast<int>(raw_output_sz / sizeof(s16) / channel_count));
|
||||
const auto out_sample_count =
|
||||
opus_decode(decoder.get(), frame, hdr.size, output.data(), frame_size, 0);
|
||||
opus_multistream_decode(decoder.get(), frame, hdr.size, output.data(), frame_size, 0);
|
||||
if (out_sample_count < 0) {
|
||||
LOG_ERROR(Audio,
|
||||
"Incorrect sample count received from opus_decode, "
|
||||
|
@ -158,7 +150,7 @@ private:
|
|||
void ResetDecoderContext() {
|
||||
ASSERT(decoder != nullptr);
|
||||
|
||||
opus_decoder_ctl(decoder.get(), OPUS_RESET_STATE);
|
||||
opus_multistream_decoder_ctl(decoder.get(), OPUS_RESET_STATE);
|
||||
}
|
||||
|
||||
OpusDecoderPtr decoder;
|
||||
|
@ -168,7 +160,7 @@ private:
|
|||
|
||||
class IHardwareOpusDecoderManager final : public ServiceFramework<IHardwareOpusDecoderManager> {
|
||||
public:
|
||||
explicit IHardwareOpusDecoderManager(std::unique_ptr<OpusDecoderStateBase> decoder_state)
|
||||
explicit IHardwareOpusDecoderManager(OpusDecoderState decoder_state)
|
||||
: ServiceFramework("IHardwareOpusDecoderManager"), decoder_state{std::move(decoder_state)} {
|
||||
// clang-format off
|
||||
static const FunctionInfo functions[] = {
|
||||
|
@ -190,35 +182,51 @@ private:
|
|||
void DecodeInterleavedOld(Kernel::HLERequestContext& ctx) {
|
||||
LOG_DEBUG(Audio, "called");
|
||||
|
||||
decoder_state->DecodeInterleaved(ctx, OpusDecoderStateBase::PerfTime::Disabled,
|
||||
OpusDecoderStateBase::ExtraBehavior::None);
|
||||
decoder_state.DecodeInterleaved(ctx, OpusDecoderState::PerfTime::Disabled,
|
||||
OpusDecoderState::ExtraBehavior::None);
|
||||
}
|
||||
|
||||
void DecodeInterleavedWithPerfOld(Kernel::HLERequestContext& ctx) {
|
||||
LOG_DEBUG(Audio, "called");
|
||||
|
||||
decoder_state->DecodeInterleaved(ctx, OpusDecoderStateBase::PerfTime::Enabled,
|
||||
OpusDecoderStateBase::ExtraBehavior::None);
|
||||
decoder_state.DecodeInterleaved(ctx, OpusDecoderState::PerfTime::Enabled,
|
||||
OpusDecoderState::ExtraBehavior::None);
|
||||
}
|
||||
|
||||
void DecodeInterleaved(Kernel::HLERequestContext& ctx) {
|
||||
LOG_DEBUG(Audio, "called");
|
||||
|
||||
IPC::RequestParser rp{ctx};
|
||||
const auto extra_behavior = rp.Pop<bool>()
|
||||
? OpusDecoderStateBase::ExtraBehavior::ResetContext
|
||||
: OpusDecoderStateBase::ExtraBehavior::None;
|
||||
const auto extra_behavior = rp.Pop<bool>() ? OpusDecoderState::ExtraBehavior::ResetContext
|
||||
: OpusDecoderState::ExtraBehavior::None;
|
||||
|
||||
decoder_state->DecodeInterleaved(ctx, OpusDecoderStateBase::PerfTime::Enabled,
|
||||
extra_behavior);
|
||||
decoder_state.DecodeInterleaved(ctx, OpusDecoderState::PerfTime::Enabled, extra_behavior);
|
||||
}
|
||||
|
||||
std::unique_ptr<OpusDecoderStateBase> decoder_state;
|
||||
OpusDecoderState decoder_state;
|
||||
};
|
||||
|
||||
std::size_t WorkerBufferSize(u32 channel_count) {
|
||||
ASSERT_MSG(channel_count == 1 || channel_count == 2, "Invalid channel count");
|
||||
return opus_decoder_get_size(static_cast<int>(channel_count));
|
||||
constexpr int num_streams = 1;
|
||||
const int num_stereo_streams = channel_count == 2 ? 1 : 0;
|
||||
return opus_multistream_decoder_get_size(num_streams, num_stereo_streams);
|
||||
}
|
||||
|
||||
// Creates the mapping table that maps the input channels to the particular
|
||||
// output channels. In the stereo case, we map the left and right input channels
|
||||
// to the left and right output channels respectively.
|
||||
//
|
||||
// However, in the monophonic case, we only map the one available channel
|
||||
// to the sole output channel. We specify 255 for the would-be right channel
|
||||
// as this is a special value defined by Opus to indicate to the decoder to
|
||||
// ignore that channel.
|
||||
std::array<u8, 2> CreateMappingTable(u32 channel_count) {
|
||||
if (channel_count == 2) {
|
||||
return {{0, 1}};
|
||||
}
|
||||
|
||||
return {{0, 255}};
|
||||
}
|
||||
} // Anonymous namespace
|
||||
|
||||
|
@ -259,9 +267,15 @@ void HwOpus::OpenOpusDecoder(Kernel::HLERequestContext& ctx) {
|
|||
const std::size_t worker_sz = WorkerBufferSize(channel_count);
|
||||
ASSERT_MSG(buffer_sz >= worker_sz, "Worker buffer too large");
|
||||
|
||||
OpusDecoderPtr decoder{static_cast<OpusDecoder*>(operator new(worker_sz))};
|
||||
if (const int err = opus_decoder_init(decoder.get(), sample_rate, channel_count)) {
|
||||
LOG_ERROR(Audio, "Failed to init opus decoder with error={}", err);
|
||||
const int num_stereo_streams = channel_count == 2 ? 1 : 0;
|
||||
const auto mapping_table = CreateMappingTable(channel_count);
|
||||
|
||||
int error = 0;
|
||||
OpusDecoderPtr decoder{
|
||||
opus_multistream_decoder_create(sample_rate, static_cast<int>(channel_count), 1,
|
||||
num_stereo_streams, mapping_table.data(), &error)};
|
||||
if (error != OPUS_OK || decoder == nullptr) {
|
||||
LOG_ERROR(Audio, "Failed to create Opus decoder (error={}).", error);
|
||||
IPC::ResponseBuilder rb{ctx, 2};
|
||||
// TODO(ogniK): Use correct error code
|
||||
rb.Push(ResultCode(-1));
|
||||
|
@ -271,7 +285,7 @@ void HwOpus::OpenOpusDecoder(Kernel::HLERequestContext& ctx) {
|
|||
IPC::ResponseBuilder rb{ctx, 2, 0, 1};
|
||||
rb.Push(RESULT_SUCCESS);
|
||||
rb.PushIpcInterface<IHardwareOpusDecoderManager>(
|
||||
std::make_unique<OpusDecoderState>(std::move(decoder), sample_rate, channel_count));
|
||||
OpusDecoderState{std::move(decoder), sample_rate, channel_count});
|
||||
}
|
||||
|
||||
HwOpus::HwOpus() : ServiceFramework("hwopus") {
|
||||
|
|
Loading…
Reference in a new issue