aboutsummaryrefslogtreecommitdiffhomepage
path: root/src/audio_core/voice_context.cpp
blob: c46ee55f1320cef48ae6b35683dc306351d9b874 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
// Copyright 2020 yuzu Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.

#include "audio_core/behavior_info.h"
#include "audio_core/voice_context.h"
#include "core/memory.h"

namespace AudioCore {

ServerVoiceChannelResource::ServerVoiceChannelResource(s32 id) : id(id) {}
ServerVoiceChannelResource::~ServerVoiceChannelResource() = default;

bool ServerVoiceChannelResource::InUse() const {
    return in_use;
}

float ServerVoiceChannelResource::GetCurrentMixVolumeAt(std::size_t i) const {
    ASSERT(i < AudioCommon::MAX_MIX_BUFFERS);
    return mix_volume.at(i);
}

float ServerVoiceChannelResource::GetLastMixVolumeAt(std::size_t i) const {
    ASSERT(i < AudioCommon::MAX_MIX_BUFFERS);
    return last_mix_volume.at(i);
}

void ServerVoiceChannelResource::Update(VoiceChannelResource::InParams& in_params) {
    in_use = in_params.in_use;
    // Update our mix volumes only if it's in use
    if (in_params.in_use) {
        mix_volume = in_params.mix_volume;
    }
}

void ServerVoiceChannelResource::UpdateLastMixVolumes() {
    last_mix_volume = mix_volume;
}

const std::array<float, AudioCommon::MAX_MIX_BUFFERS>&
ServerVoiceChannelResource::GetCurrentMixVolume() const {
    return mix_volume;
}

const std::array<float, AudioCommon::MAX_MIX_BUFFERS>&
ServerVoiceChannelResource::GetLastMixVolume() const {
    return last_mix_volume;
}

ServerVoiceInfo::ServerVoiceInfo() {
    Initialize();
}
ServerVoiceInfo::~ServerVoiceInfo() = default;

void ServerVoiceInfo::Initialize() {
    in_params.in_use = false;
    in_params.node_id = 0;
    in_params.id = 0;
    in_params.current_playstate = ServerPlayState::Stop;
    in_params.priority = 255;
    in_params.sample_rate = 0;
    in_params.sample_format = SampleFormat::Invalid;
    in_params.channel_count = 0;
    in_params.pitch = 0.0f;
    in_params.volume = 0.0f;
    in_params.last_volume = 0.0f;
    in_params.biquad_filter.fill({});
    in_params.wave_buffer_count = 0;
    in_params.wave_bufffer_head = 0;
    in_params.mix_id = AudioCommon::NO_MIX;
    in_params.splitter_info_id = AudioCommon::NO_SPLITTER;
    in_params.additional_params_address = 0;
    in_params.additional_params_size = 0;
    in_params.is_new = false;
    out_params.played_sample_count = 0;
    out_params.wave_buffer_consumed = 0;
    in_params.voice_drop_flag = false;
    in_params.buffer_mapped = false;
    in_params.wave_buffer_flush_request_count = 0;
    in_params.was_biquad_filter_enabled.fill(false);

    for (auto& wave_buffer : in_params.wave_buffer) {
        wave_buffer.start_sample_offset = 0;
        wave_buffer.end_sample_offset = 0;
        wave_buffer.is_looping = false;
        wave_buffer.end_of_stream = false;
        wave_buffer.buffer_address = 0;
        wave_buffer.buffer_size = 0;
        wave_buffer.context_address = 0;
        wave_buffer.context_size = 0;
        wave_buffer.sent_to_dsp = true;
    }

    stored_samples.clear();
}

void ServerVoiceInfo::UpdateParameters(const VoiceInfo::InParams& voice_in,
                                       BehaviorInfo& behavior_info) {
    in_params.in_use = voice_in.is_in_use;
    in_params.id = voice_in.id;
    in_params.node_id = voice_in.node_id;
    in_params.last_playstate = in_params.current_playstate;
    switch (voice_in.play_state) {
    case PlayState::Paused:
        in_params.current_playstate = ServerPlayState::Paused;
        break;
    case PlayState::Stopped:
        if (in_params.current_playstate != ServerPlayState::Stop) {
            in_params.current_playstate = ServerPlayState::RequestStop;
        }
        break;
    case PlayState::Started:
        in_params.current_playstate = ServerPlayState::Play;
        break;
    default:
        UNREACHABLE_MSG("Unknown playstate {}", voice_in.play_state);
        break;
    }

    in_params.priority = voice_in.priority;
    in_params.sorting_order = voice_in.sorting_order;
    in_params.sample_rate = voice_in.sample_rate;
    in_params.sample_format = voice_in.sample_format;
    in_params.channel_count = voice_in.channel_count;
    in_params.pitch = voice_in.pitch;
    in_params.volume = voice_in.volume;
    in_params.biquad_filter = voice_in.biquad_filter;
    in_params.wave_buffer_count = voice_in.wave_buffer_count;
    in_params.wave_bufffer_head = voice_in.wave_buffer_head;
    if (behavior_info.IsFlushVoiceWaveBuffersSupported()) {
        const auto in_request_count = in_params.wave_buffer_flush_request_count;
        const auto voice_request_count = voice_in.wave_buffer_flush_request_count;
        in_params.wave_buffer_flush_request_count =
            static_cast<u8>(in_request_count + voice_request_count);
    }
    in_params.mix_id = voice_in.mix_id;
    if (behavior_info.IsSplitterSupported()) {
        in_params.splitter_info_id = voice_in.splitter_info_id;
    } else {
        in_params.splitter_info_id = AudioCommon::NO_SPLITTER;
    }

    std::memcpy(in_params.voice_channel_resource_id.data(),
                voice_in.voice_channel_resource_ids.data(),
                sizeof(s32) * in_params.voice_channel_resource_id.size());

    if (behavior_info.IsVoicePlayedSampleCountResetAtLoopPointSupported()) {
        in_params.behavior_flags.is_played_samples_reset_at_loop_point =
            voice_in.behavior_flags.is_played_samples_reset_at_loop_point;
    } else {
        in_params.behavior_flags.is_played_samples_reset_at_loop_point.Assign(0);
    }
    if (behavior_info.IsVoicePitchAndSrcSkippedSupported()) {
        in_params.behavior_flags.is_pitch_and_src_skipped =
            voice_in.behavior_flags.is_pitch_and_src_skipped;
    } else {
        in_params.behavior_flags.is_pitch_and_src_skipped.Assign(0);
    }

    if (voice_in.is_voice_drop_flag_clear_requested) {
        in_params.voice_drop_flag = false;
    }

    if (in_params.additional_params_address != voice_in.additional_params_address ||
        in_params.additional_params_size != voice_in.additional_params_size) {
        in_params.additional_params_address = voice_in.additional_params_address;
        in_params.additional_params_size = voice_in.additional_params_size;
        // TODO(ogniK): Reattach buffer, do we actually need to? Maybe just signal to the DSP that
        // our context is new
    }
}

void ServerVoiceInfo::UpdateWaveBuffers(
    const VoiceInfo::InParams& voice_in,
    std::array<VoiceState*, AudioCommon::MAX_CHANNEL_COUNT>& voice_states,
    BehaviorInfo& behavior_info) {
    if (voice_in.is_new) {
        // Initialize our wave buffers
        for (auto& wave_buffer : in_params.wave_buffer) {
            wave_buffer.start_sample_offset = 0;
            wave_buffer.end_sample_offset = 0;
            wave_buffer.is_looping = false;
            wave_buffer.end_of_stream = false;
            wave_buffer.buffer_address = 0;
            wave_buffer.buffer_size = 0;
            wave_buffer.context_address = 0;
            wave_buffer.context_size = 0;
            wave_buffer.sent_to_dsp = true;
        }

        // Mark all our wave buffers as invalid
        for (std::size_t channel = 0; channel < static_cast<std::size_t>(in_params.channel_count);
             channel++) {
            for (auto& is_valid : voice_states[channel]->is_wave_buffer_valid) {
                is_valid = false;
            }
        }
    }

    // Update our wave buffers
    for (std::size_t i = 0; i < AudioCommon::MAX_WAVE_BUFFERS; i++) {
        // Assume that we have at least 1 channel voice state
        const auto have_valid_wave_buffer = voice_states[0]->is_wave_buffer_valid[i];

        UpdateWaveBuffer(in_params.wave_buffer[i], voice_in.wave_buffer[i], in_params.sample_format,
                         have_valid_wave_buffer, behavior_info);
    }
}

void ServerVoiceInfo::UpdateWaveBuffer(ServerWaveBuffer& out_wavebuffer,
                                       const WaveBuffer& in_wave_buffer, SampleFormat sample_format,
                                       bool is_buffer_valid, BehaviorInfo& behavior_info) {
    if (!is_buffer_valid && out_wavebuffer.sent_to_dsp) {
        out_wavebuffer.buffer_address = 0;
        out_wavebuffer.buffer_size = 0;
    }

    if (!in_wave_buffer.sent_to_server || !in_params.buffer_mapped) {
        // Validate sample offset sizings
        if (sample_format == SampleFormat::Pcm16) {
            const auto buffer_size = in_wave_buffer.buffer_size;
            if (in_wave_buffer.start_sample_offset < 0 || in_wave_buffer.end_sample_offset < 0 ||
                (buffer_size < (sizeof(s16) * in_wave_buffer.start_sample_offset)) ||
                (buffer_size < (sizeof(s16) * in_wave_buffer.end_sample_offset))) {
                // TODO(ogniK): Write error info
                return;
            }
        }
        // TODO(ogniK): ADPCM Size error

        out_wavebuffer.sent_to_dsp = false;
        out_wavebuffer.start_sample_offset = in_wave_buffer.start_sample_offset;
        out_wavebuffer.end_sample_offset = in_wave_buffer.end_sample_offset;
        out_wavebuffer.is_looping = in_wave_buffer.is_looping;
        out_wavebuffer.end_of_stream = in_wave_buffer.end_of_stream;

        out_wavebuffer.buffer_address = in_wave_buffer.buffer_address;
        out_wavebuffer.buffer_size = in_wave_buffer.buffer_size;
        out_wavebuffer.context_address = in_wave_buffer.context_address;
        out_wavebuffer.context_size = in_wave_buffer.context_size;
        in_params.buffer_mapped =
            in_wave_buffer.buffer_address != 0 && in_wave_buffer.buffer_size != 0;
        // TODO(ogniK): Pool mapper attachment
        // TODO(ogniK): IsAdpcmLoopContextBugFixed
    }
}

void ServerVoiceInfo::WriteOutStatus(
    VoiceInfo::OutParams& voice_out, VoiceInfo::InParams& voice_in,
    std::array<VoiceState*, AudioCommon::MAX_CHANNEL_COUNT>& voice_states) {
    if (voice_in.is_new) {
        in_params.is_new = true;
        voice_out.wave_buffer_consumed = 0;
        voice_out.played_sample_count = 0;
        voice_out.voice_dropped = false;
    } else if (!in_params.is_new) {
        voice_out.wave_buffer_consumed = voice_states[0]->wave_buffer_consumed;
        voice_out.played_sample_count = voice_states[0]->played_sample_count;
        voice_out.voice_dropped = in_params.voice_drop_flag;
    } else {
        voice_out.wave_buffer_consumed = 0;
        voice_out.played_sample_count = 0;
        voice_out.voice_dropped = false;
    }
}

const ServerVoiceInfo::InParams& ServerVoiceInfo::GetInParams() const {
    return in_params;
}

ServerVoiceInfo::InParams& ServerVoiceInfo::GetInParams() {
    return in_params;
}

const ServerVoiceInfo::OutParams& ServerVoiceInfo::GetOutParams() const {
    return out_params;
}

ServerVoiceInfo::OutParams& ServerVoiceInfo::GetOutParams() {
    return out_params;
}

bool ServerVoiceInfo::ShouldSkip() const {
    // TODO(ogniK): Handle unmapped wave buffers or parameters
    return !in_params.in_use || (in_params.wave_buffer_count == 0) || in_params.voice_drop_flag;
}

bool ServerVoiceInfo::UpdateForCommandGeneration(VoiceContext& voice_context) {
    std::array<VoiceState*, AudioCommon::MAX_CHANNEL_COUNT> dsp_voice_states{};
    if (in_params.is_new) {
        ResetResources(voice_context);
        in_params.last_volume = in_params.volume;
        in_params.is_new = false;
    }

    const s32 channel_count = in_params.channel_count;
    for (s32 i = 0; i < channel_count; i++) {
        const auto channel_resource = in_params.voice_channel_resource_id[i];
        dsp_voice_states[i] =
            &voice_context.GetDspSharedState(static_cast<std::size_t>(channel_resource));
    }
    return UpdateParametersForCommandGeneration(dsp_voice_states);
}

void ServerVoiceInfo::ResetResources(VoiceContext& voice_context) {
    const s32 channel_count = in_params.channel_count;
    for (s32 i = 0; i < channel_count; i++) {
        const auto channel_resource = in_params.voice_channel_resource_id[i];
        auto& dsp_state =
            voice_context.GetDspSharedState(static_cast<std::size_t>(channel_resource));
        dsp_state = {};
        voice_context.GetChannelResource(static_cast<std::size_t>(channel_resource))
            .UpdateLastMixVolumes();
    }
}

bool ServerVoiceInfo::UpdateParametersForCommandGeneration(
    std::array<VoiceState*, AudioCommon::MAX_CHANNEL_COUNT>& dsp_voice_states) {
    const s32 channel_count = in_params.channel_count;
    if (in_params.wave_buffer_flush_request_count > 0) {
        FlushWaveBuffers(in_params.wave_buffer_flush_request_count, dsp_voice_states,
                         channel_count);
        in_params.wave_buffer_flush_request_count = 0;
    }

    switch (in_params.current_playstate) {
    case ServerPlayState::Play: {
        for (std::size_t i = 0; i < AudioCommon::MAX_WAVE_BUFFERS; i++) {
            if (!in_params.wave_buffer[i].sent_to_dsp) {
                for (s32 channel = 0; channel < channel_count; channel++) {
                    dsp_voice_states[channel]->is_wave_buffer_valid[i] = true;
                }
                in_params.wave_buffer[i].sent_to_dsp = true;
            }
        }
        in_params.should_depop = false;
        return HasValidWaveBuffer(dsp_voice_states[0]);
    }
    case ServerPlayState::Paused:
    case ServerPlayState::Stop: {
        in_params.should_depop = in_params.last_playstate == ServerPlayState::Play;
        return in_params.should_depop;
    }
    case ServerPlayState::RequestStop: {
        for (std::size_t i = 0; i < AudioCommon::MAX_WAVE_BUFFERS; i++) {
            in_params.wave_buffer[i].sent_to_dsp = true;
            for (s32 channel = 0; channel < channel_count; channel++) {
                auto* dsp_state = dsp_voice_states[channel];

                if (dsp_state->is_wave_buffer_valid[i]) {
                    dsp_state->wave_buffer_index =
                        (dsp_state->wave_buffer_index + 1) % AudioCommon::MAX_WAVE_BUFFERS;
                    dsp_state->wave_buffer_consumed++;
                }

                dsp_state->is_wave_buffer_valid[i] = false;
            }
        }

        for (s32 channel = 0; channel < channel_count; channel++) {
            auto* dsp_state = dsp_voice_states[channel];
            dsp_state->offset = 0;
            dsp_state->played_sample_count = 0;
            dsp_state->fraction = 0;
            dsp_state->sample_history.fill(0);
            dsp_state->context = {};
        }

        in_params.current_playstate = ServerPlayState::Stop;
        in_params.should_depop = in_params.last_playstate == ServerPlayState::Play;
        return in_params.should_depop;
    }
    default:
        UNREACHABLE_MSG("Invalid playstate {}", in_params.current_playstate);
    }

    return false;
}

void ServerVoiceInfo::FlushWaveBuffers(
    u8 flush_count, std::array<VoiceState*, AudioCommon::MAX_CHANNEL_COUNT>& dsp_voice_states,
    s32 channel_count) {
    auto wave_head = in_params.wave_bufffer_head;

    for (u8 i = 0; i < flush_count; i++) {
        in_params.wave_buffer[wave_head].sent_to_dsp = true;
        for (s32 channel = 0; channel < channel_count; channel++) {
            auto* dsp_state = dsp_voice_states[channel];
            dsp_state->wave_buffer_consumed++;
            dsp_state->is_wave_buffer_valid[wave_head] = false;
            dsp_state->wave_buffer_index =
                (dsp_state->wave_buffer_index + 1) % AudioCommon::MAX_WAVE_BUFFERS;
        }
        wave_head = (wave_head + 1) % AudioCommon::MAX_WAVE_BUFFERS;
    }
}

bool ServerVoiceInfo::HasValidWaveBuffer(const VoiceState* state) const {
    const auto& valid_wb = state->is_wave_buffer_valid;
    return std::find(valid_wb.begin(), valid_wb.end(), true) != valid_wb.end();
}

VoiceContext::VoiceContext(std::size_t voice_count) : voice_count(voice_count) {
    for (std::size_t i = 0; i < voice_count; i++) {
        voice_channel_resources.emplace_back(static_cast<s32>(i));
        sorted_voice_info.push_back(&voice_info.emplace_back());
        voice_states.emplace_back();
        dsp_voice_states.emplace_back();
    }
}

VoiceContext::~VoiceContext() {
    sorted_voice_info.clear();
}

std::size_t VoiceContext::GetVoiceCount() const {
    return voice_count;
}

ServerVoiceChannelResource& VoiceContext::GetChannelResource(std::size_t i) {
    ASSERT(i < voice_count);
    return voice_channel_resources.at(i);
}

const ServerVoiceChannelResource& VoiceContext::GetChannelResource(std::size_t i) const {
    ASSERT(i < voice_count);
    return voice_channel_resources.at(i);
}

VoiceState& VoiceContext::GetState(std::size_t i) {
    ASSERT(i < voice_count);
    return voice_states.at(i);
}

const VoiceState& VoiceContext::GetState(std::size_t i) const {
    ASSERT(i < voice_count);
    return voice_states.at(i);
}

VoiceState& VoiceContext::GetDspSharedState(std::size_t i) {
    ASSERT(i < voice_count);
    return dsp_voice_states.at(i);
}

const VoiceState& VoiceContext::GetDspSharedState(std::size_t i) const {
    ASSERT(i < voice_count);
    return dsp_voice_states.at(i);
}

ServerVoiceInfo& VoiceContext::GetInfo(std::size_t i) {
    ASSERT(i < voice_count);
    return voice_info.at(i);
}

const ServerVoiceInfo& VoiceContext::GetInfo(std::size_t i) const {
    ASSERT(i < voice_count);
    return voice_info.at(i);
}

ServerVoiceInfo& VoiceContext::GetSortedInfo(std::size_t i) {
    ASSERT(i < voice_count);
    return *sorted_voice_info.at(i);
}

const ServerVoiceInfo& VoiceContext::GetSortedInfo(std::size_t i) const {
    ASSERT(i < voice_count);
    return *sorted_voice_info.at(i);
}

s32 VoiceContext::DecodePcm16(s32* output_buffer, ServerWaveBuffer* wave_buffer, s32 channel,
                              s32 channel_count, s32 buffer_offset, s32 sample_count,
                              Core::Memory::Memory& memory) {
    if (wave_buffer->buffer_address == 0) {
        return 0;
    }
    if (wave_buffer->buffer_size == 0) {
        return 0;
    }
    if (wave_buffer->end_sample_offset < wave_buffer->start_sample_offset) {
        return 0;
    }

    const auto samples_remaining =
        (wave_buffer->end_sample_offset - wave_buffer->start_sample_offset) - buffer_offset;
    const auto start_offset = (wave_buffer->start_sample_offset + buffer_offset) * channel_count;
    const auto buffer_pos = wave_buffer->buffer_address + start_offset;

    s16* buffer_data = reinterpret_cast<s16*>(memory.GetPointer(buffer_pos));

    const auto samples_processed = std::min(sample_count, samples_remaining);

    // Fast path
    if (channel_count == 1) {
        for (std::ptrdiff_t i = 0; i < samples_processed; i++) {
            output_buffer[i] = buffer_data[i];
        }
    } else {
        for (std::ptrdiff_t i = 0; i < samples_processed; i++) {
            output_buffer[i] = buffer_data[i * channel_count + channel];
        }
    }

    return samples_processed;
}

void VoiceContext::SortInfo() {
    for (std::size_t i = 0; i < voice_count; i++) {
        sorted_voice_info[i] = &voice_info[i];
    }

    std::sort(sorted_voice_info.begin(), sorted_voice_info.end(),
              [](const ServerVoiceInfo* lhs, const ServerVoiceInfo* rhs) {
                  const auto& lhs_in = lhs->GetInParams();
                  const auto& rhs_in = rhs->GetInParams();
                  // Sort by priority
                  if (lhs_in.priority != rhs_in.priority) {
                      return lhs_in.priority > rhs_in.priority;
                  } else {
                      // If the priorities match, sort by sorting order
                      return lhs_in.sorting_order > rhs_in.sorting_order;
                  }
              });
}

void VoiceContext::UpdateStateByDspShared() {
    voice_states = dsp_voice_states;
}

} // namespace AudioCore