summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
Diffstat (limited to '')
-rw-r--r--src/audio_core/CMakeLists.txt10
-rw-r--r--src/audio_core/codec.cpp5
-rw-r--r--src/audio_core/codec.h2
-rw-r--r--src/audio_core/command_generator.cpp17
-rw-r--r--src/audio_core/cubeb_sink.cpp4
-rw-r--r--src/audio_core/effect_context.cpp8
-rw-r--r--src/audio_core/effect_context.h33
-rw-r--r--src/audio_core/info_updater.cpp7
-rw-r--r--src/audio_core/mix_context.cpp4
-rw-r--r--src/audio_core/splitter_context.cpp6
-rw-r--r--src/audio_core/voice_context.cpp4
-rw-r--r--src/core/hle/service/nvdrv/devices/nvhost_nvdec.cpp70
-rw-r--r--src/core/hle/service/nvdrv/devices/nvhost_nvdec.h52
-rw-r--r--src/core/hle/service/nvdrv/devices/nvhost_vic.cpp74
-rw-r--r--src/core/hle/service/nvdrv/devices/nvhost_vic.h63
-rw-r--r--src/video_core/renderer_vulkan/vk_stream_buffer.cpp3
-rw-r--r--src/video_core/shader/registry.cpp50
-rw-r--r--src/video_core/shader/registry.h2
18 files changed, 339 insertions, 75 deletions
diff --git a/src/audio_core/CMakeLists.txt b/src/audio_core/CMakeLists.txt
index cb00ef60e..6a7075f73 100644
--- a/src/audio_core/CMakeLists.txt
+++ b/src/audio_core/CMakeLists.txt
@@ -44,6 +44,16 @@ add_library(audio_core STATIC
create_target_directory_groups(audio_core)
+if (NOT MSVC)
+ target_compile_options(audio_core PRIVATE
+ -Werror=ignored-qualifiers
+ -Werror=implicit-fallthrough
+ -Werror=reorder
+ -Werror=sign-compare
+ -Werror=unused-variable
+ )
+endif()
+
target_link_libraries(audio_core PUBLIC common core)
target_link_libraries(audio_core PRIVATE SoundTouch)
diff --git a/src/audio_core/codec.cpp b/src/audio_core/codec.cpp
index c5a0d98ce..2fb91c13a 100644
--- a/src/audio_core/codec.cpp
+++ b/src/audio_core/codec.cpp
@@ -16,8 +16,9 @@ std::vector<s16> DecodeADPCM(const u8* const data, std::size_t size, const ADPCM
constexpr std::size_t FRAME_LEN = 8;
constexpr std::size_t SAMPLES_PER_FRAME = 14;
- constexpr std::array<int, 16> SIGNED_NIBBLES = {
- {0, 1, 2, 3, 4, 5, 6, 7, -8, -7, -6, -5, -4, -3, -2, -1}};
+ static constexpr std::array<int, 16> SIGNED_NIBBLES{
+ 0, 1, 2, 3, 4, 5, 6, 7, -8, -7, -6, -5, -4, -3, -2, -1,
+ };
const std::size_t sample_count = (size / FRAME_LEN) * SAMPLES_PER_FRAME;
const std::size_t ret_size =
diff --git a/src/audio_core/codec.h b/src/audio_core/codec.h
index ef2ce01a8..9507abb1b 100644
--- a/src/audio_core/codec.h
+++ b/src/audio_core/codec.h
@@ -38,7 +38,7 @@ using ADPCM_Coeff = std::array<s16, 16>;
* @param state ADPCM state, this is updated with new state
* @return Decoded stereo signed PCM16 data, sample_count in length
*/
-std::vector<s16> DecodeADPCM(const u8* const data, std::size_t size, const ADPCM_Coeff& coeff,
+std::vector<s16> DecodeADPCM(const u8* data, std::size_t size, const ADPCM_Coeff& coeff,
ADPCMState& state);
}; // namespace AudioCore::Codec
diff --git a/src/audio_core/command_generator.cpp b/src/audio_core/command_generator.cpp
index 8f7da49e6..bba40d13d 100644
--- a/src/audio_core/command_generator.cpp
+++ b/src/audio_core/command_generator.cpp
@@ -152,7 +152,7 @@ void CommandGenerator::GenerateVoiceCommand(ServerVoiceInfo& voice_info) {
if (!destination_data->IsConfigured()) {
continue;
}
- if (destination_data->GetMixId() >= mix_context.GetCount()) {
+ if (destination_data->GetMixId() >= static_cast<int>(mix_context.GetCount())) {
continue;
}
@@ -435,7 +435,7 @@ void CommandGenerator::GenerateAuxCommand(s32 mix_buffer_offset, EffectBase* inf
GetMixBuffer(output_index), worker_params.sample_count, offset, write_count);
memory.WriteBlock(aux->GetRecvInfo(), &recv_info, sizeof(AuxInfoDSP));
- if (samples_read != worker_params.sample_count &&
+ if (samples_read != static_cast<int>(worker_params.sample_count) &&
samples_read <= params.sample_count) {
std::memset(GetMixBuffer(output_index), 0, params.sample_count - samples_read);
}
@@ -611,7 +611,8 @@ void CommandGenerator::GenerateMixCommands(ServerMixInfo& mix_info) {
const auto& dest_mix = mix_context.GetInfo(destination_data->GetMixId());
const auto& dest_in_params = dest_mix.GetInParams();
const auto mix_index = (base - 1) % in_params.buffer_count + in_params.buffer_offset;
- for (std::size_t i = 0; i < dest_in_params.buffer_count; i++) {
+ for (std::size_t i = 0; i < static_cast<std::size_t>(dest_in_params.buffer_count);
+ i++) {
const auto mixed_volume = in_params.volume * destination_data->GetMixVolume(i);
if (mixed_volume != 0.0f) {
GenerateMixCommand(dest_in_params.buffer_offset + i, mix_index, mixed_volume,
@@ -704,7 +705,7 @@ s32 CommandGenerator::DecodePcm16(ServerVoiceInfo& voice_info, VoiceState& dsp_s
std::vector<s16> buffer(samples_processed * channel_count);
memory.ReadBlock(buffer_pos, buffer.data(), buffer.size() * sizeof(s16));
- for (std::size_t i = 0; i < samples_processed; i++) {
+ for (std::size_t i = 0; i < static_cast<std::size_t>(samples_processed); i++) {
sample_buffer[mix_offset + i] = buffer[i * channel_count + channel];
}
}
@@ -726,8 +727,9 @@ s32 CommandGenerator::DecodeAdpcm(ServerVoiceInfo& voice_info, VoiceState& dsp_s
return 0;
}
- constexpr std::array<int, 16> SIGNED_NIBBLES = {
- {0, 1, 2, 3, 4, 5, 6, 7, -8, -7, -6, -5, -4, -3, -2, -1}};
+ static constexpr std::array<int, 16> SIGNED_NIBBLES{
+ 0, 1, 2, 3, 4, 5, 6, 7, -8, -7, -6, -5, -4, -3, -2, -1,
+ };
constexpr std::size_t FRAME_LEN = 8;
constexpr std::size_t NIBBLES_PER_SAMPLE = 16;
@@ -789,7 +791,7 @@ s32 CommandGenerator::DecodeAdpcm(ServerVoiceInfo& voice_info, VoiceState& dsp_s
position_in_frame += 2;
// Decode entire frame
- if (remaining_samples >= SAMPLES_PER_FRAME) {
+ if (remaining_samples >= static_cast<int>(SAMPLES_PER_FRAME)) {
for (std::size_t i = 0; i < SAMPLES_PER_FRAME / 2; i++) {
// Sample 1
@@ -866,7 +868,6 @@ void CommandGenerator::DecodeFromWaveBuffers(ServerVoiceInfo& voice_info, s32* o
const auto resample_rate = static_cast<s32>(
static_cast<float>(in_params.sample_rate) / static_cast<float>(target_sample_rate) *
static_cast<float>(static_cast<s32>(in_params.pitch * 32768.0f)));
- auto* output_base = output;
if (dsp_state.fraction + sample_count * resample_rate >
static_cast<s32>(SCALED_MIX_BUFFER_SIZE - 4ULL)) {
return;
diff --git a/src/audio_core/cubeb_sink.cpp b/src/audio_core/cubeb_sink.cpp
index 83c06c0ed..eb82791f6 100644
--- a/src/audio_core/cubeb_sink.cpp
+++ b/src/audio_core/cubeb_sink.cpp
@@ -192,8 +192,8 @@ SinkStream& CubebSink::AcquireSinkStream(u32 sample_rate, u32 num_channels,
long CubebSinkStream::DataCallback(cubeb_stream* stream, void* user_data, const void* input_buffer,
void* output_buffer, long num_frames) {
- CubebSinkStream* impl = static_cast<CubebSinkStream*>(user_data);
- u8* buffer = reinterpret_cast<u8*>(output_buffer);
+ auto* impl = static_cast<CubebSinkStream*>(user_data);
+ auto* buffer = static_cast<u8*>(output_buffer);
if (!impl) {
return {};
diff --git a/src/audio_core/effect_context.cpp b/src/audio_core/effect_context.cpp
index adfec3df5..4d9cdf524 100644
--- a/src/audio_core/effect_context.cpp
+++ b/src/audio_core/effect_context.cpp
@@ -184,19 +184,19 @@ void EffectAuxInfo::UpdateForCommandGeneration() {
}
}
-const VAddr EffectAuxInfo::GetSendInfo() const {
+VAddr EffectAuxInfo::GetSendInfo() const {
return send_info;
}
-const VAddr EffectAuxInfo::GetSendBuffer() const {
+VAddr EffectAuxInfo::GetSendBuffer() const {
return send_buffer;
}
-const VAddr EffectAuxInfo::GetRecvInfo() const {
+VAddr EffectAuxInfo::GetRecvInfo() const {
return recv_info;
}
-const VAddr EffectAuxInfo::GetRecvBuffer() const {
+VAddr EffectAuxInfo::GetRecvBuffer() const {
return recv_buffer;
}
diff --git a/src/audio_core/effect_context.h b/src/audio_core/effect_context.h
index 2f2da72dd..2c4ce53ef 100644
--- a/src/audio_core/effect_context.h
+++ b/src/audio_core/effect_context.h
@@ -166,13 +166,13 @@ public:
std::array<u8, 0xa0> raw;
};
};
- static_assert(sizeof(EffectInfo::InParams) == 0xc0, "InParams is an invalid size");
+ static_assert(sizeof(InParams) == 0xc0, "InParams is an invalid size");
struct OutParams {
UsageStatus status{};
INSERT_PADDING_BYTES(15);
};
- static_assert(sizeof(EffectInfo::OutParams) == 0x10, "OutParams is an invalid size");
+ static_assert(sizeof(OutParams) == 0x10, "OutParams is an invalid size");
};
struct AuxAddress {
@@ -184,8 +184,8 @@ struct AuxAddress {
class EffectBase {
public:
- EffectBase(EffectType effect_type);
- ~EffectBase();
+ explicit EffectBase(EffectType effect_type);
+ virtual ~EffectBase();
virtual void Update(EffectInfo::InParams& in_params) = 0;
virtual void UpdateForCommandGeneration() = 0;
@@ -206,8 +206,7 @@ protected:
template <typename T>
class EffectGeneric : public EffectBase {
public:
- EffectGeneric(EffectType effect_type) : EffectBase::EffectBase(effect_type) {}
- ~EffectGeneric() = default;
+ explicit EffectGeneric(EffectType effect_type) : EffectBase(effect_type) {}
T& GetParams() {
return internal_params;
@@ -224,7 +223,7 @@ private:
class EffectStubbed : public EffectBase {
public:
explicit EffectStubbed();
- ~EffectStubbed();
+ ~EffectStubbed() override;
void Update(EffectInfo::InParams& in_params) override;
void UpdateForCommandGeneration() override;
@@ -233,7 +232,7 @@ public:
class EffectI3dl2Reverb : public EffectGeneric<I3dl2ReverbParams> {
public:
explicit EffectI3dl2Reverb();
- ~EffectI3dl2Reverb();
+ ~EffectI3dl2Reverb() override;
void Update(EffectInfo::InParams& in_params) override;
void UpdateForCommandGeneration() override;
@@ -245,7 +244,7 @@ private:
class EffectBiquadFilter : public EffectGeneric<BiquadFilterParams> {
public:
explicit EffectBiquadFilter();
- ~EffectBiquadFilter();
+ ~EffectBiquadFilter() override;
void Update(EffectInfo::InParams& in_params) override;
void UpdateForCommandGeneration() override;
@@ -254,14 +253,14 @@ public:
class EffectAuxInfo : public EffectGeneric<AuxInfo> {
public:
explicit EffectAuxInfo();
- ~EffectAuxInfo();
+ ~EffectAuxInfo() override;
void Update(EffectInfo::InParams& in_params) override;
void UpdateForCommandGeneration() override;
- const VAddr GetSendInfo() const;
- const VAddr GetSendBuffer() const;
- const VAddr GetRecvInfo() const;
- const VAddr GetRecvBuffer() const;
+ VAddr GetSendInfo() const;
+ VAddr GetSendBuffer() const;
+ VAddr GetRecvInfo() const;
+ VAddr GetRecvBuffer() const;
private:
VAddr send_info{};
@@ -275,7 +274,7 @@ private:
class EffectDelay : public EffectGeneric<DelayParams> {
public:
explicit EffectDelay();
- ~EffectDelay();
+ ~EffectDelay() override;
void Update(EffectInfo::InParams& in_params) override;
void UpdateForCommandGeneration() override;
@@ -287,7 +286,7 @@ private:
class EffectBufferMixer : public EffectGeneric<BufferMixerParams> {
public:
explicit EffectBufferMixer();
- ~EffectBufferMixer();
+ ~EffectBufferMixer() override;
void Update(EffectInfo::InParams& in_params) override;
void UpdateForCommandGeneration() override;
@@ -296,7 +295,7 @@ public:
class EffectReverb : public EffectGeneric<ReverbParams> {
public:
explicit EffectReverb();
- ~EffectReverb();
+ ~EffectReverb() override;
void Update(EffectInfo::InParams& in_params) override;
void UpdateForCommandGeneration() override;
diff --git a/src/audio_core/info_updater.cpp b/src/audio_core/info_updater.cpp
index f53ce21a5..2940e53a9 100644
--- a/src/audio_core/info_updater.cpp
+++ b/src/audio_core/info_updater.cpp
@@ -64,7 +64,6 @@ bool InfoUpdater::UpdateBehaviorInfo(BehaviorInfo& in_behavior_info) {
}
bool InfoUpdater::UpdateMemoryPools(std::vector<ServerMemoryPoolInfo>& memory_pool_info) {
- const auto force_mapping = behavior_info.IsMemoryPoolForceMappingEnabled();
const auto memory_pool_count = memory_pool_info.size();
const auto total_memory_pool_in = sizeof(ServerMemoryPoolInfo::InParams) * memory_pool_count;
const auto total_memory_pool_out = sizeof(ServerMemoryPoolInfo::OutParams) * memory_pool_count;
@@ -174,7 +173,7 @@ bool InfoUpdater::UpdateVoices(VoiceContext& voice_context,
}
// Voice states for each channel
std::array<VoiceState*, AudioCommon::MAX_CHANNEL_COUNT> voice_states{};
- ASSERT(in_params.id < voice_count);
+ ASSERT(static_cast<std::size_t>(in_params.id) < voice_count);
// Grab our current voice info
auto& voice_info = voice_context.GetInfo(static_cast<std::size_t>(in_params.id));
@@ -352,8 +351,8 @@ ResultCode InfoUpdater::UpdateMixes(MixContext& mix_context, std::size_t mix_buf
for (std::size_t i = 0; i < mix_count; i++) {
const auto& in = mix_in_params[i];
total_buffer_count += in.buffer_count;
- if (in.dest_mix_id > mix_count && in.dest_mix_id != AudioCommon::NO_MIX &&
- in.mix_id != AudioCommon::FINAL_MIX) {
+ if (static_cast<std::size_t>(in.dest_mix_id) > mix_count &&
+ in.dest_mix_id != AudioCommon::NO_MIX && in.mix_id != AudioCommon::FINAL_MIX) {
LOG_ERROR(
Audio,
"Invalid mix destination, mix_id={:X}, dest_mix_id={:X}, mix_buffer_count={:X}",
diff --git a/src/audio_core/mix_context.cpp b/src/audio_core/mix_context.cpp
index 042891490..4bca72eb0 100644
--- a/src/audio_core/mix_context.cpp
+++ b/src/audio_core/mix_context.cpp
@@ -53,7 +53,7 @@ void MixContext::UpdateDistancesFromFinalMix() {
auto mix_id = in_params.mix_id;
// Needs to be referenced out of scope
s32 distance_to_final_mix{AudioCommon::FINAL_MIX};
- for (; distance_to_final_mix < info_count; distance_to_final_mix++) {
+ for (; distance_to_final_mix < static_cast<s32>(info_count); distance_to_final_mix++) {
if (mix_id == AudioCommon::FINAL_MIX) {
// If we're at the final mix, we're done
break;
@@ -77,7 +77,7 @@ void MixContext::UpdateDistancesFromFinalMix() {
}
// If we're out of range for our distance, mark it as no final mix
- if (distance_to_final_mix >= info_count) {
+ if (distance_to_final_mix >= static_cast<s32>(info_count)) {
distance_to_final_mix = AudioCommon::NO_FINAL_MIX;
}
diff --git a/src/audio_core/splitter_context.cpp b/src/audio_core/splitter_context.cpp
index 79bb2f516..f21b53147 100644
--- a/src/audio_core/splitter_context.cpp
+++ b/src/audio_core/splitter_context.cpp
@@ -306,7 +306,7 @@ bool SplitterContext::UpdateInfo(const std::vector<u8>& input, std::size_t& inpu
break;
}
- if (header.send_id < 0 || header.send_id > info_count) {
+ if (header.send_id < 0 || static_cast<std::size_t>(header.send_id) > info_count) {
LOG_ERROR(Audio, "Bad splitter data id");
break;
}
@@ -348,7 +348,7 @@ bool SplitterContext::UpdateData(const std::vector<u8>& input, std::size_t& inpu
break;
}
- if (header.splitter_id < 0 || header.splitter_id > data_count) {
+ if (header.splitter_id < 0 || static_cast<std::size_t>(header.splitter_id) > data_count) {
LOG_ERROR(Audio, "Bad splitter data id");
break;
}
@@ -434,7 +434,7 @@ const std::vector<s32>& NodeStates::GetIndexList() const {
}
void NodeStates::PushTsortResult(s32 index) {
- ASSERT(index < node_count);
+ ASSERT(index < static_cast<s32>(node_count));
index_list[index_pos++] = index;
}
diff --git a/src/audio_core/voice_context.cpp b/src/audio_core/voice_context.cpp
index 1d8f69844..863ac9267 100644
--- a/src/audio_core/voice_context.cpp
+++ b/src/audio_core/voice_context.cpp
@@ -488,11 +488,11 @@ s32 VoiceContext::DecodePcm16(s32* output_buffer, ServerWaveBuffer* wave_buffer,
// Fast path
if (channel_count == 1) {
- for (std::size_t i = 0; i < samples_processed; i++) {
+ for (std::ptrdiff_t i = 0; i < samples_processed; i++) {
output_buffer[i] = buffer_data[i];
}
} else {
- for (std::size_t i = 0; i < samples_processed; i++) {
+ for (std::ptrdiff_t i = 0; i < samples_processed; i++) {
output_buffer[i] = buffer_data[i * channel_count + channel];
}
}
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_nvdec.cpp b/src/core/hle/service/nvdrv/devices/nvhost_nvdec.cpp
index bdae8b887..fcb612864 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_nvdec.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvhost_nvdec.cpp
@@ -22,6 +22,18 @@ u32 nvhost_nvdec::ioctl(Ioctl command, const std::vector<u8>& input, const std::
switch (static_cast<IoctlCommand>(command.raw)) {
case IoctlCommand::IocSetNVMAPfdCommand:
return SetNVMAPfd(input, output);
+ case IoctlCommand::IocSubmit:
+ return Submit(input, output);
+ case IoctlCommand::IocGetSyncpoint:
+ return GetSyncpoint(input, output);
+ case IoctlCommand::IocGetWaitbase:
+ return GetWaitbase(input, output);
+ case IoctlCommand::IocMapBuffer:
+ return MapBuffer(input, output);
+ case IoctlCommand::IocMapBufferEx:
+ return MapBufferEx(input, output);
+ case IoctlCommand::IocUnmapBufferEx:
+ return UnmapBufferEx(input, output);
}
UNIMPLEMENTED_MSG("Unimplemented ioctl");
@@ -30,11 +42,67 @@ u32 nvhost_nvdec::ioctl(Ioctl command, const std::vector<u8>& input, const std::
u32 nvhost_nvdec::SetNVMAPfd(const std::vector<u8>& input, std::vector<u8>& output) {
IoctlSetNvmapFD params{};
- std::memcpy(&params, input.data(), input.size());
+ std::memcpy(&params, input.data(), sizeof(IoctlSetNvmapFD));
LOG_DEBUG(Service_NVDRV, "called, fd={}", params.nvmap_fd);
nvmap_fd = params.nvmap_fd;
return 0;
}
+u32 nvhost_nvdec::Submit(const std::vector<u8>& input, std::vector<u8>& output) {
+ IoctlSubmit params{};
+ std::memcpy(&params, input.data(), sizeof(IoctlSubmit));
+ LOG_WARNING(Service_NVDRV, "(STUBBED) called");
+ std::memcpy(output.data(), &params, sizeof(IoctlSubmit));
+ return 0;
+}
+
+u32 nvhost_nvdec::GetSyncpoint(const std::vector<u8>& input, std::vector<u8>& output) {
+ IoctlGetSyncpoint params{};
+ std::memcpy(&params, input.data(), sizeof(IoctlGetSyncpoint));
+ LOG_INFO(Service_NVDRV, "called, unknown=0x{:X}", params.unknown);
+ params.value = 0; // Seems to be hard coded at 0
+ std::memcpy(output.data(), &params, sizeof(IoctlGetSyncpoint));
+ return 0;
+}
+
+u32 nvhost_nvdec::GetWaitbase(const std::vector<u8>& input, std::vector<u8>& output) {
+ IoctlGetWaitbase params{};
+ std::memcpy(&params, input.data(), sizeof(IoctlGetWaitbase));
+ LOG_INFO(Service_NVDRV, "called, unknown=0x{:X}", params.unknown);
+ params.value = 0; // Seems to be hard coded at 0
+ std::memcpy(output.data(), &params, sizeof(IoctlGetWaitbase));
+ return 0;
+}
+
+u32 nvhost_nvdec::MapBuffer(const std::vector<u8>& input, std::vector<u8>& output) {
+ IoctlMapBuffer params{};
+ std::memcpy(&params, input.data(), sizeof(IoctlMapBuffer));
+ LOG_WARNING(Service_NVDRV, "(STUBBED) called with address={:08X}{:08X}", params.address_2,
+ params.address_1);
+ params.address_1 = 0;
+ params.address_2 = 0;
+ std::memcpy(output.data(), &params, sizeof(IoctlMapBuffer));
+ return 0;
+}
+
+u32 nvhost_nvdec::MapBufferEx(const std::vector<u8>& input, std::vector<u8>& output) {
+ IoctlMapBufferEx params{};
+ std::memcpy(&params, input.data(), sizeof(IoctlMapBufferEx));
+ LOG_WARNING(Service_NVDRV, "(STUBBED) called with address={:08X}{:08X}", params.address_2,
+ params.address_1);
+ params.address_1 = 0;
+ params.address_2 = 0;
+ std::memcpy(output.data(), &params, sizeof(IoctlMapBufferEx));
+ return 0;
+}
+
+u32 nvhost_nvdec::UnmapBufferEx(const std::vector<u8>& input, std::vector<u8>& output) {
+ IoctlUnmapBufferEx params{};
+ std::memcpy(&params, input.data(), sizeof(IoctlUnmapBufferEx));
+ LOG_WARNING(Service_NVDRV, "(STUBBED) called");
+ std::memcpy(output.data(), &params, sizeof(IoctlUnmapBufferEx));
+ return 0;
+}
+
} // namespace Service::Nvidia::Devices
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_nvdec.h b/src/core/hle/service/nvdrv/devices/nvhost_nvdec.h
index cbdac8069..4332db118 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_nvdec.h
+++ b/src/core/hle/service/nvdrv/devices/nvhost_nvdec.h
@@ -23,16 +23,66 @@ public:
private:
enum class IoctlCommand : u32_le {
IocSetNVMAPfdCommand = 0x40044801,
+ IocSubmit = 0xC0400001,
+ IocGetSyncpoint = 0xC0080002,
+ IocGetWaitbase = 0xC0080003,
+ IocMapBuffer = 0xC01C0009,
+ IocMapBufferEx = 0xC0A40009,
+ IocUnmapBufferEx = 0xC0A4000A,
};
struct IoctlSetNvmapFD {
u32_le nvmap_fd;
};
- static_assert(sizeof(IoctlSetNvmapFD) == 4, "IoctlSetNvmapFD is incorrect size");
+ static_assert(sizeof(IoctlSetNvmapFD) == 0x4, "IoctlSetNvmapFD is incorrect size");
+
+ struct IoctlSubmit {
+ INSERT_PADDING_BYTES(0x40); // TODO(DarkLordZach): RE this structure
+ };
+ static_assert(sizeof(IoctlSubmit) == 0x40, "IoctlSubmit has incorrect size");
+
+ struct IoctlGetSyncpoint {
+ u32 unknown; // seems to be ignored? Nintendo added this
+ u32 value;
+ };
+ static_assert(sizeof(IoctlGetSyncpoint) == 0x08, "IoctlGetSyncpoint has incorrect size");
+
+ struct IoctlGetWaitbase {
+ u32 unknown; // seems to be ignored? Nintendo added this
+ u32 value;
+ };
+ static_assert(sizeof(IoctlGetWaitbase) == 0x08, "IoctlGetWaitbase has incorrect size");
+
+ struct IoctlMapBuffer {
+ u32 unknown;
+ u32 address_1;
+ u32 address_2;
+ INSERT_PADDING_BYTES(0x10); // TODO(DarkLordZach): RE this structure
+ };
+ static_assert(sizeof(IoctlMapBuffer) == 0x1C, "IoctlMapBuffer is incorrect size");
+
+ struct IoctlMapBufferEx {
+ u32 unknown;
+ u32 address_1;
+ u32 address_2;
+ INSERT_PADDING_BYTES(0x98); // TODO(DarkLordZach): RE this structure
+ };
+ static_assert(sizeof(IoctlMapBufferEx) == 0xA4, "IoctlMapBufferEx has incorrect size");
+
+ struct IoctlUnmapBufferEx {
+ INSERT_PADDING_BYTES(0xA4); // TODO(DarkLordZach): RE this structure
+ };
+ static_assert(sizeof(IoctlUnmapBufferEx) == 0xA4, "IoctlUnmapBufferEx has incorrect size");
u32_le nvmap_fd{};
u32 SetNVMAPfd(const std::vector<u8>& input, std::vector<u8>& output);
+ u32 Submit(const std::vector<u8>& input, std::vector<u8>& output);
+ u32 GetSyncpoint(const std::vector<u8>& input, std::vector<u8>& output);
+ u32 GetWaitbase(const std::vector<u8>& input, std::vector<u8>& output);
+ u32 MapBuffer(const std::vector<u8>& input, std::vector<u8>& output);
+ u32 MapBufferEx(const std::vector<u8>& input, std::vector<u8>& output);
+ u32 UnmapBufferEx(const std::vector<u8>& input, std::vector<u8>& output);
};
} // namespace Service::Nvidia::Devices
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_vic.cpp b/src/core/hle/service/nvdrv/devices/nvhost_vic.cpp
index c695b8863..9da19ad56 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_vic.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvhost_vic.cpp
@@ -22,6 +22,18 @@ u32 nvhost_vic::ioctl(Ioctl command, const std::vector<u8>& input, const std::ve
switch (static_cast<IoctlCommand>(command.raw)) {
case IoctlCommand::IocSetNVMAPfdCommand:
return SetNVMAPfd(input, output);
+ case IoctlCommand::IocSubmit:
+ return Submit(input, output);
+ case IoctlCommand::IocGetSyncpoint:
+ return GetSyncpoint(input, output);
+ case IoctlCommand::IocGetWaitbase:
+ return GetWaitbase(input, output);
+ case IoctlCommand::IocMapBuffer:
+ return MapBuffer(input, output);
+ case IoctlCommand::IocMapBufferEx:
+ return MapBuffer(input, output);
+ case IoctlCommand::IocUnmapBufferEx:
+ return UnmapBufferEx(input, output);
}
UNIMPLEMENTED_MSG("Unimplemented ioctl");
@@ -30,11 +42,71 @@ u32 nvhost_vic::ioctl(Ioctl command, const std::vector<u8>& input, const std::ve
u32 nvhost_vic::SetNVMAPfd(const std::vector<u8>& input, std::vector<u8>& output) {
IoctlSetNvmapFD params{};
- std::memcpy(&params, input.data(), input.size());
+ std::memcpy(&params, input.data(), sizeof(IoctlSetNvmapFD));
LOG_DEBUG(Service_NVDRV, "called, fd={}", params.nvmap_fd);
nvmap_fd = params.nvmap_fd;
return 0;
}
+u32 nvhost_vic::Submit(const std::vector<u8>& input, std::vector<u8>& output) {
+ IoctlSubmit params{};
+ std::memcpy(&params, input.data(), sizeof(IoctlSubmit));
+ LOG_WARNING(Service_NVDRV, "(STUBBED) called");
+
+ // Workaround for Luigi's Mansion 3, as nvhost_vic is not implemented for asynch GPU
+ params.command_buffer = {};
+
+ std::memcpy(output.data(), &params, sizeof(IoctlSubmit));
+ return 0;
+}
+
+u32 nvhost_vic::GetSyncpoint(const std::vector<u8>& input, std::vector<u8>& output) {
+ IoctlGetSyncpoint params{};
+ std::memcpy(&params, input.data(), sizeof(IoctlGetSyncpoint));
+ LOG_INFO(Service_NVDRV, "called, unknown=0x{:X}", params.unknown);
+ params.value = 0; // Seems to be hard coded at 0
+ std::memcpy(output.data(), &params, sizeof(IoctlGetSyncpoint));
+ return 0;
+}
+
+u32 nvhost_vic::GetWaitbase(const std::vector<u8>& input, std::vector<u8>& output) {
+ IoctlGetWaitbase params{};
+ std::memcpy(&params, input.data(), sizeof(IoctlGetWaitbase));
+ LOG_INFO(Service_NVDRV, "called, unknown=0x{:X}", params.unknown);
+ params.value = 0; // Seems to be hard coded at 0
+ std::memcpy(output.data(), &params, sizeof(IoctlGetWaitbase));
+ return 0;
+}
+
+u32 nvhost_vic::MapBuffer(const std::vector<u8>& input, std::vector<u8>& output) {
+ IoctlMapBuffer params{};
+ std::memcpy(&params, input.data(), sizeof(IoctlMapBuffer));
+ LOG_WARNING(Service_NVDRV, "(STUBBED) called with address={:08X}{:08X}", params.address_2,
+ params.address_1);
+ params.address_1 = 0;
+ params.address_2 = 0;
+ std::memcpy(output.data(), &params, sizeof(IoctlMapBuffer));
+ return 0;
+}
+
+u32 nvhost_vic::MapBufferEx(const std::vector<u8>& input, std::vector<u8>& output) {
+ IoctlMapBufferEx params{};
+ std::memcpy(&params, input.data(), sizeof(IoctlMapBufferEx));
+ LOG_WARNING(Service_NVDRV, "(STUBBED) called with address={:08X}{:08X}", params.address_2,
+ params.address_1);
+ params.address_1 = 0;
+ params.address_2 = 0;
+ std::memcpy(output.data(), &params, sizeof(IoctlMapBufferEx));
+ return 0;
+}
+
+u32 nvhost_vic::UnmapBufferEx(const std::vector<u8>& input, std::vector<u8>& output) {
+ IoctlUnmapBufferEx params{};
+ std::memcpy(&params, input.data(), sizeof(IoctlUnmapBufferEx));
+ LOG_WARNING(Service_NVDRV, "(STUBBED) called");
+ std::memcpy(output.data(), &params, sizeof(IoctlUnmapBufferEx));
+ return 0;
+}
+
} // namespace Service::Nvidia::Devices
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_vic.h b/src/core/hle/service/nvdrv/devices/nvhost_vic.h
index bec32bea1..a7bb7bbd5 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_vic.h
+++ b/src/core/hle/service/nvdrv/devices/nvhost_vic.h
@@ -4,6 +4,7 @@
#pragma once
+#include <array>
#include <vector>
#include "common/common_types.h"
#include "common/swap.h"
@@ -23,6 +24,12 @@ public:
private:
enum class IoctlCommand : u32_le {
IocSetNVMAPfdCommand = 0x40044801,
+ IocSubmit = 0xC0400001,
+ IocGetSyncpoint = 0xC0080002,
+ IocGetWaitbase = 0xC0080003,
+ IocMapBuffer = 0xC01C0009,
+ IocMapBufferEx = 0xC03C0009,
+ IocUnmapBufferEx = 0xC03C000A,
};
struct IoctlSetNvmapFD {
@@ -30,9 +37,65 @@ private:
};
static_assert(sizeof(IoctlSetNvmapFD) == 4, "IoctlSetNvmapFD is incorrect size");
+ struct IoctlSubmitCommandBuffer {
+ u32 id;
+ u32 offset;
+ u32 count;
+ };
+ static_assert(sizeof(IoctlSubmitCommandBuffer) == 0xC,
+ "IoctlSubmitCommandBuffer is incorrect size");
+
+ struct IoctlSubmit {
+ u32 command_buffer_count;
+ u32 relocations_count;
+ u32 syncpt_count;
+ u32 wait_count;
+ std::array<IoctlSubmitCommandBuffer, 4> command_buffer;
+ };
+ static_assert(sizeof(IoctlSubmit) == 0x40, "IoctlSubmit is incorrect size");
+
+ struct IoctlGetSyncpoint {
+ u32 unknown; // seems to be ignored? Nintendo added this
+ u32 value;
+ };
+ static_assert(sizeof(IoctlGetSyncpoint) == 0x8, "IoctlGetSyncpoint is incorrect size");
+
+ struct IoctlGetWaitbase {
+ u32 unknown; // seems to be ignored? Nintendo added this
+ u32 value;
+ };
+ static_assert(sizeof(IoctlGetWaitbase) == 0x8, "IoctlGetWaitbase is incorrect size");
+
+ struct IoctlMapBuffer {
+ u32 unknown;
+ u32 address_1;
+ u32 address_2;
+ INSERT_PADDING_BYTES(0x10); // TODO(DarkLordZach): RE this structure
+ };
+ static_assert(sizeof(IoctlMapBuffer) == 0x1C, "IoctlMapBuffer is incorrect size");
+
+ struct IoctlMapBufferEx {
+ u32 unknown;
+ u32 address_1;
+ u32 address_2;
+ INSERT_PADDING_BYTES(0x30); // TODO(DarkLordZach): RE this structure
+ };
+ static_assert(sizeof(IoctlMapBufferEx) == 0x3C, "IoctlMapBufferEx is incorrect size");
+
+ struct IoctlUnmapBufferEx {
+ INSERT_PADDING_BYTES(0x3C); // TODO(DarkLordZach): RE this structure
+ };
+ static_assert(sizeof(IoctlUnmapBufferEx) == 0x3C, "IoctlUnmapBufferEx is incorrect size");
+
u32_le nvmap_fd{};
u32 SetNVMAPfd(const std::vector<u8>& input, std::vector<u8>& output);
+ u32 Submit(const std::vector<u8>& input, std::vector<u8>& output);
+ u32 GetSyncpoint(const std::vector<u8>& input, std::vector<u8>& output);
+ u32 GetWaitbase(const std::vector<u8>& input, std::vector<u8>& output);
+ u32 MapBuffer(const std::vector<u8>& input, std::vector<u8>& output);
+ u32 MapBufferEx(const std::vector<u8>& input, std::vector<u8>& output);
+ u32 UnmapBufferEx(const std::vector<u8>& input, std::vector<u8>& output);
};
} // namespace Service::Nvidia::Devices
diff --git a/src/video_core/renderer_vulkan/vk_stream_buffer.cpp b/src/video_core/renderer_vulkan/vk_stream_buffer.cpp
index 5218c875b..1b59612b9 100644
--- a/src/video_core/renderer_vulkan/vk_stream_buffer.cpp
+++ b/src/video_core/renderer_vulkan/vk_stream_buffer.cpp
@@ -120,7 +120,8 @@ void VKStreamBuffer::CreateBuffers(VkBufferUsageFlags usage) {
// Substract from the preferred heap size some bytes to avoid getting out of memory.
const VkDeviceSize heap_size = memory_properties.memoryHeaps[preferred_heap].size;
- const VkDeviceSize allocable_size = heap_size - 9 * 1024 * 1024;
+ // As per DXVK's example, using `heap_size / 2`
+ const VkDeviceSize allocable_size = heap_size / 2;
buffer = device.GetLogical().CreateBuffer({
.sType = VK_STRUCTURE_TYPE_BUFFER_CREATE_INFO,
.pNext = nullptr,
diff --git a/src/video_core/shader/registry.cpp b/src/video_core/shader/registry.cpp
index cdf274e54..148d91fcb 100644
--- a/src/video_core/shader/registry.cpp
+++ b/src/video_core/shader/registry.cpp
@@ -24,44 +24,45 @@ GraphicsInfo MakeGraphicsInfo(ShaderType shader_stage, ConstBufferEngineInterfac
if (shader_stage == ShaderType::Compute) {
return {};
}
- auto& graphics = static_cast<Tegra::Engines::Maxwell3D&>(engine);
-
- GraphicsInfo info;
- info.tfb_layouts = graphics.regs.tfb_layouts;
- info.tfb_varying_locs = graphics.regs.tfb_varying_locs;
- info.primitive_topology = graphics.regs.draw.topology;
- info.tessellation_primitive = graphics.regs.tess_mode.prim;
- info.tessellation_spacing = graphics.regs.tess_mode.spacing;
- info.tfb_enabled = graphics.regs.tfb_enabled;
- info.tessellation_clockwise = graphics.regs.tess_mode.cw;
- return info;
+
+ auto& graphics = dynamic_cast<Tegra::Engines::Maxwell3D&>(engine);
+
+ return {
+ .tfb_layouts = graphics.regs.tfb_layouts,
+ .tfb_varying_locs = graphics.regs.tfb_varying_locs,
+ .primitive_topology = graphics.regs.draw.topology,
+ .tessellation_primitive = graphics.regs.tess_mode.prim,
+ .tessellation_spacing = graphics.regs.tess_mode.spacing,
+ .tfb_enabled = graphics.regs.tfb_enabled != 0,
+ .tessellation_clockwise = graphics.regs.tess_mode.cw.Value() != 0,
+ };
}
ComputeInfo MakeComputeInfo(ShaderType shader_stage, ConstBufferEngineInterface& engine) {
if (shader_stage != ShaderType::Compute) {
return {};
}
- auto& compute = static_cast<Tegra::Engines::KeplerCompute&>(engine);
+
+ auto& compute = dynamic_cast<Tegra::Engines::KeplerCompute&>(engine);
const auto& launch = compute.launch_description;
- ComputeInfo info;
- info.workgroup_size = {launch.block_dim_x, launch.block_dim_y, launch.block_dim_z};
- info.local_memory_size_in_words = launch.local_pos_alloc;
- info.shared_memory_size_in_words = launch.shared_alloc;
- return info;
+ return {
+ .workgroup_size = {launch.block_dim_x, launch.block_dim_y, launch.block_dim_z},
+ .shared_memory_size_in_words = launch.shared_alloc,
+ .local_memory_size_in_words = launch.local_pos_alloc,
+ };
}
} // Anonymous namespace
-Registry::Registry(Tegra::Engines::ShaderType shader_stage, const SerializedRegistryInfo& info)
+Registry::Registry(ShaderType shader_stage, const SerializedRegistryInfo& info)
: stage{shader_stage}, stored_guest_driver_profile{info.guest_driver_profile},
bound_buffer{info.bound_buffer}, graphics_info{info.graphics}, compute_info{info.compute} {}
-Registry::Registry(Tegra::Engines::ShaderType shader_stage,
- Tegra::Engines::ConstBufferEngineInterface& engine)
- : stage{shader_stage}, engine{&engine}, bound_buffer{engine.GetBoundBuffer()},
- graphics_info{MakeGraphicsInfo(shader_stage, engine)}, compute_info{MakeComputeInfo(
- shader_stage, engine)} {}
+Registry::Registry(ShaderType shader_stage, ConstBufferEngineInterface& engine_)
+ : stage{shader_stage}, engine{&engine_}, bound_buffer{engine_.GetBoundBuffer()},
+ graphics_info{MakeGraphicsInfo(shader_stage, engine_)}, compute_info{MakeComputeInfo(
+ shader_stage, engine_)} {}
Registry::~Registry() = default;
@@ -113,8 +114,7 @@ std::optional<Tegra::Engines::SamplerDescriptor> Registry::ObtainSeparateSampler
return value;
}
-std::optional<Tegra::Engines::SamplerDescriptor> Registry::ObtainBindlessSampler(u32 buffer,
- u32 offset) {
+std::optional<SamplerDescriptor> Registry::ObtainBindlessSampler(u32 buffer, u32 offset) {
const std::pair key = {buffer, offset};
const auto iter = bindless_samplers.find(key);
if (iter != bindless_samplers.end()) {
diff --git a/src/video_core/shader/registry.h b/src/video_core/shader/registry.h
index 231206765..4bebefdde 100644
--- a/src/video_core/shader/registry.h
+++ b/src/video_core/shader/registry.h
@@ -94,7 +94,7 @@ public:
explicit Registry(Tegra::Engines::ShaderType shader_stage, const SerializedRegistryInfo& info);
explicit Registry(Tegra::Engines::ShaderType shader_stage,
- Tegra::Engines::ConstBufferEngineInterface& engine);
+ Tegra::Engines::ConstBufferEngineInterface& engine_);
~Registry();
generated by cgit v1.2.3 (git 2.25.1) at 2024-07-07 09:04:19 +0200