Compare commits

..

1 Commits

Author SHA1 Message Date
MonsterDruide1
115bf20448 TAS: Show all script lengths for multiplayer 2023-01-07 21:15:37 +01:00
49 changed files with 1338 additions and 1264 deletions

View File

@@ -292,6 +292,9 @@ class InputDevice {
public:
virtual ~InputDevice() = default;
// Request input device to update if necessary
virtual void SoftUpdate() {}
// Force input device to update data regardless of the current state
virtual void ForceUpdate() {}

View File

@@ -185,7 +185,6 @@ void RestoreGlobalState(bool is_powered_on) {
// Renderer
values.fsr_sharpening_slider.SetGlobal(true);
values.renderer_backend.SetGlobal(true);
values.renderer_force_max_clock.SetGlobal(true);
values.vulkan_device.SetGlobal(true);
values.aspect_ratio.SetGlobal(true);
values.max_anisotropy.SetGlobal(true);

View File

@@ -415,7 +415,6 @@ struct Values {
// Renderer
SwitchableSetting<RendererBackend, true> renderer_backend{
RendererBackend::Vulkan, RendererBackend::OpenGL, RendererBackend::Null, "backend"};
SwitchableSetting<bool> renderer_force_max_clock{true, "force_max_clock"};
Setting<bool> renderer_debug{false, "debug"};
Setting<bool> renderer_shader_feedback{false, "shader_feedback"};
Setting<bool> enable_nsight_aftermath{false, "nsight_aftermath"};

View File

@@ -229,11 +229,7 @@ std::shared_ptr<Dynarmic::A32::Jit> ARM_Dynarmic_32::MakeJit(Common::PageTable*
config.enable_cycle_counting = true;
// Code cache size
#ifdef ARCHITECTURE_arm64
config.code_cache_size = 128_MiB;
#else
config.code_cache_size = 512_MiB;
#endif
// Allow memory fault handling to work
if (system.DebuggerEnabled()) {

View File

@@ -288,11 +288,7 @@ std::shared_ptr<Dynarmic::A64::Jit> ARM_Dynarmic_64::MakeJit(Common::PageTable*
config.enable_cycle_counting = true;
// Code cache size
#ifdef ARCHITECTURE_arm64
config.code_cache_size = 128_MiB;
#else
config.code_cache_size = 512_MiB;
#endif
// Allow memory fault handling to work
if (system.DebuggerEnabled()) {

View File

@@ -1434,6 +1434,16 @@ AnalogSticks EmulatedController::GetSticks() const {
return {};
}
// Some drivers like stick from buttons need constant refreshing
for (auto& device : stick_devices) {
if (!device) {
continue;
}
lock.unlock();
device->SoftUpdate();
lock.lock();
}
return controller.analog_stick_state;
}

View File

@@ -117,8 +117,6 @@ Errno TranslateNativeError(int e) {
return Errno::NETUNREACH;
case WSAEMSGSIZE:
return Errno::MSGSIZE;
case WSAETIMEDOUT:
return Errno::TIMEDOUT;
default:
UNIMPLEMENTED_MSG("Unimplemented errno={}", e);
return Errno::OTHER;
@@ -213,8 +211,6 @@ Errno TranslateNativeError(int e) {
return Errno::NETUNREACH;
case EMSGSIZE:
return Errno::MSGSIZE;
case ETIMEDOUT:
return Errno::TIMEDOUT;
default:
UNIMPLEMENTED_MSG("Unimplemented errno={}", e);
return Errno::OTHER;
@@ -230,7 +226,7 @@ Errno GetAndLogLastError() {
int e = errno;
#endif
const Errno err = TranslateNativeError(e);
if (err == Errno::AGAIN || err == Errno::TIMEDOUT) {
if (err == Errno::AGAIN) {
return err;
}
LOG_ERROR(Network, "Socket operation error: {}", Common::NativeErrorToString(e));

View File

@@ -156,10 +156,12 @@ void Tas::RecordInput(u64 buttons, TasAnalog left_axis, TasAnalog right_axis) {
};
}
std::tuple<TasState, size_t, size_t> Tas::GetStatus() const {
std::tuple<TasState, size_t, std::array<size_t, PLAYER_NUMBER>> Tas::GetStatus() const {
TasState state;
std::array<size_t, PLAYER_NUMBER> lengths{0};
if (is_recording) {
return {TasState::Recording, 0, record_commands.size()};
lengths[0] = record_commands.size();
return {TasState::Recording, record_commands.size(), lengths};
}
if (is_running) {
@@ -168,7 +170,11 @@ std::tuple<TasState, size_t, size_t> Tas::GetStatus() const {
state = TasState::Stopped;
}
return {state, current_command, script_length};
for (size_t i = 0; i < PLAYER_NUMBER; i++) {
lengths[i] = commands[i].size();
}
return {state, current_command, lengths};
}
void Tas::UpdateThread() {

View File

@@ -124,7 +124,7 @@ public:
* Current playback progress ;
* Total length of script file currently loaded or being recorded
*/
std::tuple<TasState, size_t, size_t> GetStatus() const;
std::tuple<TasState, size_t, std::array<size_t, PLAYER_NUMBER>> GetStatus() const;
private:
enum class TasAxis : u8;

View File

@@ -13,11 +13,11 @@ class Stick final : public Common::Input::InputDevice {
public:
using Button = std::unique_ptr<Common::Input::InputDevice>;
Stick(Button up_, Button down_, Button left_, Button right_, Button modifier_, Button updater_,
Stick(Button up_, Button down_, Button left_, Button right_, Button modifier_,
float modifier_scale_, float modifier_angle_)
: up(std::move(up_)), down(std::move(down_)), left(std::move(left_)),
right(std::move(right_)), modifier(std::move(modifier_)), updater(std::move(updater_)),
modifier_scale(modifier_scale_), modifier_angle(modifier_angle_) {
right(std::move(right_)), modifier(std::move(modifier_)), modifier_scale(modifier_scale_),
modifier_angle(modifier_angle_) {
up->SetCallback({
.on_change =
[this](const Common::Input::CallbackStatus& callback_) {
@@ -48,9 +48,6 @@ public:
UpdateModButtonStatus(callback_);
},
});
updater->SetCallback({
.on_change = [this](const Common::Input::CallbackStatus& callback_) { SoftUpdate(); },
});
last_x_axis_value = 0.0f;
last_y_axis_value = 0.0f;
}
@@ -251,7 +248,7 @@ public:
modifier->ForceUpdate();
}
void SoftUpdate() {
void SoftUpdate() override {
Common::Input::CallbackStatus status{
.type = Common::Input::InputType::Stick,
.stick_status = GetStatus(),
@@ -311,7 +308,6 @@ private:
Button left;
Button right;
Button modifier;
Button updater;
float modifier_scale{};
float modifier_angle{};
float angle{};
@@ -335,12 +331,11 @@ std::unique_ptr<Common::Input::InputDevice> StickFromButton::Create(
auto left = Common::Input::CreateInputDeviceFromString(params.Get("left", null_engine));
auto right = Common::Input::CreateInputDeviceFromString(params.Get("right", null_engine));
auto modifier = Common::Input::CreateInputDeviceFromString(params.Get("modifier", null_engine));
auto updater = Common::Input::CreateInputDeviceFromString("engine:updater,button:0");
auto modifier_scale = params.Get("modifier_scale", 0.5f);
auto modifier_angle = params.Get("modifier_angle", 5.5f);
return std::make_unique<Stick>(std::move(up), std::move(down), std::move(left),
std::move(right), std::move(modifier), std::move(updater),
modifier_scale, modifier_angle);
std::move(right), std::move(modifier), modifier_scale,
modifier_angle);
}
} // namespace InputCommon

View File

@@ -28,28 +28,6 @@
namespace InputCommon {
/// Dummy engine to get periodic updates
class UpdateEngine final : public InputEngine {
public:
explicit UpdateEngine(std::string input_engine_) : InputEngine(std::move(input_engine_)) {
PreSetController(identifier);
}
void PumpEvents() {
SetButton(identifier, 0, last_state);
last_state = !last_state;
}
private:
static constexpr PadIdentifier identifier = {
.guid = Common::UUID{},
.port = 0,
.pad = 0,
};
bool last_state{};
};
struct InputSubsystem::Impl {
template <typename Engine>
void RegisterEngine(std::string name, std::shared_ptr<Engine>& engine) {
@@ -67,7 +45,6 @@ struct InputSubsystem::Impl {
void Initialize() {
mapping_factory = std::make_shared<MappingFactory>();
RegisterEngine("updater", update_engine);
RegisterEngine("keyboard", keyboard);
RegisterEngine("mouse", mouse);
RegisterEngine("touch", touch_screen);
@@ -97,7 +74,6 @@ struct InputSubsystem::Impl {
}
void Shutdown() {
UnregisterEngine(update_engine);
UnregisterEngine(keyboard);
UnregisterEngine(mouse);
UnregisterEngine(touch_screen);
@@ -276,7 +252,6 @@ struct InputSubsystem::Impl {
}
void PumpEvents() const {
update_engine->PumpEvents();
#ifdef HAVE_SDL2
sdl->PumpEvents();
#endif
@@ -288,7 +263,6 @@ struct InputSubsystem::Impl {
std::shared_ptr<MappingFactory> mapping_factory;
std::shared_ptr<UpdateEngine> update_engine;
std::shared_ptr<Keyboard> keyboard;
std::shared_ptr<Mouse> mouse;
std::shared_ptr<TouchScreen> touch_screen;

View File

@@ -321,12 +321,8 @@ Id EmitGetAttribute(EmitContext& ctx, IR::Attribute attr, Id vertex) {
case IR::Attribute::PositionY:
case IR::Attribute::PositionZ:
case IR::Attribute::PositionW:
return ctx.OpLoad(
ctx.F32[1],
ctx.need_input_position_indirect
? AttrPointer(ctx, ctx.input_f32, vertex, ctx.input_position, ctx.u32_zero_value,
ctx.Const(element))
: AttrPointer(ctx, ctx.input_f32, vertex, ctx.input_position, ctx.Const(element)));
return ctx.OpLoad(ctx.F32[1], AttrPointer(ctx, ctx.input_f32, vertex, ctx.input_position,
ctx.Const(element)));
case IR::Attribute::InstanceId:
if (ctx.profile.support_vertex_instance_id) {
return ctx.OpBitcast(ctx.F32[1], ctx.OpLoad(ctx.U32[1], ctx.instance_id));

View File

@@ -58,10 +58,11 @@ Id SelectValue(EmitContext& ctx, Id in_range, Id value, Id src_thread_id) {
ctx.OpGroupNonUniformShuffle(ctx.U32[1], SubgroupScope(ctx), value, src_thread_id), value);
}
Id AddPartitionBase(EmitContext& ctx, Id thread_id) {
const Id partition_idx{ctx.OpShiftRightLogical(ctx.U32[1], GetThreadId(ctx), ctx.Const(5u))};
const Id partition_base{ctx.OpShiftLeftLogical(ctx.U32[1], partition_idx, ctx.Const(5u))};
return ctx.OpIAdd(ctx.U32[1], thread_id, partition_base);
Id GetUpperClamp(EmitContext& ctx, Id invocation_id, Id clamp) {
const Id thirty_two{ctx.Const(32u)};
const Id is_upper_partition{ctx.OpSGreaterThanEqual(ctx.U1, invocation_id, thirty_two)};
const Id upper_clamp{ctx.OpIAdd(ctx.U32[1], thirty_two, clamp)};
return ctx.OpSelect(ctx.U32[1], is_upper_partition, upper_clamp, clamp);
}
} // Anonymous namespace
@@ -144,32 +145,35 @@ Id EmitSubgroupGeMask(EmitContext& ctx) {
Id EmitShuffleIndex(EmitContext& ctx, IR::Inst* inst, Id value, Id index, Id clamp,
Id segmentation_mask) {
const Id not_seg_mask{ctx.OpNot(ctx.U32[1], segmentation_mask)};
const Id thread_id{EmitLaneId(ctx)};
const Id thread_id{GetThreadId(ctx)};
if (ctx.profile.warp_size_potentially_larger_than_guest) {
const Id thirty_two{ctx.Const(32u)};
const Id is_upper_partition{ctx.OpSGreaterThanEqual(ctx.U1, thread_id, thirty_two)};
const Id upper_index{ctx.OpIAdd(ctx.U32[1], thirty_two, index)};
const Id upper_clamp{ctx.OpIAdd(ctx.U32[1], thirty_two, clamp)};
index = ctx.OpSelect(ctx.U32[1], is_upper_partition, upper_index, index);
clamp = ctx.OpSelect(ctx.U32[1], is_upper_partition, upper_clamp, clamp);
}
const Id min_thread_id{ComputeMinThreadId(ctx, thread_id, segmentation_mask)};
const Id max_thread_id{ComputeMaxThreadId(ctx, min_thread_id, clamp, not_seg_mask)};
const Id lhs{ctx.OpBitwiseAnd(ctx.U32[1], index, not_seg_mask)};
Id src_thread_id{ctx.OpBitwiseOr(ctx.U32[1], lhs, min_thread_id)};
const Id src_thread_id{ctx.OpBitwiseOr(ctx.U32[1], lhs, min_thread_id)};
const Id in_range{ctx.OpSLessThanEqual(ctx.U1, src_thread_id, max_thread_id)};
if (ctx.profile.warp_size_potentially_larger_than_guest) {
src_thread_id = AddPartitionBase(ctx, src_thread_id);
}
SetInBoundsFlag(inst, in_range);
return SelectValue(ctx, in_range, value, src_thread_id);
}
Id EmitShuffleUp(EmitContext& ctx, IR::Inst* inst, Id value, Id index, Id clamp,
Id segmentation_mask) {
const Id thread_id{EmitLaneId(ctx)};
const Id max_thread_id{GetMaxThreadId(ctx, thread_id, clamp, segmentation_mask)};
Id src_thread_id{ctx.OpISub(ctx.U32[1], thread_id, index)};
const Id in_range{ctx.OpSGreaterThanEqual(ctx.U1, src_thread_id, max_thread_id)};
const Id thread_id{GetThreadId(ctx)};
if (ctx.profile.warp_size_potentially_larger_than_guest) {
src_thread_id = AddPartitionBase(ctx, src_thread_id);
clamp = GetUpperClamp(ctx, thread_id, clamp);
}
const Id max_thread_id{GetMaxThreadId(ctx, thread_id, clamp, segmentation_mask)};
const Id src_thread_id{ctx.OpISub(ctx.U32[1], thread_id, index)};
const Id in_range{ctx.OpSGreaterThanEqual(ctx.U1, src_thread_id, max_thread_id)};
SetInBoundsFlag(inst, in_range);
return SelectValue(ctx, in_range, value, src_thread_id);
@@ -177,14 +181,13 @@ Id EmitShuffleUp(EmitContext& ctx, IR::Inst* inst, Id value, Id index, Id clamp,
Id EmitShuffleDown(EmitContext& ctx, IR::Inst* inst, Id value, Id index, Id clamp,
Id segmentation_mask) {
const Id thread_id{EmitLaneId(ctx)};
const Id max_thread_id{GetMaxThreadId(ctx, thread_id, clamp, segmentation_mask)};
Id src_thread_id{ctx.OpIAdd(ctx.U32[1], thread_id, index)};
const Id in_range{ctx.OpSLessThanEqual(ctx.U1, src_thread_id, max_thread_id)};
const Id thread_id{GetThreadId(ctx)};
if (ctx.profile.warp_size_potentially_larger_than_guest) {
src_thread_id = AddPartitionBase(ctx, src_thread_id);
clamp = GetUpperClamp(ctx, thread_id, clamp);
}
const Id max_thread_id{GetMaxThreadId(ctx, thread_id, clamp, segmentation_mask)};
const Id src_thread_id{ctx.OpIAdd(ctx.U32[1], thread_id, index)};
const Id in_range{ctx.OpSLessThanEqual(ctx.U1, src_thread_id, max_thread_id)};
SetInBoundsFlag(inst, in_range);
return SelectValue(ctx, in_range, value, src_thread_id);
@@ -192,14 +195,13 @@ Id EmitShuffleDown(EmitContext& ctx, IR::Inst* inst, Id value, Id index, Id clam
Id EmitShuffleButterfly(EmitContext& ctx, IR::Inst* inst, Id value, Id index, Id clamp,
Id segmentation_mask) {
const Id thread_id{EmitLaneId(ctx)};
const Id max_thread_id{GetMaxThreadId(ctx, thread_id, clamp, segmentation_mask)};
Id src_thread_id{ctx.OpBitwiseXor(ctx.U32[1], thread_id, index)};
const Id in_range{ctx.OpSLessThanEqual(ctx.U1, src_thread_id, max_thread_id)};
const Id thread_id{GetThreadId(ctx)};
if (ctx.profile.warp_size_potentially_larger_than_guest) {
src_thread_id = AddPartitionBase(ctx, src_thread_id);
clamp = GetUpperClamp(ctx, thread_id, clamp);
}
const Id max_thread_id{GetMaxThreadId(ctx, thread_id, clamp, segmentation_mask)};
const Id src_thread_id{ctx.OpBitwiseXor(ctx.U32[1], thread_id, index)};
const Id in_range{ctx.OpSLessThanEqual(ctx.U1, src_thread_id, max_thread_id)};
SetInBoundsFlag(inst, in_range);
return SelectValue(ctx, in_range, value, src_thread_id);

View File

@@ -544,7 +544,7 @@ void EmitContext::DefineCommonTypes(const Info& info) {
U16 = Name(TypeInt(16, false), "u16");
S16 = Name(TypeInt(16, true), "s16");
}
if (info.uses_int64 && profile.support_int64) {
if (info.uses_int64) {
AddCapability(spv::Capability::Int64);
U64 = Name(TypeInt(64, false), "u64");
}
@@ -721,21 +721,9 @@ void EmitContext::DefineAttributeMemAccess(const Info& info) {
size_t label_index{0};
if (info.loads.AnyComponent(IR::Attribute::PositionX)) {
AddLabel(labels[label_index]);
const Id pointer{[&]() {
if (need_input_position_indirect) {
if (is_array)
return OpAccessChain(input_f32, input_position, vertex, u32_zero_value,
masked_index);
else
return OpAccessChain(input_f32, input_position, u32_zero_value,
masked_index);
} else {
if (is_array)
return OpAccessChain(input_f32, input_position, vertex, masked_index);
else
return OpAccessChain(input_f32, input_position, masked_index);
}
}()};
const Id pointer{is_array
? OpAccessChain(input_f32, input_position, vertex, masked_index)
: OpAccessChain(input_f32, input_position, masked_index)};
const Id result{OpLoad(F32[1], pointer)};
OpReturnValue(result);
++label_index;
@@ -1379,25 +1367,12 @@ void EmitContext::DefineInputs(const IR::Program& program) {
Decorate(layer, spv::Decoration::Flat);
}
if (loads.AnyComponent(IR::Attribute::PositionX)) {
const bool is_fragment{stage == Stage::Fragment};
if (!is_fragment && profile.has_broken_spirv_position_input) {
need_input_position_indirect = true;
const Id input_position_struct = TypeStruct(F32[4]);
input_position = DefineInput(*this, input_position_struct, true);
MemberDecorate(input_position_struct, 0, spv::Decoration::BuiltIn,
static_cast<unsigned>(spv::BuiltIn::Position));
Decorate(input_position_struct, spv::Decoration::Block);
} else {
const spv::BuiltIn built_in{is_fragment ? spv::BuiltIn::FragCoord
: spv::BuiltIn::Position};
input_position = DefineInput(*this, F32[4], true, built_in);
if (profile.support_geometry_shader_passthrough) {
if (info.passthrough.AnyComponent(IR::Attribute::PositionX)) {
Decorate(input_position, spv::Decoration::PassthroughNV);
}
const bool is_fragment{stage != Stage::Fragment};
const spv::BuiltIn built_in{is_fragment ? spv::BuiltIn::Position : spv::BuiltIn::FragCoord};
input_position = DefineInput(*this, F32[4], true, built_in);
if (profile.support_geometry_shader_passthrough) {
if (info.passthrough.AnyComponent(IR::Attribute::PositionX)) {
Decorate(input_position, spv::Decoration::PassthroughNV);
}
}
}

View File

@@ -280,7 +280,6 @@ public:
Id write_global_func_u32x2{};
Id write_global_func_u32x4{};
bool need_input_position_indirect{};
Id input_position{};
std::array<Id, 32> input_generics{};

View File

@@ -171,70 +171,6 @@ std::map<IR::Attribute, IR::Attribute> GenerateLegacyToGenericMappings(
}
return mapping;
}
void EmitGeometryPassthrough(IR::IREmitter& ir, const IR::Program& program,
const Shader::VaryingState& passthrough_mask,
bool passthrough_position,
std::optional<IR::Attribute> passthrough_layer_attr) {
for (u32 i = 0; i < program.output_vertices; i++) {
// Assign generics from input
for (u32 j = 0; j < 32; j++) {
if (!passthrough_mask.Generic(j)) {
continue;
}
const IR::Attribute attr = IR::Attribute::Generic0X + (j * 4);
ir.SetAttribute(attr + 0, ir.GetAttribute(attr + 0, ir.Imm32(i)), ir.Imm32(0));
ir.SetAttribute(attr + 1, ir.GetAttribute(attr + 1, ir.Imm32(i)), ir.Imm32(0));
ir.SetAttribute(attr + 2, ir.GetAttribute(attr + 2, ir.Imm32(i)), ir.Imm32(0));
ir.SetAttribute(attr + 3, ir.GetAttribute(attr + 3, ir.Imm32(i)), ir.Imm32(0));
}
if (passthrough_position) {
// Assign position from input
const IR::Attribute attr = IR::Attribute::PositionX;
ir.SetAttribute(attr + 0, ir.GetAttribute(attr + 0, ir.Imm32(i)), ir.Imm32(0));
ir.SetAttribute(attr + 1, ir.GetAttribute(attr + 1, ir.Imm32(i)), ir.Imm32(0));
ir.SetAttribute(attr + 2, ir.GetAttribute(attr + 2, ir.Imm32(i)), ir.Imm32(0));
ir.SetAttribute(attr + 3, ir.GetAttribute(attr + 3, ir.Imm32(i)), ir.Imm32(0));
}
if (passthrough_layer_attr) {
// Assign layer
ir.SetAttribute(IR::Attribute::Layer, ir.GetAttribute(*passthrough_layer_attr),
ir.Imm32(0));
}
// Emit vertex
ir.EmitVertex(ir.Imm32(0));
}
ir.EndPrimitive(ir.Imm32(0));
}
u32 GetOutputTopologyVertices(OutputTopology output_topology) {
switch (output_topology) {
case OutputTopology::PointList:
return 1;
case OutputTopology::LineStrip:
return 2;
default:
return 3;
}
}
void LowerGeometryPassthrough(const IR::Program& program, const HostTranslateInfo& host_info) {
for (IR::Block* const block : program.blocks) {
for (IR::Inst& inst : block->Instructions()) {
if (inst.GetOpcode() == IR::Opcode::Epilogue) {
IR::IREmitter ir{*block, IR::Block::InstructionList::s_iterator_to(inst)};
EmitGeometryPassthrough(
ir, program, program.info.passthrough,
program.info.passthrough.AnyComponent(IR::Attribute::PositionX), {});
}
}
}
}
} // Anonymous namespace
IR::Program TranslateProgram(ObjectPool<IR::Inst>& inst_pool, ObjectPool<IR::Block>& block_pool,
@@ -259,14 +195,9 @@ IR::Program TranslateProgram(ObjectPool<IR::Inst>& inst_pool, ObjectPool<IR::Blo
program.is_geometry_passthrough = sph.common0.geometry_passthrough != 0;
if (program.is_geometry_passthrough) {
const auto& mask{env.GpPassthroughMask()};
for (size_t i = 0; i < mask.size() * 32; ++i) {
for (size_t i = 0; i < program.info.passthrough.mask.size(); ++i) {
program.info.passthrough.mask[i] = ((mask[i / 32] >> (i % 32)) & 1) == 0;
}
if (!host_info.support_geometry_shader_passthrough) {
program.output_vertices = GetOutputTopologyVertices(program.output_topology);
LowerGeometryPassthrough(program, host_info);
}
}
break;
}
@@ -292,7 +223,7 @@ IR::Program TranslateProgram(ObjectPool<IR::Inst>& inst_pool, ObjectPool<IR::Blo
Optimization::PositionPass(env, program);
Optimization::GlobalMemoryToStorageBufferPass(program, host_info);
Optimization::GlobalMemoryToStorageBufferPass(program);
Optimization::TexturePass(env, program, host_info);
if (Settings::values.resolution_info.active) {
@@ -411,7 +342,17 @@ IR::Program GenerateGeometryPassthrough(ObjectPool<IR::Inst>& inst_pool,
IR::Program program;
program.stage = Stage::Geometry;
program.output_topology = output_topology;
program.output_vertices = GetOutputTopologyVertices(output_topology);
switch (output_topology) {
case OutputTopology::PointList:
program.output_vertices = 1;
break;
case OutputTopology::LineStrip:
program.output_vertices = 2;
break;
default:
program.output_vertices = 3;
break;
}
program.is_geometry_passthrough = false;
program.info.loads.mask = source_program.info.stores.mask;
@@ -425,8 +366,35 @@ IR::Program GenerateGeometryPassthrough(ObjectPool<IR::Inst>& inst_pool,
node.data.block = current_block;
IR::IREmitter ir{*current_block};
EmitGeometryPassthrough(ir, program, program.info.stores, true,
source_program.info.emulated_layer);
for (u32 i = 0; i < program.output_vertices; i++) {
// Assign generics from input
for (u32 j = 0; j < 32; j++) {
if (!program.info.stores.Generic(j)) {
continue;
}
const IR::Attribute attr = IR::Attribute::Generic0X + (j * 4);
ir.SetAttribute(attr + 0, ir.GetAttribute(attr + 0, ir.Imm32(i)), ir.Imm32(0));
ir.SetAttribute(attr + 1, ir.GetAttribute(attr + 1, ir.Imm32(i)), ir.Imm32(0));
ir.SetAttribute(attr + 2, ir.GetAttribute(attr + 2, ir.Imm32(i)), ir.Imm32(0));
ir.SetAttribute(attr + 3, ir.GetAttribute(attr + 3, ir.Imm32(i)), ir.Imm32(0));
}
// Assign position from input
const IR::Attribute attr = IR::Attribute::PositionX;
ir.SetAttribute(attr + 0, ir.GetAttribute(attr + 0, ir.Imm32(i)), ir.Imm32(0));
ir.SetAttribute(attr + 1, ir.GetAttribute(attr + 1, ir.Imm32(i)), ir.Imm32(0));
ir.SetAttribute(attr + 2, ir.GetAttribute(attr + 2, ir.Imm32(i)), ir.Imm32(0));
ir.SetAttribute(attr + 3, ir.GetAttribute(attr + 3, ir.Imm32(i)), ir.Imm32(0));
// Assign layer
ir.SetAttribute(IR::Attribute::Layer, ir.GetAttribute(source_program.info.emulated_layer),
ir.Imm32(0));
// Emit vertex
ir.EmitVertex(ir.Imm32(0));
}
ir.EndPrimitive(ir.Imm32(0));
IR::Block* return_block{block_pool.Create(inst_pool)};
IR::IREmitter{*return_block}.Epilogue();

View File

@@ -15,9 +15,6 @@ struct HostTranslateInfo {
bool needs_demote_reorder{}; ///< True when the device needs DemoteToHelperInvocation reordered
bool support_snorm_render_buffer{}; ///< True when the device supports SNORM render buffers
bool support_viewport_index_layer{}; ///< True when the device supports gl_Layer in VS
u32 min_ssbo_alignment{}; ///< Minimum alignment supported by the device for SSBOs
bool support_geometry_shader_passthrough{}; ///< True when the device supports geometry
///< passthrough shaders
};
} // namespace Shader

View File

@@ -11,7 +11,6 @@
#include "shader_recompiler/frontend/ir/breadth_first_search.h"
#include "shader_recompiler/frontend/ir/ir_emitter.h"
#include "shader_recompiler/frontend/ir/value.h"
#include "shader_recompiler/host_translate_info.h"
#include "shader_recompiler/ir_opt/passes.h"
namespace Shader::Optimization {
@@ -403,7 +402,7 @@ void CollectStorageBuffers(IR::Block& block, IR::Inst& inst, StorageInfo& info)
}
/// Returns the offset in indices (not bytes) for an equivalent storage instruction
IR::U32 StorageOffset(IR::Block& block, IR::Inst& inst, StorageBufferAddr buffer, u32 alignment) {
IR::U32 StorageOffset(IR::Block& block, IR::Inst& inst, StorageBufferAddr buffer) {
IR::IREmitter ir{block, IR::Block::InstructionList::s_iterator_to(inst)};
IR::U32 offset;
if (const std::optional<LowAddrInfo> low_addr{TrackLowAddress(&inst)}) {
@@ -416,10 +415,7 @@ IR::U32 StorageOffset(IR::Block& block, IR::Inst& inst, StorageBufferAddr buffer
}
// Subtract the least significant 32 bits from the guest offset. The result is the storage
// buffer offset in bytes.
IR::U32 low_cbuf{ir.GetCbuf(ir.Imm32(buffer.index), ir.Imm32(buffer.offset))};
// Align the offset base to match the host alignment requirements
low_cbuf = ir.BitwiseAnd(low_cbuf, ir.Imm32(~(alignment - 1U)));
const IR::U32 low_cbuf{ir.GetCbuf(ir.Imm32(buffer.index), ir.Imm32(buffer.offset))};
return ir.ISub(offset, low_cbuf);
}
@@ -514,7 +510,7 @@ void Replace(IR::Block& block, IR::Inst& inst, const IR::U32& storage_index,
}
} // Anonymous namespace
void GlobalMemoryToStorageBufferPass(IR::Program& program, const HostTranslateInfo& host_info) {
void GlobalMemoryToStorageBufferPass(IR::Program& program) {
StorageInfo info;
for (IR::Block* const block : program.post_order_blocks) {
for (IR::Inst& inst : block->Instructions()) {
@@ -538,8 +534,7 @@ void GlobalMemoryToStorageBufferPass(IR::Program& program, const HostTranslateIn
const IR::U32 index{IR::Value{static_cast<u32>(info.set.index_of(it))}};
IR::Block* const block{storage_inst.block};
IR::Inst* const inst{storage_inst.inst};
const IR::U32 offset{
StorageOffset(*block, *inst, storage_buffer, host_info.min_ssbo_alignment)};
const IR::U32 offset{StorageOffset(*block, *inst, storage_buffer)};
Replace(*block, *inst, index, offset);
}
}

View File

@@ -15,7 +15,7 @@ namespace Shader::Optimization {
void CollectShaderInfoPass(Environment& env, IR::Program& program);
void ConstantPropagationPass(Environment& env, IR::Program& program);
void DeadCodeEliminationPass(IR::Program& program);
void GlobalMemoryToStorageBufferPass(IR::Program& program, const HostTranslateInfo& host_info);
void GlobalMemoryToStorageBufferPass(IR::Program& program);
void IdentityRemovalPass(IR::Program& program);
void LowerFp16ToFp32(IR::Program& program);
void LowerInt64ToInt32(IR::Program& program);

View File

@@ -55,8 +55,6 @@ struct Profile {
/// OpFClamp is broken and OpFMax + OpFMin should be used instead
bool has_broken_spirv_clamp{};
/// The Position builtin needs to be wrapped in a struct when used as an input
bool has_broken_spirv_position_input{};
/// Offset image operands with an unsigned type do not work
bool has_broken_unsigned_image_offsets{};
/// Signed instructions with unsigned data types are misinterpreted

View File

@@ -65,8 +65,6 @@ enum class Interpolation {
struct ConstantBufferDescriptor {
u32 index;
u32 count;
auto operator<=>(const ConstantBufferDescriptor&) const = default;
};
struct StorageBufferDescriptor {
@@ -74,8 +72,6 @@ struct StorageBufferDescriptor {
u32 cbuf_offset;
u32 count;
bool is_written;
auto operator<=>(const StorageBufferDescriptor&) const = default;
};
struct TextureBufferDescriptor {
@@ -88,8 +84,6 @@ struct TextureBufferDescriptor {
u32 secondary_shift_left;
u32 count;
u32 size_shift;
auto operator<=>(const TextureBufferDescriptor&) const = default;
};
using TextureBufferDescriptors = boost::container::small_vector<TextureBufferDescriptor, 6>;
@@ -101,8 +95,6 @@ struct ImageBufferDescriptor {
u32 cbuf_offset;
u32 count;
u32 size_shift;
auto operator<=>(const ImageBufferDescriptor&) const = default;
};
using ImageBufferDescriptors = boost::container::small_vector<ImageBufferDescriptor, 2>;
@@ -118,8 +110,6 @@ struct TextureDescriptor {
u32 secondary_shift_left;
u32 count;
u32 size_shift;
auto operator<=>(const TextureDescriptor&) const = default;
};
using TextureDescriptors = boost::container::small_vector<TextureDescriptor, 12>;
@@ -132,8 +122,6 @@ struct ImageDescriptor {
u32 cbuf_offset;
u32 count;
u32 size_shift;
auto operator<=>(const ImageDescriptor&) const = default;
};
using ImageDescriptors = boost::container::small_vector<ImageDescriptor, 4>;

View File

@@ -191,8 +191,6 @@ add_library(video_core STATIC
renderer_vulkan/vk_texture_cache.cpp
renderer_vulkan/vk_texture_cache.h
renderer_vulkan/vk_texture_cache_base.cpp
renderer_vulkan/vk_turbo_mode.cpp
renderer_vulkan/vk_turbo_mode.h
renderer_vulkan/vk_update_descriptor.cpp
renderer_vulkan/vk_update_descriptor.h
shader_cache.cpp

View File

@@ -1938,21 +1938,14 @@ typename BufferCache<P>::Binding BufferCache<P>::StorageBufferBinding(GPUVAddr s
bool is_written) const {
const GPUVAddr gpu_addr = gpu_memory->Read<u64>(ssbo_addr);
const u32 size = gpu_memory->Read<u32>(ssbo_addr + 8);
const u32 alignment = runtime.GetStorageBufferAlignment();
const GPUVAddr aligned_gpu_addr = Common::AlignDown(gpu_addr, alignment);
const u32 aligned_size =
Common::AlignUp(static_cast<u32>(gpu_addr - aligned_gpu_addr) + size, alignment);
const std::optional<VAddr> cpu_addr = gpu_memory->GpuToCpuAddress(aligned_gpu_addr);
const std::optional<VAddr> cpu_addr = gpu_memory->GpuToCpuAddress(gpu_addr);
if (!cpu_addr || size == 0) {
return NULL_BINDING;
}
const VAddr cpu_end = Common::AlignUp(*cpu_addr + aligned_size, Core::Memory::YUZU_PAGESIZE);
const VAddr cpu_end = Common::AlignUp(*cpu_addr + size, Core::Memory::YUZU_PAGESIZE);
const Binding binding{
.cpu_addr = *cpu_addr,
.size = is_written ? aligned_size : static_cast<u32>(cpu_end - *cpu_addr),
.size = is_written ? size : static_cast<u32>(cpu_end - *cpu_addr),
.buffer_id = BufferId{},
};
return binding;

View File

@@ -47,7 +47,6 @@ set(SHADER_FILES
vulkan_present_scaleforce_fp16.frag
vulkan_present_scaleforce_fp32.frag
vulkan_quad_indexed.comp
vulkan_turbo_mode.comp
vulkan_uint8.comp
)

View File

@@ -1,29 +0,0 @@
// SPDX-FileCopyrightText: Copyright 2022 yuzu Emulator Project
// SPDX-License-Identifier: GPL-2.0-or-later
#version 460 core
layout (local_size_x = 16, local_size_y = 8, local_size_z = 1) in;
layout (binding = 0) buffer ThreadData {
uint data[];
};
uint xorshift32(uint x) {
x ^= x << 13;
x ^= x >> 17;
x ^= x << 5;
return x;
}
uint getGlobalIndex() {
return gl_GlobalInvocationID.x + gl_GlobalInvocationID.y * gl_WorkGroupSize.y * gl_NumWorkGroups.y;
}
void main() {
uint myIndex = xorshift32(getGlobalIndex());
uint otherIndex = xorshift32(myIndex);
uint otherValue = atomicAdd(data[otherIndex % data.length()], 0) + 1;
atomicAdd(data[myIndex % data.length()], otherValue);
}

View File

@@ -160,10 +160,6 @@ public:
return device.CanReportMemoryUsage();
}
u32 GetStorageBufferAlignment() const {
return static_cast<u32>(device.GetShaderStorageBufferAlignment());
}
private:
static constexpr std::array PABO_LUT{
GL_VERTEX_PROGRAM_PARAMETER_BUFFER_NV, GL_TESS_CONTROL_PROGRAM_PARAMETER_BUFFER_NV,

View File

@@ -139,7 +139,6 @@ void RasterizerOpenGL::LoadDiskResources(u64 title_id, std::stop_token stop_load
void RasterizerOpenGL::Clear(u32 layer_count) {
MICROPROFILE_SCOPE(OpenGL_Clears);
gpu_memory->FlushCaching();
const auto& regs = maxwell3d->regs;
bool use_color{};
bool use_depth{};
@@ -208,7 +207,6 @@ void RasterizerOpenGL::PrepareDraw(bool is_indexed, Func&& draw_func) {
MICROPROFILE_SCOPE(OpenGL_Drawing);
SCOPE_EXIT({ gpu.TickWork(); });
gpu_memory->FlushCaching();
query_cache.UpdateCounters();
GraphicsPipeline* const pipeline{shader_cache.CurrentGraphicsPipeline()};
@@ -321,7 +319,6 @@ void RasterizerOpenGL::DrawIndirect() {
}
void RasterizerOpenGL::DispatchCompute() {
gpu_memory->FlushCaching();
ComputePipeline* const pipeline{shader_cache.CurrentComputePipeline()};
if (!pipeline) {
return;
@@ -529,7 +526,6 @@ void RasterizerOpenGL::TickFrame() {
}
bool RasterizerOpenGL::AccelerateConditionalRendering() {
gpu_memory->FlushCaching();
if (Settings::IsGPULevelHigh()) {
// Reimplement Host conditional rendering.
return false;

View File

@@ -236,8 +236,6 @@ ShaderCache::ShaderCache(RasterizerOpenGL& rasterizer_, Core::Frontend::EmuWindo
.needs_demote_reorder = device.IsAmd(),
.support_snorm_render_buffer = false,
.support_viewport_index_layer = device.HasVertexViewportLayer(),
.min_ssbo_alignment = static_cast<u32>(device.GetShaderStorageBufferAlignment()),
.support_geometry_shader_passthrough = device.HasGeometryShaderPassthrough(),
} {
if (use_asynchronous_shaders) {
workers = CreateWorkers();

View File

@@ -442,13 +442,7 @@ void RendererOpenGL::DrawScreen(const Layout::FramebufferLayout& layout) {
glBindTextureUnit(0, screen_info.display_texture);
auto anti_aliasing = Settings::values.anti_aliasing.GetValue();
if (anti_aliasing > Settings::AntiAliasing::LastAA) {
LOG_ERROR(Render_OpenGL, "Invalid antialiasing option selected {}", anti_aliasing);
anti_aliasing = Settings::AntiAliasing::None;
Settings::values.anti_aliasing.SetValue(anti_aliasing);
}
const auto anti_aliasing = Settings::values.anti_aliasing.GetValue();
if (anti_aliasing != Settings::AntiAliasing::None) {
glEnablei(GL_SCISSOR_TEST, 0);
auto viewport_width = screen_info.texture.width;

View File

@@ -60,12 +60,23 @@ std::string GetDriverVersion(const Device& device) {
return GetReadableVersion(version);
}
std::string BuildCommaSeparatedExtensions(
const std::set<std::string, std::less<>>& available_extensions) {
return fmt::format("{}", fmt::join(available_extensions, ","));
}
std::string BuildCommaSeparatedExtensions(std::vector<std::string> available_extensions) {
std::sort(std::begin(available_extensions), std::end(available_extensions));
} // Anonymous namespace
static constexpr std::size_t AverageExtensionSize = 64;
std::string separated_extensions;
separated_extensions.reserve(available_extensions.size() * AverageExtensionSize);
const auto end = std::end(available_extensions);
for (auto extension = std::begin(available_extensions); extension != end; ++extension) {
if (const bool is_last = extension + 1 == end; is_last) {
separated_extensions += *extension;
} else {
separated_extensions += fmt::format("{},", *extension);
}
}
return separated_extensions;
}
Device CreateDevice(const vk::Instance& instance, const vk::InstanceDispatch& dld,
VkSurfaceKHR surface) {
@@ -78,6 +89,7 @@ Device CreateDevice(const vk::Instance& instance, const vk::InstanceDispatch& dl
const vk::PhysicalDevice physical_device(devices[device_index], dld);
return Device(*instance, physical_device, surface, dld);
}
} // Anonymous namespace
RendererVulkan::RendererVulkan(Core::TelemetrySession& telemetry_session_,
Core::Frontend::EmuWindow& emu_window,
@@ -97,9 +109,6 @@ RendererVulkan::RendererVulkan(Core::TelemetrySession& telemetry_session_,
screen_info),
rasterizer(render_window, gpu, cpu_memory, screen_info, device, memory_allocator,
state_tracker, scheduler) {
if (Settings::values.renderer_force_max_clock.GetValue() && device.ShouldBoostClocks()) {
turbo_mode.emplace(instance, dld);
}
Report();
} catch (const vk::Exception& exception) {
LOG_ERROR(Render_Vulkan, "Vulkan initialization failed with error: {}", exception.what());

View File

@@ -13,7 +13,6 @@
#include "video_core/renderer_vulkan/vk_scheduler.h"
#include "video_core/renderer_vulkan/vk_state_tracker.h"
#include "video_core/renderer_vulkan/vk_swapchain.h"
#include "video_core/renderer_vulkan/vk_turbo_mode.h"
#include "video_core/vulkan_common/vulkan_device.h"
#include "video_core/vulkan_common/vulkan_memory_allocator.h"
#include "video_core/vulkan_common/vulkan_wrapper.h"
@@ -32,9 +31,6 @@ class GPU;
namespace Vulkan {
Device CreateDevice(const vk::Instance& instance, const vk::InstanceDispatch& dld,
VkSurfaceKHR surface);
class RendererVulkan final : public VideoCore::RendererBase {
public:
explicit RendererVulkan(Core::TelemetrySession& telemtry_session,
@@ -78,7 +74,6 @@ private:
Swapchain swapchain;
BlitScreen blit_screen;
RasterizerVulkan rasterizer;
std::optional<TurboMode> turbo_mode;
};
} // namespace Vulkan

View File

@@ -330,10 +330,6 @@ bool BufferCacheRuntime::CanReportMemoryUsage() const {
return device.CanReportMemoryUsage();
}
u32 BufferCacheRuntime::GetStorageBufferAlignment() const {
return static_cast<u32>(device.GetStorageBufferAlignment());
}
void BufferCacheRuntime::Finish() {
scheduler.Finish();
}

View File

@@ -73,8 +73,6 @@ public:
bool CanReportMemoryUsage() const;
u32 GetStorageBufferAlignment() const;
[[nodiscard]] StagingBufferRef UploadStagingBuffer(size_t size);
[[nodiscard]] StagingBufferRef DownloadStagingBuffer(size_t size);

View File

@@ -331,7 +331,6 @@ PipelineCache::PipelineCache(RasterizerVulkan& rasterizer_, const Device& device
.need_declared_frag_colors = false,
.has_broken_spirv_clamp = driver_id == VK_DRIVER_ID_INTEL_PROPRIETARY_WINDOWS,
.has_broken_spirv_position_input = driver_id == VK_DRIVER_ID_QUALCOMM_PROPRIETARY,
.has_broken_unsigned_image_offsets = false,
.has_broken_signed_operations = false,
.has_broken_fp16_float_controls = driver_id == VK_DRIVER_ID_NVIDIA_PROPRIETARY,
@@ -344,8 +343,6 @@ PipelineCache::PipelineCache(RasterizerVulkan& rasterizer_, const Device& device
driver_id == VK_DRIVER_ID_AMD_PROPRIETARY || driver_id == VK_DRIVER_ID_AMD_OPEN_SOURCE,
.support_snorm_render_buffer = true,
.support_viewport_index_layer = device.IsExtShaderViewportIndexLayerSupported(),
.min_ssbo_alignment = static_cast<u32>(device.GetStorageBufferAlignment()),
.support_geometry_shader_passthrough = device.IsNvGeometryShaderPassthroughSupported(),
};
if (device.GetMaxVertexInputAttributes() < Maxwell::NumVertexAttributes) {
@@ -793,8 +790,7 @@ vk::PipelineCache PipelineCache::LoadVulkanPipelineCache(const std::filesystem::
return create_pipeline_cache(0, nullptr);
}
static constexpr size_t header_size = magic_number.size() + sizeof(cache_version);
const size_t cache_size = static_cast<size_t>(end) - header_size;
const size_t cache_size = static_cast<size_t>(end) - magic_number.size();
std::vector<char> cache_data(cache_size);
file.read(cache_data.data(), cache_size);

View File

@@ -270,7 +270,6 @@ void RasterizerVulkan::Clear(u32 layer_count) {
MICROPROFILE_SCOPE(Vulkan_Clearing);
FlushWork();
gpu_memory->FlushCaching();
query_cache.UpdateCounters();
@@ -629,7 +628,6 @@ void RasterizerVulkan::TickFrame() {
}
bool RasterizerVulkan::AccelerateConditionalRendering() {
gpu_memory->FlushCaching();
if (Settings::IsGPULevelHigh()) {
// TODO(Blinkhawk): Reimplement Host conditional rendering.
return false;

View File

@@ -1,205 +0,0 @@
// SPDX-FileCopyrightText: Copyright 2022 yuzu Emulator Project
// SPDX-License-Identifier: GPL-2.0-or-later
#include "common/literals.h"
#include "video_core/host_shaders/vulkan_turbo_mode_comp_spv.h"
#include "video_core/renderer_vulkan/renderer_vulkan.h"
#include "video_core/renderer_vulkan/vk_shader_util.h"
#include "video_core/renderer_vulkan/vk_turbo_mode.h"
#include "video_core/vulkan_common/vulkan_device.h"
namespace Vulkan {
using namespace Common::Literals;
TurboMode::TurboMode(const vk::Instance& instance, const vk::InstanceDispatch& dld)
: m_device{CreateDevice(instance, dld, VK_NULL_HANDLE)}, m_allocator{m_device, false} {
m_thread = std::jthread([&](auto stop_token) { Run(stop_token); });
}
TurboMode::~TurboMode() = default;
void TurboMode::Run(std::stop_token stop_token) {
auto& dld = m_device.GetLogical();
// Allocate buffer. 2MiB should be sufficient.
auto buffer = dld.CreateBuffer(VkBufferCreateInfo{
.sType = VK_STRUCTURE_TYPE_BUFFER_CREATE_INFO,
.pNext = nullptr,
.flags = 0,
.size = 2_MiB,
.usage = VK_BUFFER_USAGE_STORAGE_BUFFER_BIT | VK_BUFFER_USAGE_TRANSFER_DST_BIT,
.sharingMode = VK_SHARING_MODE_EXCLUSIVE,
.queueFamilyIndexCount = 0,
.pQueueFamilyIndices = nullptr,
});
// Commit some device local memory for the buffer.
auto commit = m_allocator.Commit(buffer, MemoryUsage::DeviceLocal);
// Create the descriptor pool to contain our descriptor.
constexpr VkDescriptorPoolSize pool_size{
.type = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER,
.descriptorCount = 1,
};
auto descriptor_pool = dld.CreateDescriptorPool(VkDescriptorPoolCreateInfo{
.sType = VK_STRUCTURE_TYPE_DESCRIPTOR_POOL_CREATE_INFO,
.pNext = nullptr,
.flags = VK_DESCRIPTOR_POOL_CREATE_FREE_DESCRIPTOR_SET_BIT,
.maxSets = 1,
.poolSizeCount = 1,
.pPoolSizes = &pool_size,
});
// Create the descriptor set layout from the pool.
constexpr VkDescriptorSetLayoutBinding layout_binding{
.binding = 0,
.descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER,
.descriptorCount = 1,
.stageFlags = VK_SHADER_STAGE_COMPUTE_BIT,
.pImmutableSamplers = nullptr,
};
auto descriptor_set_layout = dld.CreateDescriptorSetLayout(VkDescriptorSetLayoutCreateInfo{
.sType = VK_STRUCTURE_TYPE_DESCRIPTOR_SET_LAYOUT_CREATE_INFO,
.pNext = nullptr,
.flags = 0,
.bindingCount = 1,
.pBindings = &layout_binding,
});
// Actually create the descriptor set.
auto descriptor_set = descriptor_pool.Allocate(VkDescriptorSetAllocateInfo{
.sType = VK_STRUCTURE_TYPE_DESCRIPTOR_SET_ALLOCATE_INFO,
.pNext = nullptr,
.descriptorPool = *descriptor_pool,
.descriptorSetCount = 1,
.pSetLayouts = descriptor_set_layout.address(),
});
// Create the shader.
auto shader = BuildShader(m_device, VULKAN_TURBO_MODE_COMP_SPV);
// Create the pipeline layout.
auto pipeline_layout = dld.CreatePipelineLayout(VkPipelineLayoutCreateInfo{
.sType = VK_STRUCTURE_TYPE_PIPELINE_LAYOUT_CREATE_INFO,
.pNext = nullptr,
.flags = 0,
.setLayoutCount = 1,
.pSetLayouts = descriptor_set_layout.address(),
.pushConstantRangeCount = 0,
.pPushConstantRanges = nullptr,
});
// Actually create the pipeline.
const VkPipelineShaderStageCreateInfo shader_stage{
.sType = VK_STRUCTURE_TYPE_PIPELINE_SHADER_STAGE_CREATE_INFO,
.pNext = nullptr,
.flags = 0,
.stage = VK_SHADER_STAGE_COMPUTE_BIT,
.module = *shader,
.pName = "main",
.pSpecializationInfo = nullptr,
};
auto pipeline = dld.CreateComputePipeline(VkComputePipelineCreateInfo{
.sType = VK_STRUCTURE_TYPE_COMPUTE_PIPELINE_CREATE_INFO,
.pNext = nullptr,
.flags = 0,
.stage = shader_stage,
.layout = *pipeline_layout,
.basePipelineHandle = VK_NULL_HANDLE,
.basePipelineIndex = 0,
});
// Create a fence to wait on.
auto fence = dld.CreateFence(VkFenceCreateInfo{
.sType = VK_STRUCTURE_TYPE_FENCE_CREATE_INFO,
.pNext = nullptr,
.flags = 0,
});
// Create a command pool to allocate a command buffer from.
auto command_pool = dld.CreateCommandPool(VkCommandPoolCreateInfo{
.sType = VK_STRUCTURE_TYPE_COMMAND_POOL_CREATE_INFO,
.pNext = nullptr,
.flags =
VK_COMMAND_POOL_CREATE_TRANSIENT_BIT | VK_COMMAND_POOL_CREATE_RESET_COMMAND_BUFFER_BIT,
.queueFamilyIndex = m_device.GetGraphicsFamily(),
});
// Create a single command buffer.
auto cmdbufs = command_pool.Allocate(1, VK_COMMAND_BUFFER_LEVEL_PRIMARY);
auto cmdbuf = vk::CommandBuffer{cmdbufs[0], m_device.GetDispatchLoader()};
while (!stop_token.stop_requested()) {
// Reset the fence.
fence.Reset();
// Update descriptor set.
const VkDescriptorBufferInfo buffer_info{
.buffer = *buffer,
.offset = 0,
.range = VK_WHOLE_SIZE,
};
const VkWriteDescriptorSet buffer_write{
.sType = VK_STRUCTURE_TYPE_WRITE_DESCRIPTOR_SET,
.pNext = nullptr,
.dstSet = descriptor_set[0],
.dstBinding = 0,
.dstArrayElement = 0,
.descriptorCount = 1,
.descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER,
.pImageInfo = nullptr,
.pBufferInfo = &buffer_info,
.pTexelBufferView = nullptr,
};
dld.UpdateDescriptorSets(std::array{buffer_write}, {});
// Set up the command buffer.
cmdbuf.Begin(VkCommandBufferBeginInfo{
.sType = VK_STRUCTURE_TYPE_COMMAND_BUFFER_BEGIN_INFO,
.pNext = nullptr,
.flags = VK_COMMAND_BUFFER_USAGE_ONE_TIME_SUBMIT_BIT,
.pInheritanceInfo = nullptr,
});
// Clear the buffer.
cmdbuf.FillBuffer(*buffer, 0, VK_WHOLE_SIZE, 0);
// Bind descriptor set.
cmdbuf.BindDescriptorSets(VK_PIPELINE_BIND_POINT_COMPUTE, *pipeline_layout, 0,
descriptor_set, {});
// Bind the pipeline.
cmdbuf.BindPipeline(VK_PIPELINE_BIND_POINT_COMPUTE, *pipeline);
// Dispatch.
cmdbuf.Dispatch(64, 64, 1);
// Finish.
cmdbuf.End();
const VkSubmitInfo submit_info{
.sType = VK_STRUCTURE_TYPE_SUBMIT_INFO,
.pNext = nullptr,
.waitSemaphoreCount = 0,
.pWaitSemaphores = nullptr,
.pWaitDstStageMask = nullptr,
.commandBufferCount = 1,
.pCommandBuffers = cmdbuf.address(),
.signalSemaphoreCount = 0,
.pSignalSemaphores = nullptr,
};
m_device.GetGraphicsQueue().Submit(std::array{submit_info}, *fence);
// Wait for completion.
fence.Wait();
}
}
} // namespace Vulkan

View File

@@ -1,26 +0,0 @@
// SPDX-FileCopyrightText: Copyright 2022 yuzu Emulator Project
// SPDX-License-Identifier: GPL-2.0-or-later
#pragma once
#include "common/polyfill_thread.h"
#include "video_core/vulkan_common/vulkan_device.h"
#include "video_core/vulkan_common/vulkan_memory_allocator.h"
#include "video_core/vulkan_common/vulkan_wrapper.h"
namespace Vulkan {
class TurboMode {
public:
explicit TurboMode(const vk::Instance& instance, const vk::InstanceDispatch& dld);
~TurboMode();
private:
void Run(std::stop_token stop_token);
Device m_device;
MemoryAllocator m_allocator;
std::jthread m_thread;
};
} // namespace Vulkan

File diff suppressed because it is too large Load Diff

View File

@@ -3,7 +3,6 @@
#pragma once
#include <set>
#include <span>
#include <string>
#include <unordered_map>
@@ -12,155 +11,6 @@
#include "common/common_types.h"
#include "video_core/vulkan_common/vulkan_wrapper.h"
// Define all features which may be used by the implementation here.
// Vulkan version in the macro describes the minimum version required for feature availability.
// If the Vulkan version is lower than the required version, the named extension is required.
#define FOR_EACH_VK_FEATURE_1_1(FEATURE) \
FEATURE(EXT, SubgroupSizeControl, SUBGROUP_SIZE_CONTROL, subgroup_size_control) \
FEATURE(KHR, 16BitStorage, 16BIT_STORAGE, bit16_storage) \
FEATURE(KHR, ShaderAtomicInt64, SHADER_ATOMIC_INT64, shader_atomic_int64) \
FEATURE(KHR, ShaderDrawParameters, SHADER_DRAW_PARAMETERS, shader_draw_parameters) \
FEATURE(KHR, ShaderFloat16Int8, SHADER_FLOAT16_INT8, shader_float16_int8) \
FEATURE(KHR, UniformBufferStandardLayout, UNIFORM_BUFFER_STANDARD_LAYOUT, \
uniform_buffer_standard_layout) \
FEATURE(KHR, VariablePointer, VARIABLE_POINTERS, variable_pointer)
#define FOR_EACH_VK_FEATURE_1_2(FEATURE) \
FEATURE(EXT, HostQueryReset, HOST_QUERY_RESET, host_query_reset) \
FEATURE(KHR, 8BitStorage, 8BIT_STORAGE, bit8_storage) \
FEATURE(KHR, TimelineSemaphore, TIMELINE_SEMAPHORE, timeline_semaphore)
#define FOR_EACH_VK_FEATURE_1_3(FEATURE) \
FEATURE(EXT, ShaderDemoteToHelperInvocation, SHADER_DEMOTE_TO_HELPER_INVOCATION, \
shader_demote_to_helper_invocation)
// Define all features which may be used by the implementation and require an extension here.
#define FOR_EACH_VK_FEATURE_EXT(FEATURE) \
FEATURE(EXT, CustomBorderColor, CUSTOM_BORDER_COLOR, custom_border_color) \
FEATURE(EXT, DepthClipControl, DEPTH_CLIP_CONTROL, depth_clip_control) \
FEATURE(EXT, ExtendedDynamicState, EXTENDED_DYNAMIC_STATE, extended_dynamic_state) \
FEATURE(EXT, ExtendedDynamicState2, EXTENDED_DYNAMIC_STATE_2, extended_dynamic_state2) \
FEATURE(EXT, ExtendedDynamicState3, EXTENDED_DYNAMIC_STATE_3, extended_dynamic_state3) \
FEATURE(EXT, IndexTypeUint8, INDEX_TYPE_UINT8, index_type_uint8) \
FEATURE(EXT, LineRasterization, LINE_RASTERIZATION, line_rasterization) \
FEATURE(EXT, PrimitiveTopologyListRestart, PRIMITIVE_TOPOLOGY_LIST_RESTART, \
primitive_topology_list_restart) \
FEATURE(EXT, ProvokingVertex, PROVOKING_VERTEX, provoking_vertex) \
FEATURE(EXT, Robustness2, ROBUSTNESS_2, robustness2) \
FEATURE(EXT, TransformFeedback, TRANSFORM_FEEDBACK, transform_feedback) \
FEATURE(EXT, VertexInputDynamicState, VERTEX_INPUT_DYNAMIC_STATE, vertex_input_dynamic_state) \
FEATURE(KHR, PipelineExecutableProperties, PIPELINE_EXECUTABLE_PROPERTIES, \
pipeline_executable_properties) \
FEATURE(KHR, WorkgroupMemoryExplicitLayout, WORKGROUP_MEMORY_EXPLICIT_LAYOUT, \
workgroup_memory_explicit_layout)
// Define miscellaneous extensions which may be used by the implementation here.
#define FOR_EACH_VK_EXTENSION(EXTENSION) \
EXTENSION(EXT, CONSERVATIVE_RASTERIZATION, conservative_rasterization) \
EXTENSION(EXT, DEPTH_RANGE_UNRESTRICTED, depth_range_unrestricted) \
EXTENSION(EXT, MEMORY_BUDGET, memory_budget) \
EXTENSION(EXT, ROBUSTNESS_2, robustness_2) \
EXTENSION(EXT, SAMPLER_FILTER_MINMAX, sampler_filter_minmax) \
EXTENSION(EXT, SHADER_STENCIL_EXPORT, shader_stencil_export) \
EXTENSION(EXT, SHADER_VIEWPORT_INDEX_LAYER, shader_viewport_index_layer) \
EXTENSION(EXT, TOOLING_INFO, tooling_info) \
EXTENSION(EXT, VERTEX_ATTRIBUTE_DIVISOR, vertex_attribute_divisor) \
EXTENSION(KHR, DRIVER_PROPERTIES, driver_properties) \
EXTENSION(KHR, EXTERNAL_MEMORY_FD, external_memory_fd) \
EXTENSION(KHR, PUSH_DESCRIPTOR, push_descriptor) \
EXTENSION(KHR, SAMPLER_MIRROR_CLAMP_TO_EDGE, sampler_mirror_clamp_to_edge) \
EXTENSION(KHR, SHADER_FLOAT_CONTROLS, shader_float_controls) \
EXTENSION(KHR, SPIRV_1_4, spirv_1_4) \
EXTENSION(KHR, SWAPCHAIN, swapchain) \
EXTENSION(KHR, SWAPCHAIN_MUTABLE_FORMAT, swapchain_mutable_format) \
EXTENSION(NV, DEVICE_DIAGNOSTICS_CONFIG, device_diagnostics_config) \
EXTENSION(NV, GEOMETRY_SHADER_PASSTHROUGH, geometry_shader_passthrough) \
EXTENSION(NV, VIEWPORT_ARRAY2, viewport_array2) \
EXTENSION(NV, VIEWPORT_SWIZZLE, viewport_swizzle)
#define FOR_EACH_VK_EXTENSION_WIN32(EXTENSION) \
EXTENSION(KHR, EXTERNAL_MEMORY_WIN32, external_memory_win32)
// Define extensions which must be supported.
#define FOR_EACH_VK_MANDATORY_EXTENSION(EXTENSION_NAME) \
EXTENSION_NAME(VK_EXT_ROBUSTNESS_2_EXTENSION_NAME) \
EXTENSION_NAME(VK_EXT_VERTEX_ATTRIBUTE_DIVISOR_EXTENSION_NAME) \
EXTENSION_NAME(VK_KHR_DRIVER_PROPERTIES_EXTENSION_NAME) \
EXTENSION_NAME(VK_KHR_SAMPLER_MIRROR_CLAMP_TO_EDGE_EXTENSION_NAME) \
EXTENSION_NAME(VK_KHR_SHADER_FLOAT_CONTROLS_EXTENSION_NAME)
#define FOR_EACH_VK_MANDATORY_EXTENSION_GENERIC(EXTENSION_NAME) \
EXTENSION_NAME(VK_KHR_EXTERNAL_MEMORY_FD_EXTENSION_NAME)
#define FOR_EACH_VK_MANDATORY_EXTENSION_WIN32(EXTENSION_NAME) \
EXTENSION_NAME(VK_KHR_EXTERNAL_MEMORY_WIN32_EXTENSION_NAME)
// Define extensions where the absence of the extension may result in a degraded experience.
#define FOR_EACH_VK_RECOMMENDED_EXTENSION(EXTENSION_NAME) \
EXTENSION_NAME(VK_EXT_CONSERVATIVE_RASTERIZATION_EXTENSION_NAME) \
EXTENSION_NAME(VK_EXT_DEPTH_RANGE_UNRESTRICTED_EXTENSION_NAME) \
EXTENSION_NAME(VK_EXT_EXTENDED_DYNAMIC_STATE_EXTENSION_NAME) \
EXTENSION_NAME(VK_EXT_EXTENDED_DYNAMIC_STATE_2_EXTENSION_NAME) \
EXTENSION_NAME(VK_EXT_EXTENDED_DYNAMIC_STATE_3_EXTENSION_NAME) \
EXTENSION_NAME(VK_EXT_LINE_RASTERIZATION_EXTENSION_NAME) \
EXTENSION_NAME(VK_EXT_VERTEX_INPUT_DYNAMIC_STATE_EXTENSION_NAME) \
EXTENSION_NAME(VK_NV_GEOMETRY_SHADER_PASSTHROUGH_EXTENSION_NAME) \
EXTENSION_NAME(VK_NV_VIEWPORT_ARRAY2_EXTENSION_NAME) \
EXTENSION_NAME(VK_NV_VIEWPORT_SWIZZLE_EXTENSION_NAME)
// Define features which must be supported.
#define FOR_EACH_VK_MANDATORY_FEATURE(FEATURE_NAME) \
FEATURE_NAME(bit16_storage, storageBuffer16BitAccess) \
FEATURE_NAME(bit16_storage, uniformAndStorageBuffer16BitAccess) \
FEATURE_NAME(bit8_storage, storageBuffer8BitAccess) \
FEATURE_NAME(bit8_storage, uniformAndStorageBuffer8BitAccess) \
FEATURE_NAME(features, depthBiasClamp) \
FEATURE_NAME(features, depthClamp) \
FEATURE_NAME(features, drawIndirectFirstInstance) \
FEATURE_NAME(features, dualSrcBlend) \
FEATURE_NAME(features, fillModeNonSolid) \
FEATURE_NAME(features, fragmentStoresAndAtomics) \
FEATURE_NAME(features, geometryShader) \
FEATURE_NAME(features, imageCubeArray) \
FEATURE_NAME(features, independentBlend) \
FEATURE_NAME(features, largePoints) \
FEATURE_NAME(features, logicOp) \
FEATURE_NAME(features, multiDrawIndirect) \
FEATURE_NAME(features, multiViewport) \
FEATURE_NAME(features, occlusionQueryPrecise) \
FEATURE_NAME(features, robustBufferAccess) \
FEATURE_NAME(features, samplerAnisotropy) \
FEATURE_NAME(features, sampleRateShading) \
FEATURE_NAME(features, shaderClipDistance) \
FEATURE_NAME(features, shaderCullDistance) \
FEATURE_NAME(features, shaderImageGatherExtended) \
FEATURE_NAME(features, shaderStorageImageWriteWithoutFormat) \
FEATURE_NAME(features, tessellationShader) \
FEATURE_NAME(features, vertexPipelineStoresAndAtomics) \
FEATURE_NAME(features, wideLines) \
FEATURE_NAME(host_query_reset, hostQueryReset) \
FEATURE_NAME(robustness2, nullDescriptor) \
FEATURE_NAME(robustness2, robustBufferAccess2) \
FEATURE_NAME(robustness2, robustImageAccess2) \
FEATURE_NAME(shader_demote_to_helper_invocation, shaderDemoteToHelperInvocation) \
FEATURE_NAME(shader_draw_parameters, shaderDrawParameters) \
FEATURE_NAME(timeline_semaphore, timelineSemaphore) \
FEATURE_NAME(variable_pointer, variablePointers) \
FEATURE_NAME(variable_pointer, variablePointersStorageBuffer)
// Define features where the absence of the feature may result in a degraded experience.
#define FOR_EACH_VK_RECOMMENDED_FEATURE(FEATURE_NAME) \
FEATURE_NAME(custom_border_color, customBorderColors) \
FEATURE_NAME(extended_dynamic_state, extendedDynamicState) \
FEATURE_NAME(index_type_uint8, indexTypeUint8) \
FEATURE_NAME(primitive_topology_list_restart, primitiveTopologyListRestart) \
FEATURE_NAME(provoking_vertex, provokingVertexLast) \
FEATURE_NAME(shader_float16_int8, shaderFloat16) \
FEATURE_NAME(shader_float16_int8, shaderInt8) \
FEATURE_NAME(transform_feedback, transformFeedback) \
FEATURE_NAME(uniform_buffer_standard_layout, uniformBufferStandardLayout) \
FEATURE_NAME(vertex_input_dynamic_state, vertexInputDynamicState)
namespace Vulkan {
class NsightAftermathTracker;
@@ -238,69 +88,67 @@ public:
/// Returns the current Vulkan API version provided in Vulkan-formatted version numbers.
u32 ApiVersion() const {
return properties.properties.apiVersion;
return properties.apiVersion;
}
/// Returns the current driver version provided in Vulkan-formatted version numbers.
u32 GetDriverVersion() const {
return properties.properties.driverVersion;
return properties.driverVersion;
}
/// Returns the device name.
std::string_view GetModelName() const {
return properties.properties.deviceName;
return properties.deviceName;
}
/// Returns the driver ID.
VkDriverIdKHR GetDriverID() const {
return properties.driver.driverID;
return driver_id;
}
bool ShouldBoostClocks() const;
/// Returns uniform buffer alignment requeriment.
VkDeviceSize GetUniformBufferAlignment() const {
return properties.properties.limits.minUniformBufferOffsetAlignment;
return properties.limits.minUniformBufferOffsetAlignment;
}
/// Returns storage alignment requeriment.
VkDeviceSize GetStorageBufferAlignment() const {
return properties.properties.limits.minStorageBufferOffsetAlignment;
return properties.limits.minStorageBufferOffsetAlignment;
}
/// Returns the maximum range for storage buffers.
VkDeviceSize GetMaxStorageBufferRange() const {
return properties.properties.limits.maxStorageBufferRange;
return properties.limits.maxStorageBufferRange;
}
/// Returns the maximum size for push constants.
VkDeviceSize GetMaxPushConstantsSize() const {
return properties.properties.limits.maxPushConstantsSize;
return properties.limits.maxPushConstantsSize;
}
/// Returns the maximum size for shared memory.
u32 GetMaxComputeSharedMemorySize() const {
return properties.properties.limits.maxComputeSharedMemorySize;
return properties.limits.maxComputeSharedMemorySize;
}
/// Returns float control properties of the device.
const VkPhysicalDeviceFloatControlsPropertiesKHR& FloatControlProperties() const {
return properties.float_controls;
return float_controls;
}
/// Returns true if ASTC is natively supported.
bool IsOptimalAstcSupported() const {
return features.features.textureCompressionASTC_LDR;
return is_optimal_astc_supported;
}
/// Returns true if the device supports float16 natively.
bool IsFloat16Supported() const {
return features.shader_float16_int8.shaderFloat16;
return is_float16_supported;
}
/// Returns true if the device supports int8 natively.
bool IsInt8Supported() const {
return features.shader_float16_int8.shaderInt8;
return is_int8_supported;
}
/// Returns true if the device warp size can potentially be bigger than guest's warp size.
@@ -310,32 +158,32 @@ public:
/// Returns true if the device can be forced to use the guest warp size.
bool IsGuestWarpSizeSupported(VkShaderStageFlagBits stage) const {
return properties.subgroup_size_control.requiredSubgroupSizeStages & stage;
return guest_warp_stages & stage;
}
/// Returns the maximum number of push descriptors.
u32 MaxPushDescriptors() const {
return properties.push_descriptor.maxPushDescriptors;
return max_push_descriptors;
}
/// Returns true if formatless image load is supported.
bool IsFormatlessImageLoadSupported() const {
return features.features.shaderStorageImageReadWithoutFormat;
return is_formatless_image_load_supported;
}
/// Returns true if shader int64 is supported.
bool IsShaderInt64Supported() const {
return features.features.shaderInt64;
return is_shader_int64_supported;
}
/// Returns true if shader int16 is supported.
bool IsShaderInt16Supported() const {
return features.features.shaderInt16;
return is_shader_int16_supported;
}
// Returns true if depth bounds is supported.
bool IsDepthBoundsSupported() const {
return features.features.depthBounds;
return is_depth_bounds_supported;
}
/// Returns true when blitting from and to depth stencil images is supported.
@@ -345,151 +193,151 @@ public:
/// Returns true if the device supports VK_NV_viewport_swizzle.
bool IsNvViewportSwizzleSupported() const {
return extensions.viewport_swizzle;
return nv_viewport_swizzle;
}
/// Returns true if the device supports VK_NV_viewport_array2.
bool IsNvViewportArray2Supported() const {
return extensions.viewport_array2;
return nv_viewport_array2;
}
/// Returns true if the device supports VK_NV_geometry_shader_passthrough.
bool IsNvGeometryShaderPassthroughSupported() const {
return extensions.geometry_shader_passthrough;
return nv_geometry_shader_passthrough;
}
/// Returns true if the device supports VK_KHR_uniform_buffer_standard_layout.
bool IsKhrUniformBufferStandardLayoutSupported() const {
return extensions.uniform_buffer_standard_layout;
return khr_uniform_buffer_standard_layout;
}
/// Returns true if the device supports VK_KHR_push_descriptor.
bool IsKhrPushDescriptorSupported() const {
return extensions.push_descriptor;
return khr_push_descriptor;
}
/// Returns true if VK_KHR_pipeline_executable_properties is enabled.
bool IsKhrPipelineExecutablePropertiesEnabled() const {
return extensions.pipeline_executable_properties;
return khr_pipeline_executable_properties;
}
/// Returns true if VK_KHR_swapchain_mutable_format is enabled.
bool IsKhrSwapchainMutableFormatEnabled() const {
return extensions.swapchain_mutable_format;
return khr_swapchain_mutable_format;
}
/// Returns true if the device supports VK_KHR_workgroup_memory_explicit_layout.
bool IsKhrWorkgroupMemoryExplicitLayoutSupported() const {
return extensions.workgroup_memory_explicit_layout;
return khr_workgroup_memory_explicit_layout;
}
/// Returns true if the device supports VK_EXT_primitive_topology_list_restart.
bool IsTopologyListPrimitiveRestartSupported() const {
return features.primitive_topology_list_restart.primitiveTopologyListRestart;
return is_topology_list_restart_supported;
}
/// Returns true if the device supports VK_EXT_primitive_topology_list_restart.
bool IsPatchListPrimitiveRestartSupported() const {
return features.primitive_topology_list_restart.primitiveTopologyPatchListRestart;
return is_patch_list_restart_supported;
}
/// Returns true if the device supports VK_EXT_index_type_uint8.
bool IsExtIndexTypeUint8Supported() const {
return extensions.index_type_uint8;
return ext_index_type_uint8;
}
/// Returns true if the device supports VK_EXT_sampler_filter_minmax.
bool IsExtSamplerFilterMinmaxSupported() const {
return extensions.sampler_filter_minmax;
return ext_sampler_filter_minmax;
}
/// Returns true if the device supports VK_EXT_depth_range_unrestricted.
bool IsExtDepthRangeUnrestrictedSupported() const {
return extensions.depth_range_unrestricted;
return ext_depth_range_unrestricted;
}
/// Returns true if the device supports VK_EXT_depth_clip_control.
bool IsExtDepthClipControlSupported() const {
return extensions.depth_clip_control;
return ext_depth_clip_control;
}
/// Returns true if the device supports VK_EXT_shader_viewport_index_layer.
bool IsExtShaderViewportIndexLayerSupported() const {
return extensions.shader_viewport_index_layer;
return ext_shader_viewport_index_layer;
}
/// Returns true if the device supports VK_EXT_subgroup_size_control.
bool IsExtSubgroupSizeControlSupported() const {
return extensions.subgroup_size_control;
return ext_subgroup_size_control;
}
/// Returns true if the device supports VK_EXT_transform_feedback.
bool IsExtTransformFeedbackSupported() const {
return extensions.transform_feedback;
return ext_transform_feedback;
}
/// Returns true if the device supports VK_EXT_custom_border_color.
bool IsExtCustomBorderColorSupported() const {
return extensions.custom_border_color;
return ext_custom_border_color;
}
/// Returns true if the device supports VK_EXT_extended_dynamic_state.
bool IsExtExtendedDynamicStateSupported() const {
return extensions.extended_dynamic_state;
return ext_extended_dynamic_state;
}
/// Returns true if the device supports VK_EXT_extended_dynamic_state2.
bool IsExtExtendedDynamicState2Supported() const {
return extensions.extended_dynamic_state2;
return ext_extended_dynamic_state_2;
}
bool IsExtExtendedDynamicState2ExtrasSupported() const {
return features.extended_dynamic_state2.extendedDynamicState2LogicOp;
return ext_extended_dynamic_state_2_extra;
}
/// Returns true if the device supports VK_EXT_extended_dynamic_state3.
bool IsExtExtendedDynamicState3Supported() const {
return extensions.extended_dynamic_state3;
return ext_extended_dynamic_state_3;
}
/// Returns true if the device supports VK_EXT_extended_dynamic_state3.
bool IsExtExtendedDynamicState3BlendingSupported() const {
return dynamic_state3_blending;
return ext_extended_dynamic_state_3_blend;
}
/// Returns true if the device supports VK_EXT_extended_dynamic_state3.
bool IsExtExtendedDynamicState3EnablesSupported() const {
return dynamic_state3_enables;
return ext_extended_dynamic_state_3_enables;
}
/// Returns true if the device supports VK_EXT_line_rasterization.
bool IsExtLineRasterizationSupported() const {
return extensions.line_rasterization;
return ext_line_rasterization;
}
/// Returns true if the device supports VK_EXT_vertex_input_dynamic_state.
bool IsExtVertexInputDynamicStateSupported() const {
return extensions.vertex_input_dynamic_state;
return ext_vertex_input_dynamic_state;
}
/// Returns true if the device supports VK_EXT_shader_stencil_export.
bool IsExtShaderStencilExportSupported() const {
return extensions.shader_stencil_export;
return ext_shader_stencil_export;
}
/// Returns true if the device supports VK_EXT_conservative_rasterization.
bool IsExtConservativeRasterizationSupported() const {
return extensions.conservative_rasterization;
return ext_conservative_rasterization;
}
/// Returns true if the device supports VK_EXT_provoking_vertex.
bool IsExtProvokingVertexSupported() const {
return extensions.provoking_vertex;
return ext_provoking_vertex;
}
/// Returns true if the device supports VK_KHR_shader_atomic_int64.
bool IsExtShaderAtomicInt64Supported() const {
return extensions.shader_atomic_int64;
return ext_shader_atomic_int64;
}
/// Returns the minimum supported version of SPIR-V.
@@ -497,7 +345,7 @@ public:
if (instance_version >= VK_API_VERSION_1_3) {
return 0x00010600U;
}
if (extensions.spirv_1_4) {
if (khr_spirv_1_4) {
return 0x00010400U;
}
return 0x00010000U;
@@ -515,11 +363,11 @@ public:
/// Returns the vendor name reported from Vulkan.
std::string_view GetVendorName() const {
return properties.driver.driverName;
return vendor_name;
}
/// Returns the list of available extensions.
const std::set<std::string, std::less<>>& GetAvailableExtensions() const {
const std::vector<std::string>& GetAvailableExtensions() const {
return supported_extensions;
}
@@ -528,7 +376,7 @@ public:
}
bool CanReportMemoryUsage() const {
return extensions.memory_budget;
return ext_memory_budget;
}
u64 GetDeviceMemoryUsage() const;
@@ -550,29 +398,36 @@ public:
}
bool HasNullDescriptor() const {
return features.robustness2.nullDescriptor;
return has_null_descriptor;
}
u32 GetMaxVertexInputAttributes() const {
return properties.properties.limits.maxVertexInputAttributes;
return max_vertex_input_attributes;
}
u32 GetMaxVertexInputBindings() const {
return properties.properties.limits.maxVertexInputBindings;
return max_vertex_input_bindings;
}
private:
/// Checks if the physical device is suitable and configures the object state
/// with all necessary info about its properties.
bool GetSuitability(bool requires_swapchain);
/// Checks if the physical device is suitable.
void CheckSuitability(bool requires_swapchain) const;
// Remove extensions which have incomplete feature support.
void RemoveUnsuitableExtensions();
void RemoveExtensionIfUnsuitable(bool is_suitable, const std::string& extension_name);
/// Loads extensions into a vector and stores available ones in this object.
std::vector<const char*> LoadExtensions(bool requires_surface);
/// Sets up queue families.
void SetupFamilies(VkSurfaceKHR surface);
/// Sets up device features.
void SetupFeatures();
/// Sets up device properties.
void SetupProperties();
/// Collects telemetry information from the device.
void CollectTelemetryParameters();
/// Collects information about attached tools.
void CollectToolingInfo();
@@ -583,93 +438,91 @@ private:
std::vector<VkDeviceQueueCreateInfo> GetDeviceQueueCreateInfos() const;
/// Returns true if ASTC textures are natively supported.
bool ComputeIsOptimalAstcSupported() const;
bool IsOptimalAstcSupported(const VkPhysicalDeviceFeatures& features) const;
/// Returns true if the device natively supports blitting depth stencil images.
bool TestDepthStencilBlits() const;
private:
VkInstance instance; ///< Vulkan instance.
vk::DeviceDispatch dld; ///< Device function pointers.
vk::PhysicalDevice physical; ///< Physical device.
vk::Device logical; ///< Logical device.
vk::Queue graphics_queue; ///< Main graphics queue.
vk::Queue present_queue; ///< Main present queue.
u32 instance_version{}; ///< Vulkan instance version.
u32 graphics_family{}; ///< Main graphics queue family index.
u32 present_family{}; ///< Main present queue family index.
struct Extensions {
#define EXTENSION(prefix, macro_name, var_name) bool var_name{};
#define FEATURE(prefix, struct_name, macro_name, var_name) bool var_name{};
FOR_EACH_VK_FEATURE_1_1(FEATURE);
FOR_EACH_VK_FEATURE_1_2(FEATURE);
FOR_EACH_VK_FEATURE_1_3(FEATURE);
FOR_EACH_VK_FEATURE_EXT(FEATURE);
FOR_EACH_VK_EXTENSION(EXTENSION);
FOR_EACH_VK_EXTENSION_WIN32(EXTENSION);
#undef EXTENSION
#undef FEATURE
};
struct Features {
#define FEATURE_CORE(prefix, struct_name, macro_name, var_name) \
VkPhysicalDevice##struct_name##Features var_name{};
#define FEATURE_EXT(prefix, struct_name, macro_name, var_name) \
VkPhysicalDevice##struct_name##Features##prefix var_name{};
FOR_EACH_VK_FEATURE_1_1(FEATURE_CORE);
FOR_EACH_VK_FEATURE_1_2(FEATURE_CORE);
FOR_EACH_VK_FEATURE_1_3(FEATURE_CORE);
FOR_EACH_VK_FEATURE_EXT(FEATURE_EXT);
#undef FEATURE_CORE
#undef FEATURE_EXT
VkPhysicalDeviceFeatures features{};
};
struct Properties {
VkPhysicalDeviceDriverProperties driver{};
VkPhysicalDeviceFloatControlsProperties float_controls{};
VkPhysicalDevicePushDescriptorPropertiesKHR push_descriptor{};
VkPhysicalDeviceSubgroupSizeControlProperties subgroup_size_control{};
VkPhysicalDeviceTransformFeedbackPropertiesEXT transform_feedback{};
VkPhysicalDeviceProperties properties{};
};
Extensions extensions{};
Features features{};
Properties properties{};
VkPhysicalDeviceFeatures2 features2{};
VkPhysicalDeviceProperties2 properties2{};
// Misc features
bool is_optimal_astc_supported{}; ///< Support for all guest ASTC formats.
bool is_blit_depth_stencil_supported{}; ///< Support for blitting from and to depth stencil.
bool is_warp_potentially_bigger{}; ///< Host warp size can be bigger than guest.
bool is_integrated{}; ///< Is GPU an iGPU.
bool is_virtual{}; ///< Is GPU a virtual GPU.
bool is_non_gpu{}; ///< Is SoftwareRasterizer, FPGA, non-GPU device.
bool has_broken_cube_compatibility{}; ///< Has broken cube compatiblity bit
bool has_renderdoc{}; ///< Has RenderDoc attached
bool has_nsight_graphics{}; ///< Has Nsight Graphics attached
bool supports_d24_depth{}; ///< Supports D24 depth buffers.
bool cant_blit_msaa{}; ///< Does not support MSAA<->MSAA blitting.
bool must_emulate_bgr565{}; ///< Emulates BGR565 by swizzling RGB565 format.
bool dynamic_state3_blending{}; ///< Has all blending features of dynamic_state3.
bool dynamic_state3_enables{}; ///< Has all enables features of dynamic_state3.
u64 device_access_memory{}; ///< Total size of device local memory in bytes.
u32 sets_per_pool{}; ///< Sets per Description Pool
VkInstance instance; ///< Vulkan instance.
vk::DeviceDispatch dld; ///< Device function pointers.
vk::PhysicalDevice physical; ///< Physical device.
VkPhysicalDeviceProperties properties; ///< Device properties.
VkPhysicalDeviceFloatControlsPropertiesKHR float_controls{}; ///< Float control properties.
vk::Device logical; ///< Logical device.
vk::Queue graphics_queue; ///< Main graphics queue.
vk::Queue present_queue; ///< Main present queue.
u32 instance_version{}; ///< Vulkan onstance version.
u32 graphics_family{}; ///< Main graphics queue family index.
u32 present_family{}; ///< Main present queue family index.
VkDriverIdKHR driver_id{}; ///< Driver ID.
VkShaderStageFlags guest_warp_stages{}; ///< Stages where the guest warp size can be forced.
u64 device_access_memory{}; ///< Total size of device local memory in bytes.
u32 max_push_descriptors{}; ///< Maximum number of push descriptors
u32 sets_per_pool{}; ///< Sets per Description Pool
bool is_optimal_astc_supported{}; ///< Support for native ASTC.
bool is_float16_supported{}; ///< Support for float16 arithmetic.
bool is_int8_supported{}; ///< Support for int8 arithmetic.
bool is_warp_potentially_bigger{}; ///< Host warp size can be bigger than guest.
bool is_formatless_image_load_supported{}; ///< Support for shader image read without format.
bool is_depth_bounds_supported{}; ///< Support for depth bounds.
bool is_shader_float64_supported{}; ///< Support for float64.
bool is_shader_int64_supported{}; ///< Support for int64.
bool is_shader_int16_supported{}; ///< Support for int16.
bool is_shader_storage_image_multisample{}; ///< Support for image operations on MSAA images.
bool is_blit_depth_stencil_supported{}; ///< Support for blitting from and to depth stencil.
bool is_topology_list_restart_supported{}; ///< Support for primitive restart with list
///< topologies.
bool is_patch_list_restart_supported{}; ///< Support for primitive restart with list patch.
bool is_integrated{}; ///< Is GPU an iGPU.
bool is_virtual{}; ///< Is GPU a virtual GPU.
bool is_non_gpu{}; ///< Is SoftwareRasterizer, FPGA, non-GPU device.
bool nv_viewport_swizzle{}; ///< Support for VK_NV_viewport_swizzle.
bool nv_viewport_array2{}; ///< Support for VK_NV_viewport_array2.
bool nv_geometry_shader_passthrough{}; ///< Support for VK_NV_geometry_shader_passthrough.
bool khr_draw_indirect_count{}; ///< Support for VK_KHR_draw_indirect_count.
bool khr_uniform_buffer_standard_layout{}; ///< Support for scalar uniform buffer layouts.
bool khr_spirv_1_4{}; ///< Support for VK_KHR_spirv_1_4.
bool khr_workgroup_memory_explicit_layout{}; ///< Support for explicit workgroup layouts.
bool khr_push_descriptor{}; ///< Support for VK_KHR_push_descritor.
bool khr_pipeline_executable_properties{}; ///< Support for executable properties.
bool khr_swapchain_mutable_format{}; ///< Support for VK_KHR_swapchain_mutable_format.
bool ext_index_type_uint8{}; ///< Support for VK_EXT_index_type_uint8.
bool ext_sampler_filter_minmax{}; ///< Support for VK_EXT_sampler_filter_minmax.
bool ext_depth_clip_control{}; ///< Support for VK_EXT_depth_clip_control
bool ext_depth_range_unrestricted{}; ///< Support for VK_EXT_depth_range_unrestricted.
bool ext_shader_viewport_index_layer{}; ///< Support for VK_EXT_shader_viewport_index_layer.
bool ext_tooling_info{}; ///< Support for VK_EXT_tooling_info.
bool ext_subgroup_size_control{}; ///< Support for VK_EXT_subgroup_size_control.
bool ext_transform_feedback{}; ///< Support for VK_EXT_transform_feedback.
bool ext_custom_border_color{}; ///< Support for VK_EXT_custom_border_color.
bool ext_extended_dynamic_state{}; ///< Support for VK_EXT_extended_dynamic_state.
bool ext_extended_dynamic_state_2{}; ///< Support for VK_EXT_extended_dynamic_state2.
bool ext_extended_dynamic_state_2_extra{}; ///< Support for VK_EXT_extended_dynamic_state2.
bool ext_extended_dynamic_state_3{}; ///< Support for VK_EXT_extended_dynamic_state3.
bool ext_extended_dynamic_state_3_blend{}; ///< Support for VK_EXT_extended_dynamic_state3.
bool ext_extended_dynamic_state_3_enables{}; ///< Support for VK_EXT_extended_dynamic_state3.
bool ext_line_rasterization{}; ///< Support for VK_EXT_line_rasterization.
bool ext_vertex_input_dynamic_state{}; ///< Support for VK_EXT_vertex_input_dynamic_state.
bool ext_shader_stencil_export{}; ///< Support for VK_EXT_shader_stencil_export.
bool ext_shader_atomic_int64{}; ///< Support for VK_KHR_shader_atomic_int64.
bool ext_conservative_rasterization{}; ///< Support for VK_EXT_conservative_rasterization.
bool ext_provoking_vertex{}; ///< Support for VK_EXT_provoking_vertex.
bool ext_memory_budget{}; ///< Support for VK_EXT_memory_budget.
bool nv_device_diagnostics_config{}; ///< Support for VK_NV_device_diagnostics_config.
bool has_broken_cube_compatibility{}; ///< Has broken cube compatiblity bit
bool has_renderdoc{}; ///< Has RenderDoc attached
bool has_nsight_graphics{}; ///< Has Nsight Graphics attached
bool supports_d24_depth{}; ///< Supports D24 depth buffers.
bool cant_blit_msaa{}; ///< Does not support MSAA<->MSAA blitting.
bool must_emulate_bgr565{}; ///< Emulates BGR565 by swizzling RGB565 format.
bool has_null_descriptor{}; ///< Has support for null descriptors.
u32 max_vertex_input_attributes{}; ///< Max vertex input attributes in pipeline
u32 max_vertex_input_bindings{}; ///< Max vertex input buffers in pipeline
// Telemetry parameters
std::set<std::string, std::less<>> supported_extensions; ///< Reported Vulkan extensions.
std::set<std::string, std::less<>> loaded_extensions; ///< Loaded Vulkan extensions.
std::vector<size_t> valid_heap_memory; ///< Heaps used.
std::string vendor_name; ///< Device's driver name.
std::vector<std::string> supported_extensions; ///< Reported Vulkan extensions.
std::vector<size_t> valid_heap_memory; ///< Heaps used.
/// Format properties dictionary.
std::unordered_map<VkFormat, VkFormatProperties> format_properties;

View File

@@ -690,7 +690,6 @@ void Config::ReadRendererValues() {
qt_config->beginGroup(QStringLiteral("Renderer"));
ReadGlobalSetting(Settings::values.renderer_backend);
ReadGlobalSetting(Settings::values.renderer_force_max_clock);
ReadGlobalSetting(Settings::values.vulkan_device);
ReadGlobalSetting(Settings::values.fullscreen_mode);
ReadGlobalSetting(Settings::values.aspect_ratio);
@@ -1307,9 +1306,6 @@ void Config::SaveRendererValues() {
static_cast<u32>(Settings::values.renderer_backend.GetValue(global)),
static_cast<u32>(Settings::values.renderer_backend.GetDefault()),
Settings::values.renderer_backend.UsingGlobal());
WriteSetting(QString::fromStdString(Settings::values.renderer_force_max_clock.GetLabel()),
static_cast<u32>(Settings::values.renderer_force_max_clock.GetValue(global)),
static_cast<u32>(Settings::values.renderer_force_max_clock.GetDefault()));
WriteGlobalSetting(Settings::values.vulkan_device);
WriteSetting(QString::fromStdString(Settings::values.fullscreen_mode.GetLabel()),
static_cast<u32>(Settings::values.fullscreen_mode.GetValue(global)),

View File

@@ -25,7 +25,6 @@ void ConfigureGraphicsAdvanced::SetConfiguration() {
ui->use_asynchronous_shaders->setEnabled(runtime_lock);
ui->anisotropic_filtering_combobox->setEnabled(runtime_lock);
ui->renderer_force_max_clock->setChecked(Settings::values.renderer_force_max_clock.GetValue());
ui->use_vsync->setChecked(Settings::values.use_vsync.GetValue());
ui->use_asynchronous_shaders->setChecked(Settings::values.use_asynchronous_shaders.GetValue());
ui->use_fast_gpu_time->setChecked(Settings::values.use_fast_gpu_time.GetValue());
@@ -40,8 +39,6 @@ void ConfigureGraphicsAdvanced::SetConfiguration() {
Settings::values.max_anisotropy.GetValue());
} else {
ConfigurationShared::SetPerGameSetting(ui->gpu_accuracy, &Settings::values.gpu_accuracy);
ConfigurationShared::SetPerGameSetting(ui->renderer_force_max_clock,
&Settings::values.renderer_force_max_clock);
ConfigurationShared::SetPerGameSetting(ui->anisotropic_filtering_combobox,
&Settings::values.max_anisotropy);
ConfigurationShared::SetHighlight(ui->label_gpu_accuracy,
@@ -53,9 +50,6 @@ void ConfigureGraphicsAdvanced::SetConfiguration() {
void ConfigureGraphicsAdvanced::ApplyConfiguration() {
ConfigurationShared::ApplyPerGameSetting(&Settings::values.gpu_accuracy, ui->gpu_accuracy);
ConfigurationShared::ApplyPerGameSetting(&Settings::values.renderer_force_max_clock,
ui->renderer_force_max_clock,
renderer_force_max_clock);
ConfigurationShared::ApplyPerGameSetting(&Settings::values.max_anisotropy,
ui->anisotropic_filtering_combobox);
ConfigurationShared::ApplyPerGameSetting(&Settings::values.use_vsync, ui->use_vsync, use_vsync);
@@ -87,8 +81,6 @@ void ConfigureGraphicsAdvanced::SetupPerGameUI() {
// Disable if not global (only happens during game)
if (Settings::IsConfiguringGlobal()) {
ui->gpu_accuracy->setEnabled(Settings::values.gpu_accuracy.UsingGlobal());
ui->renderer_force_max_clock->setEnabled(
Settings::values.renderer_force_max_clock.UsingGlobal());
ui->use_vsync->setEnabled(Settings::values.use_vsync.UsingGlobal());
ui->use_asynchronous_shaders->setEnabled(
Settings::values.use_asynchronous_shaders.UsingGlobal());
@@ -103,9 +95,6 @@ void ConfigureGraphicsAdvanced::SetupPerGameUI() {
return;
}
ConfigurationShared::SetColoredTristate(ui->renderer_force_max_clock,
Settings::values.renderer_force_max_clock,
renderer_force_max_clock);
ConfigurationShared::SetColoredTristate(ui->use_vsync, Settings::values.use_vsync, use_vsync);
ConfigurationShared::SetColoredTristate(ui->use_asynchronous_shaders,
Settings::values.use_asynchronous_shaders,

View File

@@ -36,7 +36,6 @@ private:
std::unique_ptr<Ui::ConfigureGraphicsAdvanced> ui;
ConfigurationShared::CheckState renderer_force_max_clock;
ConfigurationShared::CheckState use_vsync;
ConfigurationShared::CheckState use_asynchronous_shaders;
ConfigurationShared::CheckState use_fast_gpu_time;

View File

@@ -69,16 +69,6 @@
</layout>
</widget>
</item>
<item>
<widget class="QCheckBox" name="renderer_force_max_clock">
<property name="toolTip">
<string>Runs work in the background while waiting for graphics commands to keep the GPU from lowering its clock speed.</string>
</property>
<property name="text">
<string>Force maximum clocks (Vulkan only)</string>
</property>
</widget>
</item>
<item>
<widget class="QCheckBox" name="use_vsync">
<property name="toolTip">

View File

@@ -3726,15 +3726,36 @@ void GMainWindow::UpdateWindowTitle(std::string_view title_name, std::string_vie
}
}
std::string GMainWindow::CreateTASFramesString(
std::array<size_t, InputCommon::TasInput::PLAYER_NUMBER> frames) const {
std::string string = "";
size_t maxPlayerIndex = 0;
for (size_t i = 0; i < frames.size(); i++) {
if (frames[i] != 0) {
if (maxPlayerIndex != 0)
string += ", ";
while (maxPlayerIndex++ != i)
string += "0, ";
string += std::to_string(frames[i]);
}
}
return string;
}
QString GMainWindow::GetTasStateDescription() const {
auto [tas_status, current_tas_frame, total_tas_frames] = input_subsystem->GetTas()->GetStatus();
std::string tas_frames_string = CreateTASFramesString(total_tas_frames);
switch (tas_status) {
case InputCommon::TasInput::TasState::Running:
return tr("TAS state: Running %1/%2").arg(current_tas_frame).arg(total_tas_frames);
return tr("TAS state: Running %1/%2")
.arg(current_tas_frame)
.arg(QString::fromStdString(tas_frames_string));
case InputCommon::TasInput::TasState::Recording:
return tr("TAS state: Recording %1").arg(total_tas_frames);
return tr("TAS state: Recording %1").arg(total_tas_frames[0]);
case InputCommon::TasInput::TasState::Stopped:
return tr("TAS state: Idle %1/%2").arg(current_tas_frame).arg(total_tas_frames);
return tr("TAS state: Idle %1/%2")
.arg(current_tas_frame)
.arg(QString::fromStdString(tas_frames_string));
default:
return tr("TAS State: Invalid");
}

View File

@@ -12,6 +12,7 @@
#include "common/announce_multiplayer_room.h"
#include "common/common_types.h"
#include "input_common/drivers/tas_input.h"
#include "yuzu/compatibility_list.h"
#include "yuzu/hotkeys.h"
@@ -266,6 +267,9 @@ private:
void changeEvent(QEvent* event) override;
void closeEvent(QCloseEvent* event) override;
std::string CreateTASFramesString(
std::array<size_t, InputCommon::TasInput::PLAYER_NUMBER> frames) const;
#ifdef __unix__
void SetupSigInterrupts();
static void HandleSigInterrupt(int);

View File

@@ -296,7 +296,6 @@ void Config::ReadValues() {
// Renderer
ReadSetting("Renderer", Settings::values.renderer_backend);
ReadSetting("Renderer", Settings::values.renderer_force_max_clock);
ReadSetting("Renderer", Settings::values.renderer_debug);
ReadSetting("Renderer", Settings::values.renderer_shader_feedback);
ReadSetting("Renderer", Settings::values.enable_nsight_aftermath);