thek0tyara/llama-cpp-turboquant
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 4

Device Properties

Browse source
This commit is contained in:
alexander-g 2021-03-12 08:32:51 +01:00
parent 50366c6658
commit f52efcef0f
5 changed files with 198 additions and 107 deletions
Download patch file Download diff file Expand all files Collapse all files

21
python/src/main.cpp
View file

@ -217,7 +217,26 @@ PYBIND11_MODULE(kp, m) {
py::arg("spirv"),
py::arg("workgroup") = kp::Workgroup(),
py::arg("spec_consts") = kp::Constants(),
py::arg("push_consts") = kp::Constants());
py::arg("push_consts") = kp::Constants())
.def("get_device_properties", &kp::Manager::getDeviceProperties, "Return a struct containing information about the device");
py::class_<kp::DeviceProperties>(m, "DeviceProperties")
.def_readonly("device_name", &kp::DeviceProperties::deviceName)
.def_readonly("max_work_group_count", &kp::DeviceProperties::maxWorkGroupCount)
.def_readonly("max_work_group_invocations", &kp::DeviceProperties::maxWorkGroupInvocations)
.def_readonly("max_work_group_size", &kp::DeviceProperties::maxWorkGroupSize)
.def_readonly("timestamps_supported", &kp::DeviceProperties::timestampsSupported)
.def("__repr__", [](const kp::DeviceProperties &p) {
return "Device Name: " + p.deviceName + "\n"
+"Maximum Workgroup Count: " + std::to_string(p.maxWorkGroupCount[0]) + ", "
+ std::to_string(p.maxWorkGroupCount[1]) + ", "
+ std::to_string(p.maxWorkGroupCount[2]) + "\n"
+"Maximum Workgroup Invocations: " + std::to_string(p.maxWorkGroupInvocations) + "\n"
+"Maximum Workgroup Size: " + std::to_string(p.maxWorkGroupSize[0]) + ", "
+ std::to_string(p.maxWorkGroupSize[1]) + ", "
+ std::to_string(p.maxWorkGroupSize[2]) + "\n"
+"Timestamps Supported: " + (p.timestampsSupported? "True" : "False") + "\n";
});
#ifdef VERSION_INFO
m.attr("__version__") = VERSION_INFO;

240
single_include/kompute/Kompute.hpp
View file

@ -741,7 +741,6 @@ namespace kp {
class Shader
{
public:
// The default resource limit for the GLSL compiler, can be overwritten
// Has been adopted by:
// https://github.com/KhronosGroup/glslang/blob/master/StandAlone/ResourceLimits.cpp
@ -951,41 +950,33 @@ class Tensor
* @return Unsigned integer representing the total number of elements
*/
// TODO: move to cpp
uint32_t size() {
return this->mSize;
}
uint32_t size() { return this->mSize; }
// TODO: move to cpp
uint32_t dataTypeMemorySize() {
return this->mDataTypeMemorySize;
}
uint32_t dataTypeMemorySize() { return this->mDataTypeMemorySize; }
// TODO: move to cpp
uint32_t memorySize() {
return this->mSize * this->mDataTypeMemorySize;
}
uint32_t memorySize() { return this->mSize * this->mDataTypeMemorySize; }
/**
* Retrieve the underlying data type of the Tensor
*
* @return Data type of tensor of type kp::Tensor::TensorDataTypes
*/
TensorDataTypes dataType() {
return this->mDataType;
}
TensorDataTypes dataType() { return this->mDataType; }
void* rawData() {
return this->mRawData;
}
void* rawData() { return this->mRawData; }
// TODO: move to cpp
template <typename T>
T* data() {
template<typename T>
T* data()
{
return (T*)this->mRawData;
}
template <typename T>
std::vector<T> vector() {
template<typename T>
std::vector<T> vector()
{
return { (T*)this->mRawData, ((T*)this->mRawData) + this->size() };
}
@ -993,9 +984,9 @@ class Tensor
* Sets / resets the vector data of the tensor. This function does not
* perform any copies into GPU memory and is only performed on the host.
*/
void setRawData(const void* data)
void setRawData(const void* data)
{
// Copy data
// Copy data
memcpy(this->mRawData, data, this->memorySize());
}
@ -1008,7 +999,8 @@ class Tensor
void* mRawData;
private:
void mapRawData() {
void mapRawData()
{
KP_LOG_DEBUG("Kompute Tensor mapping data from host buffer");
@ -1026,14 +1018,17 @@ class Tensor
vk::DeviceSize bufferSize = this->memorySize();
// Given we request coherent host memory we don't need to invalidate / flush
// Given we request coherent host memory we don't need to invalidate /
// flush
this->mRawData = this->mDevice->mapMemory(
*hostVisibleMemory, 0, bufferSize, vk::MemoryMapFlags());
vk::MappedMemoryRange mappedMemoryRange(*hostVisibleMemory, 0, bufferSize);
vk::MappedMemoryRange mappedMemoryRange(
*hostVisibleMemory, 0, bufferSize);
}
void unmapRawData() {
void unmapRawData()
{
KP_LOG_DEBUG("Kompute Tensor mapping data from host buffer");
@ -1087,49 +1082,46 @@ class Tensor
vk::MemoryPropertyFlags getPrimaryMemoryPropertyFlags();
vk::BufferUsageFlags getStagingBufferUsageFlags();
vk::MemoryPropertyFlags getStagingMemoryPropertyFlags();
};
// TODO: Limit T to be only float, bool, double, etc
template <typename T>
class TensorT: public Tensor
template<typename T>
class TensorT : public Tensor
{
public:
TensorT(std::shared_ptr<vk::PhysicalDevice> physicalDevice,
std::shared_ptr<vk::Device> device,
const std::vector<T>& data,
const TensorTypes& tensorType = TensorTypes::eDevice)
: Tensor(physicalDevice,
device,
(void*)data.data(),
data.size(),
sizeof(T),
this->dataType(),
tensorType)
std::shared_ptr<vk::Device> device,
const std::vector<T>& data,
const TensorTypes& tensorType = TensorTypes::eDevice)
: Tensor(physicalDevice,
device,
(void*)data.data(),
data.size(),
sizeof(T),
this->dataType(),
tensorType)
{
KP_LOG_DEBUG("Kompute TensorT constructor with data size {}", data.size());
KP_LOG_DEBUG("Kompute TensorT constructor with data size {}",
data.size());
}
~TensorT() {
KP_LOG_DEBUG("Kompute TensorT destructor");
}
~TensorT() { KP_LOG_DEBUG("Kompute TensorT destructor"); }
T* data() {
return (T*)this->mRawData;
}
T* data() { return (T*)this->mRawData; }
std::vector<T> vector() {
std::vector<T> vector()
{
return { (T*)this->mRawData, ((T*)this->mRawData) + this->size() };
}
T& operator[](int index) {
return *(((T*)this->mRawData) + index);
}
T& operator[](int index) { return *(((T*)this->mRawData) + index); }
void setData(const std::vector<T>& data) {
void setData(const std::vector<T>& data)
{
KP_LOG_DEBUG("Kompute TensorT setting data with data size {}", data.size());
KP_LOG_DEBUG("Kompute TensorT setting data with data size {}",
data.size());
if (data.size() != this->mSize) {
throw std::runtime_error(
@ -1140,7 +1132,6 @@ class TensorT: public Tensor
}
TensorDataTypes dataType();
};
} // End namespace kp
@ -1159,15 +1150,17 @@ class Algorithm
* the underlying resources.
*
* @param device The Vulkan device to use for creating resources
* @param tensors (optional) The tensors to use to create the descriptor resources
* @param tensors (optional) The tensors to use to create the descriptor
* resources
* @param spirv (optional) The spirv code to use to create the algorithm
* @param workgroup (optional) The kp::Workgroup to use for the dispatch which defaults to
* kp::Workgroup(tensor[0].size(), 1, 1) if not set.
* @param specializationConstants (optional) The kp::Constants to use to initialize
* the specialization constants which cannot be changed once set.
* @param pushConstants (optional) The kp::Constants to use when initializing the
* pipeline, which set the size of the push constants - these can be modified but
* all new values must have the same vector size as this initial value.
* @param workgroup (optional) The kp::Workgroup to use for the dispatch
* which defaults to kp::Workgroup(tensor[0].size(), 1, 1) if not set.
* @param specializationConstants (optional) The kp::Constants to use to
* initialize the specialization constants which cannot be changed once set.
* @param pushConstants (optional) The kp::Constants to use when
* initializing the pipeline, which set the size of the push constants -
* these can be modified but all new values must have the same vector size
* as this initial value.
*/
Algorithm(std::shared_ptr<vk::Device> device,
const std::vector<std::shared_ptr<Tensor>>& tensors = {},
@ -1177,18 +1170,19 @@ class Algorithm
const Constants& pushConstants = {});
/**
* Rebuild function to reconstruct algorithm with configuration parameters to create
* the underlying resources.
* Rebuild function to reconstruct algorithm with configuration parameters
* to create the underlying resources.
*
* @param tensors The tensors to use to create the descriptor resources
* @param spirv The spirv code to use to create the algorithm
* @param workgroup (optional) The kp::Workgroup to use for the dispatch which defaults to
* kp::Workgroup(tensor[0].size(), 1, 1) if not set.
* @param specializationConstants (optional) The kp::Constants to use to initialize
* the specialization constants which cannot be changed once set.
* @param pushConstants (optional) The kp::Constants to use when initializing the
* pipeline, which set the size of the push constants - these can be modified but
* all new values must have the same vector size as this initial value.
* @param workgroup (optional) The kp::Workgroup to use for the dispatch
* which defaults to kp::Workgroup(tensor[0].size(), 1, 1) if not set.
* @param specializationConstants (optional) The kp::Constants to use to
* initialize the specialization constants which cannot be changed once set.
* @param pushConstants (optional) The kp::Constants to use when
* initializing the pipeline, which set the size of the push constants -
* these can be modified but all new values must have the same vector size
* as this initial value.
*/
void rebuild(const std::vector<std::shared_ptr<Tensor>>& tensors,
const std::vector<uint32_t>& spirv,
@ -1211,25 +1205,26 @@ class Algorithm
void recordDispatch(const vk::CommandBuffer& commandBuffer);
/**
* Records command that binds the "core" algorithm components which consist of
* binding the pipeline and binding the descriptorsets.
* Records command that binds the "core" algorithm components which consist
* of binding the pipeline and binding the descriptorsets.
*
* @param commandBuffer Command buffer to record the algorithm resources to
*/
void recordBindCore(const vk::CommandBuffer& commandBuffer);
/**
* Records command that binds the push constants to the command buffer provided
* - it is required that the pushConstants provided are of the same size as the
* ones provided during initialization.
* Records command that binds the push constants to the command buffer
* provided
* - it is required that the pushConstants provided are of the same size as
* the ones provided during initialization.
*
* @param commandBuffer Command buffer to record the algorithm resources to
*/
void recordBindPush(const vk::CommandBuffer& commandBuffer);
/**
* function that checks all the gpu resource components to verify if these have
* been created and returns true if all are valid.
* function that checks all the gpu resource components to verify if these
* have been created and returns true if all are valid.
*
* @returns returns true if the algorithm is currently initialized.
*/
@ -1238,26 +1233,28 @@ class Algorithm
/**
* Sets the work group to use in the recordDispatch
*
* @param workgroup The kp::Workgroup value to use to update the algorithm. It
* must have a value greater than 1 on the x value (index 1) otherwise it will
* be initialized on the size of the first tensor (ie. this->mTensor[0]->size())
* @param workgroup The kp::Workgroup value to use to update the algorithm.
* It must have a value greater than 1 on the x value (index 1) otherwise it
* will be initialized on the size of the first tensor (ie.
* this->mTensor[0]->size())
*/
void setWorkgroup(const Workgroup& workgroup, uint32_t minSize = 1);
/**
* Sets the push constants to the new value provided to use in the next bindPush()
* Sets the push constants to the new value provided to use in the next
* bindPush()
*
* @param The kp::Constant to use to set the push constants to use in the next
* bindPush(...) calls. The constants provided must be of the same size as the
* ones created during initialization.
* @param The kp::Constant to use to set the push constants to use in the
* next bindPush(...) calls. The constants provided must be of the same size
* as the ones created during initialization.
*/
void setPush(const Constants& pushConstants);
/**
* Gets the current workgroup from the algorithm.
*
* @param The kp::Constant to use to set the push constants to use in the next
* bindPush(...) calls. The constants provided must be of the same size as the
* ones created during initialization.
* @param The kp::Constant to use to set the push constants to use in the
* next bindPush(...) calls. The constants provided must be of the same size
* as the ones created during initialization.
*/
const Workgroup& getWorkgroup();
/**
@ -1690,8 +1687,8 @@ class Sequence : public std::enable_shared_from_this<Sequence>
* function also requires the Sequence to be recording, otherwise it will
* not be able to add the operation.
*
* @param op Object derived from kp::BaseOp that will be recoreded by the sequence
* which will be used when the operation is evaluated.
* @param op Object derived from kp::BaseOp that will be recoreded by the
* sequence which will be used when the operation is evaluated.
* @return shared_ptr<Sequence> of the Sequence class itself
*/
std::shared_ptr<Sequence> record(std::shared_ptr<OpBase> op);
@ -1709,7 +1706,8 @@ class Sequence : public std::enable_shared_from_this<Sequence>
*/
template<typename T, typename... TArgs>
std::shared_ptr<Sequence> record(
std::vector<std::shared_ptr<Tensor>> tensors, TArgs&&... params)
std::vector<std::shared_ptr<Tensor>> tensors,
TArgs&&... params)
{
std::shared_ptr<T> op{ new T(tensors, std::forward<TArgs>(params)...) };
return this->record(op);
@ -1744,8 +1742,9 @@ class Sequence : public std::enable_shared_from_this<Sequence>
std::shared_ptr<Sequence> eval();
/**
* Resets all the recorded and stored operations, records the operation
* provided and submits into the gpu as a submit job synchronously (with a barrier).
* Resets all the recorded and stored operations, records the operation
* provided and submits into the gpu as a submit job synchronously (with a
* barrier).
*
* @return shared_ptr<Sequence> of the Sequence class itself
*/
@ -1788,16 +1787,18 @@ class Sequence : public std::enable_shared_from_this<Sequence>
/**
* Eval Async sends all the recorded and stored operations in the vector of
* operations into the gpu as a submit job without a barrier. EvalAwait() must
* ALWAYS be called after to ensure the sequence is terminated correctly.
* operations into the gpu as a submit job without a barrier. EvalAwait()
* must ALWAYS be called after to ensure the sequence is terminated
* correctly.
*
* @return Boolean stating whether execution was successful.
*/
std::shared_ptr<Sequence> evalAsync();
/**
* Clears currnet operations to record provided one in the vector of
* operations into the gpu as a submit job without a barrier. EvalAwait() must
* ALWAYS be called after to ensure the sequence is terminated correctly.
* operations into the gpu as a submit job without a barrier. EvalAwait()
* must ALWAYS be called after to ensure the sequence is terminated
* correctly.
*
* @return Boolean stating whether execution was successful.
*/
@ -1891,9 +1892,9 @@ class Sequence : public std::enable_shared_from_this<Sequence>
bool isInit();
/**
* Clears command buffer and triggers re-record of all the current operations
* saved, which is useful if the underlying kp::Tensors or kp::Algorithms
* are modified and need to be re-recorded.
* Clears command buffer and triggers re-record of all the current
* operations saved, which is useful if the underlying kp::Tensors or
* kp::Algorithms are modified and need to be re-recorded.
*/
void rerecord();
@ -1948,6 +1949,21 @@ class Sequence : public std::enable_shared_from_this<Sequence>
namespace kp {
struct DeviceProperties
{
//Name of the device
const std::string deviceName;
//Maximum number of workgroups that can be dispatched per shader
const std::array<uint32_t,3> maxWorkGroupCount;
//Maximum number of shader invocations per local workgroup
//i.e. the product of maxWorkGroupSize must not exceed this value
const uint32_t maxWorkGroupInvocations;
//Maximum number of shader invocations per local workgroup
const std::array<uint32_t,3> maxWorkGroupSize;
//Whether timestamping is supported by this device or not
const bool timestampsSupported;
};
/**
Base orchestrator which creates and manages device and child components
*/
@ -1961,13 +1977,14 @@ class Manager
Manager();
/**
* Similar to base constructor but allows for further configuration to use when
* creating the Vulkan resources.
* Similar to base constructor but allows for further configuration to use
* when creating the Vulkan resources.
*
* @param physicalDeviceIndex The index of the physical device to use
* @param familyQueueIndices (Optional) List of queue indices to add for
* explicit allocation
* @param desiredExtensions The desired extensions to load from physicalDevice
* @param desiredExtensions The desired extensions to load from
* physicalDevice
*/
Manager(uint32_t physicalDeviceIndex,
const std::vector<uint32_t>& familyQueueIndices = {},
@ -2001,7 +2018,8 @@ class Manager
* If zero (default), disables latching of timestamps.
* @returns Shared pointer with initialised sequence
*/
std::shared_ptr<Sequence> sequence(uint32_t queueIndex = 0, uint32_t totalTimestamps = 0);
std::shared_ptr<Sequence> sequence(uint32_t queueIndex = 0,
uint32_t totalTimestamps = 0);
/**
* Create a managed tensor that will be destroyed by this manager
@ -2011,7 +2029,7 @@ class Manager
* @param tensorType The type of tensor to initialize
* @returns Shared pointer with initialised tensor
*/
template <typename T>
template<typename T>
std::shared_ptr<TensorT<T>> tensorT(
const std::vector<T>& data,
Tensor::TensorTypes tensorType = Tensor::TensorTypes::eDevice)
@ -2042,8 +2060,13 @@ class Manager
const Tensor::TensorDataTypes& dataType,
Tensor::TensorTypes tensorType = Tensor::TensorTypes::eDevice)
{
std::shared_ptr<Tensor> tensor{ new kp::Tensor(
this->mPhysicalDevice, this->mDevice, data, elementTotalCount, elementMemorySize, dataType, tensorType) };
std::shared_ptr<Tensor> tensor{ new kp::Tensor(this->mPhysicalDevice,
this->mDevice,
data,
elementTotalCount,
elementMemorySize,
dataType,
tensorType) };
if (this->mManageResources) {
this->mManagedTensors.push_back(tensor);
@ -2083,6 +2106,11 @@ class Manager
**/
void clear();
/**
* Return a struct containing information about the device.
**/
DeviceProperties getDeviceProperties() const;
private:
// -------------- OPTIONALLY OWNED RESOURCES
std::shared_ptr<vk::Instance> mInstance = nullptr;

13
src/Manager.cpp
View file

@ -447,4 +447,17 @@ Manager::sequence(uint32_t queueIndex, uint32_t totalTimestamps)
return sq;
}
DeviceProperties Manager::getDeviceProperties() const
{
const vk::PhysicalDeviceProperties properties = this->mPhysicalDevice->getProperties();
const DeviceProperties output{
std::string(properties.deviceName.data()),
properties.limits.maxComputeWorkGroupCount,
properties.limits.maxComputeWorkGroupInvocations,
properties.limits.maxComputeWorkGroupSize,
properties.limits.timestampComputeAndGraphics,
};
return output;
}
}

23
src/include/kompute/Manager.hpp
View file

@ -11,6 +11,23 @@
namespace kp {
struct DeviceProperties
{
//Name of the device
const std::string deviceName;
//Maximum number of workgroups that can be dispatched per shader
const std::array<uint32_t,3> maxWorkGroupCount;
//Maximum number of shader invocations per local workgroup
//i.e. the product of maxWorkGroupSize must not exceed this value
const uint32_t maxWorkGroupInvocations;
//Maximum number of shader invocations per local workgroup
const std::array<uint32_t,3> maxWorkGroupSize;
//Whether timestamping is supported by this device or not
const bool timestampsSupported;
};
/**
Base orchestrator which creates and manages device and child components
*/
@ -153,6 +170,11 @@ class Manager
**/
void clear();
/**
* Return a struct containing information about the device.
**/
DeviceProperties getDeviceProperties() const;
private:
// -------------- OPTIONALLY OWNED RESOURCES
std::shared_ptr<vk::Instance> mInstance = nullptr;
@ -185,4 +207,5 @@ class Manager
const std::vector<std::string>& desiredExtensions = {});
};
} // End namespace kp

8
test/TestManager.cpp
View file

@ -62,3 +62,11 @@ TEST(TestManager, TestMultipleSequences)
EXPECT_EQ(tensorOutput->vector(), std::vector<float>({ 0, 4, 12 }));
}
TEST(TestManager, TestDeviceProperties)
{
kp::Manager mgr;
const auto properties = mgr.getDeviceProperties();
EXPECT_GT(properties.deviceName.size(), 0);
}