MNN/express/Utils.hpp

95 lines
2.7 KiB
C++
Raw Normal View History

//
// Utils.hpp
// MNN
//
// Created by MNN on 2019/07/26.
// Copyright © 2018, Alibaba Group Holding Limited
//
#ifndef Utils_hpp
#define Utils_hpp
2019-12-27 22:16:57 +08:00
#include <MNN/expr/Expr.hpp>
#include "Type_generated.h"
#include "MNN_generated.h"
#include <MNN/expr/Executor.hpp>
#include "core/AutoStorage.h"
namespace MNN {
2022-12-30 15:18:58 +08:00
class Session;
namespace Express {
struct Expr::Inside {
2020-11-05 16:41:56 +08:00
Inside(int outputSize);
2021-04-08 15:34:23 +08:00
Inside(Tensor* tensor, bool own = false);
2020-11-05 16:41:56 +08:00
~ Inside();
std::vector<Variable::Info> mOutputInfos;
2020-11-05 16:41:56 +08:00
std::vector<Tensor*> mOutputTensors;
Executor::Requirement mReq;
std::shared_ptr<Executor::ComputeCache> mCache;
int mCacheOffset = 0;
bool mInfoDirty = true;
bool mContentDirty = true;
2021-01-06 16:29:37 +08:00
bool mOwnTensor = true;
Tensor* mHostTensor = nullptr;
std::shared_ptr<Backend> mHoldBackend;
};
struct Executor::DebugTools {
TensorCallBackWithInfo before = nullptr;
TensorCallBackWithInfo after = nullptr;
2022-07-22 09:59:30 +08:00
mutable float flops = 0.0f;
};
2022-12-30 15:18:58 +08:00
struct Executor::SubGraph {
std::unique_ptr<MNN::SubGraphProtoT> info;
std::vector<std::string> depends;
};
class Executor::ComputeCache {
public:
void setShapeDirty();
void setContentDirty();
void* mapOutput(int offset, Tensor* dest);
~ ComputeCache();
ComputeCache() {
// Do nothing
}
ErrorCode compute();
ErrorCode resize();
ErrorCode resizeImpl();
Session* getSession() {
return mSession.get();
}
friend class Executor;
private:
std::set<std::shared_ptr<Expr::Inside>> mInputInside;
std::set<std::shared_ptr<ComputeCache>> mInputs;
std::shared_ptr<Session> mSession;
bool mContentDirty = true;
bool mShapeDirty = true;
std::vector<std::shared_ptr<BufferStorage>> mCacheBuffers;
#ifdef MNN_EXPRESS_MEMLEAK_DEBUG
static int gInstanceCount;
#endif
};
class Utils {
public:
static void copyInfoToTensor(Tensor* dest, const Variable::Info* source);
static void copyTensorToInfo(Variable::Info* dest, const Tensor* source);
static DataType convertDataType(halide_type_t type);
- build: - unify schema building in core and converter; - add more build script for android; - add linux build script for python; - ops impl: - add floor mod support in binary; - use eltwise impl in add/max/sub/mul binary for optimization; - remove fake double support in cast; - fix 5d support for concat; - add adjX and adjY support for batch matmul; - optimize conv2d back prop filter; - add pad mode support for conv3d; - fix bug in conv2d & conv depthwise with very small feature map; - optimize binary without broacast; - add data types support for gather; - add gather ND support; - use uint8 data type in gather v2; - add transpose support for matmul; - add matrix band part; - add dim != 4 support for padding, reshape & tensor convert; - add pad type support for pool3d; - make ops based on TensorFlow Lite quantization optional; - add all & any support for reduction; - use type in parameter as output type in reduction; - add int support for unary; - add variable weight support for conv2d; - fix conv2d depthwise weights initialization; - fix type support for transpose; - fix grad outputs count for reduce grad and reshape grad; - fix priorbox & detection output; - fix metal softmax error; - python: - add runSessionWithCallBackInfo interface; - add max nodes limit (1400) for visualization tool; - fix save error in python3; - align default dim; - convert: - add extra design for optimization; - add more post converting optimizers; - add caffe v1 weights blob support; - add cast, unary, conv transpose support for onnx model; - optimize batchnorm, conv with variable weights, prelu, reshape, slice, upsample for onnx model; - add cos/sin/atan/tan support for unary for tensorflow model; - add any/all support for reduction for tensorflow model; - add elu, conv3d, pool3d support for tensorflow model; - optimize argmax, batchnorm, concat, batch to space, conv with variable weights, prelu, slice for tensorflow model; - others: - fix size computer lock; - fix thread pool deadlock; - add express & parameters in express; - rewrite blitter chooser without static map; - add tests for expr;
2019-10-29 13:37:26 +08:00
static int convertFormat(Dimensionformat format);
2019-12-27 22:16:57 +08:00
static Express::Dimensionformat revertFormat(int format);
static halide_type_t revertDataType(DataType dataType);
2020-11-05 16:41:56 +08:00
static bool allocMemoryForHostTensor(Tensor* dest);
static bool releaseMemoryForHostTensor(Tensor* dest);
2022-07-22 09:59:30 +08:00
static Tensor* getTensor(VARP var);
2022-12-30 15:18:58 +08:00
static EXPRP makeRaster(const std::vector<VARP>& vars, const std::vector<int>& regions, const std::vector<int>& shape, halide_type_t dataType, MNN_DATA_FORMAT format);
};
2025-07-23 14:10:58 +08:00
class Executor::RuntimeExecuteWrap {
public:
RuntimeExecuteWrap(const RuntimeInfo& info);
~ RuntimeExecuteWrap();
private:
const RuntimeInfo& mRt;
};
} // namespace Express
} // namespace MNN
#endif