bsf-inl-tensor.h

#pragma once

#include <errno.h>
#include <vector>
#include <deque>
#include <butil/atomicops.h>
#include <comlog/comlog.h>
#include "common/inner_common.h"
#include "framework/infer_data.h"
#include "framework/memory.h"

#include <boost/function.hpp>

namespace im {
namespace bsf {

template<>
struct Task<baidu::paddle_serving::predictor::Tensor, 
        baidu::paddle_serving::predictor::Tensor> {

    typedef Task<baidu::paddle_serving::predictor::Tensor, 
            baidu::paddle_serving::predictor::Tensor> TaskT;
    typedef baidu::paddle_serving::predictor::Tensor Tensor;
    typedef baidu::paddle_serving::predictor::Tensor InType;
    typedef baidu::paddle_serving::predictor::Tensor OutType;
    typedef baidu::paddle_serving::predictor::BatchTensor BatchTensor;
    typedef baidu::paddle_serving::predictor::BatchTensor InArrayT;
    typedef baidu::paddle_serving::predictor::BatchTensor OutArrayT;

    struct Segment {
        Segment(void* p, size_t b, size_t s) 
            : ptr(p), begin(b), size(s) {}
        void* ptr;
        size_t begin;
        size_t size;
    };

    int read_fd;
    int write_fd;

    pid_t owner_tid;

    const InArrayT* in;
    OutArrayT* out;

    size_t rem;
    size_t size;

    butil::atomic<size_t> index;

    const BatchTensor* get(bool is_in) const {
        if (is_in) {
            return in;
        } else {
            return out;
        }
    }

    BatchTensor* get(bool is_in) {
        if (is_in) {
            return const_cast<BatchTensor*>(in);
        } else {
            return out;
        }
    }

    Task() {
        read_fd = -1;
        write_fd = -1;
        owner_tid = -1;
        in = NULL;
        out = NULL;
        rem = -1;
        size = -1;
        index.store(0, butil::memory_order_relaxed); 
    }
};

template<>
class BatchTasks<Task<
            baidu::paddle_serving::predictor::Tensor, 
            baidu::paddle_serving::predictor::Tensor> > {
public:
    typedef baidu::paddle_serving::predictor::Tensor Tensor;
    typedef baidu::paddle_serving::predictor::Tensor InType;
    typedef baidu::paddle_serving::predictor::Tensor OutType;
    typedef baidu::paddle_serving::predictor::DataBuf DataBuf;
    typedef baidu::paddle_serving::predictor::MempoolWrapper MempoolWrapper;

    typedef Task<baidu::paddle_serving::predictor::Tensor, 
            baidu::paddle_serving::predictor::Tensor> TaskT;
    typedef TaskMeta<TaskT> TaskMetaT;
    typedef TaskT::InArrayT InArrayT;
    typedef TaskT::OutArrayT OutArrayT;

    BatchTasks(size_t batch_size, bool batch_align = false)
            : _batch_size(batch_size)
            , _rem_size(batch_size)
            , _batch_align(batch_align) {
        _batch_in.clear();
        _batch_out.clear();
        _tasks.clear();
    }

    ~BatchTasks() {
        _batch_in.clear();
        _batch_out.clear(); 
        _tasks.clear();
    }

    static bool check_valid(
            const InArrayT& in, OutArrayT& out, bool align) {
        if (align) {
            if (out.count() <= 0 || out.size() <= 0) {
                CFATAL_LOG("Out tensor is empty, when aligned");
                return false;
            }

            if (out.size() != in.size()) {
                CFATAL_LOG("In/Out tensor size not eq: %ld!=%ld",
                        out.size(), in.size());
                return false;
            }

            for (size_t fi = 0, shape0 = 0; fi < out.count(); ++fi) {
                if (!out[fi].valid()) {
                    CFATAL_LOG("Out[%ld] tensor not valid", fi);
                    return false;
                }

                if (out.size() != out[fi].shape0()) {
                    CFATAL_LOG("Shape0 not consistency, %ld!=%ld, %ld",
                            out.size(), out[fi].shape0(), fi);
                    return false;
                }
            }
        }

        return true;
    }

    size_t append_task(TaskT* task) {
        size_t add = std::min(task->rem, _rem_size);
        if (!_batch_align) {
            add = task->rem; 
        }
        TaskMetaT tm(task, task->in->size() - task->rem, add);
        _tasks.push_back(tm);

        task->rem -= add;
        _rem_size -= add;
        return _rem_size;
    }

    void merge_tasks() {
        merge_input();
        merge_output();
    }

    void merge_input() {
        if (_tasks.size() <= 0 || _tasks[0].task->in->count() <= 0) {
            return ;
        }

        if (_tasks.size() == 1 && !_batch_align) {
            TaskMetaT& tm = _tasks[0];
            _batch_in = *(tm.task->in);
            return ;
        }

        merge_tensor(true);
    }

    void merge_output() {
        if (_batch_align) {
            if (_tasks.size() <= 0 || _tasks[0].task->out->count() <= 0) {
                return ;
            }
        } 

        if (_tasks.size() <= 0 || _tasks[0].task->out->count() <= 0) {
            return ;
        }

        TaskMetaT& tm = _tasks[0];
        if (_tasks.size() == 1 && !_batch_align) {
            _batch_out = *(tm.task->out);
            return ;
        }

        if (tm.task->out->size() <= 0) {
            // shape is empty
            _batch_out = *(tm.task->out);
            return ;
        }

        if ((*tm.task->out)[0].data.data() == 0 
                || (*tm.task->out)[0].data.size() == 0) {
            _batch_out = *(tm.task->out);
            return ;
        }

        merge_tensor(false);
    }

    void merge_tensor(bool is_in) {
        // accumulate batch size from fetched tasks
        size_t batch_size = 0;
        for (size_t ti = 0; ti < _tasks.size(); ++ti) {
            TaskMetaT& tm = _tasks[ti];
            size_t add = tm.end - tm.begin;
            batch_size += add;
        }

        // merge all instanses in each tensor data
        size_t tensor_count = _tasks[0].task->get(is_in)->count();
        for (size_t fi = 0; fi < tensor_count; ++fi) {
            const Tensor& head = (*(_tasks[0].task->get(is_in)))[fi];
            Tensor batch_tensor;
            batch_tensor.name = head.name;
            batch_tensor.type = head.type;
            batch_tensor.shape.push_back(batch_size);
        
            size_t ins_ele_count = 1;
            for (size_t si = 1; si < head.shape.size(); ++si) {
                batch_tensor.shape.push_back(head.shape[si]);
                ins_ele_count *= head.shape[si];
            }

            size_t tensor_ele_count = ins_ele_count * batch_size;
            size_t ins_byte = ins_ele_count * head.ele_byte();

            size_t tensor_byte = tensor_ele_count * head.ele_byte();
            void* data_buf
                = MempoolWrapper::instance().malloc(tensor_byte);
            if (!data_buf) {
                CFATAL_LOG("Malloc failed, size: %ld", tensor_byte);
                return ;
            }

            size_t data_byte = 0;
            for (size_t ti = 0; ti < _tasks.size(); ++ti) {
                TaskMetaT& tm = _tasks[ti];
                size_t acc_byte = ins_byte * (tm.end - tm.begin);
                if (data_byte + acc_byte > tensor_byte) {
                    CFATAL_LOG("Invalid bytes: %ld + %ld >= %ld",
                            data_byte, acc_byte, tensor_byte);
                    return ;
                }

                const Tensor& tensor = (*(tm.task->get(is_in)))[fi];
                memcpy(data_buf + data_byte, 
                        tensor.data.data() + tm.begin * ins_byte, 
                        acc_byte);
                data_byte += acc_byte;
            }

            if (data_byte != tensor_byte) {
                CFATAL_LOG("Invalid tensor byte: %ld != %ld",
                        data_byte, tensor_byte); 
                return ;
            }

            batch_tensor.data = DataBuf(data_buf, tensor_byte);
            if (is_in) {
                _batch_in.push_back(batch_tensor);
            } else {
                _batch_out.push_back(batch_tensor);
            }
        }

        LOG(INFO) << "merge input(" << is_in << ") samples: " 
            << batch_size << " from " << _tasks.size() << " pvs";
    }

    void notify_tasks() {
        if (_batch_out.size() != _batch_in.size()) {
            CFATAL_LOG("batch size not consistency: %ld != %ld",
                    _batch_out.size(), _batch_in.size());
            return ;
        }

        size_t tensor_count = _batch_out.count();
        size_t batch_size = _batch_out.size();
        for (size_t fi = 0; fi < tensor_count; ++fi) {
            const Tensor& tensor = _batch_out[fi];
            size_t ins_byte = tensor.ele_byte();
            for (size_t si = 1; si < tensor.shape.size(); ++si) {
                ins_byte *= tensor.shape[si];
            }

            for (size_t ti = 0, bi = 0, add = 0; 
                    ti < _tasks.size(); ++ti, bi += add) {
                OutArrayT* dst = _tasks[ti].task->out;
                add = _tasks[ti].end - _tasks[ti].begin;
                size_t offset_src = ins_byte * bi;
                size_t add_byte = add * ins_byte;

                if (_batch_align) {  // merge all batchs
                    size_t offset_dst = ins_byte * _tasks[ti].begin;
                    void* ptr = const_cast<void*>((*dst)[fi].data.data());
                    memcpy(ptr + offset_dst, 
                            _batch_out[fi].data.data() + offset_src, add_byte);
                } else {  // overwrite
                    if (dst->count() <= 0) {
                        dst->push_back(_batch_out[fi]);
                    } else {
                        (*dst)[fi] = _batch_out[fi];
                    }

                    (*dst)[fi].shape[0] = add;
                    (*dst)[fi].data = DataBuf(
                            _batch_out[fi].data.data() + offset_src, add_byte);
                }
            }
        }
        
        for (size_t ti = 0; ti < _tasks.size(); ++ti) {
            TaskT* task = _tasks[ti].task;
            size_t begin = _tasks[ti].begin;
            size_t end = _tasks[ti].end;
            size_t add = end - begin;

            size_t index = task->index.fetch_add(add);
            if ((index + add) >= task->in->size()) {
                char c = 0;
                while (write(task->write_fd, &c, 1) != 1 && errno == EINTR) {
                    ;
                }
                butil::return_object(task);
            }
        }
    }

    const typename TaskT::InArrayT& in() const {
        return _batch_in;
    }

    typename TaskT::OutArrayT& out() {
        return _batch_out;
    }

    size_t task_size() {
        return _tasks.size();
    }

private:
    std::vector<TaskMetaT> _tasks;
    InArrayT _batch_in;
    OutArrayT _batch_out;
    size_t _rem_size;
    size_t _batch_size;
    bool _batch_align;
};

} // namespace bsf
} // namespace im