Unverified Commit 244d0e34 authored by groot's avatar groot Committed by GitHub
Browse files

Remove unnecessary memcpy (#2337)



* fix #2264

Signed-off-by: default avataryhmo <yihua.mo@zilliz.com>

* typo

Signed-off-by: default avataryhmo <yihua.mo@zilliz.com>

* fix ut

Signed-off-by: default avataryhmo <yihua.mo@zilliz.com>

* typo

Signed-off-by: default avataryhmo <yihua.mo@zilliz.com>

* Remove unnecessary memcpy

Signed-off-by: default avataryhmo <yihua.mo@zilliz.com>
parent 312df5af
Loading
Loading
Loading
Loading
+11 −28
Original line number Diff line number Diff line
@@ -18,6 +18,7 @@
#include "db/engine/ExecutionEngine.h"
#include "metrics/Metrics.h"
#include "utils/Log.h"
#include "utils/TimeRecorder.h"

namespace milvus {
namespace engine {
@@ -36,9 +37,8 @@ VectorSource::VectorSource(milvus::engine::VectorsData vectors,
}

Status
VectorSource::Add(/*const ExecutionEnginePtr& execution_engine,*/ const segment::SegmentWriterPtr& segment_writer_ptr,
                  const meta::SegmentSchema& table_file_schema, const size_t& num_vectors_to_add,
                  size_t& num_vectors_added) {
VectorSource::Add(const segment::SegmentWriterPtr& segment_writer_ptr, const meta::SegmentSchema& table_file_schema,
                  const size_t& num_vectors_to_add, size_t& num_vectors_added) {
    uint64_t n = vectors_.vector_count_;
    server::CollectAddMetrics metrics(n, table_file_schema.dimension_);

@@ -61,35 +61,17 @@ VectorSource::Add(/*const ExecutionEnginePtr& execution_engine,*/ const segment:

    Status status;
    if (!vectors_.float_data_.empty()) {
        /*
        status = execution_engine->AddWithIds(
            num_vectors_added, vectors_.float_data_.data() + current_num_vectors_added * table_file_schema.dimension_,
            vector_ids_to_add.data());
        */
        std::vector<uint8_t> vectors;
        LOG_ENGINE_DEBUG_ << LogOut("[%s][%ld]", "insert", 0) << "Insert float data into segment";
        auto size = num_vectors_added * table_file_schema.dimension_ * sizeof(float);
        vectors.resize(size);
        memcpy(vectors.data(), vectors_.float_data_.data() + current_num_vectors_added * table_file_schema.dimension_,
               size);
        LOG_ENGINE_DEBUG_ << LogOut("[%s][%ld]", "insert", 0) << "Insert into segment";
        status = segment_writer_ptr->AddVectors(table_file_schema.file_id_, vectors, vector_ids_to_add);

        float* ptr = vectors_.float_data_.data() + current_num_vectors_added * table_file_schema.dimension_;
        status = segment_writer_ptr->AddVectors(table_file_schema.file_id_, (uint8_t*)ptr, size, vector_ids_to_add);
    } else if (!vectors_.binary_data_.empty()) {
        /*
        status = execution_engine->AddWithIds(
            num_vectors_added,
            vectors_.binary_data_.data() + current_num_vectors_added * SingleVectorSize(table_file_schema.dimension_),
            vector_ids_to_add.data());
        */
        LOG_ENGINE_DEBUG_ << LogOut("[%s][%ld]", "insert", 0) << "Insert binary data into segment";
        std::vector<uint8_t> vectors;
        auto size = num_vectors_added * SingleVectorSize(table_file_schema.dimension_) * sizeof(uint8_t);
        vectors.resize(size);
        memcpy(
            vectors.data(),
            vectors_.binary_data_.data() + current_num_vectors_added * SingleVectorSize(table_file_schema.dimension_),
            size);
        LOG_ENGINE_DEBUG_ << LogOut("[%s][%ld]", "insert", 0) << "Insert into segment";
        status = segment_writer_ptr->AddVectors(table_file_schema.file_id_, vectors, vector_ids_to_add);
        uint8_t* ptr =
            vectors_.binary_data_.data() + current_num_vectors_added * SingleVectorSize(table_file_schema.dimension_);
        status = segment_writer_ptr->AddVectors(table_file_schema.file_id_, ptr, size, vector_ids_to_add);
    }

    // Clear vector data
@@ -175,6 +157,7 @@ VectorSource::SingleVectorSize(uint16_t dimension) {

    return 0;
}

size_t
VectorSource::SingleEntitySize(uint16_t dimension) {
    // TODO(yukun) add entity type and size compute
+2 −2
Original line number Diff line number Diff line
@@ -36,8 +36,8 @@ class VectorSource {
                 const std::unordered_map<std::string, std::vector<uint8_t>>& attr_data);

    Status
    Add(/*const ExecutionEnginePtr& execution_engine,*/ const segment::SegmentWriterPtr& segment_writer_ptr,
        const meta::SegmentSchema& table_file_schema, const size_t& num_vectors_to_add, size_t& num_vectors_added);
    Add(const segment::SegmentWriterPtr& segment_writer_ptr, const meta::SegmentSchema& table_file_schema,
        const size_t& num_vectors_to_add, size_t& num_vectors_added);

    Status
    AddEntities(const segment::SegmentWriterPtr& segment_writer_ptr, const meta::SegmentSchema& collection_file_schema,
+1 −1
Original line number Diff line number Diff line
@@ -20,7 +20,7 @@
namespace milvus {
namespace engine {

const int64_t FORCE_MERGE_THREASHOLD = 10;  // force merge files older this time(in second)
const int64_t FORCE_MERGE_THREASHOLD = 30;  // force merge files older this time(in second)

Status
MergeLayeredStrategy::RegroupFiles(meta::FilesHolder& files_holder, MergeFilesGroups& files_groups) {
+10 −0
Original line number Diff line number Diff line
@@ -52,6 +52,16 @@ SegmentWriter::AddVectors(const std::string& name, const std::vector<uint8_t>& d
    return Status::OK();
}

Status
SegmentWriter::AddVectors(const std::string& name, const uint8_t* data, uint64_t size,
                          const std::vector<doc_id_t>& uids) {
    segment_ptr_->vectors_ptr_->AddData(data, size);
    segment_ptr_->vectors_ptr_->AddUids(uids);
    segment_ptr_->vectors_ptr_->SetName(name);

    return Status::OK();
}

Status
SegmentWriter::AddAttrs(const std::string& name, const std::unordered_map<std::string, uint64_t>& attr_nbytes,
                        const std::unordered_map<std::string, std::vector<uint8_t>>& attr_data,
+3 −0
Original line number Diff line number Diff line
@@ -36,6 +36,9 @@ class SegmentWriter {
    Status
    AddVectors(const std::string& name, const std::vector<uint8_t>& data, const std::vector<doc_id_t>& uids);

    Status
    AddVectors(const std::string& name, const uint8_t* data, uint64_t size, const std::vector<doc_id_t>& uids);

    Status
    AddAttrs(const std::string& name, const std::unordered_map<std::string, uint64_t>& attr_nbytes,
             const std::unordered_map<std::string, std::vector<uint8_t>>& attr_data, const std::vector<doc_id_t>& uids);
Loading