Hendrik/xgboost
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Partial rewrite EllpackPage (#5352)

Browse Source
This commit is contained in:
Rory Mitchell 2020-03-11 10:15:53 +13:00 committed by GitHub
parent 7a99f8f27f
commit 3ad4333b0e
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
23 changed files with 496 additions and 733 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
src/common/compressed_iterator.h
View File

@ -181,13 +181,13 @@ class CompressedIterator {
typedef value_type reference; // NOLINT typedef value_type reference; // NOLINT
private: private:
CompressedByteT *buffer_; const CompressedByteT *buffer_;
size_t symbol_bits_; size_t symbol_bits_;
size_t offset_; size_t offset_;
public: public:
CompressedIterator() : buffer_(nullptr), symbol_bits_(0), offset_(0) {} CompressedIterator() : buffer_(nullptr), symbol_bits_(0), offset_(0) {}
CompressedIterator(CompressedByteT *buffer, size_t num_symbols) CompressedIterator(const CompressedByteT *buffer, size_t num_symbols)
: buffer_(buffer), offset_(0) { : buffer_(buffer), offset_(0) {
symbol_bits_ = detail::SymbolBits(num_symbols); symbol_bits_ = detail::SymbolBits(num_symbols);
} }

74
src/common/hist_util.cc
View File

@ -31,7 +31,7 @@ namespace common {
HistogramCuts::HistogramCuts() { HistogramCuts::HistogramCuts() {
monitor_.Init(__FUNCTION__); monitor_.Init(__FUNCTION__);
cut_ptrs_.emplace_back(0); cut_ptrs_.HostVector().emplace_back(0);
} }
// Dispatch to specific builder. // Dispatch to specific builder.
@ -52,7 +52,7 @@ void HistogramCuts::Build(DMatrix* dmat, uint32_t const max_num_bins) {
DenseCuts cuts(this); DenseCuts cuts(this);
cuts.Build(dmat, max_num_bins); cuts.Build(dmat, max_num_bins);
} }
LOG(INFO) << "Total number of hist bins: " << cut_ptrs_.back(); LOG(INFO) << "Total number of hist bins: " << cut_ptrs_.HostVector().back();
} }
bool CutsBuilder::UseGroup(DMatrix* dmat) { bool CutsBuilder::UseGroup(DMatrix* dmat) {
@ -75,7 +75,10 @@ void SparseCuts::SingleThreadBuild(SparsePage const& page, MetaInfo const& info,
// Data groups, used in ranking. // Data groups, used in ranking.
std::vector<bst_uint> const& group_ptr = info.group_ptr_; std::vector<bst_uint> const& group_ptr = info.group_ptr_;
p_cuts_->min_vals_.resize(end_col - beg_col, 0); auto &local_min_vals = p_cuts_->min_vals_.HostVector();
auto &local_cuts = p_cuts_->cut_values_.HostVector();
auto &local_ptrs = p_cuts_->cut_ptrs_.HostVector();
local_min_vals.resize(end_col - beg_col, 0);
for (uint32_t col_id = beg_col; col_id < page.Size() && col_id < end_col; ++col_id) { for (uint32_t col_id = beg_col; col_id < page.Size() && col_id < end_col; ++col_id) {
// Using a local variable makes things easier, but at the cost of memory trashing. // Using a local variable makes things easier, but at the cost of memory trashing.
@ -85,7 +88,7 @@ void SparseCuts::SingleThreadBuild(SparsePage const& page, MetaInfo const& info,
max_num_bins); max_num_bins);
if (n_bins == 0) { if (n_bins == 0) {
// cut_ptrs_ is initialized with a zero, so there's always an element at the back // cut_ptrs_ is initialized with a zero, so there's always an element at the back
p_cuts_->cut_ptrs_.emplace_back(p_cuts_->cut_ptrs_.back()); local_ptrs.emplace_back(local_ptrs.back());
continue; continue;
} }
@ -112,17 +115,17 @@ void SparseCuts::SingleThreadBuild(SparsePage const& page, MetaInfo const& info,
// Can be use data[1] as the min values so that we don't need to // Can be use data[1] as the min values so that we don't need to
// store another array? // store another array?
float mval = summary.data[0].value; float mval = summary.data[0].value;
p_cuts_->min_vals_[col_id - beg_col] = mval - (fabs(mval) + 1e-5); local_min_vals[col_id - beg_col] = mval - (fabs(mval) + 1e-5);
this->AddCutPoint(summary, max_num_bins); this->AddCutPoint(summary, max_num_bins);
bst_float cpt = (summary.size > 0) ? bst_float cpt = (summary.size > 0) ?
summary.data[summary.size - 1].value : summary.data[summary.size - 1].value :
p_cuts_->min_vals_[col_id - beg_col]; local_min_vals[col_id - beg_col];
cpt += fabs(cpt) + 1e-5; cpt += fabs(cpt) + 1e-5;
p_cuts_->cut_values_.emplace_back(cpt); local_cuts.emplace_back(cpt);
p_cuts_->cut_ptrs_.emplace_back(p_cuts_->cut_values_.size()); local_ptrs.emplace_back(local_cuts.size());
} }
} }
@ -196,33 +199,40 @@ void SparseCuts::Concat(
std::vector<std::unique_ptr<SparseCuts>> const& cuts, uint32_t n_cols) { std::vector<std::unique_ptr<SparseCuts>> const& cuts, uint32_t n_cols) {
monitor_.Start(__FUNCTION__); monitor_.Start(__FUNCTION__);
uint32_t nthreads = omp_get_max_threads(); uint32_t nthreads = omp_get_max_threads();
p_cuts_->min_vals_.resize(n_cols, std::numeric_limits<float>::max()); auto &local_min_vals = p_cuts_->min_vals_.HostVector();
auto &local_cuts = p_cuts_->cut_values_.HostVector();
auto &local_ptrs = p_cuts_->cut_ptrs_.HostVector();
local_min_vals.resize(n_cols, std::numeric_limits<float>::max());
size_t min_vals_tail = 0; size_t min_vals_tail = 0;
for (uint32_t t = 0; t < nthreads; ++t) { for (uint32_t t = 0; t < nthreads; ++t) {
auto& thread_min_vals = cuts[t]->p_cuts_->min_vals_.HostVector();
auto& thread_cuts = cuts[t]->p_cuts_->cut_values_.HostVector();
auto& thread_ptrs = cuts[t]->p_cuts_->cut_ptrs_.HostVector();
// concat csc pointers. // concat csc pointers.
size_t const old_ptr_size = p_cuts_->cut_ptrs_.size(); size_t const old_ptr_size = local_ptrs.size();
p_cuts_->cut_ptrs_.resize( local_ptrs.resize(
cuts[t]->p_cuts_->cut_ptrs_.size() + p_cuts_->cut_ptrs_.size() - 1); thread_ptrs.size() + local_ptrs.size() - 1);
size_t const new_icp_size = p_cuts_->cut_ptrs_.size(); size_t const new_icp_size = local_ptrs.size();
auto tail = p_cuts_->cut_ptrs_[old_ptr_size-1]; auto tail = local_ptrs[old_ptr_size-1];
for (size_t j = old_ptr_size; j < new_icp_size; ++j) { for (size_t j = old_ptr_size; j < new_icp_size; ++j) {
p_cuts_->cut_ptrs_[j] = tail + cuts[t]->p_cuts_->cut_ptrs_[j-old_ptr_size+1]; local_ptrs[j] = tail + thread_ptrs[j-old_ptr_size+1];
} }
// concat csc values // concat csc values
size_t const old_iv_size = p_cuts_->cut_values_.size(); size_t const old_iv_size = local_cuts.size();
p_cuts_->cut_values_.resize( local_cuts.resize(
cuts[t]->p_cuts_->cut_values_.size() + p_cuts_->cut_values_.size()); thread_cuts.size() + local_cuts.size());
size_t const new_iv_size = p_cuts_->cut_values_.size(); size_t const new_iv_size = local_cuts.size();
for (size_t j = old_iv_size; j < new_iv_size; ++j) { for (size_t j = old_iv_size; j < new_iv_size; ++j) {
p_cuts_->cut_values_[j] = cuts[t]->p_cuts_->cut_values_[j-old_iv_size]; local_cuts[j] = thread_cuts[j-old_iv_size];
} }
// merge min values // merge min values
for (size_t j = 0; j < cuts[t]->p_cuts_->min_vals_.size(); ++j) { for (size_t j = 0; j < thread_min_vals.size(); ++j) {
p_cuts_->min_vals_.at(min_vals_tail + j) = local_min_vals.at(min_vals_tail + j) =
std::min(p_cuts_->min_vals_.at(min_vals_tail + j), cuts.at(t)->p_cuts_->min_vals_.at(j)); std::min(local_min_vals.at(min_vals_tail + j), thread_min_vals.at(j));
} }
min_vals_tail += cuts[t]->p_cuts_->min_vals_.size(); min_vals_tail += thread_min_vals.size();
} }
monitor_.Stop(__FUNCTION__); monitor_.Stop(__FUNCTION__);
} }
@ -323,27 +333,27 @@ void DenseCuts::Init
// TODO(chenqin): rabit failure recovery assumes no boostrap onetime call after loadcheckpoint // TODO(chenqin): rabit failure recovery assumes no boostrap onetime call after loadcheckpoint
// we need to move this allreduce before loadcheckpoint call in future // we need to move this allreduce before loadcheckpoint call in future
sreducer.Allreduce(dmlc::BeginPtr(summary_array), nbytes, summary_array.size()); sreducer.Allreduce(dmlc::BeginPtr(summary_array), nbytes, summary_array.size());
p_cuts_->min_vals_.resize(sketchs.size()); p_cuts_->min_vals_.HostVector().resize(sketchs.size());
for (size_t fid = 0; fid < summary_array.size(); ++fid) { for (size_t fid = 0; fid < summary_array.size(); ++fid) {
WQSketch::SummaryContainer a; WQSketch::SummaryContainer a;
a.Reserve(max_num_bins + 1); a.Reserve(max_num_bins + 1);
a.SetPrune(summary_array[fid], max_num_bins + 1); a.SetPrune(summary_array[fid], max_num_bins + 1);
const bst_float mval = a.data[0].value; const bst_float mval = a.data[0].value;
p_cuts_->min_vals_[fid] = mval - (fabs(mval) + 1e-5); p_cuts_->min_vals_.HostVector()[fid] = mval - (fabs(mval) + 1e-5);
AddCutPoint(a, max_num_bins); AddCutPoint(a, max_num_bins);
// push a value that is greater than anything // push a value that is greater than anything
const bst_float cpt const bst_float cpt
= (a.size > 0) ? a.data[a.size - 1].value : p_cuts_->min_vals_[fid]; = (a.size > 0) ? a.data[a.size - 1].value : p_cuts_->min_vals_.HostVector()[fid];
// this must be bigger than last value in a scale // this must be bigger than last value in a scale
const bst_float last = cpt + (fabs(cpt) + 1e-5); const bst_float last = cpt + (fabs(cpt) + 1e-5);
p_cuts_->cut_values_.push_back(last); p_cuts_->cut_values_.HostVector().push_back(last);
// Ensure that every feature gets at least one quantile point // Ensure that every feature gets at least one quantile point
CHECK_LE(p_cuts_->cut_values_.size(), std::numeric_limits<uint32_t>::max()); CHECK_LE(p_cuts_->cut_values_.HostVector().size(), std::numeric_limits<uint32_t>::max());
auto cut_size = static_cast<uint32_t>(p_cuts_->cut_values_.size()); auto cut_size = static_cast<uint32_t>(p_cuts_->cut_values_.HostVector().size());
CHECK_GT(cut_size, p_cuts_->cut_ptrs_.back()); CHECK_GT(cut_size, p_cuts_->cut_ptrs_.HostVector().back());
p_cuts_->cut_ptrs_.push_back(cut_size); p_cuts_->cut_ptrs_.HostVector().push_back(cut_size);
} }
monitor_.Stop(__func__); monitor_.Stop(__func__);
} }

58
src/common/hist_util.h
View File

@ -44,17 +44,35 @@ class HistogramCuts {
using BinIdx = uint32_t; using BinIdx = uint32_t;
common::Monitor monitor_; common::Monitor monitor_;
std::vector<bst_float> cut_values_;
std::vector<uint32_t> cut_ptrs_;
std::vector<float> min_vals_; // storing minimum value in a sketch set.
public: public:
HostDeviceVector<bst_float> cut_values_;
HostDeviceVector<uint32_t> cut_ptrs_;
HostDeviceVector<float> min_vals_; // storing minimum value in a sketch set.
HistogramCuts(); HistogramCuts();
HistogramCuts(HistogramCuts const& that) = delete; HistogramCuts(HistogramCuts const& that) {
cut_values_.Resize(that.cut_values_.Size());
cut_ptrs_.Resize(that.cut_ptrs_.Size());
min_vals_.Resize(that.min_vals_.Size());
cut_values_.Copy(that.cut_values_);
cut_ptrs_.Copy(that.cut_ptrs_);
min_vals_.Copy(that.min_vals_);
}
HistogramCuts(HistogramCuts&& that) noexcept(true) { HistogramCuts(HistogramCuts&& that) noexcept(true) {
*this = std::forward<HistogramCuts&&>(that); *this = std::forward<HistogramCuts&&>(that);
} }
HistogramCuts& operator=(HistogramCuts const& that) = delete;
HistogramCuts& operator=(HistogramCuts const& that) {
cut_values_.Resize(that.cut_values_.Size());
cut_ptrs_.Resize(that.cut_ptrs_.Size());
min_vals_.Resize(that.min_vals_.Size());
cut_values_.Copy(that.cut_values_);
cut_ptrs_.Copy(that.cut_ptrs_);
min_vals_.Copy(that.min_vals_);
return *this;
}
HistogramCuts& operator=(HistogramCuts&& that) noexcept(true) { HistogramCuts& operator=(HistogramCuts&& that) noexcept(true) {
monitor_ = std::move(that.monitor_); monitor_ = std::move(that.monitor_);
cut_ptrs_ = std::move(that.cut_ptrs_); cut_ptrs_ = std::move(that.cut_ptrs_);
@ -67,28 +85,30 @@ class HistogramCuts {
void Build(DMatrix* dmat, uint32_t const max_num_bins); void Build(DMatrix* dmat, uint32_t const max_num_bins);
/* \brief How many bins a feature has. */ /* \brief How many bins a feature has. */
uint32_t FeatureBins(uint32_t feature) const { uint32_t FeatureBins(uint32_t feature) const {
return cut_ptrs_.at(feature+1) - cut_ptrs_[feature]; return cut_ptrs_.ConstHostVector().at(feature + 1) -
cut_ptrs_.ConstHostVector()[feature];
} }
// Getters. Cuts should be of no use after building histogram indices, but currently // Getters. Cuts should be of no use after building histogram indices, but currently
// it's deeply linked with quantile_hist, gpu sketcher and gpu_hist. So we preserve // it's deeply linked with quantile_hist, gpu sketcher and gpu_hist. So we preserve
// these for now. // these for now.
std::vector<uint32_t> const& Ptrs() const { return cut_ptrs_; } std::vector<uint32_t> const& Ptrs() const { return cut_ptrs_.ConstHostVector(); }
std::vector<float> const& Values() const { return cut_values_; } std::vector<float> const& Values() const { return cut_values_.ConstHostVector(); }
std::vector<float> const& MinValues() const { return min_vals_; } std::vector<float> const& MinValues() const { return min_vals_.ConstHostVector(); }
size_t TotalBins() const { return cut_ptrs_.back(); } size_t TotalBins() const { return cut_ptrs_.ConstHostVector().back(); }
// Return the index of a cut point that is strictly greater than the input // Return the index of a cut point that is strictly greater than the input
// value, or the last available index if none exists // value, or the last available index if none exists
BinIdx SearchBin(float value, uint32_t column_id) const { BinIdx SearchBin(float value, uint32_t column_id) const {
auto beg = cut_ptrs_.at(column_id); auto beg = cut_ptrs_.ConstHostVector().at(column_id);
auto end = cut_ptrs_.at(column_id + 1); auto end = cut_ptrs_.ConstHostVector().at(column_id + 1);
auto it = std::upper_bound(cut_values_.cbegin() + beg, cut_values_.cbegin() + end, value); const auto &values = cut_values_.ConstHostVector();
if (it == cut_values_.cend()) { auto it = std::upper_bound(values.cbegin() + beg, values.cbegin() + end, value);
it = cut_values_.cend() - 1; if (it == values.cend()) {
it = values.cend() - 1;
} }
BinIdx idx = it - cut_values_.cbegin(); BinIdx idx = it - values.cbegin();
return idx; return idx;
} }
@ -133,8 +153,8 @@ class CutsBuilder {
size_t required_cuts = std::min(summary.size, static_cast<size_t>(max_bin)); size_t required_cuts = std::min(summary.size, static_cast<size_t>(max_bin));
for (size_t i = 1; i < required_cuts; ++i) { for (size_t i = 1; i < required_cuts; ++i) {
bst_float cpt = summary.data[i].value; bst_float cpt = summary.data[i].value;
if (i == 1 || cpt > p_cuts_->cut_values_.back()) { if (i == 1 || cpt > p_cuts_->cut_values_.ConstHostVector().back()) {
p_cuts_->cut_values_.push_back(cpt); p_cuts_->cut_values_.HostVector().push_back(cpt);
} }
} }
} }

1
src/common/host_device_vector.cu
View File

@ -371,6 +371,7 @@ void HostDeviceVector<T>::Resize(size_t new_size, T v) {
template class HostDeviceVector<bst_float>; template class HostDeviceVector<bst_float>;
template class HostDeviceVector<GradientPair>; template class HostDeviceVector<GradientPair>;
template class HostDeviceVector<int32_t>; // bst_node_t template class HostDeviceVector<int32_t>; // bst_node_t
template class HostDeviceVector<uint8_t>;
template class HostDeviceVector<Entry>; template class HostDeviceVector<Entry>;
template class HostDeviceVector<uint64_t>; // bst_row_t template class HostDeviceVector<uint64_t>; // bst_row_t
template class HostDeviceVector<uint32_t>; // bst_feature_t template class HostDeviceVector<uint32_t>; // bst_feature_t

17
src/data/ellpack_page.cc
View File

@ -13,11 +13,24 @@ class EllpackPageImpl {};
EllpackPage::EllpackPage() = default; EllpackPage::EllpackPage() = default;
EllpackPage::EllpackPage(DMatrix* dmat, const BatchParam& param) { EllpackPage::EllpackPage(DMatrix* dmat, const BatchParam& param) {
LOG(FATAL) << "Internal Error: XGBoost is not compiled with CUDA but EllpackPage is required"; LOG(FATAL) << "Internal Error: XGBoost is not compiled with CUDA but "
"EllpackPage is required";
} }
EllpackPage::~EllpackPage() { EllpackPage::~EllpackPage() {
LOG(FATAL) << "Internal Error: XGBoost is not compiled with CUDA but EllpackPage is required"; LOG(FATAL) << "Internal Error: XGBoost is not compiled with CUDA but "
"EllpackPage is required";
}
void EllpackPage::SetBaseRowId(size_t row_id) {
LOG(FATAL) << "Internal Error: XGBoost is not compiled with CUDA but "
"EllpackPage is required";
}
size_t EllpackPage::Size() const {
LOG(FATAL) << "Internal Error: XGBoost is not compiled with CUDA but "
"EllpackPage is required";
return 0;
} }
} // namespace xgboost } // namespace xgboost

295
src/data/ellpack_page.cu
View File

@ -4,9 +4,9 @@
#include <xgboost/data.h> #include <xgboost/data.h>
#include "./ellpack_page.cuh"
#include "../common/hist_util.h" #include "../common/hist_util.h"
#include "../common/random.h" #include "../common/random.h"
#include "./ellpack_page.cuh"
namespace xgboost { namespace xgboost {
@ -17,13 +17,9 @@ EllpackPage::EllpackPage(DMatrix* dmat, const BatchParam& param)
EllpackPage::~EllpackPage() = default; EllpackPage::~EllpackPage() = default;
size_t EllpackPage::Size() const { size_t EllpackPage::Size() const { return impl_->Size(); }
return impl_->Size();
}
void EllpackPage::SetBaseRowId(size_t row_id) { void EllpackPage::SetBaseRowId(size_t row_id) { impl_->SetBaseRowId(row_id); }
impl_->SetBaseRowId(row_id);
}
// Bin each input data entry, store the bin indices in compressed form. // Bin each input data entry, store the bin indices in compressed form.
__global__ void CompressBinEllpackKernel( __global__ void CompressBinEllpackKernel(
@ -65,16 +61,18 @@ __global__ void CompressBinEllpackKernel(
} }
// Construct an ELLPACK matrix with the given number of empty rows. // Construct an ELLPACK matrix with the given number of empty rows.
EllpackPageImpl::EllpackPageImpl(int device, EllpackInfo info, size_t n_rows) { EllpackPageImpl::EllpackPageImpl(int device, common::HistogramCuts cuts,
bool is_dense, size_t row_stride,
size_t n_rows)
: is_dense(is_dense),
cuts_(std::move(cuts)),
row_stride(row_stride),
n_rows(n_rows) {
monitor_.Init("ellpack_page"); monitor_.Init("ellpack_page");
dh::safe_cuda(cudaSetDevice(device)); dh::safe_cuda(cudaSetDevice(device));
matrix.info = info;
matrix.base_rowid = 0;
matrix.n_rows = n_rows;
monitor_.StartCuda("InitCompressedData"); monitor_.StartCuda("InitCompressedData");
InitCompressedData(device, n_rows); InitCompressedData(device);
monitor_.StopCuda("InitCompressedData"); monitor_.StopCuda("InitCompressedData");
} }
@ -93,33 +91,27 @@ size_t GetRowStride(DMatrix* dmat) {
} }
// Construct an ELLPACK matrix in memory. // Construct an ELLPACK matrix in memory.
EllpackPageImpl::EllpackPageImpl(DMatrix* dmat, const BatchParam& param) { EllpackPageImpl::EllpackPageImpl(DMatrix* dmat, const BatchParam& param)
: is_dense(dmat->IsDense()) {
monitor_.Init("ellpack_page"); monitor_.Init("ellpack_page");
dh::safe_cuda(cudaSetDevice(param.gpu_id)); dh::safe_cuda(cudaSetDevice(param.gpu_id));
matrix.n_rows = dmat->Info().num_row_; n_rows = dmat->Info().num_row_;
monitor_.StartCuda("Quantiles"); monitor_.StartCuda("Quantiles");
// Create the quantile sketches for the dmatrix and initialize HistogramCuts. // Create the quantile sketches for the dmatrix and initialize HistogramCuts.
size_t row_stride = GetRowStride(dmat); row_stride = GetRowStride(dmat);
auto cuts = common::DeviceSketch(param.gpu_id, dmat, param.max_bin, cuts_ = common::DeviceSketch(param.gpu_id, dmat, param.max_bin,
param.gpu_batch_nrows); param.gpu_batch_nrows);
monitor_.StopCuda("Quantiles"); monitor_.StopCuda("Quantiles");
monitor_.StartCuda("InitEllpackInfo");
InitInfo(param.gpu_id, dmat->IsDense(), row_stride, cuts);
monitor_.StopCuda("InitEllpackInfo");
monitor_.StartCuda("InitCompressedData"); monitor_.StartCuda("InitCompressedData");
InitCompressedData(param.gpu_id, dmat->Info().num_row_); InitCompressedData(param.gpu_id);
monitor_.StopCuda("InitCompressedData"); monitor_.StopCuda("InitCompressedData");
monitor_.StartCuda("BinningCompression"); monitor_.StartCuda("BinningCompression");
DeviceHistogramBuilderState hist_builder_row_state(dmat->Info().num_row_);
for (const auto& batch : dmat->GetBatches<SparsePage>()) { for (const auto& batch : dmat->GetBatches<SparsePage>()) {
hist_builder_row_state.BeginBatch(batch); CreateHistIndices(param.gpu_id, batch);
CreateHistIndices(param.gpu_id, batch, hist_builder_row_state.GetRowStateOnDevice());
hist_builder_row_state.EndBatch();
} }
monitor_.StopCuda("BinningCompression"); monitor_.StopCuda("BinningCompression");
} }
@ -133,23 +125,26 @@ struct CopyPage {
size_t offset; size_t offset;
CopyPage(EllpackPageImpl* dst, EllpackPageImpl* src, size_t offset) CopyPage(EllpackPageImpl* dst, EllpackPageImpl* src, size_t offset)
: cbw{dst->matrix.info.NumSymbols()}, : cbw{dst->NumSymbols()},
dst_data_d{dst->gidx_buffer.data()}, dst_data_d{dst->gidx_buffer.DevicePointer()},
src_iterator_d{src->gidx_buffer.data(), src->matrix.info.NumSymbols()}, src_iterator_d{src->gidx_buffer.DevicePointer(), src->NumSymbols()},
offset(offset) {} offset(offset) {}
__device__ void operator()(size_t element_id) { __device__ void operator()(size_t element_id) {
cbw.AtomicWriteSymbol(dst_data_d, src_iterator_d[element_id], element_id + offset); cbw.AtomicWriteSymbol(dst_data_d, src_iterator_d[element_id],
element_id + offset);
} }
}; };
// Copy the data from the given EllpackPage to the current page. // Copy the data from the given EllpackPage to the current page.
size_t EllpackPageImpl::Copy(int device, EllpackPageImpl* page, size_t offset) { size_t EllpackPageImpl::Copy(int device, EllpackPageImpl* page, size_t offset) {
monitor_.StartCuda("Copy"); monitor_.StartCuda("Copy");
size_t num_elements = page->matrix.n_rows * page->matrix.info.row_stride; size_t num_elements = page->n_rows * page->row_stride;
CHECK_EQ(matrix.info.row_stride, page->matrix.info.row_stride); CHECK_EQ(row_stride, page->row_stride);
CHECK_EQ(matrix.info.NumSymbols(), page->matrix.info.NumSymbols()); CHECK_EQ(NumSymbols(), page->NumSymbols());
CHECK_GE(matrix.n_rows * matrix.info.row_stride, offset + num_elements); CHECK_GE(n_rows * row_stride, offset + num_elements);
gidx_buffer.SetDevice(device);
page->gidx_buffer.SetDevice(device);
dh::LaunchN(device, num_elements, CopyPage(this, page, offset)); dh::LaunchN(device, num_elements, CopyPage(this, page, offset));
monitor_.StopCuda("Copy"); monitor_.StopCuda("Copy");
return num_elements; return num_elements;
@ -160,26 +155,29 @@ struct CompactPage {
common::CompressedBufferWriter cbw; common::CompressedBufferWriter cbw;
common::CompressedByteT* dst_data_d; common::CompressedByteT* dst_data_d;
common::CompressedIterator<uint32_t> src_iterator_d; common::CompressedIterator<uint32_t> src_iterator_d;
/*! \brief An array that maps the rows from the full DMatrix to the compacted page. /*! \brief An array that maps the rows from the full DMatrix to the compacted
* page.
* *
* The total size is the number of rows in the original, uncompacted DMatrix. Elements are the * The total size is the number of rows in the original, uncompacted DMatrix.
* row ids in the compacted page. Rows not needed are set to SIZE_MAX. * Elements are the row ids in the compacted page. Rows not needed are set to
* SIZE_MAX.
* *
* An example compacting 16 rows to 8 rows: * An example compacting 16 rows to 8 rows:
* [SIZE_MAX, 0, 1, SIZE_MAX, SIZE_MAX, 2, SIZE_MAX, 3, 4, 5, SIZE_MAX, 6, SIZE_MAX, 7, SIZE_MAX, * [SIZE_MAX, 0, 1, SIZE_MAX, SIZE_MAX, 2, SIZE_MAX, 3, 4, 5, SIZE_MAX, 6,
* SIZE_MAX] * SIZE_MAX, 7, SIZE_MAX, SIZE_MAX]
*/ */
common::Span<size_t> row_indexes; common::Span<size_t> row_indexes;
size_t base_rowid; size_t base_rowid;
size_t row_stride; size_t row_stride;
CompactPage(EllpackPageImpl* dst, EllpackPageImpl* src, common::Span<size_t> row_indexes) CompactPage(EllpackPageImpl* dst, EllpackPageImpl* src,
: cbw{dst->matrix.info.NumSymbols()}, common::Span<size_t> row_indexes)
dst_data_d{dst->gidx_buffer.data()}, : cbw{dst->NumSymbols()},
src_iterator_d{src->gidx_buffer.data(), src->matrix.info.NumSymbols()}, dst_data_d{dst->gidx_buffer.DevicePointer()},
src_iterator_d{src->gidx_buffer.DevicePointer(), src->NumSymbols()},
row_indexes(row_indexes), row_indexes(row_indexes),
base_rowid{src->matrix.base_rowid}, base_rowid{src->base_rowid},
row_stride{src->matrix.info.row_stride} {} row_stride{src->row_stride} {}
__device__ void operator()(size_t row_id) { __device__ void operator()(size_t row_id) {
size_t src_row = base_rowid + row_id; size_t src_row = base_rowid + row_id;
@ -188,100 +186,72 @@ struct CompactPage {
size_t dst_offset = dst_row * row_stride; size_t dst_offset = dst_row * row_stride;
size_t src_offset = row_id * row_stride; size_t src_offset = row_id * row_stride;
for (size_t j = 0; j < row_stride; j++) { for (size_t j = 0; j < row_stride; j++) {
cbw.AtomicWriteSymbol(dst_data_d, src_iterator_d[src_offset + j], dst_offset + j); cbw.AtomicWriteSymbol(dst_data_d, src_iterator_d[src_offset + j],
dst_offset + j);
} }
} }
}; };
// Compacts the data from the given EllpackPage into the current page. // Compacts the data from the given EllpackPage into the current page.
void EllpackPageImpl::Compact(int device, EllpackPageImpl* page, common::Span<size_t> row_indexes) { void EllpackPageImpl::Compact(int device, EllpackPageImpl* page,
common::Span<size_t> row_indexes) {
monitor_.StartCuda("Compact"); monitor_.StartCuda("Compact");
CHECK_EQ(matrix.info.row_stride, page->matrix.info.row_stride); CHECK_EQ(row_stride, page->row_stride);
CHECK_EQ(matrix.info.NumSymbols(), page->matrix.info.NumSymbols()); CHECK_EQ(NumSymbols(), page->NumSymbols());
CHECK_LE(page->matrix.base_rowid + page->matrix.n_rows, row_indexes.size()); CHECK_LE(page->base_rowid + page->n_rows, row_indexes.size());
dh::LaunchN(device, page->matrix.n_rows, CompactPage(this, page, row_indexes)); gidx_buffer.SetDevice(device);
page->gidx_buffer.SetDevice(device);
dh::LaunchN(device, page->n_rows, CompactPage(this, page, row_indexes));
monitor_.StopCuda("Compact"); monitor_.StopCuda("Compact");
} }
// Construct an EllpackInfo based on histogram cuts of features.
EllpackInfo::EllpackInfo(int device,
bool is_dense,
size_t row_stride,
const common::HistogramCuts& hmat,
dh::BulkAllocator* ba)
: is_dense(is_dense), row_stride(row_stride), n_bins(hmat.Ptrs().back()) {
ba->Allocate(device,
&feature_segments, hmat.Ptrs().size(),
&gidx_fvalue_map, hmat.Values().size(),
&min_fvalue, hmat.MinValues().size());
dh::CopyVectorToDeviceSpan(gidx_fvalue_map, hmat.Values());
dh::CopyVectorToDeviceSpan(min_fvalue, hmat.MinValues());
dh::CopyVectorToDeviceSpan(feature_segments, hmat.Ptrs());
}
// Initialize the EllpackInfo for this page.
void EllpackPageImpl::InitInfo(int device,
bool is_dense,
size_t row_stride,
const common::HistogramCuts& hmat) {
matrix.info = EllpackInfo(device, is_dense, row_stride, hmat, &ba_);
}
// Initialize the buffer to stored compressed features. // Initialize the buffer to stored compressed features.
void EllpackPageImpl::InitCompressedData(int device, size_t num_rows) { void EllpackPageImpl::InitCompressedData(int device) {
size_t num_symbols = matrix.info.NumSymbols(); size_t num_symbols = NumSymbols();
// Required buffer size for storing data matrix in ELLPack format. // Required buffer size for storing data matrix in ELLPack format.
size_t compressed_size_bytes = common::CompressedBufferWriter::CalculateBufferSize( size_t compressed_size_bytes =
matrix.info.row_stride * num_rows, num_symbols); common::CompressedBufferWriter::CalculateBufferSize(row_stride * n_rows,
ba_.Allocate(device, &gidx_buffer, compressed_size_bytes); num_symbols);
gidx_buffer.SetDevice(device);
// Don't call fill unnecessarily
if (gidx_buffer.Size() == 0) {
gidx_buffer.Resize(compressed_size_bytes, 0);
} else {
gidx_buffer.Resize(compressed_size_bytes, 0);
thrust::fill(dh::tbegin(gidx_buffer), dh::tend(gidx_buffer), 0); thrust::fill(dh::tbegin(gidx_buffer), dh::tend(gidx_buffer), 0);
}
matrix.gidx_iter = common::CompressedIterator<uint32_t>(gidx_buffer.data(), num_symbols);
} }
// Compress a CSR page into ELLPACK. // Compress a CSR page into ELLPACK.
void EllpackPageImpl::CreateHistIndices(int device, void EllpackPageImpl::CreateHistIndices(int device,
const SparsePage& row_batch, const SparsePage& row_batch) {
const RowStateOnDevice& device_row_state) { if (row_batch.Size() == 0) return;
// Has any been allocated for me in this batch? unsigned int null_gidx_value = NumSymbols() - 1;
if (!device_row_state.rows_to_process_from_batch) return;
unsigned int null_gidx_value = matrix.info.n_bins;
size_t row_stride = matrix.info.row_stride;
const auto& offset_vec = row_batch.offset.ConstHostVector(); const auto& offset_vec = row_batch.offset.ConstHostVector();
// bin and compress entries in batches of rows // bin and compress entries in batches of rows
size_t gpu_batch_nrows = std::min( size_t gpu_batch_nrows =
dh::TotalMemory(device) / (16 * row_stride * sizeof(Entry)), std::min(dh::TotalMemory(device) / (16 * row_stride * sizeof(Entry)),
static_cast<size_t>(device_row_state.rows_to_process_from_batch)); static_cast<size_t>(row_batch.Size()));
const std::vector<Entry>& data_vec = row_batch.data.ConstHostVector(); const std::vector<Entry>& data_vec = row_batch.data.ConstHostVector();
size_t gpu_nbatches = common::DivRoundUp(device_row_state.rows_to_process_from_batch, size_t gpu_nbatches = common::DivRoundUp(row_batch.Size(), gpu_batch_nrows);
gpu_batch_nrows);
for (size_t gpu_batch = 0; gpu_batch < gpu_nbatches; ++gpu_batch) { for (size_t gpu_batch = 0; gpu_batch < gpu_nbatches; ++gpu_batch) {
size_t batch_row_begin = gpu_batch * gpu_batch_nrows; size_t batch_row_begin = gpu_batch * gpu_batch_nrows;
size_t batch_row_end = (gpu_batch + 1) * gpu_batch_nrows; size_t batch_row_end =
if (batch_row_end > device_row_state.rows_to_process_from_batch) { std::min((gpu_batch + 1) * gpu_batch_nrows, row_batch.Size());
batch_row_end = device_row_state.rows_to_process_from_batch;
}
size_t batch_nrows = batch_row_end - batch_row_begin; size_t batch_nrows = batch_row_end - batch_row_begin;
const auto ent_cnt_begin = const auto ent_cnt_begin = offset_vec[batch_row_begin];
offset_vec[device_row_state.row_offset_in_current_batch + batch_row_begin]; const auto ent_cnt_end = offset_vec[batch_row_end];
const auto ent_cnt_end =
offset_vec[device_row_state.row_offset_in_current_batch + batch_row_end];
/*! \brief row offset in SparsePage (the input data). */ /*! \brief row offset in SparsePage (the input data). */
dh::device_vector<size_t> row_ptrs(batch_nrows + 1); dh::device_vector<size_t> row_ptrs(batch_nrows + 1);
thrust::copy( thrust::copy(offset_vec.data() + batch_row_begin,
offset_vec.data() + device_row_state.row_offset_in_current_batch + batch_row_begin, offset_vec.data() + batch_row_end + 1, row_ptrs.begin());
offset_vec.data() + device_row_state.row_offset_in_current_batch + batch_row_end + 1,
row_ptrs.begin());
// number of entries in this batch. // number of entries in this batch.
size_t n_entries = ent_cnt_end - ent_cnt_begin; size_t n_entries = ent_cnt_end - ent_cnt_begin;
@ -289,97 +259,50 @@ void EllpackPageImpl::CreateHistIndices(int device,
// copy data entries to device. // copy data entries to device.
dh::safe_cuda(cudaMemcpy(entries_d.data().get(), dh::safe_cuda(cudaMemcpy(entries_d.data().get(),
data_vec.data() + ent_cnt_begin, data_vec.data() + ent_cnt_begin,
n_entries * sizeof(Entry), n_entries * sizeof(Entry), cudaMemcpyDefault));
cudaMemcpyDefault));
const dim3 block3(32, 8, 1); // 256 threads const dim3 block3(32, 8, 1); // 256 threads
const dim3 grid3(common::DivRoundUp(batch_nrows, block3.x), const dim3 grid3(common::DivRoundUp(batch_nrows, block3.x),
common::DivRoundUp(row_stride, block3.y), common::DivRoundUp(row_stride, block3.y), 1);
1); auto device_accessor = GetDeviceAccessor(device);
dh::LaunchKernel {grid3, block3}( dh::LaunchKernel {grid3, block3}(
CompressBinEllpackKernel, CompressBinEllpackKernel, common::CompressedBufferWriter(NumSymbols()),
common::CompressedBufferWriter(matrix.info.NumSymbols()), gidx_buffer.DevicePointer(), row_ptrs.data().get(),
gidx_buffer.data(), entries_d.data().get(), device_accessor.gidx_fvalue_map.data(),
row_ptrs.data().get(), device_accessor.feature_segments.data(),
entries_d.data().get(), row_batch.base_rowid + batch_row_begin, batch_nrows, row_stride,
matrix.info.gidx_fvalue_map.data(),
matrix.info.feature_segments.data(),
device_row_state.total_rows_processed + batch_row_begin,
batch_nrows,
row_stride,
null_gidx_value); null_gidx_value);
} }
} }
// Return the number of rows contained in this page. // Return the number of rows contained in this page.
size_t EllpackPageImpl::Size() const { size_t EllpackPageImpl::Size() const { return n_rows; }
return matrix.n_rows;
}
// Clear the current page.
void EllpackPageImpl::Clear() {
ba_.Clear();
gidx_buffer = {};
idx_buffer.clear();
sparse_page_.Clear();
matrix.base_rowid = 0;
matrix.n_rows = 0;
device_initialized_ = false;
}
// Push a CSR page to the current page.
//
// The CSR pages are accumulated in memory until they reach a certain size, then written out as
// compressed ELLPACK.
void EllpackPageImpl::Push(int device, const SparsePage& batch) {
sparse_page_.Push(batch);
matrix.n_rows += batch.Size();
}
// Compress the accumulated SparsePage.
void EllpackPageImpl::CompressSparsePage(int device) {
monitor_.StartCuda("InitCompressedData");
InitCompressedData(device, matrix.n_rows);
monitor_.StopCuda("InitCompressedData");
monitor_.StartCuda("BinningCompression");
DeviceHistogramBuilderState hist_builder_row_state(matrix.n_rows);
hist_builder_row_state.BeginBatch(sparse_page_);
CreateHistIndices(device, sparse_page_, hist_builder_row_state.GetRowStateOnDevice());
hist_builder_row_state.EndBatch();
monitor_.StopCuda("BinningCompression");
monitor_.StartCuda("CopyDeviceToHost");
idx_buffer.resize(gidx_buffer.size());
dh::CopyDeviceSpanToVector(&idx_buffer, gidx_buffer);
ba_.Clear();
gidx_buffer = {};
monitor_.StopCuda("CopyDeviceToHost");
}
// Return the memory cost for storing the compressed features. // Return the memory cost for storing the compressed features.
size_t EllpackPageImpl::MemCostBytes() const { size_t EllpackPageImpl::MemCostBytes(size_t num_rows, size_t row_stride,
// Required buffer size for storing data matrix in ELLPack format. const common::HistogramCuts& cuts) {
size_t compressed_size_bytes = common::CompressedBufferWriter::CalculateBufferSize( // Required buffer size for storing data matrix in EtoLLPack format.
matrix.info.row_stride * matrix.n_rows, matrix.info.NumSymbols()); size_t compressed_size_bytes =
common::CompressedBufferWriter::CalculateBufferSize(row_stride * num_rows,
cuts.TotalBins() + 1);
return compressed_size_bytes; return compressed_size_bytes;
} }
// Copy the compressed features to GPU. EllpackDeviceAccessor EllpackPageImpl::GetDeviceAccessor(int device) const {
void EllpackPageImpl::InitDevice(int device, EllpackInfo info) { gidx_buffer.SetDevice(device);
if (device_initialized_) return; return EllpackDeviceAccessor(
device, cuts_, is_dense, row_stride, base_rowid, n_rows,
common::CompressedIterator<uint32_t>(gidx_buffer.ConstDevicePointer(),
NumSymbols()));
}
monitor_.StartCuda("CopyPageToDevice"); EllpackPageImpl::EllpackPageImpl(int device, common::HistogramCuts cuts,
dh::safe_cuda(cudaSetDevice(device)); const SparsePage& page, bool is_dense,
size_t row_stride)
gidx_buffer = {}; : cuts_(std::move(cuts)),
ba_.Allocate(device, &gidx_buffer, idx_buffer.size()); is_dense(is_dense),
dh::CopyVectorToDeviceSpan(gidx_buffer, idx_buffer); n_rows(page.Size()),
row_stride(row_stride) {
matrix.info = info; this->InitCompressedData(device);
matrix.gidx_iter = common::CompressedIterator<uint32_t>(gidx_buffer.data(), info.n_bins + 1); this->CreateHistIndices(device, page);
monitor_.StopCuda("CopyPageToDevice");
device_initialized_ = true;
} }
} // namespace xgboost } // namespace xgboost

265
src/data/ellpack_page.cuh
View File

@ -40,71 +40,53 @@ __forceinline__ __device__ int BinarySearchRow(
return -1; return -1;
} }
/** \brief Meta information about the ELLPACK matrix. */ /** \brief Struct for accessing and manipulating an ellpack matrix on the
struct EllpackInfo { * device. Does not own underlying memory and may be trivially copied into
* kernels.*/
struct EllpackDeviceAccessor {
/*! \brief Whether or not if the matrix is dense. */ /*! \brief Whether or not if the matrix is dense. */
bool is_dense; bool is_dense;
/*! \brief Row length for ELLPack, equal to number of features. */ /*! \brief Row length for ELLPack, equal to number of features. */
size_t row_stride; size_t row_stride;
/*! \brief Total number of bins, also used as the null index value, . */
size_t n_bins;
/*! \brief Minimum value for each feature. Size equals to number of features. */
common::Span<bst_float> min_fvalue;
/*! \brief Histogram cut pointers. Size equals to (number of features + 1). */
common::Span<uint32_t> feature_segments;
/*! \brief Histogram cut values. Size equals to (bins per feature * number of features). */
common::Span<bst_float> gidx_fvalue_map;
EllpackInfo() = default;
/*!
* \brief Constructor.
*
* @param device The GPU device to use.
* @param is_dense Whether the matrix is dense.
* @param row_stride The number of features between starts of consecutive rows.
* @param hmat The histogram cuts of all the features.
* @param ba The BulkAllocator that owns the GPU memory.
*/
explicit EllpackInfo(int device,
bool is_dense,
size_t row_stride,
const common::HistogramCuts& hmat,
dh::BulkAllocator* ba);
/*! \brief Return the total number of symbols (total number of bins plus 1 for not found). */
size_t NumSymbols() const {
return n_bins + 1;
}
size_t NumFeatures() const {
return min_fvalue.size();
}
};
/** \brief Struct for accessing and manipulating an ellpack matrix on the
* device. Does not own underlying memory and may be trivially copied into
* kernels.*/
struct EllpackMatrix {
EllpackInfo info;
size_t base_rowid{}; size_t base_rowid{};
size_t n_rows{}; size_t n_rows{};
common::CompressedIterator<uint32_t> gidx_iter; common::CompressedIterator<uint32_t> gidx_iter;
/*! \brief Minimum value for each feature. Size equals to number of features. */
common::Span<const bst_float> min_fvalue;
/*! \brief Histogram cut pointers. Size equals to (number of features + 1). */
common::Span<const uint32_t> feature_segments;
/*! \brief Histogram cut values. Size equals to (bins per feature * number of features). */
common::Span<const bst_float> gidx_fvalue_map;
EllpackDeviceAccessor(int device, const common::HistogramCuts& cuts,
bool is_dense, size_t row_stride, size_t base_rowid,
size_t n_rows,common::CompressedIterator<uint32_t> gidx_iter)
: is_dense(is_dense),
row_stride(row_stride),
base_rowid(base_rowid),
n_rows(n_rows) ,gidx_iter(gidx_iter){
cuts.cut_values_.SetDevice(device);
cuts.cut_ptrs_.SetDevice(device);
cuts.min_vals_.SetDevice(device);
gidx_fvalue_map = cuts.cut_values_.ConstDeviceSpan();
feature_segments = cuts.cut_ptrs_.ConstDeviceSpan();
min_fvalue = cuts.min_vals_.ConstDeviceSpan();
}
// Get a matrix element, uses binary search for look up Return NaN if missing // Get a matrix element, uses binary search for look up Return NaN if missing
// Given a row index and a feature index, returns the corresponding cut value // Given a row index and a feature index, returns the corresponding cut value
__device__ int32_t GetBinIndex(size_t ridx, size_t fidx) const { __device__ int32_t GetBinIndex(size_t ridx, size_t fidx) const {
ridx -= base_rowid; ridx -= base_rowid;
auto row_begin = info.row_stride * ridx; auto row_begin = row_stride * ridx;
auto row_end = row_begin + info.row_stride; auto row_end = row_begin + row_stride;
auto gidx = -1; auto gidx = -1;
if (info.is_dense) { if (is_dense) {
gidx = gidx_iter[row_begin + fidx]; gidx = gidx_iter[row_begin + fidx];
} else { } else {
gidx = BinarySearchRow(row_begin, gidx = BinarySearchRow(row_begin,
row_end, row_end,
gidx_iter, gidx_iter,
info.feature_segments[fidx], feature_segments[fidx],
info.feature_segments[fidx + 1]); feature_segments[fidx + 1]);
} }
return gidx; return gidx;
} }
@ -113,97 +95,27 @@ struct EllpackMatrix {
if (gidx == -1) { if (gidx == -1) {
return nan(""); return nan("");
} }
return info.gidx_fvalue_map[gidx]; return gidx_fvalue_map[gidx];
} }
// Check if the row id is withing range of the current batch. // Check if the row id is withing range of the current batch.
__device__ bool IsInRange(size_t row_id) const { __device__ bool IsInRange(size_t row_id) const {
return row_id >= base_rowid && row_id < base_rowid + n_rows; return row_id >= base_rowid && row_id < base_rowid + n_rows;
} }
/*! \brief Return the total number of symbols (total number of bins plus 1 for
* not found). */
size_t NumSymbols() const { return gidx_fvalue_map.size() + 1; }
size_t NullValue() const { return gidx_fvalue_map.size(); }
XGBOOST_DEVICE size_t NumBins() const { return gidx_fvalue_map.size(); }
XGBOOST_DEVICE size_t NumFeatures() const { return min_fvalue.size(); }
}; };
// Instances of this type are created while creating the histogram bins for the
// entire dataset across multiple sparse page batches. This keeps track of the number
// of rows to process from a batch and the position from which to process on each device.
struct RowStateOnDevice {
// Number of rows assigned to this device
size_t total_rows_assigned_to_device;
// Number of rows processed thus far
size_t total_rows_processed;
// Number of rows to process from the current sparse page batch
size_t rows_to_process_from_batch;
// Offset from the current sparse page batch to begin processing
size_t row_offset_in_current_batch;
explicit RowStateOnDevice(size_t total_rows)
: total_rows_assigned_to_device(total_rows), total_rows_processed(0),
rows_to_process_from_batch(0), row_offset_in_current_batch(0) {
}
explicit RowStateOnDevice(size_t total_rows, size_t batch_rows)
: total_rows_assigned_to_device(total_rows), total_rows_processed(0),
rows_to_process_from_batch(batch_rows), row_offset_in_current_batch(0) {
}
// Advance the row state by the number of rows processed
void Advance() {
total_rows_processed += rows_to_process_from_batch;
CHECK_LE(total_rows_processed, total_rows_assigned_to_device);
rows_to_process_from_batch = row_offset_in_current_batch = 0;
}
};
// An instance of this type is created which keeps track of total number of rows to process,
// rows processed thus far, rows to process and the offset from the current sparse page batch
// to begin processing on each device
class DeviceHistogramBuilderState {
public:
explicit DeviceHistogramBuilderState(size_t n_rows) : device_row_state_(n_rows) {}
const RowStateOnDevice& GetRowStateOnDevice() const {
return device_row_state_;
}
// This method is invoked at the beginning of each sparse page batch. This distributes
// the rows in the sparse page to the device.
// TODO(sriramch): Think of a way to utilize *all* the GPUs to build the compressed bins.
void BeginBatch(const SparsePage &batch) {
size_t rem_rows = batch.Size();
size_t row_offset_in_current_batch = 0;
// Do we have anymore left to process from this batch on this device?
if (device_row_state_.total_rows_assigned_to_device > device_row_state_.total_rows_processed) {
// There are still some rows that needs to be assigned to this device
device_row_state_.rows_to_process_from_batch =
std::min(
device_row_state_.total_rows_assigned_to_device - device_row_state_.total_rows_processed,
rem_rows);
} else {
// All rows have been assigned to this device
device_row_state_.rows_to_process_from_batch = 0;
}
device_row_state_.row_offset_in_current_batch = row_offset_in_current_batch;
row_offset_in_current_batch += device_row_state_.rows_to_process_from_batch;
rem_rows -= device_row_state_.rows_to_process_from_batch;
}
// This method is invoked after completion of each sparse page batch
void EndBatch() {
device_row_state_.Advance();
}
private:
RowStateOnDevice device_row_state_{0};
};
class EllpackPageImpl { class EllpackPageImpl {
public: public:
EllpackMatrix matrix;
/*! \brief global index of histogram, which is stored in ELLPack format. */
common::Span<common::CompressedByteT> gidx_buffer;
std::vector<common::CompressedByteT> idx_buffer;
/*! /*!
* \brief Default constructor. * \brief Default constructor.
* *
@ -218,7 +130,12 @@ class EllpackPageImpl {
* This is used in the sampling case. The ELLPACK page is constructed from an existing EllpackInfo * This is used in the sampling case. The ELLPACK page is constructed from an existing EllpackInfo
* and the given number of rows. * and the given number of rows.
*/ */
explicit EllpackPageImpl(int device, EllpackInfo info, size_t n_rows); EllpackPageImpl(int device, common::HistogramCuts cuts, bool is_dense,
size_t row_stride, size_t n_rows);
EllpackPageImpl(int device, common::HistogramCuts cuts,
const SparsePage& page,
bool is_dense,size_t row_stride);
/*! /*!
* \brief Constructor from an existing DMatrix. * \brief Constructor from an existing DMatrix.
@ -245,77 +162,53 @@ class EllpackPageImpl {
*/ */
void Compact(int device, EllpackPageImpl* page, common::Span<size_t> row_indexes); void Compact(int device, EllpackPageImpl* page, common::Span<size_t> row_indexes);
/*!
* \brief Initialize the EllpackInfo contained in the EllpackMatrix.
*
* This is used in the in-memory case. The current page owns the BulkAllocator, which in turn owns
* the GPU memory used by the EllpackInfo.
*
* @param device The GPU device to use.
* @param is_dense Whether the matrix is dense.
* @param row_stride The number of features between starts of consecutive rows.
* @param hmat The histogram cuts of all the features.
*/
void InitInfo(int device, bool is_dense, size_t row_stride, const common::HistogramCuts& hmat);
/*!
* \brief Initialize the buffer to store compressed features.
*
* @param device The GPU device to use.
* @param num_rows The number of rows we are storing in the buffer.
*/
void InitCompressedData(int device, size_t num_rows);
/*!
* \brief Compress a single page of CSR data into ELLPACK.
*
* @param device The GPU device to use.
* @param row_batch The CSR page.
* @param device_row_state On-device data for maintaining state.
*/
void CreateHistIndices(int device,
const SparsePage& row_batch,
const RowStateOnDevice& device_row_state);
/*! \return Number of instances in the page. */ /*! \return Number of instances in the page. */
size_t Size() const; size_t Size() const;
/*! \brief Set the base row id for this page. */ /*! \brief Set the base row id for this page. */
inline void SetBaseRowId(size_t row_id) { void SetBaseRowId(size_t row_id) {
matrix.base_rowid = row_id; base_rowid = row_id;
} }
/*! \brief clear the page. */
void Clear();
/*!
* \brief Push a sparse page.
* \param batch The row page.
*/
void Push(int device, const SparsePage& batch);
/*! \return Estimation of memory cost of this page. */ /*! \return Estimation of memory cost of this page. */
size_t MemCostBytes() const; static size_t MemCostBytes(size_t num_rows, size_t row_stride, const common::HistogramCuts&cuts) ;
/*!
* \brief Copy the ELLPACK matrix to GPU.
*
* @param device The GPU device to use.
* @param info The EllpackInfo for the matrix.
*/
void InitDevice(int device, EllpackInfo info);
/*! \brief Compress the accumulated SparsePage into ELLPACK format. /*! \brief Return the total number of symbols (total number of bins plus 1 for
* * not found). */
* @param device The GPU device to use. size_t NumSymbols() const { return cuts_.TotalBins() + 1; }
*/
void CompressSparsePage(int device); EllpackDeviceAccessor GetDeviceAccessor(int device) const;
private:
/*!
* \brief Compress a single page of CSR data into ELLPACK.
*
* @param device The GPU device to use.
* @param row_batch The CSR page.
*/
void CreateHistIndices(int device,
const SparsePage& row_batch
);
/*!
* \brief Initialize the buffer to store compressed features.
*/
void InitCompressedData(int device);
public:
/*! \brief Whether or not if the matrix is dense. */
bool is_dense;
/*! \brief Row length for ELLPack. */
size_t row_stride;
size_t base_rowid{0};
size_t n_rows{};
/*! \brief global index of histogram, which is stored in ELLPack format. */
HostDeviceVector<common::CompressedByteT> gidx_buffer;
common::HistogramCuts cuts_;
private: private:
common::Monitor monitor_; common::Monitor monitor_;
dh::BulkAllocator ba_;
bool device_initialized_{false};
SparsePage sparse_page_{};
}; };
} // namespace xgboost } // namespace xgboost

31
src/data/ellpack_page_raw_format.cu
View File

@ -17,26 +17,35 @@ class EllpackPageRawFormat : public SparsePageFormat<EllpackPage> {
public: public:
bool Read(EllpackPage* page, dmlc::SeekStream* fi) override { bool Read(EllpackPage* page, dmlc::SeekStream* fi) override {
auto* impl = page->Impl(); auto* impl = page->Impl();
impl->Clear(); fi->Read(&impl->cuts_.cut_values_.HostVector());
if (!fi->Read(&impl->matrix.n_rows)) return false; fi->Read(&impl->cuts_.cut_ptrs_.HostVector());
return fi->Read(&impl->idx_buffer); fi->Read(&impl->cuts_.min_vals_.HostVector());
fi->Read(&impl->n_rows);
fi->Read(&impl->is_dense);
fi->Read(&impl->row_stride);
if (!fi->Read(&impl->gidx_buffer.HostVector())) {
return false;
}
return true;
} }
bool Read(EllpackPage* page, bool Read(EllpackPage* page,
dmlc::SeekStream* fi, dmlc::SeekStream* fi,
const std::vector<bst_uint>& sorted_index_set) override { const std::vector<bst_uint>& sorted_index_set) override {
auto* impl = page->Impl(); LOG(FATAL) << "Not implemented";
impl->Clear(); return false;
if (!fi->Read(&impl->matrix.n_rows)) return false;
return fi->Read(&page->Impl()->idx_buffer);
} }
void Write(const EllpackPage& page, dmlc::Stream* fo) override { void Write(const EllpackPage& page, dmlc::Stream* fo) override {
auto* impl = page.Impl(); auto* impl = page.Impl();
fo->Write(impl->matrix.n_rows); fo->Write(impl->cuts_.cut_values_.ConstHostVector());
auto buffer = impl->idx_buffer; fo->Write(impl->cuts_.cut_ptrs_.ConstHostVector());
CHECK(!buffer.empty()); fo->Write(impl->cuts_.min_vals_.ConstHostVector());
fo->Write(buffer); fo->Write(impl->n_rows);
fo->Write(impl->is_dense);
fo->Write(impl->row_stride);
CHECK(!impl->gidx_buffer.ConstHostVector().empty());
fo->Write(impl->gidx_buffer.HostVector());
} }
}; };

34
src/data/ellpack_page_source.cc
View File

@ -2,45 +2,23 @@
* Copyright 2019 XGBoost contributors * Copyright 2019 XGBoost contributors
*/ */
#ifndef XGBOOST_USE_CUDA #ifndef XGBOOST_USE_CUDA
#include <dmlc/base.h>
#if DMLC_ENABLE_STD_THREAD
#include <xgboost/data.h>
#include "ellpack_page_source.h" #include "ellpack_page_source.h"
#include <xgboost/data.h>
namespace xgboost { namespace xgboost {
namespace data { namespace data {
EllpackPageSource::EllpackPageSource(DMatrix* dmat, EllpackPageSource::EllpackPageSource(DMatrix* dmat,
const std::string& cache_info, const std::string& cache_info,
const BatchParam& param) noexcept(false) { const BatchParam& param) noexcept(false) {
LOG(FATAL) << "Internal Error: " LOG(FATAL)
<< "Internal Error: "
"XGBoost is not compiled with CUDA but EllpackPageSource is required"; "XGBoost is not compiled with CUDA but EllpackPageSource is required";
} }
void EllpackPageSource::BeforeFirst() {
LOG(FATAL) << "Internal Error: "
"XGBoost is not compiled with CUDA but EllpackPageSource is required";
}
bool EllpackPageSource::Next() {
LOG(FATAL) << "Internal Error: "
"XGBoost is not compiled with CUDA but EllpackPageSource is required";
return false;
}
EllpackPage& EllpackPageSource::Value() {
LOG(FATAL) << "Internal Error: "
"XGBoost is not compiled with CUDA but EllpackPageSource is required";
EllpackPage* page { nullptr };
return *page;
}
const EllpackPage& EllpackPageSource::Value() const {
LOG(FATAL) << "Internal Error: "
"XGBoost is not compiled with CUDA but EllpackPageSource is required";
EllpackPage* page { nullptr };
return *page;
}
} // namespace data } // namespace data
} // namespace xgboost } // namespace xgboost
#endif // DMLC_ENABLE_STD_THREAD
#endif // XGBOOST_USE_CUDA #endif // XGBOOST_USE_CUDA

132
src/data/ellpack_page_source.cu
View File

@ -3,73 +3,16 @@
*/ */
#include <memory> #include <memory>
#include <utility> #include <utility>
#include <vector>
#include "../common/hist_util.h" #include "../common/hist_util.h"
#include "ellpack_page.cuh"
#include "ellpack_page_source.h" #include "ellpack_page_source.h"
#include "sparse_page_source.h" #include "sparse_page_source.h"
#include "ellpack_page.cuh"
namespace xgboost { namespace xgboost {
namespace data { namespace data {
class EllpackPageSourceImpl : public DataSource<EllpackPage> {
public:
/*!
* \brief Create source from cache files the cache_prefix.
* \param cache_prefix The prefix of cache we want to solve.
*/
explicit EllpackPageSourceImpl(DMatrix* dmat,
const std::string& cache_info,
const BatchParam& param) noexcept(false);
/*! \brief destructor */
~EllpackPageSourceImpl() override = default;
void BeforeFirst() override;
bool Next() override;
EllpackPage& Value();
const EllpackPage& Value() const override;
private:
/*! \brief Write Ellpack pages after accumulating them in memory. */
void WriteEllpackPages(DMatrix* dmat, const std::string& cache_info) const;
/*! \brief The page type string for ELLPACK. */
const std::string kPageType_{".ellpack.page"};
int device_{-1};
size_t page_size_{DMatrix::kPageSize};
common::Monitor monitor_;
dh::BulkAllocator ba_;
/*! \brief The EllpackInfo, with the underlying GPU memory shared by all pages. */
EllpackInfo ellpack_info_;
std::unique_ptr<ExternalMemoryPrefetcher<EllpackPage>> source_;
std::string cache_info_;
};
EllpackPageSource::EllpackPageSource(DMatrix* dmat,
const std::string& cache_info,
const BatchParam& param) noexcept(false)
: impl_{new EllpackPageSourceImpl(dmat, cache_info, param)} {}
void EllpackPageSource::BeforeFirst() {
impl_->BeforeFirst();
}
bool EllpackPageSource::Next() {
return impl_->Next();
}
EllpackPage& EllpackPageSource::Value() {
return impl_->Value();
}
const EllpackPage& EllpackPageSource::Value() const {
return impl_->Value();
}
size_t GetRowStride(DMatrix* dmat) { size_t GetRowStride(DMatrix* dmat) {
if (dmat->IsDense()) return dmat->Info().num_col_; if (dmat->IsDense()) return dmat->Info().num_col_;
@ -86,17 +29,19 @@ size_t GetRowStride(DMatrix* dmat) {
// Build the quantile sketch across the whole input data, then use the histogram cuts to compress // Build the quantile sketch across the whole input data, then use the histogram cuts to compress
// each CSR page, and write the accumulated ELLPACK pages to disk. // each CSR page, and write the accumulated ELLPACK pages to disk.
EllpackPageSourceImpl::EllpackPageSourceImpl(DMatrix* dmat, EllpackPageSource::EllpackPageSource(DMatrix* dmat,
const std::string& cache_info, const std::string& cache_info,
const BatchParam& param) noexcept(false) const BatchParam& param) noexcept(false) {
: device_(param.gpu_id), cache_info_(cache_info) { cache_info_ = ParseCacheInfo(cache_info, kPageType_);
for (auto file : cache_info_.name_shards) {
CheckCacheFileExists(file);
}
if (param.gpu_page_size > 0) { if (param.gpu_page_size > 0) {
page_size_ = param.gpu_page_size; page_size_ = param.gpu_page_size;
} }
monitor_.Init("ellpack_page_source"); monitor_.Init("ellpack_page_source");
dh::safe_cuda(cudaSetDevice(device_)); dh::safe_cuda(cudaSetDevice(param.gpu_id));
monitor_.StartCuda("Quantiles"); monitor_.StartCuda("Quantiles");
size_t row_stride = GetRowStride(dmat); size_t row_stride = GetRowStride(dmat);
@ -104,75 +49,52 @@ EllpackPageSourceImpl::EllpackPageSourceImpl(DMatrix* dmat,
param.gpu_batch_nrows); param.gpu_batch_nrows);
monitor_.StopCuda("Quantiles"); monitor_.StopCuda("Quantiles");
monitor_.StartCuda("CreateEllpackInfo");
ellpack_info_ = EllpackInfo(device_, dmat->IsDense(), row_stride, cuts, &ba_);
monitor_.StopCuda("CreateEllpackInfo");
monitor_.StartCuda("WriteEllpackPages"); monitor_.StartCuda("WriteEllpackPages");
WriteEllpackPages(dmat, cache_info); WriteEllpackPages(param.gpu_id, dmat, cuts, cache_info, row_stride);
monitor_.StopCuda("WriteEllpackPages"); monitor_.StopCuda("WriteEllpackPages");
source_.reset(new ExternalMemoryPrefetcher<EllpackPage>( external_prefetcher_.reset(
ParseCacheInfo(cache_info_, kPageType_))); new ExternalMemoryPrefetcher<EllpackPage>(cache_info_));
}
void EllpackPageSourceImpl::BeforeFirst() {
source_.reset(new ExternalMemoryPrefetcher<EllpackPage>(
ParseCacheInfo(cache_info_, kPageType_)));
source_->BeforeFirst();
}
bool EllpackPageSourceImpl::Next() {
return source_->Next();
}
EllpackPage& EllpackPageSourceImpl::Value() {
EllpackPage& page = source_->Value();
page.Impl()->InitDevice(device_, ellpack_info_);
return page;
}
const EllpackPage& EllpackPageSourceImpl::Value() const {
EllpackPage& page = source_->Value();
page.Impl()->InitDevice(device_, ellpack_info_);
return page;
} }
// Compress each CSR page to ELLPACK, and write the accumulated pages to disk. // Compress each CSR page to ELLPACK, and write the accumulated pages to disk.
void EllpackPageSourceImpl::WriteEllpackPages(DMatrix* dmat, const std::string& cache_info) const { void EllpackPageSource::WriteEllpackPages(int device, DMatrix* dmat,
const common::HistogramCuts& cuts,
const std::string& cache_info,
size_t row_stride) const {
auto cinfo = ParseCacheInfo(cache_info, kPageType_); auto cinfo = ParseCacheInfo(cache_info, kPageType_);
const size_t extra_buffer_capacity = 6; const size_t extra_buffer_capacity = 6;
SparsePageWriter<EllpackPage> writer( SparsePageWriter<EllpackPage> writer(cinfo.name_shards, cinfo.format_shards,
cinfo.name_shards, cinfo.format_shards, extra_buffer_capacity); extra_buffer_capacity);
std::shared_ptr<EllpackPage> page; std::shared_ptr<EllpackPage> page;
SparsePage temp_host_page;
writer.Alloc(&page); writer.Alloc(&page);
auto* impl = page->Impl(); auto* impl = page->Impl();
impl->matrix.info = ellpack_info_;
impl->Clear();
const MetaInfo& info = dmat->Info();
size_t bytes_write = 0; size_t bytes_write = 0;
double tstart = dmlc::GetTime(); double tstart = dmlc::GetTime();
for (const auto& batch : dmat->GetBatches<SparsePage>()) { for (const auto& batch : dmat->GetBatches<SparsePage>()) {
impl->Push(device_, batch); temp_host_page.Push(batch);
size_t mem_cost_bytes = impl->MemCostBytes(); size_t mem_cost_bytes =
EllpackPageImpl::MemCostBytes(temp_host_page.Size(), row_stride, cuts);
if (mem_cost_bytes >= page_size_) { if (mem_cost_bytes >= page_size_) {
bytes_write += mem_cost_bytes; bytes_write += mem_cost_bytes;
impl->CompressSparsePage(device_); *impl = EllpackPageImpl(device, cuts, temp_host_page, dmat->IsDense(),
row_stride);
writer.PushWrite(std::move(page)); writer.PushWrite(std::move(page));
writer.Alloc(&page); writer.Alloc(&page);
impl = page->Impl(); impl = page->Impl();
impl->matrix.info = ellpack_info_; temp_host_page.Clear();
impl->Clear();
double tdiff = dmlc::GetTime() - tstart; double tdiff = dmlc::GetTime() - tstart;
LOG(INFO) << "Writing " << kPageType_ << " to " << cache_info << " in " LOG(INFO) << "Writing " << kPageType_ << " to " << cache_info << " in "
<< ((bytes_write >> 20UL) / tdiff) << " MB/s, " << ((bytes_write >> 20UL) / tdiff) << " MB/s, "
<< (bytes_write >> 20UL) << " written"; << (bytes_write >> 20UL) << " written";
} }
} }
if (impl->Size() != 0) { if (temp_host_page.Size() != 0) {
impl->CompressSparsePage(device_); *impl = EllpackPageImpl(device, cuts, temp_host_page, dmat->IsDense(),
row_stride);
writer.PushWrite(std::move(page)); writer.PushWrite(std::move(page));
} }
} }

42
src/data/ellpack_page_source.h
View File

@ -10,19 +10,17 @@
#include <string> #include <string>
#include "../common/timer.h" #include "../common/timer.h"
#include "../common/hist_util.h"
#include "sparse_page_source.h"
namespace xgboost { namespace xgboost {
namespace data { namespace data {
class EllpackPageSourceImpl;
/*! /*!
* \brief External memory data source for ELLPACK format. * \brief External memory data source for ELLPACK format.
* *
* This class uses the PImpl idiom (https://en.cppreference.com/w/cpp/language/pimpl) to avoid
* including CUDA-specific implementation details in the header.
*/ */
class EllpackPageSource : public DataSource<EllpackPage> { class EllpackPageSource {
public: public:
/*! /*!
* \brief Create source from cache files the cache_prefix. * \brief Create source from cache files the cache_prefix.
@ -32,19 +30,33 @@ class EllpackPageSource : public DataSource<EllpackPage> {
const std::string& cache_info, const std::string& cache_info,
const BatchParam& param) noexcept(false); const BatchParam& param) noexcept(false);
/*! \brief destructor */ BatchSet<EllpackPage> GetBatchSet() {
~EllpackPageSource() override = default; auto begin_iter = BatchIterator<EllpackPage>(
new SparseBatchIteratorImpl<ExternalMemoryPrefetcher<EllpackPage>,
EllpackPage>(external_prefetcher_.get()));
return BatchSet<EllpackPage>(begin_iter);
}
void BeforeFirst() override; ~EllpackPageSource() {
bool Next() override; external_prefetcher_.reset();
EllpackPage& Value(); for (auto file : cache_info_.name_shards) {
const EllpackPage& Value() const override; TryDeleteCacheFile(file);
}
const EllpackPageSourceImpl* Impl() const { return impl_.get(); } }
EllpackPageSourceImpl* Impl() { return impl_.get(); }
private: private:
std::shared_ptr<EllpackPageSourceImpl> impl_; void WriteEllpackPages(int device, DMatrix* dmat,
const common::HistogramCuts& cuts,
const std::string& cache_info,
size_t row_stride) const;
/*! \brief The page type string for ELLPACK. */
const std::string kPageType_{".ellpack.page"};
size_t page_size_{DMatrix::kPageSize};
common::Monitor monitor_;
std::unique_ptr<ExternalMemoryPrefetcher<EllpackPage>> external_prefetcher_;
CacheInfo cache_info_;
}; };
} // namespace data } // namespace data

6
src/data/sparse_page_dmatrix.cc
View File

@ -51,11 +51,7 @@ BatchSet<EllpackPage> SparsePageDMatrix::GetEllpackBatches(const BatchParam& par
ellpack_source_.reset(new EllpackPageSource(this, cache_info_, param)); ellpack_source_.reset(new EllpackPageSource(this, cache_info_, param));
batch_param_ = param; batch_param_ = param;
} }
ellpack_source_->BeforeFirst(); return ellpack_source_->GetBatchSet();
ellpack_source_->Next();
auto begin_iter = BatchIterator<EllpackPage>(
new SparseBatchIteratorImpl<EllpackPageSource, EllpackPage>(ellpack_source_.get()));
return BatchSet<EllpackPage>(begin_iter);
} }
} // namespace data } // namespace data

26
src/predictor/gpu_predictor.cu
View File

@ -97,9 +97,11 @@ struct SparsePageLoader {
}; };
struct EllpackLoader { struct EllpackLoader {
EllpackMatrix const& matrix; EllpackDeviceAccessor const& matrix;
XGBOOST_DEVICE EllpackLoader(EllpackMatrix const& m, bool use_shared, bst_feature_t num_features, XGBOOST_DEVICE EllpackLoader(EllpackDeviceAccessor const& m, bool use_shared,
bst_row_t num_rows, size_t entry_start) : matrix{m} {} bst_feature_t num_features, bst_row_t num_rows,
size_t entry_start)
: matrix{m} {}
__device__ __forceinline__ float GetFvalue(int ridx, int fidx) const { __device__ __forceinline__ float GetFvalue(int ridx, int fidx) const {
auto gidx = matrix.GetBinIndex(ridx, fidx); auto gidx = matrix.GetBinIndex(ridx, fidx);
if (gidx == -1) { if (gidx == -1) {
@ -107,10 +109,10 @@ struct EllpackLoader {
} }
// The gradient index needs to be shifted by one as min values are not included in the // The gradient index needs to be shifted by one as min values are not included in the
// cuts. // cuts.
if (gidx == matrix.info.feature_segments[fidx]) { if (gidx == matrix.feature_segments[fidx]) {
return matrix.info.min_fvalue[fidx]; return matrix.min_fvalue[fidx];
} }
return matrix.info.gidx_fvalue_map[gidx - 1]; return matrix.gidx_fvalue_map[gidx - 1];
} }
}; };
@ -217,7 +219,7 @@ class GPUPredictor : public xgboost::Predictor {
this->tree_begin_, this->tree_end_, num_features, num_rows, this->tree_begin_, this->tree_end_, num_features, num_rows,
entry_start, use_shared, this->num_group_); entry_start, use_shared, this->num_group_);
} }
void PredictInternal(EllpackMatrix const& batch, HostDeviceVector<bst_float>* out_preds, void PredictInternal(EllpackDeviceAccessor const& batch, HostDeviceVector<bst_float>* out_preds,
size_t batch_offset) { size_t batch_offset) {
const uint32_t BLOCK_THREADS = 256; const uint32_t BLOCK_THREADS = 256;
size_t num_rows = batch.n_rows; size_t num_rows = batch.n_rows;
@ -226,11 +228,11 @@ class GPUPredictor : public xgboost::Predictor {
bool use_shared = false; bool use_shared = false;
size_t entry_start = 0; size_t entry_start = 0;
dh::LaunchKernel {GRID_SIZE, BLOCK_THREADS} ( dh::LaunchKernel {GRID_SIZE, BLOCK_THREADS} (
PredictKernel<EllpackLoader, EllpackMatrix>, PredictKernel<EllpackLoader, EllpackDeviceAccessor>,
batch, batch,
dh::ToSpan(nodes_), out_preds->DeviceSpan().subspan(batch_offset), dh::ToSpan(nodes_), out_preds->DeviceSpan().subspan(batch_offset),
dh::ToSpan(tree_segments_), dh::ToSpan(tree_group_), dh::ToSpan(tree_segments_), dh::ToSpan(tree_group_),
this->tree_begin_, this->tree_end_, batch.info.NumFeatures(), num_rows, this->tree_begin_, this->tree_end_, batch.NumFeatures(), num_rows,
entry_start, use_shared, this->num_group_); entry_start, use_shared, this->num_group_);
} }
@ -269,8 +271,10 @@ class GPUPredictor : public xgboost::Predictor {
if (dmat->PageExists<EllpackPage>()) { if (dmat->PageExists<EllpackPage>()) {
size_t batch_offset = 0; size_t batch_offset = 0;
for (auto const& page : dmat->GetBatches<EllpackPage>()) { for (auto const& page : dmat->GetBatches<EllpackPage>()) {
this->PredictInternal(page.Impl()->matrix, out_preds, batch_offset); this->PredictInternal(
batch_offset += page.Impl()->matrix.n_rows; page.Impl()->GetDeviceAccessor(generic_param_->gpu_id), out_preds,
batch_offset);
batch_offset += page.Impl()->n_rows;
} }
} else { } else {
size_t batch_offset = 0; size_t batch_offset = 0;

17
src/tree/gpu_hist/gradient_based_sampler.cu
View File

@ -153,7 +153,8 @@ ExternalMemoryNoSampling::ExternalMemoryNoSampling(EllpackPageImpl* page,
size_t n_rows, size_t n_rows,
const BatchParam& batch_param) const BatchParam& batch_param)
: batch_param_(batch_param), : batch_param_(batch_param),
page_(new EllpackPageImpl(batch_param.gpu_id, page->matrix.info, n_rows)) {} page_(new EllpackPageImpl(batch_param.gpu_id, page->cuts_, page->is_dense,
page->row_stride, n_rows)) {}
GradientBasedSample ExternalMemoryNoSampling::Sample(common::Span<GradientPair> gpair, GradientBasedSample ExternalMemoryNoSampling::Sample(common::Span<GradientPair> gpair,
DMatrix* dmat) { DMatrix* dmat) {
@ -217,9 +218,9 @@ GradientBasedSample ExternalMemoryUniformSampling::Sample(common::Span<GradientP
// Create a new ELLPACK page with empty rows. // Create a new ELLPACK page with empty rows.
page_.reset(); // Release the device memory first before reallocating page_.reset(); // Release the device memory first before reallocating
page_.reset(new EllpackPageImpl(batch_param_.gpu_id, page_.reset(new EllpackPageImpl(
original_page_->matrix.info, batch_param_.gpu_id, original_page_->cuts_, original_page_->is_dense,
sample_rows)); original_page_->row_stride, sample_rows));
// Compact the ELLPACK pages into the single sample page. // Compact the ELLPACK pages into the single sample page.
thrust::fill(dh::tbegin(page_->gidx_buffer), dh::tend(page_->gidx_buffer), 0); thrust::fill(dh::tbegin(page_->gidx_buffer), dh::tend(page_->gidx_buffer), 0);
@ -298,9 +299,9 @@ GradientBasedSample ExternalMemoryGradientBasedSampling::Sample(common::Span<Gra
// Create a new ELLPACK page with empty rows. // Create a new ELLPACK page with empty rows.
page_.reset(); // Release the device memory first before reallocating page_.reset(); // Release the device memory first before reallocating
page_.reset(new EllpackPageImpl(batch_param_.gpu_id, page_.reset(new EllpackPageImpl(batch_param_.gpu_id, original_page_->cuts_,
original_page_->matrix.info, original_page_->is_dense,
sample_rows)); original_page_->row_stride, sample_rows));
// Compact the ELLPACK pages into the single sample page. // Compact the ELLPACK pages into the single sample page.
thrust::fill(dh::tbegin(page_->gidx_buffer), dh::tend(page_->gidx_buffer), 0); thrust::fill(dh::tbegin(page_->gidx_buffer), dh::tend(page_->gidx_buffer), 0);
@ -319,7 +320,7 @@ GradientBasedSampler::GradientBasedSampler(EllpackPageImpl* page,
monitor_.Init("gradient_based_sampler"); monitor_.Init("gradient_based_sampler");
bool is_sampling = subsample < 1.0; bool is_sampling = subsample < 1.0;
bool is_external_memory = page->matrix.n_rows != n_rows; bool is_external_memory = page->n_rows != n_rows;
if (is_sampling) { if (is_sampling) {
switch (sampling_method) { switch (sampling_method) {

22
src/tree/gpu_hist/histogram.cu
View File

@ -101,7 +101,7 @@ template GradientPairPrecise CreateRoundingFactor(common::Span<GradientPair cons
template GradientPair CreateRoundingFactor(common::Span<GradientPair const> gpair); template GradientPair CreateRoundingFactor(common::Span<GradientPair const> gpair);
template <typename GradientSumT> template <typename GradientSumT>
__global__ void SharedMemHistKernel(xgboost::EllpackMatrix matrix, __global__ void SharedMemHistKernel(EllpackDeviceAccessor matrix,
common::Span<const RowPartitioner::RowIndexT> d_ridx, common::Span<const RowPartitioner::RowIndexT> d_ridx,
GradientSumT* __restrict__ d_node_hist, GradientSumT* __restrict__ d_node_hist,
const GradientPair* __restrict__ d_gpair, const GradientPair* __restrict__ d_gpair,
@ -112,14 +112,14 @@ __global__ void SharedMemHistKernel(xgboost::EllpackMatrix matrix,
extern __shared__ char smem[]; extern __shared__ char smem[];
GradientSumT* smem_arr = reinterpret_cast<GradientSumT*>(smem); // NOLINT GradientSumT* smem_arr = reinterpret_cast<GradientSumT*>(smem); // NOLINT
if (use_shared_memory_histograms) { if (use_shared_memory_histograms) {
dh::BlockFill(smem_arr, matrix.info.n_bins, GradientSumT()); dh::BlockFill(smem_arr, matrix.NumBins(), GradientSumT());
__syncthreads(); __syncthreads();
} }
for (auto idx : dh::GridStrideRange(static_cast<size_t>(0), n_elements)) { for (auto idx : dh::GridStrideRange(static_cast<size_t>(0), n_elements)) {
int ridx = d_ridx[idx / matrix.info.row_stride]; int ridx = d_ridx[idx / matrix.row_stride];
int gidx = int gidx =
matrix.gidx_iter[ridx * matrix.info.row_stride + idx % matrix.info.row_stride]; matrix.gidx_iter[ridx * matrix.row_stride + idx % matrix.row_stride];
if (gidx != matrix.info.n_bins) { if (gidx != matrix.NumBins()) {
GradientSumT truncated { GradientSumT truncated {
TruncateWithRoundingFactor<T>(rounding.GetGrad(), d_gpair[ridx].GetGrad()), TruncateWithRoundingFactor<T>(rounding.GetGrad(), d_gpair[ridx].GetGrad()),
TruncateWithRoundingFactor<T>(rounding.GetHess(), d_gpair[ridx].GetHess()), TruncateWithRoundingFactor<T>(rounding.GetHess(), d_gpair[ridx].GetHess()),
@ -135,7 +135,7 @@ __global__ void SharedMemHistKernel(xgboost::EllpackMatrix matrix,
if (use_shared_memory_histograms) { if (use_shared_memory_histograms) {
// Write shared memory back to global memory // Write shared memory back to global memory
__syncthreads(); __syncthreads();
for (auto i : dh::BlockStrideRange(static_cast<size_t>(0), matrix.info.n_bins)) { for (auto i : dh::BlockStrideRange(static_cast<size_t>(0), matrix.NumBins())) {
GradientSumT truncated { GradientSumT truncated {
TruncateWithRoundingFactor<T>(rounding.GetGrad(), smem_arr[i].GetGrad()), TruncateWithRoundingFactor<T>(rounding.GetGrad(), smem_arr[i].GetGrad()),
TruncateWithRoundingFactor<T>(rounding.GetHess(), smem_arr[i].GetHess()), TruncateWithRoundingFactor<T>(rounding.GetHess(), smem_arr[i].GetHess()),
@ -146,16 +146,16 @@ __global__ void SharedMemHistKernel(xgboost::EllpackMatrix matrix,
} }
template <typename GradientSumT> template <typename GradientSumT>
void BuildGradientHistogram(EllpackMatrix const& matrix, void BuildGradientHistogram(EllpackDeviceAccessor const& matrix,
common::Span<GradientPair const> gpair, common::Span<GradientPair const> gpair,
common::Span<const uint32_t> d_ridx, common::Span<const uint32_t> d_ridx,
common::Span<GradientSumT> histogram, common::Span<GradientSumT> histogram,
GradientSumT rounding, bool shared) { GradientSumT rounding, bool shared) {
const size_t smem_size = const size_t smem_size =
shared shared
? sizeof(GradientSumT) * matrix.info.n_bins ? sizeof(GradientSumT) * matrix.NumBins()
: 0; : 0;
auto n_elements = d_ridx.size() * matrix.info.row_stride; auto n_elements = d_ridx.size() * matrix.row_stride;
uint32_t items_per_thread = 8; uint32_t items_per_thread = 8;
uint32_t block_threads = 256; uint32_t block_threads = 256;
@ -168,14 +168,14 @@ void BuildGradientHistogram(EllpackMatrix const& matrix,
} }
template void BuildGradientHistogram<GradientPair>( template void BuildGradientHistogram<GradientPair>(
EllpackMatrix const& matrix, EllpackDeviceAccessor const& matrix,
common::Span<GradientPair const> gpair, common::Span<GradientPair const> gpair,
common::Span<const uint32_t> ridx, common::Span<const uint32_t> ridx,
common::Span<GradientPair> histogram, common::Span<GradientPair> histogram,
GradientPair rounding, bool shared); GradientPair rounding, bool shared);
template void BuildGradientHistogram<GradientPairPrecise>( template void BuildGradientHistogram<GradientPairPrecise>(
EllpackMatrix const& matrix, EllpackDeviceAccessor const& matrix,
common::Span<GradientPair const> gpair, common::Span<GradientPair const> gpair,
common::Span<const uint32_t> ridx, common::Span<const uint32_t> ridx,
common::Span<GradientPairPrecise> histogram, common::Span<GradientPairPrecise> histogram,

2
src/tree/gpu_hist/histogram.cuh
View File

@ -18,7 +18,7 @@ DEV_INLINE T TruncateWithRoundingFactor(T const rounding_factor, float const x)
} }
template <typename GradientSumT> template <typename GradientSumT>
void BuildGradientHistogram(EllpackMatrix const& matrix, void BuildGradientHistogram(EllpackDeviceAccessor const& matrix,
common::Span<GradientPair const> gpair, common::Span<GradientPair const> gpair,
common::Span<const uint32_t> ridx, common::Span<const uint32_t> ridx,
common::Span<GradientSumT> histogram, common::Span<GradientSumT> histogram,

32
src/tree/updater_gpu_hist.cu
View File

@ -180,15 +180,15 @@ template <int BLOCK_THREADS, typename ReduceT, typename ScanT,
typename MaxReduceT, typename TempStorageT, typename GradientSumT> typename MaxReduceT, typename TempStorageT, typename GradientSumT>
__device__ void EvaluateFeature( __device__ void EvaluateFeature(
int fidx, common::Span<const GradientSumT> node_histogram, int fidx, common::Span<const GradientSumT> node_histogram,
const xgboost::EllpackMatrix& matrix, const EllpackDeviceAccessor& matrix,
DeviceSplitCandidate* best_split, // shared memory storing best split DeviceSplitCandidate* best_split, // shared memory storing best split
const DeviceNodeStats& node, const GPUTrainingParam& param, const DeviceNodeStats& node, const GPUTrainingParam& param,
TempStorageT* temp_storage, // temp memory for cub operations TempStorageT* temp_storage, // temp memory for cub operations
int constraint, // monotonic_constraints int constraint, // monotonic_constraints
const ValueConstraint& value_constraint) { const ValueConstraint& value_constraint) {
// Use pointer from cut to indicate begin and end of bins for each feature. // Use pointer from cut to indicate begin and end of bins for each feature.
uint32_t gidx_begin = matrix.info.feature_segments[fidx]; // begining bin uint32_t gidx_begin = matrix.feature_segments[fidx]; // begining bin
uint32_t gidx_end = matrix.info.feature_segments[fidx + 1]; // end bin for i^th feature uint32_t gidx_end = matrix.feature_segments[fidx + 1]; // end bin for i^th feature
// Sum histogram bins for current feature // Sum histogram bins for current feature
GradientSumT const feature_sum = ReduceFeature<BLOCK_THREADS, ReduceT>( GradientSumT const feature_sum = ReduceFeature<BLOCK_THREADS, ReduceT>(
@ -236,9 +236,9 @@ __device__ void EvaluateFeature(
int split_gidx = (scan_begin + threadIdx.x) - 1; int split_gidx = (scan_begin + threadIdx.x) - 1;
float fvalue; float fvalue;
if (split_gidx < static_cast<int>(gidx_begin)) { if (split_gidx < static_cast<int>(gidx_begin)) {
fvalue = matrix.info.min_fvalue[fidx]; fvalue = matrix.min_fvalue[fidx];
} else { } else {
fvalue = matrix.info.gidx_fvalue_map[split_gidx]; fvalue = matrix.gidx_fvalue_map[split_gidx];
} }
GradientSumT left = missing_left ? bin + missing : bin; GradientSumT left = missing_left ? bin + missing : bin;
GradientSumT right = parent_sum - left; GradientSumT right = parent_sum - left;
@ -254,7 +254,7 @@ __global__ void EvaluateSplitKernel(
common::Span<const GradientSumT> node_histogram, // histogram for gradients common::Span<const GradientSumT> node_histogram, // histogram for gradients
common::Span<const bst_feature_t> feature_set, // Selected features common::Span<const bst_feature_t> feature_set, // Selected features
DeviceNodeStats node, DeviceNodeStats node,
xgboost::EllpackMatrix matrix, xgboost::EllpackDeviceAccessor matrix,
GPUTrainingParam gpu_param, GPUTrainingParam gpu_param,
common::Span<DeviceSplitCandidate> split_candidates, // resulting split common::Span<DeviceSplitCandidate> split_candidates, // resulting split
ValueConstraint value_constraint, ValueConstraint value_constraint,
@ -601,7 +601,7 @@ struct GPUHistMakerDevice {
uint32_t constexpr kBlockThreads = 256; uint32_t constexpr kBlockThreads = 256;
dh::LaunchKernel {uint32_t(d_feature_set.size()), kBlockThreads, 0, streams[i]} ( dh::LaunchKernel {uint32_t(d_feature_set.size()), kBlockThreads, 0, streams[i]} (
EvaluateSplitKernel<kBlockThreads, GradientSumT>, EvaluateSplitKernel<kBlockThreads, GradientSumT>,
hist.GetNodeHistogram(nidx), d_feature_set, node, page->matrix, hist.GetNodeHistogram(nidx), d_feature_set, node, page->GetDeviceAccessor(device_id),
gpu_param, d_split_candidates, node_value_constraints[nidx], gpu_param, d_split_candidates, node_value_constraints[nidx],
monotone_constraints); monotone_constraints);
@ -625,9 +625,7 @@ struct GPUHistMakerDevice {
hist.AllocateHistogram(nidx); hist.AllocateHistogram(nidx);
auto d_node_hist = hist.GetNodeHistogram(nidx); auto d_node_hist = hist.GetNodeHistogram(nidx);
auto d_ridx = row_partitioner->GetRows(nidx); auto d_ridx = row_partitioner->GetRows(nidx);
auto d_gpair = gpair.data(); BuildGradientHistogram(page->GetDeviceAccessor(device_id), gpair, d_ridx, d_node_hist,
BuildGradientHistogram(page->matrix, gpair, d_ridx, d_node_hist,
histogram_rounding, use_shared_memory_histograms); histogram_rounding, use_shared_memory_histograms);
} }
@ -637,7 +635,7 @@ struct GPUHistMakerDevice {
auto d_node_hist_histogram = hist.GetNodeHistogram(nidx_histogram); auto d_node_hist_histogram = hist.GetNodeHistogram(nidx_histogram);
auto d_node_hist_subtraction = hist.GetNodeHistogram(nidx_subtraction); auto d_node_hist_subtraction = hist.GetNodeHistogram(nidx_subtraction);
dh::LaunchN(device_id, page->matrix.info.n_bins, [=] __device__(size_t idx) { dh::LaunchN(device_id, page->cuts_.TotalBins(), [=] __device__(size_t idx) {
d_node_hist_subtraction[idx] = d_node_hist_subtraction[idx] =
d_node_hist_parent[idx] - d_node_hist_histogram[idx]; d_node_hist_parent[idx] - d_node_hist_histogram[idx];
}); });
@ -652,7 +650,7 @@ struct GPUHistMakerDevice {
} }
void UpdatePosition(int nidx, RegTree::Node split_node) { void UpdatePosition(int nidx, RegTree::Node split_node) {
auto d_matrix = page->matrix; auto d_matrix = page->GetDeviceAccessor(device_id);
row_partitioner->UpdatePosition( row_partitioner->UpdatePosition(
nidx, split_node.LeftChild(), split_node.RightChild(), nidx, split_node.LeftChild(), split_node.RightChild(),
@ -689,7 +687,7 @@ struct GPUHistMakerDevice {
row_partitioner.reset(); // Release the device memory first before reallocating row_partitioner.reset(); // Release the device memory first before reallocating
row_partitioner.reset(new RowPartitioner(device_id, p_fmat->Info().num_row_)); row_partitioner.reset(new RowPartitioner(device_id, p_fmat->Info().num_row_));
} }
if (page->matrix.n_rows == p_fmat->Info().num_row_) { if (page->n_rows == p_fmat->Info().num_row_) {
FinalisePositionInPage(page, d_nodes); FinalisePositionInPage(page, d_nodes);
} else { } else {
for (auto& batch : p_fmat->GetBatches<EllpackPage>(batch_param)) { for (auto& batch : p_fmat->GetBatches<EllpackPage>(batch_param)) {
@ -699,7 +697,7 @@ struct GPUHistMakerDevice {
} }
void FinalisePositionInPage(EllpackPageImpl* page, const common::Span<RegTree::Node> d_nodes) { void FinalisePositionInPage(EllpackPageImpl* page, const common::Span<RegTree::Node> d_nodes) {
auto d_matrix = page->matrix; auto d_matrix = page->GetDeviceAccessor(device_id);
row_partitioner->FinalisePosition( row_partitioner->FinalisePosition(
[=] __device__(size_t row_id, int position) { [=] __device__(size_t row_id, int position) {
if (!d_matrix.IsInRange(row_id)) { if (!d_matrix.IsInRange(row_id)) {
@ -765,7 +763,7 @@ struct GPUHistMakerDevice {
reducer->AllReduceSum( reducer->AllReduceSum(
reinterpret_cast<typename GradientSumT::ValueT*>(d_node_hist), reinterpret_cast<typename GradientSumT::ValueT*>(d_node_hist),
reinterpret_cast<typename GradientSumT::ValueT*>(d_node_hist), reinterpret_cast<typename GradientSumT::ValueT*>(d_node_hist),
page->matrix.info.n_bins * (sizeof(GradientSumT) / sizeof(typename GradientSumT::ValueT))); page->cuts_.TotalBins() * (sizeof(GradientSumT) / sizeof(typename GradientSumT::ValueT)));
reducer->Synchronize(); reducer->Synchronize();
monitor.StopCuda("AllReduce"); monitor.StopCuda("AllReduce");
@ -954,14 +952,14 @@ inline void GPUHistMakerDevice<GradientSumT>::InitHistogram() {
// check if we can use shared memory for building histograms // check if we can use shared memory for building histograms
// (assuming atleast we need 2 CTAs per SM to maintain decent latency // (assuming atleast we need 2 CTAs per SM to maintain decent latency
// hiding) // hiding)
auto histogram_size = sizeof(GradientSumT) * page->matrix.info.n_bins; auto histogram_size = sizeof(GradientSumT) * page->cuts_.TotalBins();
auto max_smem = dh::MaxSharedMemory(device_id); auto max_smem = dh::MaxSharedMemory(device_id);
if (histogram_size <= max_smem) { if (histogram_size <= max_smem) {
use_shared_memory_histograms = true; use_shared_memory_histograms = true;
} }
// Init histogram // Init histogram
hist.Init(device_id, page->matrix.info.n_bins); hist.Init(device_id, page->cuts_.TotalBins());
} }
template <typename GradientSumT> template <typename GradientSumT>

46
tests/cpp/data/test_ellpack_page.cu
View File

@ -19,23 +19,19 @@ TEST(EllpackPage, EmptyDMatrix) {
auto dmat = *CreateDMatrix(kNRows, kNCols, kSparsity); auto dmat = *CreateDMatrix(kNRows, kNCols, kSparsity);
auto& page = *dmat->GetBatches<EllpackPage>({0, kMaxBin, kGpuBatchNRows}).begin(); auto& page = *dmat->GetBatches<EllpackPage>({0, kMaxBin, kGpuBatchNRows}).begin();
auto impl = page.Impl(); auto impl = page.Impl();
ASSERT_EQ(impl->matrix.info.feature_segments.size(), 1); ASSERT_EQ(impl->row_stride, 0);
ASSERT_EQ(impl->matrix.info.min_fvalue.size(), 0); ASSERT_EQ(impl->cuts_.TotalBins(), 0);
ASSERT_EQ(impl->matrix.info.gidx_fvalue_map.size(), 0); ASSERT_EQ(impl->gidx_buffer.Size(), 4);
ASSERT_EQ(impl->matrix.info.row_stride, 0);
ASSERT_EQ(impl->matrix.info.n_bins, 0);
ASSERT_EQ(impl->gidx_buffer.size(), 4);
} }
TEST(EllpackPage, BuildGidxDense) { TEST(EllpackPage, BuildGidxDense) {
int constexpr kNRows = 16, kNCols = 8; int constexpr kNRows = 16, kNCols = 8;
auto page = BuildEllpackPage(kNRows, kNCols); auto page = BuildEllpackPage(kNRows, kNCols);
std::vector<common::CompressedByteT> h_gidx_buffer(page->gidx_buffer.size()); std::vector<common::CompressedByteT> h_gidx_buffer(page->gidx_buffer.HostVector());
dh::CopyDeviceSpanToVector(&h_gidx_buffer, page->gidx_buffer); common::CompressedIterator<uint32_t> gidx(h_gidx_buffer.data(), page->NumSymbols());
common::CompressedIterator<uint32_t> gidx(h_gidx_buffer.data(), 25);
ASSERT_EQ(page->matrix.info.row_stride, kNCols); ASSERT_EQ(page->row_stride, kNCols);
std::vector<uint32_t> solution = { std::vector<uint32_t> solution = {
0, 3, 8, 9, 14, 17, 20, 21, 0, 3, 8, 9, 14, 17, 20, 21,
@ -64,11 +60,10 @@ TEST(EllpackPage, BuildGidxSparse) {
int constexpr kNRows = 16, kNCols = 8; int constexpr kNRows = 16, kNCols = 8;
auto page = BuildEllpackPage(kNRows, kNCols, 0.9f); auto page = BuildEllpackPage(kNRows, kNCols, 0.9f);
std::vector<common::CompressedByteT> h_gidx_buffer(page->gidx_buffer.size()); std::vector<common::CompressedByteT> h_gidx_buffer(page->gidx_buffer.HostVector());
dh::CopyDeviceSpanToVector(&h_gidx_buffer, page->gidx_buffer);
common::CompressedIterator<uint32_t> gidx(h_gidx_buffer.data(), 25); common::CompressedIterator<uint32_t> gidx(h_gidx_buffer.data(), 25);
ASSERT_LE(page->matrix.info.row_stride, 3); ASSERT_LE(page->row_stride, 3);
// row_stride = 3, 16 rows, 48 entries for ELLPack // row_stride = 3, 16 rows, 48 entries for ELLPack
std::vector<uint32_t> solution = { std::vector<uint32_t> solution = {
@ -76,16 +71,16 @@ TEST(EllpackPage, BuildGidxSparse) {
24, 24, 24, 24, 24, 5, 24, 24, 0, 16, 24, 15, 24, 24, 24, 24, 24, 24, 24, 24, 24, 5, 24, 24, 0, 16, 24, 15, 24, 24, 24, 24,
24, 7, 14, 16, 4, 24, 24, 24, 24, 24, 9, 24, 24, 1, 24, 24 24, 7, 14, 16, 4, 24, 24, 24, 24, 24, 9, 24, 24, 1, 24, 24
}; };
for (size_t i = 0; i < kNRows * page->matrix.info.row_stride; ++i) { for (size_t i = 0; i < kNRows * page->row_stride; ++i) {
ASSERT_EQ(solution[i], gidx[i]); ASSERT_EQ(solution[i], gidx[i]);
} }
} }
struct ReadRowFunction { struct ReadRowFunction {
EllpackMatrix matrix; EllpackDeviceAccessor matrix;
int row; int row;
bst_float* row_data_d; bst_float* row_data_d;
ReadRowFunction(EllpackMatrix matrix, int row, bst_float* row_data_d) ReadRowFunction(EllpackDeviceAccessor matrix, int row, bst_float* row_data_d)
: matrix(std::move(matrix)), row(row), row_data_d(row_data_d) {} : matrix(std::move(matrix)), row(row), row_data_d(row_data_d) {}
__device__ void operator()(size_t col) { __device__ void operator()(size_t col) {
@ -110,7 +105,8 @@ TEST(EllpackPage, Copy) {
auto page = (*dmat->GetBatches<EllpackPage>(param).begin()).Impl(); auto page = (*dmat->GetBatches<EllpackPage>(param).begin()).Impl();
// Create an empty result page. // Create an empty result page.
EllpackPageImpl result(0, page->matrix.info, kRows); EllpackPageImpl result(0, page->cuts_, page->is_dense, page->row_stride,
kRows);
// Copy batch pages into the result page. // Copy batch pages into the result page.
size_t offset = 0; size_t offset = 0;
@ -126,13 +122,13 @@ TEST(EllpackPage, Copy) {
std::vector<bst_float> row_result(kCols); std::vector<bst_float> row_result(kCols);
for (auto& page : dmat->GetBatches<EllpackPage>(param)) { for (auto& page : dmat->GetBatches<EllpackPage>(param)) {
auto impl = page.Impl(); auto impl = page.Impl();
EXPECT_EQ(impl->matrix.base_rowid, current_row); EXPECT_EQ(impl->base_rowid, current_row);
for (size_t i = 0; i < impl->Size(); i++) { for (size_t i = 0; i < impl->Size(); i++) {
dh::LaunchN(0, kCols, ReadRowFunction(impl->matrix, current_row, row_d.data().get())); dh::LaunchN(0, kCols, ReadRowFunction(impl->GetDeviceAccessor(0), current_row, row_d.data().get()));
thrust::copy(row_d.begin(), row_d.end(), row.begin()); thrust::copy(row_d.begin(), row_d.end(), row.begin());
dh::LaunchN(0, kCols, ReadRowFunction(result.matrix, current_row, row_result_d.data().get())); dh::LaunchN(0, kCols, ReadRowFunction(result.GetDeviceAccessor(0), current_row, row_result_d.data().get()));
thrust::copy(row_result_d.begin(), row_result_d.end(), row_result.begin()); thrust::copy(row_result_d.begin(), row_result_d.end(), row_result.begin());
EXPECT_EQ(row, row_result); EXPECT_EQ(row, row_result);
@ -155,7 +151,8 @@ TEST(EllpackPage, Compact) {
auto page = (*dmat->GetBatches<EllpackPage>(param).begin()).Impl(); auto page = (*dmat->GetBatches<EllpackPage>(param).begin()).Impl();
// Create an empty result page. // Create an empty result page.
EllpackPageImpl result(0, page->matrix.info, kCompactedRows); EllpackPageImpl result(0, page->cuts_, page->is_dense, page->row_stride,
kCompactedRows);
// Compact batch pages into the result page. // Compact batch pages into the result page.
std::vector<size_t> row_indexes_h { std::vector<size_t> row_indexes_h {
@ -174,7 +171,7 @@ TEST(EllpackPage, Compact) {
std::vector<bst_float> row_result(kCols); std::vector<bst_float> row_result(kCols);
for (auto& page : dmat->GetBatches<EllpackPage>(param)) { for (auto& page : dmat->GetBatches<EllpackPage>(param)) {
auto impl = page.Impl(); auto impl = page.Impl();
EXPECT_EQ(impl->matrix.base_rowid, current_row); EXPECT_EQ(impl->base_rowid, current_row);
for (size_t i = 0; i < impl->Size(); i++) { for (size_t i = 0; i < impl->Size(); i++) {
size_t compacted_row = row_indexes_h[current_row]; size_t compacted_row = row_indexes_h[current_row];
@ -183,11 +180,12 @@ TEST(EllpackPage, Compact) {
continue; continue;
} }
dh::LaunchN(0, kCols, ReadRowFunction(impl->matrix, current_row, row_d.data().get())); dh::LaunchN(0, kCols, ReadRowFunction(impl->GetDeviceAccessor(0), current_row, row_d.data().get()));
dh::safe_cuda (cudaDeviceSynchronize());
thrust::copy(row_d.begin(), row_d.end(), row.begin()); thrust::copy(row_d.begin(), row_d.end(), row.begin());
dh::LaunchN(0, kCols, dh::LaunchN(0, kCols,
ReadRowFunction(result.matrix, compacted_row, row_result_d.data().get())); ReadRowFunction(result.GetDeviceAccessor(0), compacted_row, row_result_d.data().get()));
thrust::copy(row_result_d.begin(), row_result_d.end(), row_result.begin()); thrust::copy(row_result_d.begin(), row_result_d.end(), row_result.begin());
EXPECT_EQ(row, row_result); EXPECT_EQ(row, row_result);

45
tests/cpp/data/test_sparse_page_dmatrix.cu
View File

@ -3,6 +3,7 @@
#include <dmlc/filesystem.h> #include <dmlc/filesystem.h>
#include "../helpers.h" #include "../helpers.h"
#include "../../../src/common/compressed_iterator.h" #include "../../../src/common/compressed_iterator.h"
#include "../../../src/data/ellpack_page.cuh"
namespace xgboost { namespace xgboost {
@ -58,31 +59,29 @@ TEST(SparsePageDMatrix, EllpackPageContent) {
BatchParam param{0, 2, 0, 0}; BatchParam param{0, 2, 0, 0};
auto impl = (*dmat->GetBatches<EllpackPage>(param).begin()).Impl(); auto impl = (*dmat->GetBatches<EllpackPage>(param).begin()).Impl();
EXPECT_EQ(impl->matrix.base_rowid, 0); EXPECT_EQ(impl->base_rowid, 0);
EXPECT_EQ(impl->matrix.n_rows, kRows); EXPECT_EQ(impl->n_rows, kRows);
EXPECT_FALSE(impl->matrix.info.is_dense); EXPECT_FALSE(impl->is_dense);
EXPECT_EQ(impl->matrix.info.row_stride, 2); EXPECT_EQ(impl->row_stride, 2);
EXPECT_EQ(impl->matrix.info.n_bins, 4); EXPECT_EQ(impl->cuts_.TotalBins(), 4);
auto impl_ext = (*dmat_ext->GetBatches<EllpackPage>(param).begin()).Impl(); auto impl_ext = (*dmat_ext->GetBatches<EllpackPage>(param).begin()).Impl();
EXPECT_EQ(impl_ext->matrix.base_rowid, 0); EXPECT_EQ(impl_ext->base_rowid, 0);
EXPECT_EQ(impl_ext->matrix.n_rows, kRows); EXPECT_EQ(impl_ext->n_rows, kRows);
EXPECT_FALSE(impl_ext->matrix.info.is_dense); EXPECT_FALSE(impl_ext->is_dense);
EXPECT_EQ(impl_ext->matrix.info.row_stride, 2); EXPECT_EQ(impl_ext->row_stride, 2);
EXPECT_EQ(impl_ext->matrix.info.n_bins, 4); EXPECT_EQ(impl_ext->cuts_.TotalBins(), 4);
std::vector<common::CompressedByteT> buffer(impl->gidx_buffer.size()); std::vector<common::CompressedByteT> buffer(impl->gidx_buffer.HostVector());
std::vector<common::CompressedByteT> buffer_ext(impl_ext->gidx_buffer.size()); std::vector<common::CompressedByteT> buffer_ext(impl_ext->gidx_buffer.HostVector());
dh::CopyDeviceSpanToVector(&buffer, impl->gidx_buffer);
dh::CopyDeviceSpanToVector(&buffer_ext, impl_ext->gidx_buffer);
EXPECT_EQ(buffer, buffer_ext); EXPECT_EQ(buffer, buffer_ext);
} }
struct ReadRowFunction { struct ReadRowFunction {
EllpackMatrix matrix; EllpackDeviceAccessor matrix;
int row; int row;
bst_float* row_data_d; bst_float* row_data_d;
ReadRowFunction(EllpackMatrix matrix, int row, bst_float* row_data_d) ReadRowFunction(EllpackDeviceAccessor matrix, int row, bst_float* row_data_d)
: matrix(std::move(matrix)), row(row), row_data_d(row_data_d) {} : matrix(std::move(matrix)), row(row), row_data_d(row_data_d) {}
__device__ void operator()(size_t col) { __device__ void operator()(size_t col) {
@ -110,8 +109,8 @@ TEST(SparsePageDMatrix, MultipleEllpackPageContent) {
BatchParam param{0, kMaxBins, 0, kPageSize}; BatchParam param{0, kMaxBins, 0, kPageSize};
auto impl = (*dmat->GetBatches<EllpackPage>(param).begin()).Impl(); auto impl = (*dmat->GetBatches<EllpackPage>(param).begin()).Impl();
EXPECT_EQ(impl->matrix.base_rowid, 0); EXPECT_EQ(impl->base_rowid, 0);
EXPECT_EQ(impl->matrix.n_rows, kRows); EXPECT_EQ(impl->n_rows, kRows);
size_t current_row = 0; size_t current_row = 0;
thrust::device_vector<bst_float> row_d(kCols); thrust::device_vector<bst_float> row_d(kCols);
@ -120,13 +119,13 @@ TEST(SparsePageDMatrix, MultipleEllpackPageContent) {
std::vector<bst_float> row_ext(kCols); std::vector<bst_float> row_ext(kCols);
for (auto& page : dmat_ext->GetBatches<EllpackPage>(param)) { for (auto& page : dmat_ext->GetBatches<EllpackPage>(param)) {
auto impl_ext = page.Impl(); auto impl_ext = page.Impl();
EXPECT_EQ(impl_ext->matrix.base_rowid, current_row); EXPECT_EQ(impl_ext->base_rowid, current_row);
for (size_t i = 0; i < impl_ext->Size(); i++) { for (size_t i = 0; i < impl_ext->Size(); i++) {
dh::LaunchN(0, kCols, ReadRowFunction(impl->matrix, current_row, row_d.data().get())); dh::LaunchN(0, kCols, ReadRowFunction(impl->GetDeviceAccessor(0), current_row, row_d.data().get()));
thrust::copy(row_d.begin(), row_d.end(), row.begin()); thrust::copy(row_d.begin(), row_d.end(), row.begin());
dh::LaunchN(0, kCols, ReadRowFunction(impl_ext->matrix, current_row, row_ext_d.data().get())); dh::LaunchN(0, kCols, ReadRowFunction(impl_ext->GetDeviceAccessor(0), current_row, row_ext_d.data().get()));
thrust::copy(row_ext_d.begin(), row_ext_d.end(), row_ext.begin()); thrust::copy(row_ext_d.begin(), row_ext_d.end(), row_ext.begin());
EXPECT_EQ(row, row_ext); EXPECT_EQ(row, row_ext);
@ -155,8 +154,8 @@ TEST(SparsePageDMatrix, EllpackPageMultipleLoops) {
size_t current_row = 0; size_t current_row = 0;
for (auto& page : dmat_ext->GetBatches<EllpackPage>(param)) { for (auto& page : dmat_ext->GetBatches<EllpackPage>(param)) {
auto impl_ext = page.Impl(); auto impl_ext = page.Impl();
EXPECT_EQ(impl_ext->matrix.base_rowid, current_row); EXPECT_EQ(impl_ext->base_rowid, current_row);
current_row += impl_ext->matrix.n_rows; current_row += impl_ext->n_rows;
} }
} }

12
tests/cpp/helpers.h
View File

@ -244,13 +244,13 @@ class HistogramCutsWrapper : public common::HistogramCuts {
public: public:
using SuperT = common::HistogramCuts; using SuperT = common::HistogramCuts;
void SetValues(std::vector<float> cuts) { void SetValues(std::vector<float> cuts) {
SuperT::cut_values_ = std::move(cuts); SuperT::cut_values_.HostVector() = std::move(cuts);
} }
void SetPtrs(std::vector<uint32_t> ptrs) { void SetPtrs(std::vector<uint32_t> ptrs) {
SuperT::cut_ptrs_ = std::move(ptrs); SuperT::cut_ptrs_.HostVector() = std::move(ptrs);
} }
void SetMins(std::vector<float> mins) { void SetMins(std::vector<float> mins) {
SuperT::min_vals_ = std::move(mins); SuperT::min_vals_.HostVector() = std::move(mins);
} }
}; };
} // anonymous namespace } // anonymous namespace
@ -279,10 +279,8 @@ inline std::unique_ptr<EllpackPageImpl> BuildEllpackPage(
row_stride = std::max(row_stride, offset_vec[i] - offset_vec[i-1]); row_stride = std::max(row_stride, offset_vec[i] - offset_vec[i-1]);
} }
auto page = std::unique_ptr<EllpackPageImpl>(new EllpackPageImpl(dmat->get(), {0, 256, 0})); auto page = std::unique_ptr<EllpackPageImpl>(
page->InitInfo(0, (*dmat)->IsDense(), row_stride, cmat); new EllpackPageImpl(0, cmat, batch, (*dmat)->IsDense(), row_stride));
page->InitCompressedData(0, n_rows);
page->CreateHistIndices(0, batch, RowStateOnDevice(batch.Size(), batch.Size()));
delete dmat; delete dmat;

25
tests/cpp/tree/gpu_hist/test_gradient_based_sampler.cu
View File

@ -3,6 +3,7 @@
#include "../../../../src/data/ellpack_page.cuh" #include "../../../../src/data/ellpack_page.cuh"
#include "../../../../src/tree/gpu_hist/gradient_based_sampler.cuh" #include "../../../../src/tree/gpu_hist/gradient_based_sampler.cuh"
#include "../../helpers.h" #include "../../helpers.h"
#include "dmlc/filesystem.h"
namespace xgboost { namespace xgboost {
namespace tree { namespace tree {
@ -29,7 +30,7 @@ void VerifySampling(size_t page_size,
BatchParam param{0, 256, 0, page_size}; BatchParam param{0, 256, 0, page_size};
auto page = (*dmat->GetBatches<EllpackPage>(param).begin()).Impl(); auto page = (*dmat->GetBatches<EllpackPage>(param).begin()).Impl();
if (page_size != 0) { if (page_size != 0) {
EXPECT_NE(page->matrix.n_rows, kRows); EXPECT_NE(page->n_rows, kRows);
} }
GradientBasedSampler sampler(page, kRows, param, subsample, sampling_method); GradientBasedSampler sampler(page, kRows, param, subsample, sampling_method);
@ -37,11 +38,11 @@ void VerifySampling(size_t page_size,
if (fixed_size_sampling) { if (fixed_size_sampling) {
EXPECT_EQ(sample.sample_rows, kRows); EXPECT_EQ(sample.sample_rows, kRows);
EXPECT_EQ(sample.page->matrix.n_rows, kRows); EXPECT_EQ(sample.page->n_rows, kRows);
EXPECT_EQ(sample.gpair.size(), kRows); EXPECT_EQ(sample.gpair.size(), kRows);
} else { } else {
EXPECT_NEAR(sample.sample_rows, sample_rows, kRows * 0.016f); EXPECT_NEAR(sample.sample_rows, sample_rows, kRows * 0.016);
EXPECT_NEAR(sample.page->matrix.n_rows, sample_rows, kRows * 0.016f); EXPECT_NEAR(sample.page->n_rows, sample_rows, kRows * 0.016f);
EXPECT_NEAR(sample.gpair.size(), sample_rows, kRows * 0.016f); EXPECT_NEAR(sample.gpair.size(), sample_rows, kRows * 0.016f);
} }
@ -83,7 +84,7 @@ TEST(GradientBasedSampler, NoSampling_ExternalMemory) {
BatchParam param{0, 256, 0, kPageSize}; BatchParam param{0, 256, 0, kPageSize};
auto page = (*dmat->GetBatches<EllpackPage>(param).begin()).Impl(); auto page = (*dmat->GetBatches<EllpackPage>(param).begin()).Impl();
EXPECT_NE(page->matrix.n_rows, kRows); EXPECT_NE(page->n_rows, kRows);
GradientBasedSampler sampler(page, kRows, param, kSubsample, TrainParam::kUniform); GradientBasedSampler sampler(page, kRows, param, kSubsample, TrainParam::kUniform);
auto sample = sampler.Sample(gpair.DeviceSpan(), dmat.get()); auto sample = sampler.Sample(gpair.DeviceSpan(), dmat.get());
@ -91,21 +92,19 @@ TEST(GradientBasedSampler, NoSampling_ExternalMemory) {
EXPECT_EQ(sample.sample_rows, kRows); EXPECT_EQ(sample.sample_rows, kRows);
EXPECT_EQ(sample.gpair.size(), gpair.Size()); EXPECT_EQ(sample.gpair.size(), gpair.Size());
EXPECT_EQ(sample.gpair.data(), gpair.DevicePointer()); EXPECT_EQ(sample.gpair.data(), gpair.DevicePointer());
EXPECT_EQ(sampled_page->matrix.n_rows, kRows); EXPECT_EQ(sampled_page->n_rows, kRows);
std::vector<common::CompressedByteT> buffer(sampled_page->gidx_buffer.size()); std::vector<common::CompressedByteT> buffer(sampled_page->gidx_buffer.HostVector());
dh::CopyDeviceSpanToVector(&buffer, sampled_page->gidx_buffer);
common::CompressedIterator<common::CompressedByteT> common::CompressedIterator<common::CompressedByteT>
ci(buffer.data(), sampled_page->matrix.info.NumSymbols()); ci(buffer.data(), sampled_page->NumSymbols());
size_t offset = 0; size_t offset = 0;
for (auto& batch : dmat->GetBatches<EllpackPage>(param)) { for (auto& batch : dmat->GetBatches<EllpackPage>(param)) {
auto page = batch.Impl(); auto page = batch.Impl();
std::vector<common::CompressedByteT> page_buffer(page->gidx_buffer.size()); std::vector<common::CompressedByteT> page_buffer(page->gidx_buffer.HostVector());
dh::CopyDeviceSpanToVector(&page_buffer, page->gidx_buffer);
common::CompressedIterator<common::CompressedByteT> common::CompressedIterator<common::CompressedByteT>
page_ci(page_buffer.data(), page->matrix.info.NumSymbols()); page_ci(page_buffer.data(), page->NumSymbols());
size_t num_elements = page->matrix.n_rows * page->matrix.info.row_stride; size_t num_elements = page->n_rows * page->row_stride;
for (size_t i = 0; i < num_elements; i++) { for (size_t i = 0; i < num_elements; i++) {
EXPECT_EQ(ci[i + offset], page_ci[i]); EXPECT_EQ(ci[i + offset], page_ci[i]);
} }

6
tests/cpp/tree/gpu_hist/test_histogram.cu
View File

@ -27,7 +27,7 @@ void TestDeterminsticHistogram() {
gpair.SetDevice(0); gpair.SetDevice(0);
auto rounding = CreateRoundingFactor<Gradient>(gpair.DeviceSpan()); auto rounding = CreateRoundingFactor<Gradient>(gpair.DeviceSpan());
BuildGradientHistogram(page->matrix, gpair.DeviceSpan(), ridx, BuildGradientHistogram(page->GetDeviceAccessor(0), gpair.DeviceSpan(), ridx,
d_histogram, rounding, true); d_histogram, rounding, true);
for (size_t i = 0; i < kRounds; ++i) { for (size_t i = 0; i < kRounds; ++i) {
@ -35,7 +35,7 @@ void TestDeterminsticHistogram() {
auto d_histogram = dh::ToSpan(new_histogram); auto d_histogram = dh::ToSpan(new_histogram);
auto rounding = CreateRoundingFactor<Gradient>(gpair.DeviceSpan()); auto rounding = CreateRoundingFactor<Gradient>(gpair.DeviceSpan());
BuildGradientHistogram(page->matrix, gpair.DeviceSpan(), ridx, BuildGradientHistogram(page->GetDeviceAccessor(0), gpair.DeviceSpan(), ridx,
d_histogram, rounding, true); d_histogram, rounding, true);
for (size_t j = 0; j < new_histogram.size(); ++j) { for (size_t j = 0; j < new_histogram.size(); ++j) {
@ -50,7 +50,7 @@ void TestDeterminsticHistogram() {
auto gpair = GenerateRandomGradients(kRows, kLower, kUpper); auto gpair = GenerateRandomGradients(kRows, kLower, kUpper);
gpair.SetDevice(0); gpair.SetDevice(0);
dh::device_vector<Gradient> baseline(kBins * kCols); dh::device_vector<Gradient> baseline(kBins * kCols);
BuildGradientHistogram(page->matrix, gpair.DeviceSpan(), ridx, BuildGradientHistogram(page->GetDeviceAccessor(0), gpair.DeviceSpan(), ridx,
dh::ToSpan(baseline), rounding, true); dh::ToSpan(baseline), rounding, true);
for (size_t i = 0; i < baseline.size(); ++i) { for (size_t i = 0; i < baseline.size(); ++i) {
EXPECT_NEAR(((Gradient)baseline[i]).GetGrad(), ((Gradient)histogram[i]).GetGrad(), EXPECT_NEAR(((Gradient)baseline[i]).GetGrad(), ((Gradient)histogram[i]).GetGrad(),

29
tests/cpp/tree/test_gpu_hist.cu
View File

@ -97,12 +97,8 @@ void TestBuildHist(bool use_shared_memory_histograms) {
} }
gpair.SetDevice(0); gpair.SetDevice(0);
thrust::host_vector<common::CompressedByteT> h_gidx_buffer (page->gidx_buffer.size()); thrust::host_vector<common::CompressedByteT> h_gidx_buffer (page->gidx_buffer.HostVector());
common::CompressedByteT* d_gidx_buffer_ptr = page->gidx_buffer.data();
dh::safe_cuda(cudaMemcpy(h_gidx_buffer.data(), d_gidx_buffer_ptr,
sizeof(common::CompressedByteT) * page->gidx_buffer.size(),
cudaMemcpyDeviceToHost));
maker.row_partitioner.reset(new RowPartitioner(0, kNRows)); maker.row_partitioner.reset(new RowPartitioner(0, kNRows));
maker.hist.AllocateHistogram(0); maker.hist.AllocateHistogram(0);
@ -196,15 +192,10 @@ TEST(GpuHist, EvaluateSplits) {
auto cmat = GetHostCutMatrix(); auto cmat = GetHostCutMatrix();
// Copy cut matrix to device. // Copy cut matrix to device.
maker.ba.Allocate(0, page->cuts_ = cmat;
&(page->matrix.info.feature_segments), cmat.Ptrs().size(), maker.ba.Allocate(0, &(maker.monotone_constraints), kNCols);
&(page->matrix.info.min_fvalue), cmat.MinValues().size(), dh::CopyVectorToDeviceSpan(maker.monotone_constraints,
&(page->matrix.info.gidx_fvalue_map), 24, param.monotone_constraints);
&(maker.monotone_constraints), kNCols);
dh::CopyVectorToDeviceSpan(page->matrix.info.feature_segments, cmat.Ptrs());
dh::CopyVectorToDeviceSpan(page->matrix.info.gidx_fvalue_map, cmat.Values());
dh::CopyVectorToDeviceSpan(maker.monotone_constraints, param.monotone_constraints);
dh::CopyVectorToDeviceSpan(page->matrix.info.min_fvalue, cmat.MinValues());
// Initialize GPUHistMakerDevice::hist // Initialize GPUHistMakerDevice::hist
maker.hist.Init(0, (max_bins - 1) * kNCols); maker.hist.Init(0, (max_bins - 1) * kNCols);
@ -274,15 +265,13 @@ void TestHistogramIndexImpl() {
// Extract the device maker from the histogram makers and from that its compressed // Extract the device maker from the histogram makers and from that its compressed
// histogram index // histogram index
const auto &maker = hist_maker.maker; const auto &maker = hist_maker.maker;
std::vector<common::CompressedByteT> h_gidx_buffer(maker->page->gidx_buffer.size()); std::vector<common::CompressedByteT> h_gidx_buffer(maker->page->gidx_buffer.HostVector());
dh::CopyDeviceSpanToVector(&h_gidx_buffer, maker->page->gidx_buffer);
const auto &maker_ext = hist_maker_ext.maker; const auto &maker_ext = hist_maker_ext.maker;
std::vector<common::CompressedByteT> h_gidx_buffer_ext(maker_ext->page->gidx_buffer.size()); std::vector<common::CompressedByteT> h_gidx_buffer_ext(maker_ext->page->gidx_buffer.HostVector());
dh::CopyDeviceSpanToVector(&h_gidx_buffer_ext, maker_ext->page->gidx_buffer);
ASSERT_EQ(maker->page->matrix.info.n_bins, maker_ext->page->matrix.info.n_bins); ASSERT_EQ(maker->page->cuts_.TotalBins(), maker_ext->page->cuts_.TotalBins());
ASSERT_EQ(maker->page->gidx_buffer.size(), maker_ext->page->gidx_buffer.size()); ASSERT_EQ(maker->page->gidx_buffer.Size(), maker_ext->page->gidx_buffer.Size());
} }