Skip to content

Enable HNSW index arrays with transparent huge pages #72

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
gouyt13 merged 1 commit into from
Aug 26, 2026
Merged

Enable HNSW index arrays with transparent huge pages #72

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 3 additions & 3 deletions include/rabitqlib/index/hnsw/hnsw.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -423,7 +423,7 @@ inline HierarchicalNSW::HierarchicalNSW(
offsetExData_ + size_ex_data_; // (# of edges + edges) + (cluster_id) + (external
// label) + (BinData) + (ExData)
data_level0_memory_ =
reinterpret_cast<char*>(malloc(max_elements_ * size_data_per_element_));
memory::huge_page_allocate<char>(max_elements_ * size_data_per_element_);
if (data_level0_memory_ == nullptr) {
throw std::runtime_error("Not enough memory");
}
Expand Down Expand Up @@ -549,7 +549,7 @@ inline void HierarchicalNSW::load(const char* filename) {
input.read(reinterpret_cast<char*>(&ef_construction_), sizeof(size_t));

const size_t centroids_bytes = num_cluster_ * padded_dim_ * sizeof(float);
centroids_memory_ = reinterpret_cast<char*>(malloc(centroids_bytes));
centroids_memory_ = memory::huge_page_allocate<char>(centroids_bytes);
if (centroids_memory_ == nullptr) {
throw std::runtime_error("Not enough memory: loadIndex failed to allocate centroids"
);
Expand All @@ -561,7 +561,7 @@ inline void HierarchicalNSW::load(const char* filename) {
);

data_level0_memory_ =
reinterpret_cast<char*>(malloc(max_elements_ * size_data_per_element_));
memory::huge_page_allocate<char>(max_elements_ * size_data_per_element_);

input.read(
data_level0_memory_,
Expand Down
8 changes: 8 additions & 0 deletions include/rabitqlib/utils/memory.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -86,6 +86,14 @@ inline T* align_allocate(size_t nbytes) {
return static_cast<T*>(ptr);
}

inline constexpr size_t kHugePageSize = 2UL << 20;

// Allocate a large buffer backed by transparent huge pages.
template <typename T>
inline T* huge_page_allocate(size_t nbytes) {
return align_allocate<kHugePageSize, T, true>(nbytes);
}

static inline void prefetch_l1(const void* addr) {
#if defined(__SSE2__)
_mm_prefetch(addr, _MM_HINT_T0);
Expand Down
Loading