This PR exposes a copy-on-write mmap option through the public StorageOptions API and fixes a bug where the MAP_POPULATE flag was applied to the wrong mmap() argument.
* feat(entity/search): add LinearPool/BlockHeap, split contiguous entity layout and add direct-pointer fast search path
Introduce single-heap candidate structures (LinearPool/BlockHeap) and
refactor greedy search to dispatch between pool and dual-heap paths.
Split node layout into separate vector/graph arrays for better cache
locality, add zero-copy get_vector_ptr() on the hot path, and extract
huge-page allocation into MemoryHelper. Also adds pyglass attribution.
Add a fast path that copies the first segment's index file as the merge base and only merges the tail segments into it. Limited to streaming indexes (HNSW, HNSW_RABITQ, FLAT) with matching index_type + quantize_type and no filter; IVF/VAMANA always rebuild (their Merge is dump-then-reopen and would drop the base docs).
Also, fix the incorrect concurrency of the compaction task.
* fix(vamana): medoid entry point, concurrent build crash and prune quality
- Add DiskANN-standard medoid (centroid-closest point) as the persisted
entry point, computed at dump time, replacing the fixed node-0 start.
- Fix data races on node_chunks_ / dist_chunks_ / node_chunk_bases_ during
concurrent build: add mutexes with double-checked locking and pre-reserve
capacity so the lock-free read fast path never hits reallocation.
- Fix RobustPrune for metrics with signed internal distance (quantized int8
cosine): introduce IndexMetric::build_distance_offset() to shift distances
into a non-negative range, restoring a geometrically meaningful
occlude_factor and improving graph quality / low-ef recall.
- Remove redundant `virtual` keyword from methods already marked `override`
- Replace `virtual ~Derived()` with `~Derived() override` for derived class destructors
- Add missing `override` on methods that override base class virtuals
When a RabitqReformer trained with one total_bits is reused for an entity
configured with a different total_bits, the quantized data layout (ex_code
size) silently mismatches. This causes get_full_est() to interpret binary
quantization codes as floating-point factors, producing garbage distances
that sporadically drop search results.
Add an ex_bits() accessor to RabitqReformer and validate it against the
entity's ex_bits in HnswRabitqStreamer::open(), returning IndexError_Mismatch
on inconsistency. Fix the HNSWRabitqGeneral test to re-create a converter
and reformer with matching total_bits=2 for the third invocation. Add
TestExBitsMismatch to verify the mismatch is correctly rejected.
Add modernize-use-override to .clang-tidy and apply fixes across
src/ and tests/: replace redundant virtual with override, annotate
missing override on derived methods, and drop redundant virtual on
already-overridden methods.
fixes#328
Problem:
SQ8 metadata (squared_sum, sum) was computed from pre-rounded float
values, causing mismatch with actual stored int8 values. On asymmetric
datasets (e.g. OpenAI 1536D where |x_min| >> x_max), this leads to
severe recall drop.
Solution:
Move std::round before accumulating squared_sum and sum.
Co-authored-by: rayx <rui.xing@alibaba-inc.com>