Delete old pq4_accumulate_loop files (replaced by dispatching.h)#4905
Closed
algoriddle wants to merge 4 commits into
Closed
Delete old pq4_accumulate_loop files (replaced by dispatching.h)#4905algoriddle wants to merge 4 commits into
algoriddle wants to merge 4 commits into
Conversation
Contributor
|
@algoriddle has exported this pull request. If you are a Meta employee, you can view the originating Diff in D96131176. |
algoriddle
added a commit
to algoriddle/faiss
that referenced
this pull request
Mar 11, 2026
…ebookresearch#4905) Summary: Pull Request resolved: facebookresearch#4905 Now that all search paths use FastScanCodeScanner (via dispatching.h), the old pq4_accumulate_loop and pq4_accumulate_loop_qbs free functions have no callers. Delete pq4_fast_scan_search_1.cpp and pq4_fast_scan_search_qbs.cpp entirely. Relocate the remaining utility functions (pq4_qbs_to_nq, pq4_preferred_qbs, accumulate_to_mem) to pq4_fast_scan.cpp. Remove dead declarations from pq4_fast_scan.h. This completes the code deduplication: dispatching.h uses shared helpers from accumulate_loops.h and decompose_qbs.h, and no duplicate implementations remain. Reviewed By: mdouze Differential Revision: D96131176
algoriddle
force-pushed
the
export-D96131176
branch
from
c7e3969 to
5486d05
Compare
algoriddle
added a commit
to algoriddle/faiss
that referenced
this pull request
Mar 11, 2026
…ebookresearch#4905) Summary: Pull Request resolved: facebookresearch#4905 Now that all search paths use FastScanCodeScanner (via dispatching.h), the old pq4_accumulate_loop and pq4_accumulate_loop_qbs free functions have no callers. Delete pq4_fast_scan_search_1.cpp and pq4_fast_scan_search_qbs.cpp entirely. Relocate the remaining utility functions (pq4_qbs_to_nq, pq4_preferred_qbs, accumulate_to_mem) to pq4_fast_scan.cpp. Remove dead declarations from pq4_fast_scan.h. This completes the code deduplication: dispatching.h uses shared helpers from accumulate_loops.h and decompose_qbs.h, and no duplicate implementations remain. Reviewed By: mdouze Differential Revision: D96131176
algoriddle
force-pushed
the
export-D96131176
branch
from
5486d05 to
560b9b6
Compare
algoriddle
added a commit
to algoriddle/faiss
that referenced
this pull request
Mar 11, 2026
…ebookresearch#4905) Summary: Pull Request resolved: facebookresearch#4905 Now that all search paths use FastScanCodeScanner (via dispatching.h), the old pq4_accumulate_loop and pq4_accumulate_loop_qbs free functions have no callers. Delete pq4_fast_scan_search_1.cpp and pq4_fast_scan_search_qbs.cpp entirely. Relocate the remaining utility functions (pq4_qbs_to_nq, pq4_preferred_qbs, accumulate_to_mem) to pq4_fast_scan.cpp. Remove dead declarations from pq4_fast_scan.h. This completes the code deduplication: dispatching.h uses shared helpers from accumulate_loops.h and decompose_qbs.h, and no duplicate implementations remain. Reviewed By: mdouze Differential Revision: D96131176
algoriddle
force-pushed
the
export-D96131176
branch
from
560b9b6 to
efbc5cc
Compare
algoriddle
added a commit
to algoriddle/faiss
that referenced
this pull request
Mar 11, 2026
…ebookresearch#4905) Summary: Pull Request resolved: facebookresearch#4905 Now that all search paths use FastScanCodeScanner (via dispatching.h), the old pq4_accumulate_loop and pq4_accumulate_loop_qbs free functions have no callers. Delete pq4_fast_scan_search_1.cpp and pq4_fast_scan_search_qbs.cpp entirely. Relocate the remaining utility functions (pq4_qbs_to_nq, pq4_preferred_qbs, accumulate_to_mem) to pq4_fast_scan.cpp. Remove dead declarations from pq4_fast_scan.h. This completes the code deduplication: dispatching.h uses shared helpers from accumulate_loops.h and decompose_qbs.h, and no duplicate implementations remain. Reviewed By: mdouze Differential Revision: D96131176
algoriddle
force-pushed
the
export-D96131176
branch
from
efbc5cc to
8a4c6a6
Compare
Summary: Add `FastScanCodeScanner`, a virtual base that bundles handler + kernel behind the SIMD dispatch boundary. In DD mode, `SINGLE_SIMD_LEVEL = NONE` so the existing fast scan code path uses emulated SIMD types. The new scanner provides per-SIMD translation units (AVX2, AVX512, ARM_NEON) compiled with the correct ISA flags, and a factory function (`make_fast_scan_knn_scanner`) that uses `DISPATCH_SIMDLevel` to select the right TU at runtime. This follows the proven `THE_LEVEL_TO_DISPATCH` pattern from the scalar quantizer per-SIMD TUs (`sq-dispatch.h`). Each per-SIMD TU includes `dispatching.h` which provides: - `ScannerMixIn<Handler>`: wraps a concrete handler and calls accumulation kernels (both search_1 multi-BB and QBS paths) - Factory specialization `make_fast_scan_scanner_impl<SL>()` with combinatorial dispatch over `is_max × with_id_map × handler_type` (SingleResultHandler for k=1, HeapHandler for k≤20, ReservoirHandler for k>20) New files: - `impl/fast_scan/dispatching.h` — dispatch template header - `impl/fast_scan/impl-avx2.cpp` — AVX2 per-SIMD TU - `impl/fast_scan/impl-avx512.cpp` — AVX512 per-SIMD TU - `impl/fast_scan/impl-neon.cpp` — ARM NEON TU (with ARM_SVE forwarding) Modified files: - `impl/fast_scan/pq4_fast_scan.h` — FastScanCodeScanner base + factory decl - `impl/fast_scan/pq4_fast_scan.cpp` — NONE specialization + dispatch wrapper - `xplat.bzl` / `CMakeLists.txt` — register SIMD files and header Note: RaBitQ handler is not wired through FastScanCodeScanner in this diff. That comes in later diffs when callers are switched. Differential Revision: D95950483
Summary: Add make_knn_scanner() virtual to IndexFastScan and IndexIVFFastScan that returns a SIMD-dispatched FastScanCodeScanner. Override in RaBitQ classes to produce rabitq-specific scanners via per-SIMD TU factories (rabitq_dispatching.h). Move RaBitQ handler method bodies into headers, add context pointer, and update rabitq_result_handler.h for SL template param. Wire IVFRaBitQFastScanScanner::scan_codes to use the new scanner. Update build files and SWIG ignores. Callers (search_implem_12/14, search_dispatch_implem) are unchanged and still use make_knn_handler. Differential Revision: D96116557
…ndler Summary: Now that every make_knn_scanner() returns a real scanner (never nullptr), switch all callers to use scanners directly and remove make_knn_handler. Add impl param to make_knn_scanner and its factory so heap vs reservoir selection respects the implem setting. Switch IndexFastScan search_implem_12/14, IndexIVFFastScan search_dispatch_implem/range_search, and IndexIVFPQFastScan scan_codes to scanner-only paths. Add range scanner factories (make_range_scanner, make_partial_range_scanner). Remove make_knn_handler from all classes. Change search_implem_10/12 scanner param from pointer to reference. Differential Revision: D96116555
algoriddle
added a commit
to algoriddle/faiss
that referenced
this pull request
Mar 11, 2026
…ebookresearch#4905) Summary: Pull Request resolved: facebookresearch#4905 Now that all search paths use FastScanCodeScanner (via dispatching.h), the old pq4_accumulate_loop and pq4_accumulate_loop_qbs free functions have no callers. Delete pq4_fast_scan_search_1.cpp and pq4_fast_scan_search_qbs.cpp entirely. Relocate the remaining utility functions (pq4_qbs_to_nq, pq4_preferred_qbs, accumulate_to_mem) to pq4_fast_scan.cpp. Remove dead declarations from pq4_fast_scan.h. This completes the code deduplication: dispatching.h uses shared helpers from accumulate_loops.h and decompose_qbs.h, and no duplicate implementations remain. Reviewed By: mdouze Differential Revision: D96131176
algoriddle
force-pushed
the
export-D96131176
branch
from
8a4c6a6 to
a2dfd91
Compare
…ebookresearch#4905) Summary: Pull Request resolved: facebookresearch#4905 Now that all search paths use FastScanCodeScanner (via dispatching.h), the old pq4_accumulate_loop and pq4_accumulate_loop_qbs free functions have no callers. Delete pq4_fast_scan_search_1.cpp and pq4_fast_scan_search_qbs.cpp entirely. Relocate the remaining utility functions (pq4_qbs_to_nq, pq4_preferred_qbs, accumulate_to_mem) to pq4_fast_scan.cpp. Remove dead declarations from pq4_fast_scan.h. This completes the code deduplication: dispatching.h uses shared helpers from accumulate_loops.h and decompose_qbs.h, and no duplicate implementations remain. Reviewed By: mdouze Differential Revision: D96131176
algoriddle
force-pushed
the
export-D96131176
branch
from
a2dfd91 to
b98404f
Compare
Contributor
|
This pull request has been merged in 000eec2. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary:
Now that all search paths use FastScanCodeScanner (via dispatching.h),
the old pq4_accumulate_loop and pq4_accumulate_loop_qbs free functions
have no callers. Delete pq4_fast_scan_search_1.cpp and
pq4_fast_scan_search_qbs.cpp entirely. Relocate the remaining utility
functions (pq4_qbs_to_nq, pq4_preferred_qbs, accumulate_to_mem) to
pq4_fast_scan.cpp. Remove dead declarations from pq4_fast_scan.h.
This completes the code deduplication: dispatching.h uses shared helpers
from accumulate_loops.h and decompose_qbs.h, and no duplicate
implementations remain.
Differential Revision: D96131176