Use DSA

Spinlocks should be held only for a few instructions, for multiple
reasons:

- You have to be very careful not to elog() out while holding a
  spinlock, because there is no mechanism to release the spinlock on
  error.

- Waiters can waste a lot of cycles spinning if the lock is
  contended. I you wait on a spinlock for too long, the PostgreSQL
  implementation will actually PANIC, see s_lock_stuck().

The flushLock is particularly problematic. It is held in exclusive
mode, which means it holds a spinlock, over the call to
FlushPages(). FlushPages() performs lots of I/O so it can take a very
long time (>= minutes), and can also easily error out for various
reasons.

allocatorLock would perhaps be OK as a spinlocks, but even that feels
a bit heavy, so I converted that to an LWLock, too.

entryLock is usually held for a very short time, in shared mode, so
that would be fine as a spinlock. However, in the rare case that the
entry point is updated, it's held for a very long time. An LWLock used
in shared mode is about as fast a spinlock, that path is pretty
heavily optimized.

I think we have some problems with the per-element spinlocks too. In
HnswUpdateNeighborPagesInMemory(), it's held over a call to
HnswUpdateConnection(), but HnswUpdateConnection() can error out at
least in case of an out-of-memory error (it uses lappend(), which
calls palloc()). It also calls the distance function, and I don't
think they are guaranteed to be ereport-free either. However, I didn't
address that in this PR, it needs a bit more thinking.

.github/workflows/build.yml

@@ -20,6 +20,8 @@ jobs:
             os: ubuntu-20.04
           - postgres: 12
             os: ubuntu-20.04
+          - postgres: 11
+            os: ubuntu-20.04
     steps:
       - uses: actions/checkout@v4
       - uses: ankane/setup-postgres@v1
@@ -28,7 +30,7 @@ jobs:
           dev-files: true
       - run: make
         env:
-          PG_CFLAGS: -DUSE_ASSERT_CHECKING -Wall -Wextra -Werror -Wno-unused-parameter -Wno-sign-compare
+          PG_CFLAGS: -Wall -Wextra -Werror -Wno-unused-parameter -Wno-sign-compare
       - run: |
           export PG_CONFIG=`which pg_config`
           sudo --preserve-env=PG_CONFIG make install
@@ -49,7 +51,7 @@ jobs:
           postgres-version: 14
       - run: make
         env:
-          PG_CFLAGS: -DUSE_ASSERT_CHECKING -Wall -Wextra -Werror -Wno-unused-parameter
+          PG_CFLAGS: -Wall -Wextra -Werror -Wno-unused-parameter
       - run: make install
       - run: make installcheck
       - if: ${{ failure() }}
@@ -57,12 +59,10 @@ jobs:
       - run: |
           brew install cpanm
           cpanm --notest IPC::Run
-          wget -q https://github.com/postgres/postgres/archive/refs/tags/REL_14_10.tar.gz
+          wget -q https://github.com/postgres/postgres/archive/refs/tags/REL_14_5.tar.gz
-          tar xf REL_14_10.tar.gz
+          tar xf REL_14_5.tar.gz
-      - run: make prove_installcheck PROVE_FLAGS="-I ./postgres-REL_14_10/src/test/perl" PERL5LIB="/Users/runner/perl5/lib/perl5"
+      - run: make prove_installcheck PROVE_FLAGS="-I ./postgres-REL_14_5/src/test/perl" PERL5LIB="/Users/runner/perl5/lib/perl5"
-      - run: make clean && /usr/local/opt/llvm@15/bin/scan-build --status-bugs make
+      - run: make clean && /usr/local/opt/llvm@15/bin/scan-build --status-bugs make PG_CFLAGS="-DUSE_ASSERT_CHECKING"
-        env:
-          PG_CFLAGS: -DUSE_ASSERT_CHECKING
   windows:
     runs-on: windows-latest
     if: ${{ !startsWith(github.ref_name, 'mac') }}
@@ -73,7 +73,6 @@ jobs:
           postgres-version: 14
       - run: |
           call "C:\Program Files\Microsoft Visual Studio\2022\Enterprise\VC\Auxiliary\Build\vcvars64.bat" && ^
-          cd %TEMP% && ^
           nmake /NOLOGO /F Makefile.win && ^
           nmake /NOLOGO /F Makefile.win install && ^
           nmake /NOLOGO /F Makefile.win installcheck && ^
@@ -84,10 +83,10 @@ jobs:
     if: ${{ !startsWith(github.ref_name, 'mac') && !startsWith(github.ref_name, 'windows') }}
     runs-on: ubuntu-latest
     container:
-      image: debian:12
+      image: debian:11
       options: --platform linux/386
     steps:
-      - run: apt-get update && apt-get install -y build-essential git libipc-run-perl postgresql-15 postgresql-server-dev-15 sudo
+      - run: apt-get update && apt-get install -y build-essential git libipc-run-perl postgresql-13 postgresql-server-dev-13 sudo
       - run: service postgresql start
       - run: |
           git clone https://github.com/${{ github.repository }}.git pgvector
@@ -100,15 +99,4 @@ jobs:
           sudo -u postgres make installcheck
           sudo -u postgres make prove_installcheck
         env:
-          PG_CFLAGS: -DUSE_ASSERT_CHECKING -Wall -Wextra -Werror -Wno-unused-parameter -Wno-sign-compare
+          PG_CFLAGS: -Wall -Wextra -Werror -Wno-unused-parameter -Wno-sign-compare
-  valgrind:
-    if: ${{ !startsWith(github.ref_name, 'mac') && !startsWith(github.ref_name, 'windows') }}
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-      - uses: ankane/setup-postgres-valgrind@v1
-        with:
-          postgres-version: 16
-      - run: make
-      - run: sudo --preserve-env=PG_CONFIG make install
-      - run: make installcheck

CHANGELOG.md

@@ -1,35 +1,10 @@
-## 0.7.0 (unreleased)
+## 0.5.2 (unreleased)
 
-- Added support for binary vectors to HNSW
-- Added `hamming_distance` function
-- Added `jaccard_distance` function
-- Added `quantize_binary` function
-
-## 0.6.2 (2024-03-18)
-
-- Reduced lock contention with parallel HNSW index builds
-
-## 0.6.1 (2024-03-04)
-
-- Fixed error with `ANALYZE` and vectors with different dimensions
-- Fixed segmentation fault with `shared_preload_libraries`
-- Fixed vector subtraction being marked as commutative
-
-## 0.6.0 (2024-01-29)
-
-If upgrading with Postgres 12 or Docker, see [these notes](https://github.com/pgvector/pgvector#060).
-
-- Added support for parallel index builds for HNSW
-- Added validation for GUC parameters
-- Changed storage for vector from `extended` to `external`
 - Improved performance of HNSW
+- Added support for parallel index builds for HNSW
 - Reduced memory usage for HNSW index builds
 - Reduced WAL generation for HNSW index builds
-- Fixed error with logical replication
+- Fixed `invalid memory alloc request size` error with HNSW index build
-- Fixed `invalid memory alloc request size` error with HNSW index builds
-- Moved Docker image to `pgvector` org
-- Added Docker tags for each supported version of Postgres
-- Dropped support for Postgres 11
 
 ## 0.5.1 (2023-10-10)
 
@@ -77,7 +52,7 @@ If upgrading with Postgres 12 or Docker, see [these notes](https://github.com/pg
 
 ## 0.4.0 (2023-01-11)
 
-If upgrading with Postgres < 13, see [this note](https://github.com/pgvector/pgvector/blob/v0.4.0/README.md#040).
+If upgrading with Postgres < 13, see [this note](https://github.com/pgvector/pgvector#040).
 
 - Changed text representation for vector elements to match `real`
 - Changed storage for vector from `plain` to `extended`
@@ -94,7 +69,7 @@ If upgrading with Postgres < 13, see [this note](https://github.com/pgvector/pgv
 
 ## 0.3.1 (2022-11-02)
 
-If upgrading from 0.2.7 or 0.3.0, [recreate](https://github.com/pgvector/pgvector/blob/v0.3.1/README.md#031) all `ivfflat` indexes after upgrading to ensure all data is indexed.
+If upgrading from 0.2.7 or 0.3.0, [recreate](https://github.com/pgvector/pgvector#031) all `ivfflat` indexes after upgrading to ensure all data is indexed.
 
 - Fixed issue with inserts silently corrupting `ivfflat` indexes (introduced in 0.2.7)
 - Fixed segmentation fault with index creation when lists > 6500

Dockerfile

@@ -1,4 +1,4 @@
-ARG PG_MAJOR=16
+ARG PG_MAJOR=15
 FROM postgres:$PG_MAJOR
 ARG PG_MAJOR
 

LICENSE

@@ -1,4 +1,4 @@
-Portions Copyright (c) 1996-2024, PostgreSQL Global Development Group
+Portions Copyright (c) 1996-2023, PostgreSQL Global Development Group
 
 Portions Copyright (c) 1994, The Regents of the University of California
 

META.json

@@ -2,7 +2,7 @@
 	"name": "vector",
 	"abstract": "Open-source vector similarity search for Postgres",
 	"description": "Supports L2 distance, inner product, and cosine distance",
-	"version": "0.6.2",
+	"version": "0.5.1",
 	"maintainer": [
 		"Andrew Kane <andrew@ankane.org>"
 	],
@@ -12,7 +12,7 @@
 	"prereqs": {
 		"runtime": {
 			"requires": {
-				"PostgreSQL": "12.0.0"
+				"PostgreSQL": "11.0.0"
 			}
 		}
 	},
@@ -20,7 +20,7 @@
 		"vector": {
 			"file": "sql/vector.sql",
 			"docfile": "README.md",
-			"version": "0.6.2",
+			"version": "0.5.1",
 			"abstract": "Open-source vector similarity search for Postgres"
 		}
 	},

Makefile

@@ -1,9 +1,9 @@
 EXTENSION = vector
-EXTVERSION = 0.6.2
+EXTVERSION = 0.5.1
 
 MODULE_big = vector
 DATA = $(wildcard sql/*--*.sql)
-OBJS = src/bitvector.o src/hnsw.o src/hnswbuild.o src/hnswinsert.o src/hnswscan.o src/hnswutils.o src/hnswvacuum.o src/ivfbuild.o src/ivfflat.o src/ivfinsert.o src/ivfkmeans.o src/ivfscan.o src/ivfutils.o src/ivfvacuum.o src/vector.o
+OBJS = src/hnsw.o src/hnswbuild.o src/hnswinsert.o src/hnswscan.o src/hnswutils.o src/hnswvacuum.o src/ivfbuild.o src/ivfflat.o src/ivfinsert.o src/ivfkmeans.o src/ivfscan.o src/ivfutils.o src/ivfvacuum.o src/vector.o
 HEADERS = src/vector.h
 
 TESTS = $(wildcard test/sql/*.sql)
@@ -65,15 +65,13 @@ dist:
 	mkdir -p dist
 	git archive --format zip --prefix=$(EXTENSION)-$(EXTVERSION)/ --output dist/$(EXTENSION)-$(EXTVERSION).zip master
 
-# for Docker
-PG_MAJOR ?= 16
-
 .PHONY: docker
 
 docker:
-	docker build --pull --no-cache --build-arg PG_MAJOR=$(PG_MAJOR) -t pgvector/pgvector:pg$(PG_MAJOR) -t pgvector/pgvector:$(EXTVERSION)-pg$(PG_MAJOR) .
+	docker build --pull --no-cache --platform linux/amd64 -t ankane/pgvector:latest .
 
 .PHONY: docker-release
 
 docker-release:
-	docker buildx build --push --pull --no-cache --platform linux/amd64,linux/arm64 --build-arg PG_MAJOR=$(PG_MAJOR) -t pgvector/pgvector:pg$(PG_MAJOR) -t pgvector/pgvector:$(EXTVERSION)-pg$(PG_MAJOR) .
+	docker buildx build --push --pull --no-cache --platform linux/amd64,linux/arm64 -t ankane/pgvector:latest .
+	docker buildx build --push --platform linux/amd64,linux/arm64 -t ankane/pgvector:v$(EXTVERSION) .

Makefile.win

@@ -1,7 +1,7 @@
 EXTENSION = vector
-EXTVERSION = 0.6.2
+EXTVERSION = 0.5.1
 
-OBJS = src\bitvector.obj src\hnsw.obj src\hnswbuild.obj src\hnswinsert.obj src\hnswscan.obj src\hnswutils.obj src\hnswvacuum.obj src\ivfbuild.obj src\ivfflat.obj src\ivfinsert.obj src\ivfkmeans.obj src\ivfscan.obj src\ivfutils.obj src\ivfvacuum.obj src\vector.obj
+OBJS = src\hnsw.obj src\hnswbuild.obj src\hnswinsert.obj src\hnswscan.obj src\hnswutils.obj src\hnswvacuum.obj src\ivfbuild.obj src\ivfflat.obj src\ivfinsert.obj src\ivfkmeans.obj src\ivfscan.obj src\ivfutils.obj src\ivfvacuum.obj src\vector.obj
 HEADERS = src\vector.h
 
 REGRESS = btree cast copy functions input ivfflat_cosine ivfflat_ip ivfflat_l2 ivfflat_options ivfflat_unlogged

README.md

@@ -5,7 +5,7 @@ Open-source vector similarity search for Postgres
 Store your vectors with the rest of your data. Supports:
 
 - exact and approximate nearest neighbor search
-- L2 distance, inner product, cosine distance, and more
+- L2 distance, inner product, and cosine distance
 - any [language](#languages) with a Postgres client
 
 Plus [ACID](https://en.wikipedia.org/wiki/ACID) compliance, point-in-time recovery, JOINs, and all of the other [great features](https://www.postgresql.org/about/) of Postgres
@@ -14,46 +14,19 @@ Plus [ACID](https://en.wikipedia.org/wiki/ACID) compliance, point-in-time recove
 
 ## Installation
 
-### Linux and Mac
+Compile and install the extension (supports Postgres 11+)
-
-Compile and install the extension (supports Postgres 12+)
 
 ```sh
 cd /tmp
-git clone --branch v0.6.2 https://github.com/pgvector/pgvector.git
+git clone --branch v0.5.1 https://github.com/pgvector/pgvector.git
 cd pgvector
 make
 make install # may need sudo
 ```
 
-See the [installation notes](#installation-notes---linux-and-mac) if you run into issues
+See the [installation notes](#installation-notes) if you run into issues
 
-You can also install it with [Docker](#docker), [Homebrew](#homebrew), [PGXN](#pgxn), [APT](#apt), [Yum](#yum), [pkg](#pkg), or [conda-forge](#conda-forge), and it comes preinstalled with [Postgres.app](#postgresapp) and many [hosted providers](#hosted-postgres). There are also instructions for [GitHub Actions](https://github.com/pgvector/setup-pgvector).
+You can also install it with [Docker](#docker), [Homebrew](#homebrew), [PGXN](#pgxn), [APT](#apt), [Yum](#yum), or [conda-forge](#conda-forge), and it comes preinstalled with [Postgres.app](#postgresapp) and many [hosted providers](#hosted-postgres). There are also instructions for [GitHub Actions](https://github.com/pgvector/setup-pgvector).
-
-### Windows
-
-Ensure [C++ support in Visual Studio](https://learn.microsoft.com/en-us/cpp/build/building-on-the-command-line?view=msvc-170#download-and-install-the-tools) is installed, and run:
-
-```cmd
-call "C:\Program Files\Microsoft Visual Studio\2022\Community\VC\Auxiliary\Build\vcvars64.bat"
-```
-
-Note: The exact path will vary depending on your Visual Studio version and edition
-
-Then use `nmake` to build:
-
-```cmd
-set "PGROOT=C:\Program Files\PostgreSQL\16"
-cd %TEMP%
-git clone --branch v0.6.2 https://github.com/pgvector/pgvector.git
-cd pgvector
-nmake /F Makefile.win
-nmake /F Makefile.win install
-```
-
-See the [installation notes](#installation-notes---windows) if you run into issues
-
-You can also install it with [Docker](#docker) or [conda-forge](#conda-forge).
 
 ## Getting Started
 
@@ -105,12 +78,6 @@ Insert vectors
 INSERT INTO items (embedding) VALUES ('[1,2,3]'), ('[4,5,6]');
 ```
 
-Or load vectors in bulk using `COPY` ([example](https://github.com/pgvector/pgvector-python/blob/master/examples/bulk_loading.py))
-
-```sql
-COPY items (embedding) FROM STDIN WITH (FORMAT BINARY);
-```
-
 Upsert vectors
 
 ```sql
@@ -221,19 +188,7 @@ Cosine distance
 CREATE INDEX ON items USING hnsw (embedding vector_cosine_ops);
 ```
 
-Hamming distance - added in 0.7.0
+Vectors with up to 2,000 dimensions can be indexed.
-
-```sql
-CREATE INDEX ON items USING hnsw (embedding bit_hamming_ops);
-```
-
-Jaccard distance - added in 0.7.0
-
-```sql
-CREATE INDEX ON items USING hnsw (embedding bit_jaccard_ops);
-```
-
-Vectors with up to 2,000 dimensions can be indexed, or bit vectors with up to 64,000 dimensions.
 
 ### Index Options
 
@@ -285,16 +240,6 @@ HINT:  Increase maintenance_work_mem to speed up builds.
 
 Note: Do not set `maintenance_work_mem` so high that it exhausts the memory on the server
 
-Like other index types, it’s faster to create an index after loading your initial data
-
-Starting with 0.6.0, you can also speed up index creation by increasing the number of parallel workers (2 by default)
-
-```sql
-SET max_parallel_maintenance_workers = 7; -- plus leader
-```
-
-For a large number of workers, you may also need to increase `max_parallel_workers` (8 by default)
-
 ### Indexing Progress
 
 Check [indexing progress](https://www.postgresql.org/docs/current/progress-reporting.html#CREATE-INDEX-PROGRESS-REPORTING) with Postgres 12+
@@ -359,16 +304,6 @@ SELECT ...
 COMMIT;
 ```
 
-### Index Build Time
-
-Speed up index creation on large tables by increasing the number of parallel workers (2 by default)
-
-```sql
-SET max_parallel_maintenance_workers = 7; -- plus leader
-```
-
-For a large number of workers, you may also need to increase `max_parallel_workers` (8 by default)
-
 ### Indexing Progress
 
 Check [indexing progress](https://www.postgresql.org/docs/current/progress-reporting.html#CREATE-INDEX-PROGRESS-REPORTING) with Postgres 12+
@@ -425,39 +360,13 @@ You can use [Reciprocal Rank Fusion](https://github.com/pgvector/pgvector-python
 
 ## Performance
 
-### Tuning
-
-Use a tool like [PgTune](https://pgtune.leopard.in.ua/) to set initial values for Postgres server parameters.
-
-### Loading
-
-Use `COPY` for bulk loading data ([example](https://github.com/pgvector/pgvector-python/blob/master/examples/bulk_loading.py)).
-
-```sql
-COPY items (embedding) FROM STDIN WITH (FORMAT BINARY);
-```
-
-Add any indexes *after* loading the initial data for best performance.
-
-### Indexing
-
-See index build time for [HNSW](#index-build-time) and [IVFFlat](#index-build-time-1).
-
-In production environments, create indexes concurrently to avoid blocking writes.
-
-```sql
-CREATE INDEX CONCURRENTLY ...
-```
-
-### Querying
-
 Use `EXPLAIN ANALYZE` to debug performance.
 
 ```sql
 EXPLAIN ANALYZE SELECT * FROM items ORDER BY embedding <-> '[3,1,2]' LIMIT 5;
 ```
 
-#### Exact Search
+### Exact Search
 
 To speed up queries without an index, increase `max_parallel_workers_per_gather`.
 
@@ -471,7 +380,7 @@ If vectors are normalized to length 1 (like [OpenAI embeddings](https://platform
 SELECT * FROM items ORDER BY embedding <#> '[3,1,2]' LIMIT 5;
 ```
 
-#### Approximate Search
+### Approximate Search
 
 To speed up queries with an IVFFlat index, increase the number of inverted lists (at the expense of recall).
 
@@ -479,50 +388,6 @@ To speed up queries with an IVFFlat index, increase the number of inverted lists
 CREATE INDEX ON items USING ivfflat (embedding vector_l2_ops) WITH (lists = 1000);
 ```
 
-### Vacuuming
-
-Vacuuming can take a while for HNSW indexes. Speed it up by reindexing first.
-
-```sql
-REINDEX INDEX CONCURRENTLY index_name;
-VACUUM table_name;
-```
-
-## Monitoring
-
-Monitor performance with [pg_stat_statements](https://www.postgresql.org/docs/current/pgstatstatements.html) (be sure to add it to `shared_preload_libraries`).
-
-```sql
-CREATE EXTENSION pg_stat_statements;
-```
-
-Get the most time-consuming queries with:
-
-```sql
-SELECT query, calls, ROUND((total_plan_time + total_exec_time) / calls) AS avg_time_ms,
-    ROUND((total_plan_time + total_exec_time) / 60000) AS total_time_min
-    FROM pg_stat_statements ORDER BY total_plan_time + total_exec_time DESC LIMIT 20;
-```
-
-Note: Replace `total_plan_time + total_exec_time` with `total_time` for Postgres < 13
-
-Monitor recall by comparing results from approximate search with exact search.
-
-```sql
-BEGIN;
-SET LOCAL enable_indexscan = off; -- use exact search
-SELECT ...
-COMMIT;
-```
-
-## Scaling
-
-Scale pgvector the same way you scale Postgres.
-
-Scale vertically by increasing memory, CPU, and storage on a single instance. Use existing tools to [tune parameters](#tuning) and [monitor performance](#monitoring).
-
-Scale horizontally with [replicas](https://www.postgresql.org/docs/current/hot-standby.html), or use [Citus](https://github.com/citusdata/citus) or another approach for sharding ([example](https://github.com/pgvector/pgvector-python/blob/master/examples/citus.py)).
-
 ## Languages
 
 Use pgvector from any language with a Postgres client. You can even generate and store vectors in one language and query them in another.
@@ -616,18 +481,6 @@ and query with:
 SELECT * FROM items ORDER BY embedding::vector(3) <-> '[3,1,2]' LIMIT 5;
 ```
 
-#### Are binary vectors supported?
-
-You can store binary vectors and perform exact nearest neighbor search by Hamming distance in Postgres without an extension ([example](https://github.com/pgvector/pgvector-python/blob/master/examples/hash_image_search.py)).
-
-```tsql
-CREATE TABLE items (id bigserial PRIMARY KEY, embedding bit(3));
-INSERT INTO items (embedding) VALUES (B'000'), (B'111');
-SELECT * FROM items ORDER BY bit_count(embedding # B'101') LIMIT 5;
-```
-
-Indexing is not currently supported.
-
 #### Do indexes need to fit into memory?
 
 No, but like other index types, you’ll likely see better performance if they do. You can get the size of an index with:
@@ -640,17 +493,7 @@ SELECT pg_size_pretty(pg_relation_size('index_name'));
 
 #### Why isn’t a query using an index?
 
-The query needs to have an `ORDER BY` and `LIMIT`, and the `ORDER BY` must be the result of a distance operator, not an expression.
+The cost estimation in pgvector < 0.4.3 does not always work well with the planner. You can encourage the planner to use an index for a query with:
-
-```sql
--- index
-ORDER BY embedding <=> '[3,1,2]' LIMIT 5;
-
--- no index
-ORDER BY 1 - (embedding <=> '[3,1,2]') DESC LIMIT 5;
-```
-
-You can encourage the planner to use an index for a query with:
 
 ```sql
 BEGIN;
@@ -679,10 +522,6 @@ or choose to store vectors inline:
 ALTER TABLE items ALTER COLUMN embedding SET STORAGE PLAIN;
 ```
 
-#### Why are there less results for a query after adding an HNSW index?
-
-Results are limited by the size of the dynamic candidate list (`hnsw.ef_search`). There may be even less results due to dead tuples or filtering conditions in the query. We recommend setting `hnsw.ef_search` to at least twice the `LIMIT` of the query. If you need more than 500 results, use an IVFFlat index instead.
-
 #### Why are there less results for a query after adding an IVFFlat index?
 
 The index was likely created with too little data for the number of lists. Drop the index until the table has more data.
@@ -691,8 +530,6 @@ The index was likely created with too little data for the number of lists. Drop
 DROP INDEX index_name;
 ```
 
-Results can also be limited by the number of probes (`ivfflat.probes`).
-
 ## Reference
 
 ### Vector Type
@@ -718,7 +555,6 @@ cosine_distance(vector, vector) → double precision | cosine distance |
 inner_product(vector, vector) → double precision | inner product |
 l2_distance(vector, vector) → double precision | Euclidean distance |
 l1_distance(vector, vector) → double precision | taxicab distance | 0.5.0
-quantize_binary(vector) → bit | quantize | 0.7.0
 vector_dims(vector) → integer | number of dimensions |
 vector_norm(vector) → double precision | Euclidean norm |
 
@@ -729,28 +565,14 @@ Function | Description | Added
 avg(vector) → vector | average |
 sum(vector) → vector | sum | 0.5.0
 
-### Bit Operators
+## Installation Notes
-
-Operator | Description | Added
---- | --- | ---
-<~> | Hamming distance | 0.7.0
-<%> | Jaccard distance | 0.7.0
-
-### Bit Functions
-
-Function | Description | Added
---- | --- | ---
-hamming_distance(bit, bit) → double precision | Hamming distance | 0.7.0
-jaccard_distance(bit, bit) → double precision | Jaccard distance | 0.7.0
-
-## Installation Notes - Linux and Mac
 
 ### Postgres Location
 
 If your machine has multiple Postgres installations, specify the path to [pg_config](https://www.postgresql.org/docs/current/app-pgconfig.html) with:
 
 ```sh
-export PG_CONFIG=/Library/PostgreSQL/16/bin/pg_config
+export PG_CONFIG=/Applications/Postgres.app/Contents/Versions/latest/bin/pg_config
 ```
 
 Then re-run the installation instructions (run `make clean` before `make` if needed). If `sudo` is needed for `make install`, use:
@@ -759,14 +581,6 @@ Then re-run the installation instructions (run `make clean` before `make` if nee
 sudo --preserve-env=PG_CONFIG make install
 ```
 
-A few common paths on Mac are:
-
-- EDB installer - `/Library/PostgreSQL/16/bin/pg_config`
-- Homebrew (arm64) - `/opt/homebrew/opt/postgresql@16/bin/pg_config`
-- Homebrew (x86-64) - `/usr/local/opt/postgresql@16/bin/pg_config`
-
-Note: Replace `16` with your Postgres server version
-
 ### Missing Header
 
 If compilation fails with `fatal error: postgres.h: No such file or directory`, make sure Postgres development files are installed on the server.
@@ -779,48 +593,44 @@ sudo apt install postgresql-server-dev-16
 
 Note: Replace `16` with your Postgres server version
 
-### Missing SDK
+### Windows
 
-If compilation fails and the output includes `warning: no such sysroot directory` on Mac, reinstall Xcode Command Line Tools.
+Support for Windows is currently experimental. Ensure [C++ support in Visual Studio](https://learn.microsoft.com/en-us/cpp/build/building-on-the-command-line?view=msvc-170#download-and-install-the-tools) is installed, and run:
 
-### Portability
+```cmd
-
+call "C:\Program Files\Microsoft Visual Studio\2022\Community\VC\Auxiliary\Build\vcvars64.bat"
-By default, pgvector compiles with `-march=native` on some platforms for best performance. However, this can lead to `Illegal instruction` errors if trying to run the compiled extension on a different machine.
-
-To compile for portability, use:
-
-```sh
-make OPTFLAGS=""
 ```
 
-## Installation Notes - Windows
+Note: The exact path will vary depending on your Visual Studio version and edition
 
-### Missing Header
+Then use `nmake` to build:
 
-If compilation fails with `Cannot open include file: 'postgres.h': No such file or directory`, make sure `PGROOT` is correct.
+```cmd
-
+set "PGROOT=C:\Program Files\PostgreSQL\16"
-### Permissions
+git clone --branch v0.5.1 https://github.com/pgvector/pgvector.git
-
+cd pgvector
-If installation fails with `Access is denied`, re-run the installation instructions as an administrator.
+nmake /F Makefile.win
+nmake /F Makefile.win install
+```
 
 ## Additional Installation Methods
 
 ### Docker
 
-Get the [Docker image](https://hub.docker.com/r/pgvector/pgvector) with:
+Get the [Docker image](https://hub.docker.com/r/ankane/pgvector) with:
 
 ```sh
-docker pull pgvector/pgvector:pg16
+docker pull ankane/pgvector
 ```
 
-This adds pgvector to the [Postgres image](https://hub.docker.com/_/postgres) (replace `16` with your Postgres server version, and run it the same way).
+This adds pgvector to the [Postgres image](https://hub.docker.com/_/postgres) (run it the same way).
 
 You can also build the image manually:
 
 ```sh
-git clone --branch v0.6.2 https://github.com/pgvector/pgvector.git
+git clone --branch v0.5.1 https://github.com/pgvector/pgvector.git
 cd pgvector
-docker build --build-arg PG_MAJOR=16 -t myuser/pgvector .
+docker build --build-arg PG_MAJOR=15 -t myuser/pgvector .
 ```
 
 ### Homebrew
@@ -863,21 +673,6 @@ sudo dnf install pgvector_16
 
 Note: Replace `16` with your Postgres server version
 
-### pkg
-
-Install the FreeBSD package with:
-
-```sh
-pkg install postgresql15-pg_vector
-```
-
-or the port with:
-
-```sh
-cd /usr/ports/databases/pgvector
-make install
-```
-
 ### conda-forge
 
 With Conda Postgres, install from [conda-forge](https://anaconda.org/conda-forge/pgvector) with:
@@ -912,32 +707,28 @@ SELECT extversion FROM pg_extension WHERE extname = 'vector';
 
 ## Upgrade Notes
 
-### 0.6.0
+### 0.4.0
 
-#### Postgres 12
+If upgrading with Postgres < 13, remove this line from `sql/vector--0.3.2--0.4.0.sql`:
-
-If upgrading with Postgres 12, remove this line from `sql/vector--0.5.1--0.6.0.sql`:
 
 ```sql
-ALTER TYPE vector SET (STORAGE = external);
+ALTER TYPE vector SET (STORAGE = extended);
 ```
 
 Then run `make install` and `ALTER EXTENSION vector UPDATE;`.
 
-#### Docker
+### 0.3.1
 
-The Docker image is now published in the `pgvector` org, and there are tags for each supported version of Postgres (rather than a `latest` tag).
+If upgrading from 0.2.7 or 0.3.0, recreate all `ivfflat` indexes after upgrading to ensure all data is indexed.
 
-```sh
+```sql
-docker pull pgvector/pgvector:pg16
+-- Postgres 12+
-# or
+REINDEX INDEX CONCURRENTLY index_name;
-docker pull pgvector/pgvector:0.6.0-pg16
-```
 
-Also, if you’ve increased `maintenance_work_mem`, make sure `--shm-size` is at least that size to avoid an error with parallel HNSW index builds.
+-- Postgres < 12
-
+CREATE INDEX CONCURRENTLY temp_name ON table USING ivfflat (column opclass);
-```sh
+DROP INDEX CONCURRENTLY index_name;
-docker run --shm-size=1g ...
+ALTER INDEX temp_name RENAME TO index_name;
 ```
 
 ## Thanks
@@ -984,31 +775,13 @@ To run single tests:
 
 ```sh
 make installcheck REGRESS=functions                    # regression test
-make prove_installcheck PROVE_TESTS=test/t/001_ivfflat_wal.pl  # TAP test
+make prove_installcheck PROVE_TESTS=test/t/001_wal.pl  # TAP test
-```
-
-To enable assertions:
-
-```sh
-make clean && PG_CFLAGS="-DUSE_ASSERT_CHECKING" make && make install
 ```
 
 To enable benchmarking:
 
 ```sh
-make clean && PG_CFLAGS="-DIVFFLAT_BENCH" make && make install
+make clean && PG_CFLAGS=-DIVFFLAT_BENCH make && make install
-```
-
-To show memory usage:
-
-```sh
-make clean && PG_CFLAGS="-DHNSW_MEMORY -DIVFFLAT_MEMORY" make && make install
-```
-
-To get k-means metrics:
-
-```sh
-make clean && PG_CFLAGS="-DIVFFLAT_KMEANS_DEBUG" make && make install
 ```
 
 Resources for contributors

sql/vector--0.5.1--0.6.0.sql

@@ -1,5 +0,0 @@
--- complain if script is sourced in psql, rather than via CREATE EXTENSION
-\echo Use "ALTER EXTENSION vector UPDATE TO '0.6.0'" to load this file. \quit
-
--- remove this single line for Postgres < 13
-ALTER TYPE vector SET (STORAGE = external);

sql/vector--0.6.0--0.6.1.sql

@@ -1,16 +0,0 @@
--- complain if script is sourced in psql, rather than via CREATE EXTENSION
-\echo Use "ALTER EXTENSION vector UPDATE TO '0.6.1'" to load this file. \quit
-
-DROP OPERATOR - (vector, vector);
-
-CREATE OPERATOR - (
-	LEFTARG = vector, RIGHTARG = vector, PROCEDURE = vector_sub
-);
-
-ALTER OPERATOR <= (vector, vector) SET (
-	RESTRICT = scalarlesel, JOIN = scalarlejoinsel
-);
-
-ALTER OPERATOR >= (vector, vector) SET (
-	RESTRICT = scalargesel, JOIN = scalargejoinsel
-);

sql/vector--0.6.1--0.6.2.sql

@@ -1,2 +0,0 @@
--- complain if script is sourced in psql, rather than via CREATE EXTENSION
-\echo Use "ALTER EXTENSION vector UPDATE TO '0.6.2'" to load this file. \quit

sql/vector--0.6.2--0.7.0.sql

@@ -1,31 +0,0 @@
--- complain if script is sourced in psql, rather than via CREATE EXTENSION
-\echo Use "ALTER EXTENSION vector UPDATE TO '0.7.0'" to load this file. \quit
-
-CREATE FUNCTION quantize_binary(vector) RETURNS bit
-	AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
-
-CREATE FUNCTION hamming_distance(bit, bit) RETURNS float8
-	AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
-
-CREATE FUNCTION jaccard_distance(bit, bit) RETURNS float8
-	AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
-
-CREATE OPERATOR <~> (
-	LEFTARG = bit, RIGHTARG = bit, PROCEDURE = hamming_distance,
-	COMMUTATOR = '<~>'
-);
-
-CREATE OPERATOR <%> (
-	LEFTARG = bit, RIGHTARG = bit, PROCEDURE = jaccard_distance,
-	COMMUTATOR = '<%>'
-);
-
-CREATE OPERATOR CLASS bit_hamming_ops
-	FOR TYPE bit USING hnsw AS
-	OPERATOR 1 <~> (bit, bit) FOR ORDER BY float_ops,
-	FUNCTION 1 hamming_distance(bit, bit);
-
-CREATE OPERATOR CLASS bit_jaccard_ops
-	FOR TYPE bit USING hnsw AS
-	OPERATOR 1 <%> (bit, bit) FOR ORDER BY float_ops,
-	FUNCTION 1 jaccard_distance(bit, bit);

sql/vector.sql

@@ -26,7 +26,7 @@ CREATE TYPE vector (
 	TYPMOD_IN = vector_typmod_in,
 	RECEIVE   = vector_recv,
 	SEND      = vector_send,
-	STORAGE   = external
+	STORAGE   = extended
 );
 
 -- functions
@@ -58,9 +58,6 @@ CREATE FUNCTION vector_sub(vector, vector) RETURNS vector
 CREATE FUNCTION vector_mul(vector, vector) RETURNS vector
 	AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
 
-CREATE FUNCTION quantize_binary(vector) RETURNS bit
-	AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
-
 -- private functions
 
 CREATE FUNCTION vector_lt(vector, vector) RETURNS bool
@@ -183,7 +180,8 @@ CREATE OPERATOR + (
 );
 
 CREATE OPERATOR - (
-	LEFTARG = vector, RIGHTARG = vector, PROCEDURE = vector_sub
+	LEFTARG = vector, RIGHTARG = vector, PROCEDURE = vector_sub,
+	COMMUTATOR = -
 );
 
 CREATE OPERATOR * (
@@ -197,10 +195,11 @@ CREATE OPERATOR < (
 	RESTRICT = scalarltsel, JOIN = scalarltjoinsel
 );
 
+-- should use scalarlesel and scalarlejoinsel, but not supported in Postgres < 11
 CREATE OPERATOR <= (
 	LEFTARG = vector, RIGHTARG = vector, PROCEDURE = vector_le,
 	COMMUTATOR = >= , NEGATOR = > ,
-	RESTRICT = scalarlesel, JOIN = scalarlejoinsel
+	RESTRICT = scalarltsel, JOIN = scalarltjoinsel
 );
 
 CREATE OPERATOR = (
@@ -215,10 +214,11 @@ CREATE OPERATOR <> (
 	RESTRICT = eqsel, JOIN = eqjoinsel
 );
 
+-- should use scalargesel and scalargejoinsel, but not supported in Postgres < 11
 CREATE OPERATOR >= (
 	LEFTARG = vector, RIGHTARG = vector, PROCEDURE = vector_ge,
 	COMMUTATOR = <= , NEGATOR = < ,
-	RESTRICT = scalargesel, JOIN = scalargejoinsel
+	RESTRICT = scalargtsel, JOIN = scalargtjoinsel
 );
 
 CREATE OPERATOR > (
@@ -290,31 +290,3 @@ CREATE OPERATOR CLASS vector_cosine_ops
 	OPERATOR 1 <=> (vector, vector) FOR ORDER BY float_ops,
 	FUNCTION 1 vector_negative_inner_product(vector, vector),
 	FUNCTION 2 vector_norm(vector);
-
--- bit functions
-
-CREATE FUNCTION hamming_distance(bit, bit) RETURNS float8
-	AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
-
-CREATE FUNCTION jaccard_distance(bit, bit) RETURNS float8
-	AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
-
-CREATE OPERATOR <~> (
-	LEFTARG = bit, RIGHTARG = bit, PROCEDURE = hamming_distance,
-	COMMUTATOR = '<~>'
-);
-
-CREATE OPERATOR <%> (
-	LEFTARG = bit, RIGHTARG = bit, PROCEDURE = jaccard_distance,
-	COMMUTATOR = '<%>'
-);
-
-CREATE OPERATOR CLASS bit_hamming_ops
-	FOR TYPE bit USING hnsw AS
-	OPERATOR 1 <~> (bit, bit) FOR ORDER BY float_ops,
-	FUNCTION 1 hamming_distance(bit, bit);
-
-CREATE OPERATOR CLASS bit_jaccard_ops
-	FOR TYPE bit USING hnsw AS
-	OPERATOR 1 <%> (bit, bit) FOR ORDER BY float_ops,
-	FUNCTION 1 jaccard_distance(bit, bit);

src/bitvector.c

@@ -1,90 +0,0 @@
-#include "postgres.h"
-
-#include "bitvector.h"
-#include "port/pg_bitutils.h"
-#include "utils/varbit.h"
-
-#if PG_VERSION_NUM >= 160000
-#include "varatt.h"
-#endif
-
-/*
- * Allocate and initialize a new bit vector
- */
-VarBit *
-InitBitVector(int dim)
-{
-	VarBit	   *result;
-	int			size;
-
-	size = VARBITTOTALLEN(dim);
-	result = (VarBit *) palloc0(size);
-	SET_VARSIZE(result, size);
-	VARBITLEN(result) = dim;
-
-	return result;
-}
-
-/*
- * Ensure same number of bits
- */
-static inline void
-CheckBitLengths(uint32 aLen, uint32 bLen)
-{
-	if (aLen != bLen)
-		ereport(ERROR,
-				(errcode(ERRCODE_DATA_EXCEPTION),
-				 errmsg("different bit lengths %u and %u", aLen, bLen)));
-}
-
-/*
- * Get the Hamming distance between two bit strings
- */
-PGDLLEXPORT PG_FUNCTION_INFO_V1(hamming_distance);
-Datum
-hamming_distance(PG_FUNCTION_ARGS)
-{
-	VarBit	   *a = PG_GETARG_VARBIT_P(0);
-	VarBit	   *b = PG_GETARG_VARBIT_P(1);
-	unsigned char *ax = VARBITS(a);
-	unsigned char *bx = VARBITS(b);
-	uint64		distance = 0;
-
-	CheckBitLengths(VARBITLEN(a), VARBITLEN(b));
-
-	/* TODO Improve performance */
-	for (uint32 i = 0; i < VARBITBYTES(a); i++)
-		distance += pg_number_of_ones[ax[i] ^ bx[i]];
-
-	PG_RETURN_FLOAT8((double) distance);
-}
-
-/*
- * Get the Jaccard distance between two bit strings
- */
-PGDLLEXPORT PG_FUNCTION_INFO_V1(jaccard_distance);
-Datum
-jaccard_distance(PG_FUNCTION_ARGS)
-{
-	VarBit	   *a = PG_GETARG_VARBIT_P(0);
-	VarBit	   *b = PG_GETARG_VARBIT_P(1);
-	unsigned char *ax = VARBITS(a);
-	unsigned char *bx = VARBITS(b);
-	uint64		ab = 0;
-	uint64		aa;
-	uint64		bb;
-
-	CheckBitLengths(VARBITLEN(a), VARBITLEN(b));
-
-	/* TODO Improve performance */
-	for (uint32 i = 0; i < VARBITBYTES(a); i++)
-		ab += pg_number_of_ones[ax[i] & bx[i]];
-
-	if (ab == 0)
-		PG_RETURN_FLOAT8(1);
-
-	aa = pg_popcount((char *) ax, VARBITBYTES(a));
-	bb = pg_popcount((char *) bx, VARBITBYTES(b));
-
-	PG_RETURN_FLOAT8(1 - (ab / ((double) (aa + bb - ab))));
-}

src/bitvector.h

@@ -1,8 +0,0 @@
-#ifndef BITVECTOR_H
-#define BITVECTOR_H
-
-#include "utils/varbit.h"
-
-VarBit	   *InitBitVector(int dim);
-
-#endif

src/hnsw.c

@@ -4,16 +4,13 @@
 #include <math.h>
 
 #include "access/amapi.h"
-#include "access/reloptions.h"
-#include "commands/progress.h"
 #include "commands/vacuum.h"
 #include "hnsw.h"
-#include "miscadmin.h"
 #include "utils/guc.h"
 #include "utils/selfuncs.h"
 
-#if PG_VERSION_NUM < 150000
+#if PG_VERSION_NUM >= 120000
-#define MarkGUCPrefixReserved(x) EmitWarningsOnPlaceholders(x)
+#include "commands/progress.h"
 #endif
 
 int			hnsw_ef_search;
@@ -29,7 +26,7 @@ static relopt_kind hnsw_relopt_kind;
  * this grows bigger, we should use a shmem_request_hook and
  * RequestAddinShmemSpace() to pre-reserve space for this.
  */
-void
+static void
 HnswInitLockTranche(void)
 {
 	int		   *tranche_ids;
@@ -54,7 +51,6 @@ HnswInitLockTranche(void)
 void
 HnswInit(void)
 {
-	if (!process_shared_preload_libraries_in_progress)
 	HnswInitLockTranche();
 
 	hnsw_relopt_kind = add_reloption_kind();
@@ -74,13 +70,12 @@ HnswInit(void)
 	DefineCustomIntVariable("hnsw.ef_search", "Sets the size of the dynamic candidate list for search",
 							"Valid range is 1..1000.", &hnsw_ef_search,
 							HNSW_DEFAULT_EF_SEARCH, HNSW_MIN_EF_SEARCH, HNSW_MAX_EF_SEARCH, PGC_USERSET, 0, NULL, NULL, NULL);
-
-	MarkGUCPrefixReserved("hnsw");
 }
 
 /*
  * Get the name of index build phase
  */
+#if PG_VERSION_NUM >= 120000
 static char *
 hnswbuildphasename(int64 phasenum)
 {
@@ -94,6 +89,7 @@ hnswbuildphasename(int64 phasenum)
 			return NULL;
 	}
 }
+#endif
 
 /*
  * Estimate the cost of an index scan
@@ -108,6 +104,9 @@ hnswcostestimate(PlannerInfo *root, IndexPath *path, double loop_count,
 	int			m;
 	int			entryLevel;
 	Relation	index;
+#if PG_VERSION_NUM < 120000
+	List	   *qinfos;
+#endif
 
 	/* Never use index without order */
 	if (path->indexorderbys == NULL)
@@ -133,7 +132,12 @@ hnswcostestimate(PlannerInfo *root, IndexPath *path, double loop_count,
 	/* Account for number of tuples (or entry level), m, and ef_search */
 	costs.numIndexTuples = (entryLevel + 2) * m;
 
+#if PG_VERSION_NUM >= 120000
 	genericcostestimate(root, path, loop_count, &costs);
+#else
+	qinfos = deconstruct_indexquals(path);
+	genericcostestimate(root, path, loop_count, qinfos, &costs);
+#endif
 
 	/* Use total cost since most work happens before first tuple is returned */
 	*indexStartupCost = costs.indexTotalCost;
@@ -227,7 +231,9 @@ hnswhandler(PG_FUNCTION_ARGS)
 	amroutine->amcostestimate = hnswcostestimate;
 	amroutine->amoptions = hnswoptions;
 	amroutine->amproperty = NULL;	/* TODO AMPROP_DISTANCE_ORDERABLE */
+#if PG_VERSION_NUM >= 120000
 	amroutine->ambuildphasename = hnswbuildphasename;
+#endif
 	amroutine->amvalidate = hnswvalidate;
 #if PG_VERSION_NUM >= 140000
 	amroutine->amadjustmembers = NULL;

src/hnsw.h

@@ -3,17 +3,21 @@
 
 #include "postgres.h"
 
-#include "access/genam.h"
+#include "access/generic_xlog.h"
 #include "access/parallel.h"
-#include "lib/pairingheap.h"
+#include "access/reloptions.h"
 #include "nodes/execnodes.h"
 #include "port.h"				/* for random() */
 #include "utils/relptr.h"
 #include "utils/sampling.h"
 #include "vector.h"
 
+#if PG_VERSION_NUM < 110000
+#error "Requires PostgreSQL 11+"
+#endif
+
 #if PG_VERSION_NUM < 120000
-#error "Requires PostgreSQL 12+"
+#include "access/relscan.h"
 #endif
 
 #define HNSW_MAX_DIM 2000
@@ -60,13 +64,10 @@
 #define PROGRESS_HNSW_PHASE_LOAD		2
 
 #define HNSW_MAX_SIZE (BLCKSZ - MAXALIGN(SizeOfPageHeaderData) - MAXALIGN(sizeof(HnswPageOpaqueData)) - sizeof(ItemIdData))
-#define HNSW_TUPLE_ALLOC_SIZE BLCKSZ
 
 #define HNSW_ELEMENT_TUPLE_SIZE(size)	MAXALIGN(offsetof(HnswElementTupleData, data) + (size))
 #define HNSW_NEIGHBOR_TUPLE_SIZE(level, m)	MAXALIGN(offsetof(HnswNeighborTupleData, indextids) + ((level) + 2) * (m) * sizeof(ItemPointerData))
 
-#define HNSW_NEIGHBOR_ARRAY_SIZE(lm)	(offsetof(HnswNeighborArray, items) + sizeof(HnswCandidate) * (lm))
-
 #define HnswPageGetOpaque(page)	((HnswPageOpaque) PageGetSpecialPointer(page))
 #define HnswPageGetMeta(page)	((HnswMetaPageData *) PageGetContents(page))
 
@@ -80,7 +81,7 @@
 
 #if PG_VERSION_NUM < 130000
 #define list_delete_last(list) list_truncate(list, list_length(list) - 1)
-#define list_sort(list, cmp) ((list) = list_qsort(list, cmp))
+#define list_sort(list, cmp) list_qsort(list, cmp)
 #endif
 
 #define HnswIsElementTuple(tup) ((tup)->type == HNSW_ELEMENT_TUPLE_TYPE)
@@ -113,6 +114,8 @@
 
 /* Variables */
 extern int	hnsw_ef_search;
+
+/* This is initialized when the module is loaded */
 extern int	hnsw_lock_tranche_id;
 
 typedef struct HnswElementData HnswElementData;
@@ -129,7 +132,7 @@ HnswPtrDeclare(HnswNeighborArray, HnswNeighborArrayRelptr, HnswNeighborArrayPtr)
 HnswPtrDeclare(HnswNeighborArrayPtr, HnswNeighborsRelptr, HnswNeighborsPtr);
 HnswPtrDeclare(char, DatumRelptr, DatumPtr);
 
-struct HnswElementData
+typedef struct HnswElementData
 {
 	HnswElementPtr next;
 	ItemPointerData heaptids[HNSW_HEAPTIDS];
@@ -144,7 +147,7 @@ struct HnswElementData
 	BlockNumber neighborPage;
 	DatumPtr	value;
 	LWLock		lock;
-};
+}			HnswElementData;
 
 typedef HnswElementData * HnswElement;
 
@@ -155,12 +158,12 @@ typedef struct HnswCandidate
 	bool		closer;
 }			HnswCandidate;
 
-struct HnswNeighborArray
+typedef struct HnswNeighborArray
 {
 	int			length;
 	bool		closerSet;
 	HnswCandidate items[FLEXIBLE_ARRAY_MEMBER];
-};
+}			HnswNeighborArray;
 
 typedef struct HnswPairingHeapNode
 {
@@ -185,7 +188,6 @@ typedef struct HnswGraph
 
 	/* Entry state */
 	LWLock		entryLock;
-	LWLock		entryWaitLock;
 	HnswElementPtr entryPoint;
 
 	/* Allocations state */
@@ -198,12 +200,20 @@ typedef struct HnswGraph
 	bool		flushed;
 }			HnswGraph;
 
+typedef struct HnswSpool
+{
+	Relation	heap;
+	Relation	index;
+}			HnswSpool;
+
 typedef struct HnswShared
 {
 	/* Immutable state */
 	Oid			heaprelid;
 	Oid			indexrelid;
 	bool		isconcurrent;
+	int			scantuplesortstates;
+	dsa_pointer dsaptr;
 
 	/* Worker progress */
 	ConditionVariable workersdonecv;
@@ -215,10 +225,16 @@ typedef struct HnswShared
 	int			nparticipantsdone;
 	double		reltuples;
 	HnswGraph	graphData;
+
+#if PG_VERSION_NUM < 120000
+	ParallelHeapScanDescData heapdesc;	/* must come last */
+#endif
 }			HnswShared;
 
+#if PG_VERSION_NUM >= 120000
 #define ParallelTableScanFromHnswShared(shared) \
 	(ParallelTableScanDesc) ((char *) (shared) + BUFFERALIGN(sizeof(HnswShared)))
+#endif
 
 typedef struct HnswLeader
 {
@@ -377,20 +393,19 @@ void		HnswGetMetaPageInfo(Relation index, int *m, HnswElement * entryPoint);
 void	   *HnswAlloc(HnswAllocator * allocator, Size size);
 HnswElement HnswInitElement(char *base, ItemPointer tid, int m, double ml, int maxLevel, HnswAllocator * alloc);
 HnswElement HnswInitElementFromBlock(BlockNumber blkno, OffsetNumber offno);
-void		HnswFindElementNeighbors(char *base, HnswElement element, HnswElement entryPoint, Relation index, FmgrInfo *procinfo, Oid collation, int m, int efConstruction, bool existing);
+void		HnswInsertElement(char *base, HnswElement element, HnswElement entryPoint, Relation index, FmgrInfo *procinfo, Oid collation, int m, int efConstruction, bool existing);
 HnswCandidate *HnswEntryCandidate(char *base, HnswElement em, Datum q, Relation rel, FmgrInfo *procinfo, Oid collation, bool loadVec);
 void		HnswUpdateMetaPage(Relation index, int updateEntry, HnswElement entryPoint, BlockNumber insertPage, ForkNumber forkNum, bool building);
 void		HnswSetNeighborTuple(char *base, HnswNeighborTuple ntup, HnswElement e, int m);
 void		HnswAddHeapTid(HnswElement element, ItemPointer heaptid);
 void		HnswInitNeighbors(char *base, HnswElement element, int m, HnswAllocator * alloc);
-bool		HnswInsertTupleOnDisk(Relation index, Datum value, Datum *values, bool *isnull, ItemPointer heap_tid, bool building);
+bool		HnswInsertTupleOnDisk(Relation index, Datum value, Datum *values, bool *isnull, ItemPointer heap_tid, Relation heapRel, bool building);
-void		HnswUpdateNeighborsOnDisk(Relation index, FmgrInfo *procinfo, Oid collation, HnswElement e, int m, bool checkExisting, bool building);
+void		HnswUpdateNeighborPages(Relation index, FmgrInfo *procinfo, Oid collation, HnswElement e, int m, bool checkExisting, bool building);
 void		HnswLoadElementFromTuple(HnswElement element, HnswElementTuple etup, bool loadHeaptids, bool loadVec);
 void		HnswLoadElement(HnswElement element, float *distance, Datum *q, Relation index, FmgrInfo *procinfo, Oid collation, bool loadVec);
 void		HnswSetElementTuple(char *base, HnswElementTuple etup, HnswElement element);
-void		HnswUpdateConnection(char *base, HnswElement element, HnswCandidate * hc, int lm, int lc, int *updateIdx, Relation index, FmgrInfo *procinfo, Oid collation);
+void		HnswUpdateConnection(char *base, HnswElement element, HnswCandidate * hc, int m, int lc, int *updateIdx, Relation index, FmgrInfo *procinfo, Oid collation);
 void		HnswLoadNeighbors(HnswElement element, Relation index, int m);
-void		HnswInitLockTranche(void);
 PGDLLEXPORT void HnswParallelBuildMain(dsm_segment *seg, shm_toc *toc);
 
 /* Index access methods */

src/hnswbuild.c

@@ -1,54 +1,14 @@
-/*
- * The HNSW build happens in two phases:
- *
- * 1. In-memory phase
- *
- * In this first phase, the graph is held completely in memory. When the graph
- * is fully built, or we run out of memory reserved for the build (determined
- * by maintenance_work_mem), we materialize the graph to disk (see
- * FlushPages()), and switch to the on-disk phase.
- *
- * In a parallel build, a large contiguous chunk of shared memory is allocated
- * to hold the graph. Each worker process has its own HnswBuildState struct in
- * private memory, which contains information that doesn't change throughout
- * the build, and pointers to the shared structs in shared memory. The shared
- * memory area is mapped to a different address in each worker process, and
- * 'HnswBuildState.hnswarea' points to the beginning of the shared area in the
- * worker process's address space. All pointers used in the graph are
- * "relative pointers", stored as an offset from 'hnswarea'.
- *
- * Each element is protected by an LWLock. It must be held when reading or
- * modifying the element's neighbors or 'heaptids'.
- *
- * In a non-parallel build, the graph is held in backend-private memory. All
- * the elements are allocated in a dedicated memory context, 'graphCtx', and
- * the pointers used in the graph are regular pointers.
- *
- * 2. On-disk phase
- *
- * In the on-disk phase, the index is built by inserting each vector to the
- * index one by one, just like on INSERT. The only difference is that we don't
- * WAL-log the individual inserts. If the graph fit completely in memory and
- * was fully built in the in-memory phase, the on-disk phase is skipped.
- *
- * After we have finished building the graph, we perform one more scan through
- * the index and write all the pages to the WAL.
- */
 #include "postgres.h"
 
 #include <math.h>
 
 #include "access/parallel.h"
-#include "access/table.h"
-#include "access/tableam.h"
 #include "access/xact.h"
-#include "access/xloginsert.h"
 #include "catalog/index.h"
-#include "catalog/pg_type_d.h"
-#include "commands/progress.h"
 #include "hnsw.h"
 #include "miscadmin.h"
-#include "optimizer/optimizer.h"
+#include "lib/pairingheap.h"
+#include "nodes/pg_list.h"
 #include "storage/bufmgr.h"
 #include "tcop/tcopprot.h"
 #include "utils/datum.h"
@@ -56,21 +16,43 @@
 
 #if PG_VERSION_NUM >= 140000
 #include "utils/backend_progress.h"
-#else
+#elif PG_VERSION_NUM >= 120000
 #include "pgstat.h"
 #endif
 
+#if PG_VERSION_NUM >= 120000
+#include "access/tableam.h"
+#include "commands/progress.h"
+#else
+#define PROGRESS_CREATEIDX_TUPLES_DONE 0
+#endif
+
 #if PG_VERSION_NUM >= 130000
 #define CALLBACK_ITEM_POINTER ItemPointer tid
 #else
 #define CALLBACK_ITEM_POINTER HeapTuple hup
 #endif
 
+#if PG_VERSION_NUM >= 120000
+#define UpdateProgress(index, val) pgstat_progress_update_param(index, val)
+#else
+#define UpdateProgress(index, val) ((void)val)
+#endif
+
 #if PG_VERSION_NUM >= 140000
 #include "utils/backend_status.h"
 #include "utils/wait_event.h"
 #endif
 
+#if PG_VERSION_NUM >= 120000
+#include "access/table.h"
+#include "optimizer/optimizer.h"
+#else
+#include "access/heapam.h"
+#include "optimizer/planner.h"
+#include "pgstat.h"
+#endif
+
 #define PARALLEL_KEY_HNSW_SHARED		UINT64CONST(0xA000000000000001)
 #define PARALLEL_KEY_HNSW_AREA			UINT64CONST(0xA000000000000002)
 #define PARALLEL_KEY_QUERY_TEXT			UINT64CONST(0xA000000000000003)
@@ -142,13 +124,14 @@ HnswBuildAppendPage(Relation index, Buffer *buf, Page *page, ForkNumber forkNum)
 }
 
 /*
- * Create graph pages
+ * Create element pages
  */
 static void
-CreateGraphPages(HnswBuildState * buildstate)
+CreateElementPages(HnswBuildState * buildstate)
 {
 	Relation	index = buildstate->index;
 	ForkNumber	forkNum = buildstate->forkNum;
+	Size		etupAllocSize;
 	Size		maxSize;
 	HnswElementTuple etup;
 	HnswNeighborTuple ntup;
@@ -160,11 +143,12 @@ CreateGraphPages(HnswBuildState * buildstate)
 	char	   *base = buildstate->hnswarea;
 
 	/* Calculate sizes */
+	etupAllocSize = BLCKSZ;
 	maxSize = HNSW_MAX_SIZE;
 
 	/* Allocate once */
-	etup = palloc0(HNSW_TUPLE_ALLOC_SIZE);
+	etup = palloc0(etupAllocSize);
-	ntup = palloc0(HNSW_TUPLE_ALLOC_SIZE);
+	ntup = palloc0(BLCKSZ);
 
 	/* Prepare first page */
 	buf = HnswNewBuffer(index, forkNum);
@@ -177,13 +161,13 @@ CreateGraphPages(HnswBuildState * buildstate)
 		Size		etupSize;
 		Size		ntupSize;
 		Size		combinedSize;
-		Pointer		valuePtr = HnswPtrAccess(base, element->value);
+		void	   *valuePtr = HnswPtrAccess(base, element->value);
 
 		/* Update iterator */
 		iter = element->next;
 
 		/* Zero memory for each element */
-		MemSet(etup, 0, HNSW_TUPLE_ALLOC_SIZE);
+		MemSet(etup, 0, etupAllocSize);
 
 		/* Calculate sizes */
 		etupSize = HNSW_ELEMENT_TUPLE_SIZE(VARSIZE_ANY(valuePtr));
@@ -191,7 +175,7 @@ CreateGraphPages(HnswBuildState * buildstate)
 		combinedSize = etupSize + ntupSize + sizeof(ItemIdData);
 
 		/* Initial size check */
-		if (etupSize > HNSW_TUPLE_ALLOC_SIZE)
+		if (etupSize > etupAllocSize)
 			elog(ERROR, "index tuple too large");
 
 		HnswSetElementTuple(base, etup, element);
@@ -243,10 +227,10 @@ CreateGraphPages(HnswBuildState * buildstate)
 }
 
 /*
- * Write neighbor tuples
+ * Create neighbor pages
  */
 static void
-WriteNeighborTuples(HnswBuildState * buildstate)
+CreateNeighborPages(HnswBuildState * buildstate)
 {
 	Relation	index = buildstate->index;
 	ForkNumber	forkNum = buildstate->forkNum;
@@ -256,32 +240,29 @@ WriteNeighborTuples(HnswBuildState * buildstate)
 	HnswNeighborTuple ntup;
 
 	/* Allocate once */
-	ntup = palloc0(HNSW_TUPLE_ALLOC_SIZE);
+	ntup = palloc0(BLCKSZ);
 
 	while (!HnswPtrIsNull(base, iter))
 	{
-		HnswElement element = HnswPtrAccess(base, iter);
+		HnswElement e = HnswPtrAccess(base, iter);
 		Buffer		buf;
 		Page		page;
-		Size		ntupSize = HNSW_NEIGHBOR_TUPLE_SIZE(element->level, m);
+		Size		ntupSize = HNSW_NEIGHBOR_TUPLE_SIZE(e->level, m);
 
 		/* Update iterator */
-		iter = element->next;
+		iter = e->next;
-
-		/* Zero memory for each element */
-		MemSet(ntup, 0, HNSW_TUPLE_ALLOC_SIZE);
 
 		/* Can take a while, so ensure we can interrupt */
 		/* Needs to be called when no buffer locks are held */
 		CHECK_FOR_INTERRUPTS();
 
-		buf = ReadBufferExtended(index, forkNum, element->neighborPage, RBM_NORMAL, NULL);
+		buf = ReadBufferExtended(index, forkNum, e->neighborPage, RBM_NORMAL, NULL);
 		LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);
 		page = BufferGetPage(buf);
 
-		HnswSetNeighborTuple(base, ntup, element, m);
+		HnswSetNeighborTuple(base, ntup, e, m);
 
-		if (!PageIndexTupleOverwrite(page, element->neighborOffno, (Item) ntup, ntupSize))
+		if (!PageIndexTupleOverwrite(page, e->neighborOffno, (Item) ntup, ntupSize))
 			elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
 
 		/* Commit */
@@ -303,8 +284,8 @@ FlushPages(HnswBuildState * buildstate)
 #endif
 
 	CreateMetaPage(buildstate);
-	CreateGraphPages(buildstate);
+	CreateElementPages(buildstate);
-	WriteNeighborTuples(buildstate);
+	CreateNeighborPages(buildstate);
 
 	buildstate->graph->flushed = true;
 	MemoryContextReset(buildstate->graphCtx);
@@ -314,7 +295,7 @@ FlushPages(HnswBuildState * buildstate)
  * Add a heap TID to an existing element
  */
 static bool
-AddDuplicateInMemory(HnswElement element, HnswElement dup)
+HnswAddDuplicateInMemory(HnswElement element, HnswElement dup)
 {
 	LWLockAcquire(&dup->lock, LW_EXCLUSIVE);
 
@@ -335,15 +316,15 @@ AddDuplicateInMemory(HnswElement element, HnswElement dup)
  * Find duplicate element
  */
 static bool
-FindDuplicateInMemory(char *base, HnswElement element)
+HnswFindDuplicateInMemory(char *base, HnswElement element)
 {
 	HnswNeighborArray *neighbors = HnswGetNeighbors(base, element, 0);
-	Datum		value = HnswGetValue(base, element);
 
 	for (int i = 0; i < neighbors->length; i++)
 	{
 		HnswCandidate *neighbor = &neighbors->items[i];
 		HnswElement neighborElement = HnswPtrAccess(base, neighbor->element);
+		Datum		value = HnswGetValue(base, element);
 		Datum		neighborValue = HnswGetValue(base, neighborElement);
 
 		/* Exit early since ordered by distance */
@@ -351,7 +332,7 @@ FindDuplicateInMemory(char *base, HnswElement element)
 			return false;
 
 		/* Check for space */
-		if (AddDuplicateInMemory(element, neighborElement))
+		if (HnswAddDuplicateInMemory(element, neighborElement))
 			return true;
 	}
 
@@ -359,10 +340,10 @@ FindDuplicateInMemory(char *base, HnswElement element)
 }
 
 /*
- * Add to element list
+ * Add to element and neighbor pages
  */
 static void
-AddElementInMemory(char *base, HnswGraph * graph, HnswElement element)
+WriteNewElementPagesInMemory(char *base, HnswGraph * graph, HnswElement element)
 {
 	SpinLockAcquire(&graph->lock);
 	element->next = graph->head;
@@ -374,7 +355,7 @@ AddElementInMemory(char *base, HnswGraph * graph, HnswElement element)
  * Update neighbors
  */
 static void
-UpdateNeighborsInMemory(char *base, FmgrInfo *procinfo, Oid collation, HnswElement e, int m)
+HnswUpdateNeighborPagesInMemory(char *base, FmgrInfo *procinfo, Oid collation, HnswElement e, int m)
 {
 	for (int lc = e->level; lc >= 0; lc--)
 	{
@@ -398,89 +379,45 @@ UpdateNeighborsInMemory(char *base, FmgrInfo *procinfo, Oid collation, HnswEleme
 }
 
 /*
- * Update graph in memory
+ * Write changes in memory
  */
 static void
-UpdateGraphInMemory(FmgrInfo *procinfo, Oid collation, HnswElement element, int m, int efConstruction, HnswElement entryPoint, HnswBuildState * buildstate)
+WriteElementInMemory(Relation index, FmgrInfo *procinfo, Oid collation, HnswElement element, int m, int efConstruction, HnswElement entryPoint, HnswBuildState * buildstate, HnswGraph * graph, bool updateEntryPoint)
 {
-	HnswGraph  *graph = buildstate->graph;
 	char	   *base = buildstate->hnswarea;
 
-	/* Look for duplicate */
+	/* Try to add to existing page */
-	if (FindDuplicateInMemory(base, element))
+	if (HnswFindDuplicateInMemory(base, element))
 		return;
 
-	/* Add element */
+	/* Write element and neighbor tuples */
-	AddElementInMemory(base, graph, element);
+	WriteNewElementPagesInMemory(base, graph, element);
 
 	/* Update neighbors */
-	UpdateNeighborsInMemory(base, procinfo, collation, element, m);
+	HnswUpdateNeighborPagesInMemory(base, procinfo, collation, element, m);
 
 	/* Update entry point if needed (already have lock) */
-	if (entryPoint == NULL || element->level > entryPoint->level)
+	if (updateEntryPoint)
 		HnswPtrStore(base, graph->entryPoint, element);
 }
 
-/*
- * Insert tuple in memory
- */
-static void
-InsertTupleInMemory(HnswBuildState * buildstate, HnswElement element)
-{
-	FmgrInfo   *procinfo = buildstate->procinfo;
-	Oid			collation = buildstate->collation;
-	HnswGraph  *graph = buildstate->graph;
-	HnswElement entryPoint;
-	LWLock	   *entryLock = &graph->entryLock;
-	LWLock	   *entryWaitLock = &graph->entryWaitLock;
-	int			efConstruction = buildstate->efConstruction;
-	int			m = buildstate->m;
-	char	   *base = buildstate->hnswarea;
-
-	/* Wait if another process needs exclusive lock on entry lock */
-	LWLockAcquire(entryWaitLock, LW_EXCLUSIVE);
-	LWLockRelease(entryWaitLock);
-
-	/* Get entry point */
-	LWLockAcquire(entryLock, LW_SHARED);
-	entryPoint = HnswPtrAccess(base, graph->entryPoint);
-
-	/* Prevent concurrent inserts when likely updating entry point */
-	if (entryPoint == NULL || element->level > entryPoint->level)
-	{
-		/* Release shared lock */
-		LWLockRelease(entryLock);
-
-		/* Tell other processes to wait and get exclusive lock */
-		LWLockAcquire(entryWaitLock, LW_EXCLUSIVE);
-		LWLockAcquire(entryLock, LW_EXCLUSIVE);
-		LWLockRelease(entryWaitLock);
-
-		/* Get latest entry point after lock is acquired */
-		entryPoint = HnswPtrAccess(base, graph->entryPoint);
-	}
-
-	/* Find neighbors for element */
-	HnswFindElementNeighbors(base, element, entryPoint, NULL, procinfo, collation, m, efConstruction, false);
-
-	/* Update graph in memory */
-	UpdateGraphInMemory(procinfo, collation, element, m, efConstruction, entryPoint, buildstate);
-
-	/* Release entry lock */
-	LWLockRelease(entryLock);
-}
-
 /*
  * Insert tuple
  */
 static bool
 InsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heaptid, HnswBuildState * buildstate)
 {
+	FmgrInfo   *procinfo = buildstate->procinfo;
+	Oid			collation = buildstate->collation;
 	HnswGraph  *graph = buildstate->graph;
+	HnswElement entryPoint;
+	int			efConstruction = buildstate->efConstruction;
+	int			m = buildstate->m;
 	HnswElement element;
 	HnswAllocator *allocator = &buildstate->allocator;
 	Size		valueSize;
 	Pointer		valuePtr;
+	bool		updateEntryPoint;
 	LWLock	   *flushLock = &graph->flushLock;
 	char	   *base = buildstate->hnswarea;
 
@@ -490,7 +427,7 @@ InsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heaptid, Hn
 	/* Normalize if needed */
 	if (buildstate->normprocinfo != NULL)
 	{
-		if (!HnswNormValue(buildstate->normprocinfo, buildstate->collation, &value, buildstate->normvec))
+		if (!HnswNormValue(buildstate->normprocinfo, collation, &value, buildstate->normvec))
 			return false;
 	}
 
@@ -500,24 +437,17 @@ InsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heaptid, Hn
 	/* Ensure graph not flushed when inserting */
 	LWLockAcquire(flushLock, LW_SHARED);
 
-	/* Are we in the on-disk phase? */
 	if (graph->flushed)
 	{
 		LWLockRelease(flushLock);
 
-		return HnswInsertTupleOnDisk(index, value, values, isnull, heaptid, true);
+		return HnswInsertTupleOnDisk(index, value, values, isnull, heaptid, buildstate->heap, true);
 	}
 
-	/*
+	/* Get lock for allocator */
-	 * In a parallel build, the HnswElement is allocated from the shared
-	 * memory area, so we need to coordinate with other processes.
-	 */
 	LWLockAcquire(&graph->allocatorLock, LW_EXCLUSIVE);
 
-	/*
+	/* Flush pages if needed */
-	 * Check that we have enough memory available for the new element now that
-	 * we have the allocator lock, and flush pages if needed.
-	 */
 	if (graph->memoryUsed >= graph->memoryTotal)
 	{
 		LWLockRelease(&graph->allocatorLock);
@@ -537,29 +467,41 @@ InsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heaptid, Hn
 
 		LWLockRelease(flushLock);
 
-		return HnswInsertTupleOnDisk(index, value, values, isnull, heaptid, true);
+		return HnswInsertTupleOnDisk(index, value, values, isnull, heaptid, buildstate->heap, true);
 	}
 
-	/* Ok, we can proceed to allocate the element */
+	/* Create an element */
 	element = HnswInitElement(base, heaptid, buildstate->m, buildstate->ml, buildstate->maxLevel, allocator);
 	valuePtr = HnswAlloc(allocator, valueSize);
 
-	/*
+	/* Release allocator lock */
-	 * We have now allocated the space needed for the element, so we don't
-	 * need the allocator lock anymore. Release it and initialize the rest of
-	 * the element.
-	 */
 	LWLockRelease(&graph->allocatorLock);
 
-	/* Copy the datum */
+	/* Copy datum */
 	memcpy(valuePtr, DatumGetPointer(value), valueSize);
 	HnswPtrStore(base, element->value, valuePtr);
 
-	/* Create a lock for the element */
+	/* Create element lock */
 	LWLockInitialize(&element->lock, hnsw_lock_tranche_id);
 
-	/* Insert tuple */
+	/* Get entry point */
-	InsertTupleInMemory(buildstate, element);
+	LWLockAcquire(&graph->entryLock, LW_EXCLUSIVE);
+	entryPoint = HnswPtrAccess(base, graph->entryPoint);
+	updateEntryPoint = entryPoint == NULL || element->level > entryPoint->level;
+
+	/* Release lock if not updating entry point */
+	if (!updateEntryPoint)
+		LWLockRelease(&graph->entryLock);
+
+	/* Insert element in graph */
+	HnswInsertElement(base, element, entryPoint, NULL, procinfo, collation, m, efConstruction, false);
+
+	/* Write to memory */
+	WriteElementInMemory(index, procinfo, collation, element, m, efConstruction, entryPoint, buildstate, graph, updateEntryPoint);
+
+	/* Release lock if needed */
+	if (updateEntryPoint)
+		LWLockRelease(&graph->entryLock);
 
 	/* Release flush lock */
 	LWLockRelease(flushLock);
@@ -594,7 +536,7 @@ BuildCallback(Relation index, CALLBACK_ITEM_POINTER, Datum *values,
 	{
 		/* Update progress */
 		SpinLockAcquire(&graph->lock);
-		pgstat_progress_update_param(PROGRESS_CREATEIDX_TUPLES_DONE, ++graph->indtuples);
+		UpdateProgress(PROGRESS_CREATEIDX_TUPLES_DONE, ++graph->indtuples);
 		SpinLockRelease(&graph->lock);
 	}
 
@@ -609,9 +551,6 @@ BuildCallback(Relation index, CALLBACK_ITEM_POINTER, Datum *values,
 static void
 InitGraph(HnswGraph * graph, char *base, long memoryTotal)
 {
-	/* Initialize the lock tranche if needed */
-	HnswInitLockTranche();
-
 	HnswPtrStore(base, graph->head, (HnswElement) NULL);
 	HnswPtrStore(base, graph->entryPoint, (HnswElement) NULL);
 	graph->memoryUsed = 0;
@@ -620,7 +559,6 @@ InitGraph(HnswGraph * graph, char *base, long memoryTotal)
 	graph->indtuples = 0;
 	SpinLockInit(&graph->lock);
 	LWLockInitialize(&graph->entryLock, hnsw_lock_tranche_id);
-	LWLockInitialize(&graph->entryWaitLock, hnsw_lock_tranche_id);
 	LWLockInitialize(&graph->allocatorLock, hnsw_lock_tranche_id);
 	LWLockInitialize(&graph->flushLock, hnsw_lock_tranche_id);
 }
@@ -672,12 +610,6 @@ HnswSharedMemoryAlloc(Size size, void *state)
 static void
 InitBuildState(HnswBuildState * buildstate, Relation heap, Relation index, IndexInfo *indexInfo, ForkNumber forkNum)
 {
-	int			maxDimensions = HNSW_MAX_DIM;
-	Oid			typid = TupleDescAttr(index->rd_att, 0)->atttypid;
-
-	if (typid == BITOID || typid == VARBITOID)
-		maxDimensions *= 32;
-
 	buildstate->heap = heap;
 	buildstate->index = index;
 	buildstate->indexInfo = indexInfo;
@@ -691,8 +623,8 @@ InitBuildState(HnswBuildState * buildstate, Relation heap, Relation index, Index
 	if (buildstate->dimensions < 0)
 		elog(ERROR, "column does not have dimensions");
 
-	if (buildstate->dimensions > maxDimensions)
+	if (buildstate->dimensions > HNSW_MAX_DIM)
-		elog(ERROR, "column cannot have more than %d dimensions for hnsw index", maxDimensions);
+		elog(ERROR, "column cannot have more than %d dimensions for hnsw index", HNSW_MAX_DIM);
 
 	if (buildstate->efConstruction < 2 * buildstate->m)
 		elog(ERROR, "ef_construction must be greater than or equal to 2 * m");
@@ -778,25 +710,36 @@ ParallelHeapScan(HnswBuildState * buildstate)
  * Perform a worker's portion of a parallel insert
  */
 static void
-HnswParallelScanAndInsert(Relation heapRel, Relation indexRel, HnswShared * hnswshared, char *hnswarea, bool progress)
+HnswParallelScanAndInsert(HnswSpool * hnswspool, HnswShared * hnswshared, char *hnswarea, bool progress)
 {
 	HnswBuildState buildstate;
+#if PG_VERSION_NUM >= 120000
 	TableScanDesc scan;
+#else
+	HeapScanDesc scan;
+#endif
 	double		reltuples;
 	IndexInfo  *indexInfo;
 
 	/* Join parallel scan */
-	indexInfo = BuildIndexInfo(indexRel);
+	indexInfo = BuildIndexInfo(hnswspool->index);
 	indexInfo->ii_Concurrent = hnswshared->isconcurrent;
-	InitBuildState(&buildstate, heapRel, indexRel, indexInfo, MAIN_FORKNUM);
+	InitBuildState(&buildstate, hnswspool->heap, hnswspool->index, indexInfo, MAIN_FORKNUM);
 	buildstate.graph = &hnswshared->graphData;
 	buildstate.hnswarea = hnswarea;
 	InitAllocator(&buildstate.allocator, &HnswSharedMemoryAlloc, &buildstate);
-	scan = table_beginscan_parallel(heapRel,
+#if PG_VERSION_NUM >= 120000
+	scan = table_beginscan_parallel(hnswspool->heap,
 									ParallelTableScanFromHnswShared(hnswshared));
-	reltuples = table_index_build_scan(heapRel, indexRel, indexInfo,
+	reltuples = table_index_build_scan(hnswspool->heap, hnswspool->index, indexInfo,
 									   true, progress, BuildCallback,
 									   (void *) &buildstate, scan);
+#else
+	scan = heap_beginscan_parallel(hnswspool->heap, &hnswshared->heapdesc);
+	reltuples = IndexBuildHeapScan(hnswspool->heap, hnswspool->index, indexInfo,
+								   true, BuildCallback,
+								   (void *) &buildstate, scan);
+#endif
 
 	/* Record statistics */
 	SpinLockAcquire(&hnswshared->mutex);
@@ -823,7 +766,10 @@ void
 HnswParallelBuildMain(dsm_segment *seg, shm_toc *toc)
 {
 	char	   *sharedquery;
+	HnswSpool  *hnswspool;
 	HnswShared *hnswshared;
+	dsa_area   *dsa;
+	void	   *dsaspace;
 	char	   *hnswarea;
 	Relation	heapRel;
 	Relation	indexRel;
@@ -853,17 +799,32 @@ HnswParallelBuildMain(dsm_segment *seg, shm_toc *toc)
 	}
 
 	/* Open relations within worker */
+#if PG_VERSION_NUM >= 120000
 	heapRel = table_open(hnswshared->heaprelid, heapLockmode);
+#else
+	heapRel = heap_open(hnswshared->heaprelid, heapLockmode);
+#endif
 	indexRel = index_open(hnswshared->indexrelid, indexLockmode);
 
-	hnswarea = shm_toc_lookup(toc, PARALLEL_KEY_HNSW_AREA, false);
+	/* Initialize worker's own spool */
+	hnswspool = (HnswSpool *) palloc0(sizeof(HnswSpool));
+	hnswspool->heap = heapRel;
+	hnswspool->index = indexRel;
+
+	dsaspace = shm_toc_lookup(toc, PARALLEL_KEY_HNSW_AREA, false);
+	dsa = dsa_attach_in_place(dsaspace, seg);
+	hnswarea = dsa_get_address(dsa, hnswshared->dsaptr);
 
 	/* Perform inserts */
-	HnswParallelScanAndInsert(heapRel, indexRel, hnswshared, hnswarea, false);
+	HnswParallelScanAndInsert(hnswspool, hnswshared, hnswarea, false);
 
 	/* Close relations within worker */
 	index_close(indexRel, indexLockmode);
+#if PG_VERSION_NUM >= 120000
 	table_close(heapRel, heapLockmode);
+#else
+	heap_close(heapRel, heapLockmode);
+#endif
 }
 
 /*
@@ -888,7 +849,19 @@ HnswEndParallel(HnswLeader * hnswleader)
 static Size
 ParallelEstimateShared(Relation heap, Snapshot snapshot)
 {
+#if PG_VERSION_NUM >= 120000
 	return add_size(BUFFERALIGN(sizeof(HnswShared)), table_parallelscan_estimate(heap, snapshot));
+#else
+	if (!IsMVCCSnapshot(snapshot))
+	{
+		Assert(snapshot == SnapshotAny);
+		return sizeof(HnswShared);
+	}
+
+	return add_size(offsetof(HnswShared, heapdesc) +
+					offsetof(ParallelHeapScanDescData, phs_snapshot_data),
+					EstimateSnapshotSpace(snapshot));
+#endif
 }
 
 /*
@@ -898,9 +871,15 @@ static void
 HnswLeaderParticipateAsWorker(HnswBuildState * buildstate)
 {
 	HnswLeader *hnswleader = buildstate->hnswleader;
+	HnswSpool  *leaderworker;
+
+	/* Allocate memory and initialize private spool */
+	leaderworker = (HnswSpool *) palloc0(sizeof(HnswSpool));
+	leaderworker->heap = buildstate->heap;
+	leaderworker->index = buildstate->index;
 
 	/* Perform work common to all participants */
-	HnswParallelScanAndInsert(buildstate->heap, buildstate->index, hnswleader->hnswshared, hnswleader->hnswarea, true);
+	HnswParallelScanAndInsert(leaderworker, hnswleader->hnswshared, hnswleader->hnswarea, true);
 }
 
 /*
@@ -910,11 +889,14 @@ static void
 HnswBeginParallel(HnswBuildState * buildstate, bool isconcurrent, int request)
 {
 	ParallelContext *pcxt;
+	int			scantuplesortstates;
 	Snapshot	snapshot;
 	Size		esthnswshared;
 	Size		esthnswarea;
-	Size		estother;
+	Size		estdsaspace;
 	HnswShared *hnswshared;
+	void	   *dsaspace;
+	dsa_area   *dsa;
 	char	   *hnswarea;
 	HnswLeader *hnswleader = (HnswLeader *) palloc0(sizeof(HnswLeader));
 	bool		leaderparticipates = true;
@@ -927,7 +909,13 @@ HnswBeginParallel(HnswBuildState * buildstate, bool isconcurrent, int request)
 	/* Enter parallel mode and create context */
 	EnterParallelMode();
 	Assert(request > 0);
+#if PG_VERSION_NUM >= 120000
 	pcxt = CreateParallelContext("vector", "HnswParallelBuildMain", request);
+#else
+	pcxt = CreateParallelContext("vector", "HnswParallelBuildMain", request, true);
+#endif
+
+	scantuplesortstates = leaderparticipates ? request + 1 : request;
 
 	/* Get snapshot for table scan */
 	if (!isconcurrent)
@@ -939,15 +927,9 @@ HnswBeginParallel(HnswBuildState * buildstate, bool isconcurrent, int request)
 	esthnswshared = ParallelEstimateShared(buildstate->heap, snapshot);
 	shm_toc_estimate_chunk(&pcxt->estimator, esthnswshared);
 
-	/* Leave space for other objects in shared memory */
+	/* Start with a minimal DSA so InitializeParallelDSM does not fail */
-	/* Docker has a default limit of 64 MB for shm_size */
+	estdsaspace = dsa_minimum_size();
-	/* which happens to be the default value of maintenance_work_mem */
+	shm_toc_estimate_chunk(&pcxt->estimator, estdsaspace);
-	esthnswarea = maintenance_work_mem * 1024L;
-	estother = 3 * 1024 * 1024;
-	if (esthnswarea > estother)
-		esthnswarea -= estother;
-
-	shm_toc_estimate_chunk(&pcxt->estimator, esthnswarea);
 	shm_toc_estimate_keys(&pcxt->estimator, 2);
 
 	/* Finally, estimate PARALLEL_KEY_QUERY_TEXT space */
@@ -979,29 +961,42 @@ HnswBeginParallel(HnswBuildState * buildstate, bool isconcurrent, int request)
 	hnswshared->heaprelid = RelationGetRelid(buildstate->heap);
 	hnswshared->indexrelid = RelationGetRelid(buildstate->index);
 	hnswshared->isconcurrent = isconcurrent;
+	hnswshared->scantuplesortstates = scantuplesortstates;
 	ConditionVariableInit(&hnswshared->workersdonecv);
 	SpinLockInit(&hnswshared->mutex);
 	/* Initialize mutable state */
 	hnswshared->nparticipantsdone = 0;
 	hnswshared->reltuples = 0;
+#if PG_VERSION_NUM >= 120000
 	table_parallelscan_initialize(buildstate->heap,
 								  ParallelTableScanFromHnswShared(hnswshared),
 								  snapshot);
+#else
+	heap_parallelscan_initialize(&hnswshared->heapdesc, buildstate->heap, snapshot);
+#endif
 
-	hnswarea = (char *) shm_toc_allocate(pcxt->toc, esthnswarea);
+	dsaspace = shm_toc_allocate(pcxt->toc, estdsaspace);
+	dsa = dsa_create_in_place(dsaspace, estdsaspace, hnsw_lock_tranche_id, pcxt->seg);
+
+	esthnswarea = maintenance_work_mem * 1024L;
+	hnswshared->dsaptr = dsa_allocate_extended(dsa, esthnswarea, DSA_ALLOC_HUGE | DSA_ALLOC_NO_OOM);
+
+	/* If not enough shared memory, back out (do serial build) */
+	if (!DsaPointerIsValid(hnswshared->dsaptr))
+	{
+		if (IsMVCCSnapshot(snapshot))
+			UnregisterSnapshot(snapshot);
+		DestroyParallelContext(pcxt);
+		ExitParallelMode();
+		return;
+	}
+
+	hnswarea = dsa_get_address(dsa, hnswshared->dsaptr);
 	/* Report less than allocated so never fails */
 	InitGraph(&hnswshared->graphData, hnswarea, esthnswarea - 1024 * 1024);
 
-	/*
-	 * Avoid base address for relptr for Postgres < 14.5
-	 * https://github.com/postgres/postgres/commit/7201cd18627afc64850537806da7f22150d1a83b
-	 */
-#if PG_VERSION_NUM < 140005
-	hnswshared->graphData.memoryUsed += MAXALIGN(1);
-#endif
-
 	shm_toc_insert(pcxt->toc, PARALLEL_KEY_HNSW_SHARED, hnswshared);
-	shm_toc_insert(pcxt->toc, PARALLEL_KEY_HNSW_AREA, hnswarea);
+	shm_toc_insert(pcxt->toc, PARALLEL_KEY_HNSW_AREA, dsaspace);
 
 	/* Store query string for workers */
 	if (debug_query_string)
@@ -1073,7 +1068,7 @@ BuildGraph(HnswBuildState * buildstate, ForkNumber forkNum)
 {
 	int			parallel_workers = 0;
 
-	pgstat_progress_update_param(PROGRESS_CREATEIDX_SUBPHASE, PROGRESS_HNSW_PHASE_LOAD);
+	UpdateProgress(PROGRESS_CREATEIDX_SUBPHASE, PROGRESS_HNSW_PHASE_LOAD);
 
 	/* Calculate parallel workers */
 	if (buildstate->heap != NULL)
@@ -1089,8 +1084,15 @@ BuildGraph(HnswBuildState * buildstate, ForkNumber forkNum)
 		if (buildstate->hnswleader)
 			buildstate->reltuples = ParallelHeapScan(buildstate);
 		else
+		{
+#if PG_VERSION_NUM >= 120000
 			buildstate->reltuples = table_index_build_scan(buildstate->heap, buildstate->index, buildstate->indexInfo,
 														   true, true, BuildCallback, (void *) buildstate, NULL);
+#else
+			buildstate->reltuples = IndexBuildHeapScan(buildstate->heap, buildstate->index, buildstate->indexInfo,
+													   true, BuildCallback, (void *) buildstate, NULL);
+#endif
+		}
 
 		buildstate->indtuples = buildstate->graph->indtuples;
 	}
@@ -1104,6 +1106,22 @@ BuildGraph(HnswBuildState * buildstate, ForkNumber forkNum)
 		HnswEndParallel(buildstate->hnswleader);
 }
 
+#if PG_VERSION_NUM < 110008
+void
+log_newpage_range(Relation rel, ForkNumber forkNum, BlockNumber startblk, BlockNumber endblk, bool page_std)
+{
+	for (BlockNumber blkno = startblk; blkno < endblk; blkno++)
+	{
+		Buffer		buf = ReadBufferExtended(rel, forkNum, blkno, RBM_NORMAL, NULL);
+
+		LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);
+		MarkBufferDirty(buf);
+		log_newpage_buffer(buf, page_std);
+		UnlockReleaseBuffer(buf);
+	}
+}
+#endif
+
 /*
  * Build the index
  */

src/hnswinsert.c

@@ -2,7 +2,6 @@
 
 #include <math.h>
 
-#include "access/generic_xlog.h"
 #include "hnsw.h"
 #include "storage/bufmgr.h"
 #include "storage/lmgr.h"
@@ -117,7 +116,7 @@ HnswInsertAppendPage(Relation index, Buffer *nbuf, Page *npage, GenericXLogState
  * Add to element and neighbor pages
  */
 static void
-AddElementOnDisk(Relation index, HnswElement e, int m, BlockNumber insertPage, BlockNumber *updatedInsertPage, bool building)
+WriteNewElementPages(Relation index, HnswElement e, int m, BlockNumber insertPage, BlockNumber *updatedInsertPage, bool building)
 {
 	Buffer		buf;
 	Page		page;
@@ -340,7 +339,7 @@ ConnectionExists(HnswElement e, HnswNeighborTuple ntup, int startIdx, int lm)
  * Update neighbors
  */
 void
-HnswUpdateNeighborsOnDisk(Relation index, FmgrInfo *procinfo, Oid collation, HnswElement e, int m, bool checkExisting, bool building)
+HnswUpdateNeighborPages(Relation index, FmgrInfo *procinfo, Oid collation, HnswElement e, int m, bool checkExisting, bool building)
 {
 	char	   *base = NULL;
 
@@ -355,7 +354,9 @@ HnswUpdateNeighborsOnDisk(Relation index, FmgrInfo *procinfo, Oid collation, Hns
 			Buffer		buf;
 			Page		page;
 			GenericXLogState *state;
+			ItemId		itemid;
 			HnswNeighborTuple ntup;
+			Size		ntupSize;
 			int			idx = -1;
 			int			startIdx;
 			HnswElement neighborElement = HnswPtrAccess(base, hc->element);
@@ -394,7 +395,9 @@ HnswUpdateNeighborsOnDisk(Relation index, FmgrInfo *procinfo, Oid collation, Hns
 			}
 
 			/* Get tuple */
-			ntup = (HnswNeighborTuple) PageGetItem(page, PageGetItemId(page, offno));
+			itemid = PageGetItemId(page, offno);
+			ntup = (HnswNeighborTuple) PageGetItem(page, itemid);
+			ntupSize = ItemIdGetLength(itemid);
 
 			/* Calculate index for update */
 			startIdx = (neighborElement->level - lc) * m;
@@ -423,9 +426,13 @@ HnswUpdateNeighborsOnDisk(Relation index, FmgrInfo *procinfo, Oid collation, Hns
 			{
 				ItemPointer indextid = &ntup->indextids[idx];
 
-				/* Update neighbor on the buffer */
+				/* Update neighbor */
 				ItemPointerSet(indextid, e->blkno, e->offno);
 
+				/* Overwrite tuple */
+				if (!PageIndexTupleOverwrite(page, offno, (Item) ntup, ntupSize))
+					elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
+
 				/* Commit */
 				if (building)
 					MarkBufferDirty(buf);
@@ -444,12 +451,14 @@ HnswUpdateNeighborsOnDisk(Relation index, FmgrInfo *procinfo, Oid collation, Hns
  * Add a heap TID to an existing element
  */
 static bool
-AddDuplicateOnDisk(Relation index, HnswElement element, HnswElement dup, bool building)
+HnswAddDuplicate(Relation index, HnswElement element, HnswElement dup, bool building)
 {
 	Buffer		buf;
 	Page		page;
 	GenericXLogState *state;
+	ItemId		itemid;
 	HnswElementTuple etup;
+	Size		etupSize;
 	int			i;
 
 	/* Read page */
@@ -467,7 +476,9 @@ AddDuplicateOnDisk(Relation index, HnswElement element, HnswElement dup, bool bu
 	}
 
 	/* Find space */
-	etup = (HnswElementTuple) PageGetItem(page, PageGetItemId(page, dup->offno));
+	itemid = PageGetItemId(page, dup->offno);
+	etup = (HnswElementTuple) PageGetItem(page, itemid);
+	etupSize = ItemIdGetLength(itemid);
 	for (i = 0; i < HNSW_HEAPTIDS; i++)
 	{
 		if (!ItemPointerIsValid(&etup->heaptids[i]))
@@ -483,9 +494,13 @@ AddDuplicateOnDisk(Relation index, HnswElement element, HnswElement dup, bool bu
 		return false;
 	}
 
-	/* Add heap TID, modifying the tuple on the page directly */
+	/* Add heap TID */
 	etup->heaptids[i] = element->heaptids[0];
 
+	/* Overwrite tuple */
+	if (!PageIndexTupleOverwrite(page, dup->offno, (Item) etup, etupSize))
+		elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
+
 	/* Commit */
 	if (building)
 		MarkBufferDirty(buf);
@@ -500,23 +515,23 @@ AddDuplicateOnDisk(Relation index, HnswElement element, HnswElement dup, bool bu
  * Find duplicate element
  */
 static bool
-FindDuplicateOnDisk(Relation index, HnswElement element, bool building)
+HnswFindDuplicate(Relation index, HnswElement element, bool building)
 {
 	char	   *base = NULL;
 	HnswNeighborArray *neighbors = HnswGetNeighbors(base, element, 0);
-	Datum		value = HnswGetValue(base, element);
 
 	for (int i = 0; i < neighbors->length; i++)
 	{
 		HnswCandidate *neighbor = &neighbors->items[i];
 		HnswElement neighborElement = HnswPtrAccess(base, neighbor->element);
+		Datum		value = HnswGetValue(base, element);
 		Datum		neighborValue = HnswGetValue(base, neighborElement);
 
 		/* Exit early since ordered by distance */
 		if (!datumIsEqual(value, neighborValue, false, -1))
 			return false;
 
-		if (AddDuplicateOnDisk(index, element, neighborElement, building))
+		if (HnswAddDuplicate(index, element, neighborElement, building))
 			return true;
 	}
 
@@ -524,26 +539,26 @@ FindDuplicateOnDisk(Relation index, HnswElement element, bool building)
 }
 
 /*
- * Update graph on disk
+ * Write changes to disk
  */
 static void
-UpdateGraphOnDisk(Relation index, FmgrInfo *procinfo, Oid collation, HnswElement element, int m, int efConstruction, HnswElement entryPoint, bool building)
+WriteElement(Relation index, FmgrInfo *procinfo, Oid collation, HnswElement element, int m, int efConstruction, HnswElement entryPoint, bool building)
 {
 	BlockNumber newInsertPage = InvalidBlockNumber;
 
 	/* Look for duplicate */
-	if (FindDuplicateOnDisk(index, element, building))
+	if (HnswFindDuplicate(index, element, building))
 		return;
 
-	/* Add element */
+	/* Write element and neighbor tuples */
-	AddElementOnDisk(index, element, m, GetInsertPage(index), &newInsertPage, building);
+	WriteNewElementPages(index, element, m, GetInsertPage(index), &newInsertPage, building);
 
 	/* Update insert page if needed */
 	if (BlockNumberIsValid(newInsertPage))
 		HnswUpdateMetaPage(index, 0, NULL, newInsertPage, MAIN_FORKNUM, building);
 
 	/* Update neighbors */
-	HnswUpdateNeighborsOnDisk(index, procinfo, collation, element, m, false, building);
+	HnswUpdateNeighborPages(index, procinfo, collation, element, m, false, building);
 
 	/* Update entry point if needed */
 	if (entryPoint == NULL || element->level > entryPoint->level)
@@ -554,7 +569,7 @@ UpdateGraphOnDisk(Relation index, FmgrInfo *procinfo, Oid collation, HnswElement
  * Insert a tuple into the index
  */
 bool
-HnswInsertTupleOnDisk(Relation index, Datum value, Datum *values, bool *isnull, ItemPointer heap_tid, bool building)
+HnswInsertTupleOnDisk(Relation index, Datum value, Datum *values, bool *isnull, ItemPointer heap_tid, Relation heapRel, bool building)
 {
 	HnswElement entryPoint;
 	HnswElement element;
@@ -593,11 +608,11 @@ HnswInsertTupleOnDisk(Relation index, Datum value, Datum *values, bool *isnull,
 		entryPoint = HnswGetEntryPoint(index);
 	}
 
-	/* Find neighbors for element */
+	/* Insert element in graph */
-	HnswFindElementNeighbors(base, element, entryPoint, index, procinfo, collation, m, efConstruction, false);
+	HnswInsertElement(base, element, entryPoint, index, procinfo, collation, m, efConstruction, false);
 
-	/* Update graph on disk */
+	/* Write to disk */
-	UpdateGraphOnDisk(index, procinfo, collation, element, m, efConstruction, entryPoint, building);
+	WriteElement(index, procinfo, collation, element, m, efConstruction, entryPoint, building);
 
 	/* Release lock */
 	UnlockPage(index, HNSW_UPDATE_LOCK, lockmode);
@@ -609,7 +624,7 @@ HnswInsertTupleOnDisk(Relation index, Datum value, Datum *values, bool *isnull,
  * Insert a tuple into the index
  */
 static void
-HnswInsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heap_tid)
+HnswInsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heap_tid, Relation heapRel)
 {
 	Datum		value;
 	FmgrInfo   *normprocinfo;
@@ -626,7 +641,7 @@ HnswInsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heap_ti
 			return;
 	}
 
-	HnswInsertTupleOnDisk(index, value, values, isnull, heap_tid, false);
+	HnswInsertTupleOnDisk(index, value, values, isnull, heap_tid, heapRel, false);
 }
 
 /*
@@ -655,7 +670,7 @@ hnswinsert(Relation index, Datum *values, bool *isnull, ItemPointer heap_tid,
 	oldCtx = MemoryContextSwitchTo(insertCtx);
 
 	/* Insert tuple */
-	HnswInsertTuple(index, values, isnull, heap_tid);
+	HnswInsertTuple(index, values, isnull, heap_tid, heap);
 
 	/* Delete memory context */
 	MemoryContextSwitchTo(oldCtx);

src/hnswscan.c

@@ -1,8 +1,6 @@
 #include "postgres.h"
 
 #include "access/relscan.h"
-#include "bitvector.h"
-#include "catalog/pg_type_d.h"
 #include "hnsw.h"
 #include "pgstat.h"
 #include "storage/bufmgr.h"
@@ -75,15 +73,7 @@ GetScanValue(IndexScanDesc scan)
 	Datum		value;
 
 	if (scan->orderByData->sk_flags & SK_ISNULL)
-	{
+		value = PointerGetDatum(InitVector(GetDimensions(scan->indexRelation)));
-		Oid			typid = TupleDescAttr(scan->indexRelation->rd_att, 0)->atttypid;
-		int			dimensions = GetDimensions(scan->indexRelation);
-
-		if (typid == BITOID || typid == VARBITOID)
-			value = PointerGetDatum(InitBitVector(dimensions));
-		else
-			value = PointerGetDatum(InitVector(dimensions));
-	}
 	else
 	{
 		value = scan->orderByData->sk_argument;
@@ -211,8 +201,12 @@ hnswgettuple(IndexScanDesc scan, ScanDirection dir)
 
 		MemoryContextSwitchTo(oldCtx);
 
+#if PG_VERSION_NUM >= 120000
 		scan->xs_heaptid = *heaptid;
-		scan->xs_recheck = false;
+#else
+		scan->xs_ctup.t_self = *heaptid;
+#endif
+
 		scan->xs_recheckorderby = false;
 		return true;
 	}

src/hnswutils.c

@@ -2,13 +2,9 @@
 
 #include <math.h>
 
-#include "access/generic_xlog.h"
 #include "hnsw.h"
-#include "lib/pairingheap.h"
 #include "storage/bufmgr.h"
 #include "utils/datum.h"
-#include "utils/memdebug.h"
-#include "utils/rel.h"
 #include "vector.h"
 
 #if PG_VERSION_NUM >= 130000
@@ -60,6 +56,13 @@ hash_tid(ItemPointerData tid)
 #define SH_DEFINE
 #include "lib/simplehash.h"
 
+/* Needed to include simplehash.h twice */
+#if PG_VERSION_NUM < 120000
+#undef SH_EQUAL
+#define sh_log2 pointerhash_sh_log2
+#define sh_pow2 pointerhash_sh_pow2
+#endif
+
 /* Pointer hash table */
 static uint32
 hash_pointer(uintptr_t ptr)
@@ -81,6 +84,15 @@ hash_pointer(uintptr_t ptr)
 #define SH_DEFINE
 #include "lib/simplehash.h"
 
+/* Needed to include simplehash.h again */
+#if PG_VERSION_NUM < 120000
+#undef SH_EQUAL
+#undef sh_log2
+#undef sh_pow2
+#define sh_log2 offsethash_sh_log2
+#define sh_pow2 offsethash_sh_pow2
+#endif
+
 /* Offset hash table */
 static uint32
 hash_offset(Size offset)
@@ -203,19 +215,6 @@ HnswInitPage(Buffer buf, Page page)
 	HnswPageGetOpaque(page)->page_id = HNSW_PAGE_ID;
 }
 
-/*
- * Allocate a neighbor array
- */
-static HnswNeighborArray *
-HnswInitNeighborArray(int lm, HnswAllocator * allocator)
-{
-	HnswNeighborArray *a = HnswAlloc(allocator, HNSW_NEIGHBOR_ARRAY_SIZE(lm));
-
-	a->length = 0;
-	a->closerSet = false;
-	return a;
-}
-
 /*
  * Allocate neighbors
  */
@@ -223,12 +222,22 @@ void
 HnswInitNeighbors(char *base, HnswElement element, int m, HnswAllocator * allocator)
 {
 	int			level = element->level;
+
 	HnswNeighborArrayPtr *neighborList = (HnswNeighborArrayPtr *) HnswAlloc(allocator, sizeof(HnswNeighborArrayPtr) * (level + 1));
 
 	HnswPtrStore(base, element->neighbors, neighborList);
 
 	for (int lc = 0; lc <= level; lc++)
-		HnswPtrStore(base, neighborList[lc], HnswInitNeighborArray(HnswGetLayerM(m, lc), allocator));
+	{
+		HnswNeighborArray *a;
+		int			lm = HnswGetLayerM(m, lc);
+
+		HnswPtrStore(base, neighborList[lc], (HnswNeighborArray *) HnswAlloc(allocator, offsetof(HnswNeighborArray, items) + sizeof(HnswCandidate) * lm));
+
+		a = HnswGetNeighbors(base, element, lc);
+		a->length = 0;
+		a->closerSet = false;
+	}
 }
 
 /*
@@ -316,10 +325,7 @@ HnswGetMetaPageInfo(Relation index, int *m, HnswElement * entryPoint)
 	if (entryPoint != NULL)
 	{
 		if (BlockNumberIsValid(metap->entryBlkno))
-		{
 			*entryPoint = HnswInitElementFromBlock(metap->entryBlkno, metap->entryOffno);
-			(*entryPoint)->level = metap->entryLevel;
-		}
 		else
 			*entryPoint = NULL;
 	}
@@ -650,25 +656,11 @@ CreatePairingHeapNode(HnswCandidate * c)
 	return node;
 }
 
-/*
- * Init visited
- */
-static inline void
-InitVisited(char *base, visited_hash * v, Relation index, int ef, int m)
-{
-	if (index != NULL)
-		v->tids = tidhash_create(CurrentMemoryContext, ef * m * 2, NULL);
-	else if (base != NULL)
-		v->offsets = offsethash_create(CurrentMemoryContext, ef * m * 2, NULL);
-	else
-		v->pointers = pointerhash_create(CurrentMemoryContext, ef * m * 2, NULL);
-}
-
 /*
  * Add to visited
  */
 static inline void
-AddToVisited(char *base, visited_hash * v, HnswCandidate * hc, Relation index, bool *found)
+AddToVisited(char *base, visited_hash v, HnswCandidate * hc, Relation index, bool *found)
 {
 	if (index != NULL)
 	{
@@ -676,16 +668,16 @@ AddToVisited(char *base, visited_hash * v, HnswCandidate * hc, Relation index, b
 		ItemPointerData indextid;
 
 		ItemPointerSet(&indextid, element->blkno, element->offno);
-		tidhash_insert(v->tids, indextid, found);
+		tidhash_insert(v.tids, indextid, found);
 	}
 	else if (base != NULL)
 	{
 #if PG_VERSION_NUM >= 130000
 		HnswElement element = HnswPtrAccess(base, hc->element);
 
-		offsethash_insert_hash(v->offsets, HnswPtrOffset(hc->element), element->hash, found);
+		offsethash_insert_hash(v.offsets, HnswPtrOffset(hc->element), element->hash, found);
 #else
-		offsethash_insert(v->offsets, HnswPtrOffset(hc->element), found);
+		offsethash_insert(v.offsets, HnswPtrOffset(hc->element), found);
 #endif
 	}
 	else
@@ -693,9 +685,9 @@ AddToVisited(char *base, visited_hash * v, HnswCandidate * hc, Relation index, b
 #if PG_VERSION_NUM >= 130000
 		HnswElement element = HnswPtrAccess(base, hc->element);
 
-		pointerhash_insert_hash(v->pointers, (uintptr_t) HnswPtrPointer(hc->element), element->hash, found);
+		pointerhash_insert_hash(v.pointers, (uintptr_t) HnswPtrPointer(hc->element), element->hash, found);
 #else
-		pointerhash_insert(v->pointers, (uintptr_t) HnswPtrPointer(hc->element), found);
+		pointerhash_insert(v.pointers, (uintptr_t) HnswPtrPointer(hc->element), found);
 #endif
 	}
 }
@@ -733,12 +725,18 @@ HnswSearchLayer(char *base, Datum q, List *ep, int ef, int lc, Relation index, F
 	HnswNeighborArray *neighborhoodData = NULL;
 	Size		neighborhoodSize;
 
-	InitVisited(base, &v, index, ef, m);
+	/* Create hash table */
+	if (index != NULL)
+		v.tids = tidhash_create(CurrentMemoryContext, ef * m * 2, NULL);
+	else if (base != NULL)
+		v.offsets = offsethash_create(CurrentMemoryContext, ef * m * 2, NULL);
+	else
+		v.pointers = pointerhash_create(CurrentMemoryContext, ef * m * 2, NULL);
 
 	/* Create local memory for neighborhood if needed */
 	if (index == NULL)
 	{
-		neighborhoodSize = HNSW_NEIGHBOR_ARRAY_SIZE(HnswGetLayerM(m, lc));
+		neighborhoodSize = offsetof(HnswNeighborArray, items) + sizeof(HnswCandidate) * HnswGetLayerM(m, lc);
 		neighborhoodData = palloc(neighborhoodSize);
 	}
 
@@ -748,7 +746,7 @@ HnswSearchLayer(char *base, Datum q, List *ep, int ef, int lc, Relation index, F
 		HnswCandidate *hc = (HnswCandidate *) lfirst(lc2);
 		bool		found;
 
-		AddToVisited(base, &v, hc, index, &found);
+		AddToVisited(base, v, hc, index, &found);
 
 		pairingheap_add(C, &(CreatePairingHeapNode(hc)->ph_node));
 		pairingheap_add(W, &(CreatePairingHeapNode(hc)->ph_node));
@@ -783,7 +781,7 @@ HnswSearchLayer(char *base, Datum q, List *ep, int ef, int lc, Relation index, F
 		/* Copy neighborhood to local memory if needed */
 		if (index == NULL)
 		{
-			LWLockAcquire(&cElement->lock, LW_SHARED);
+			LWLockAcquire(&cElement->lock, LW_EXCLUSIVE);
 			memcpy(neighborhoodData, neighborhood, neighborhoodSize);
 			LWLockRelease(&cElement->lock);
 			neighborhood = neighborhoodData;
@@ -794,7 +792,7 @@ HnswSearchLayer(char *base, Datum q, List *ep, int ef, int lc, Relation index, F
 			HnswCandidate *e = &neighborhood->items[i];
 			bool		visited;
 
-			AddToVisited(base, &v, e, index, &visited);
+			AddToVisited(base, v, e, index, &visited);
 
 			if (!visited)
 			{
@@ -860,15 +858,12 @@ HnswSearchLayer(char *base, Datum q, List *ep, int ef, int lc, Relation index, F
 static int
 #if PG_VERSION_NUM >= 130000
 CompareCandidateDistances(const ListCell *a, const ListCell *b)
-{
-	HnswCandidate *hca = lfirst(a);
-	HnswCandidate *hcb = lfirst(b);
 #else
 CompareCandidateDistances(const void *a, const void *b)
-{
-	HnswCandidate *hca = lfirst(*(ListCell **) a);
-	HnswCandidate *hcb = lfirst(*(ListCell **) b);
 #endif
+{
+	HnswCandidate *hca = lfirst((ListCell *) a);
+	HnswCandidate *hcb = lfirst((ListCell *) b);
 
 	if (hca->distance < hcb->distance)
 		return 1;
@@ -891,15 +886,12 @@ CompareCandidateDistances(const void *a, const void *b)
 static int
 #if PG_VERSION_NUM >= 130000
 CompareCandidateDistancesOffset(const ListCell *a, const ListCell *b)
-{
-	HnswCandidate *hca = lfirst(a);
-	HnswCandidate *hcb = lfirst(b);
 #else
 CompareCandidateDistancesOffset(const void *a, const void *b)
-{
-	HnswCandidate *hca = lfirst(*(ListCell **) a);
-	HnswCandidate *hcb = lfirst(*(ListCell **) b);
 #endif
+{
+	HnswCandidate *hca = lfirst((ListCell *) a);
+	HnswCandidate *hcb = lfirst((ListCell *) b);
 
 	if (hca->distance < hcb->distance)
 		return 1;
@@ -920,10 +912,40 @@ CompareCandidateDistancesOffset(const void *a, const void *b)
  * Calculate the distance between elements
  */
 static float
-HnswGetDistance(char *base, HnswElement a, HnswElement b, FmgrInfo *procinfo, Oid collation)
+HnswGetDistance(char *base, HnswElement a, HnswElement b, int lc, FmgrInfo *procinfo, Oid collation)
 {
-	Datum		aValue = HnswGetValue(base, a);
+	Datum		aValue;
-	Datum		bValue = HnswGetValue(base, b);
+	Datum		bValue;
+
+	/* Look for cached distance */
+	if (!HnswPtrIsNull(base, a->neighbors))
+	{
+		HnswNeighborArray *neighbors = HnswGetNeighbors(base, a, lc);
+
+		for (int i = 0; i < neighbors->length; i++)
+		{
+			HnswElement element = HnswPtrAccess(base, neighbors->items[i].element);
+
+			if (element == b)
+				return neighbors->items[i].distance;
+		}
+	}
+
+	if (!HnswPtrIsNull(base, b->neighbors))
+	{
+		HnswNeighborArray *neighbors = HnswGetNeighbors(base, b, lc);
+
+		for (int i = 0; i < neighbors->length; i++)
+		{
+			HnswElement element = HnswPtrAccess(base, neighbors->items[i].element);
+
+			if (element == a)
+				return neighbors->items[i].distance;
+		}
+	}
+
+	aValue = HnswGetValue(base, a);
+	bValue = HnswGetValue(base, b);
 
 	return DatumGetFloat8(FunctionCall2Coll(procinfo, collation, aValue, bValue));
 }
@@ -932,7 +954,7 @@ HnswGetDistance(char *base, HnswElement a, HnswElement b, FmgrInfo *procinfo, Oi
  * Check if an element is closer to q than any element from R
  */
 static bool
-CheckElementCloser(char *base, HnswCandidate * e, List *r, FmgrInfo *procinfo, Oid collation)
+CheckElementCloser(char *base, HnswCandidate * e, List *r, int lc, FmgrInfo *procinfo, Oid collation)
 {
 	HnswElement eElement = HnswPtrAccess(base, e->element);
 	ListCell   *lc2;
@@ -941,7 +963,8 @@ CheckElementCloser(char *base, HnswCandidate * e, List *r, FmgrInfo *procinfo, O
 	{
 		HnswCandidate *ri = lfirst(lc2);
 		HnswElement riElement = HnswPtrAccess(base, ri->element);
-		float		distance = HnswGetDistance(base, eElement, riElement, procinfo, collation);
+
+		float		distance = HnswGetDistance(base, eElement, riElement, lc, procinfo, collation);
 
 		if (distance <= e->distance)
 			return false;
@@ -954,22 +977,20 @@ CheckElementCloser(char *base, HnswCandidate * e, List *r, FmgrInfo *procinfo, O
  * Algorithm 4 from paper
  */
 static List *
-SelectNeighbors(char *base, List *c, int lm, int lc, FmgrInfo *procinfo, Oid collation, HnswElement e2, HnswCandidate * newCandidate, HnswCandidate * *pruned, bool sortCandidates)
+SelectNeighbors(char *base, List *c, int m, int lc, FmgrInfo *procinfo, Oid collation, HnswElement e2, HnswCandidate * newCandidate, HnswCandidate * *pruned, bool sortCandidates)
 {
 	List	   *r = NIL;
 	List	   *w = list_copy(c);
-	HnswCandidate **wd;
+	pairingheap *wd;
-	int			wdlen = 0;
-	int			wdoff = 0;
 	HnswNeighborArray *neighbors = HnswGetNeighbors(base, e2, lc);
 	bool		mustCalculate = !neighbors->closerSet;
 	List	   *added = NIL;
 	bool		removedAny = false;
 
-	if (list_length(w) <= lm)
+	if (list_length(w) <= m)
 		return w;
 
-	wd = palloc(sizeof(HnswCandidate *) * list_length(w));
+	wd = pairingheap_allocate(CompareNearestCandidates, NULL);
 
 	/* Ensure order of candidates is deterministic for closer caching */
 	if (sortCandidates)
@@ -980,7 +1001,7 @@ SelectNeighbors(char *base, List *c, int lm, int lc, FmgrInfo *procinfo, Oid col
 			list_sort(w, CompareCandidateDistancesOffset);
 	}
 
-	while (list_length(w) > 0 && list_length(r) < lm)
+	while (list_length(w) > 0 && list_length(r) < m)
 	{
 		/* Assumes w is already ordered desc */
 		HnswCandidate *e = llast(w);
@@ -989,20 +1010,16 @@ SelectNeighbors(char *base, List *c, int lm, int lc, FmgrInfo *procinfo, Oid col
 
 		/* Use previous state of r and wd to skip work when possible */
 		if (mustCalculate)
-			e->closer = CheckElementCloser(base, e, r, procinfo, collation);
+			e->closer = CheckElementCloser(base, e, r, lc, procinfo, collation);
 		else if (list_length(added) > 0)
 		{
-			/* Keep Valgrind happy for in-memory, parallel builds */
-			if (base != NULL)
-				VALGRIND_MAKE_MEM_DEFINED(&e->closer, 1);
-
 			/*
 			 * If the current candidate was closer, we only need to compare it
 			 * with the other candidates that we have added.
 			 */
 			if (e->closer)
 			{
-				e->closer = CheckElementCloser(base, e, added, procinfo, collation);
+				e->closer = CheckElementCloser(base, e, added, lc, procinfo, collation);
 
 				if (!e->closer)
 					removedAny = true;
@@ -1015,7 +1032,7 @@ SelectNeighbors(char *base, List *c, int lm, int lc, FmgrInfo *procinfo, Oid col
 				 */
 				if (removedAny)
 				{
-					e->closer = CheckElementCloser(base, e, r, procinfo, collation);
+					e->closer = CheckElementCloser(base, e, r, lc, procinfo, collation);
 					if (e->closer)
 						added = lappend(added, e);
 				}
@@ -1023,33 +1040,29 @@ SelectNeighbors(char *base, List *c, int lm, int lc, FmgrInfo *procinfo, Oid col
 		}
 		else if (e == newCandidate)
 		{
-			e->closer = CheckElementCloser(base, e, r, procinfo, collation);
+			e->closer = CheckElementCloser(base, e, r, lc, procinfo, collation);
 			if (e->closer)
 				added = lappend(added, e);
 		}
 
-		/* Keep Valgrind happy for in-memory, parallel builds */
-		if (base != NULL)
-			VALGRIND_MAKE_MEM_DEFINED(&e->closer, 1);
-
 		if (e->closer)
 			r = lappend(r, e);
 		else
-			wd[wdlen++] = e;
+			pairingheap_add(wd, &(CreatePairingHeapNode(e)->ph_node));
 	}
 
 	/* Cached value can only be used in future if sorted deterministically */
 	neighbors->closerSet = sortCandidates;
 
 	/* Keep pruned connections */
-	while (wdoff < wdlen && list_length(r) < lm)
+	while (!pairingheap_is_empty(wd) && list_length(r) < m)
-		r = lappend(r, wd[wdoff++]);
+		r = lappend(r, ((HnswPairingHeapNode *) pairingheap_remove_first(wd))->inner);
 
 	/* Return pruned for update connections */
 	if (pruned != NULL)
 	{
-		if (wdoff < wdlen)
+		if (!pairingheap_is_empty(wd))
-			*pruned = wd[wdoff];
+			*pruned = ((HnswPairingHeapNode *) pairingheap_first(wd))->inner;
 		else
 			*pruned = linitial(w);
 	}
@@ -1061,7 +1074,7 @@ SelectNeighbors(char *base, List *c, int lm, int lc, FmgrInfo *procinfo, Oid col
  * Add connections
  */
 static void
-AddConnections(char *base, HnswElement element, List *neighbors, int lc)
+AddConnections(char *base, HnswElement element, List *neighbors, int m, int lc)
 {
 	ListCell   *lc2;
 	HnswNeighborArray *a = HnswGetNeighbors(base, element, lc);
@@ -1074,7 +1087,7 @@ AddConnections(char *base, HnswElement element, List *neighbors, int lc)
  * Update connections
  */
 void
-HnswUpdateConnection(char *base, HnswElement element, HnswCandidate * hc, int lm, int lc, int *updateIdx, Relation index, FmgrInfo *procinfo, Oid collation)
+HnswUpdateConnection(char *base, HnswElement element, HnswCandidate * hc, int m, int lc, int *updateIdx, Relation index, FmgrInfo *procinfo, Oid collation)
 {
 	HnswElement hce = HnswPtrAccess(base, hc->element);
 	HnswNeighborArray *currentNeighbors = HnswGetNeighbors(base, hce, lc);
@@ -1083,7 +1096,7 @@ HnswUpdateConnection(char *base, HnswElement element, HnswCandidate * hc, int lm
 	HnswPtrStore(base, hc2.element, element);
 	hc2.distance = hc->distance;
 
-	if (currentNeighbors->length < lm)
+	if (currentNeighbors->length < m)
 	{
 		currentNeighbors->items[currentNeighbors->length++] = hc2;
 
@@ -1129,7 +1142,7 @@ HnswUpdateConnection(char *base, HnswElement element, HnswCandidate * hc, int lm
 				c = lappend(c, &currentNeighbors->items[i]);
 			c = lappend(c, &hc2);
 
-			SelectNeighbors(base, c, lm, lc, procinfo, collation, hce, &hc2, &pruned, true);
+			SelectNeighbors(base, c, m, lc, procinfo, collation, hce, &hc2, &pruned, true);
 
 			/* Should not happen */
 			if (pruned == NULL)
@@ -1203,7 +1216,7 @@ PrecomputeHash(char *base, HnswElement element)
  * Algorithm 1 from paper
  */
 void
-HnswFindElementNeighbors(char *base, HnswElement element, HnswElement entryPoint, Relation index, FmgrInfo *procinfo, Oid collation, int m, int efConstruction, bool existing)
+HnswInsertElement(char *base, HnswElement element, HnswElement entryPoint, Relation index, FmgrInfo *procinfo, Oid collation, int m, int efConstruction, bool existing)
 {
 	List	   *ep;
 	List	   *w;
@@ -1263,7 +1276,7 @@ HnswFindElementNeighbors(char *base, HnswElement element, HnswElement entryPoint
 		 */
 		neighbors = SelectNeighbors(base, lw, lm, lc, procinfo, collation, element, NULL, NULL, false);
 
-		AddConnections(base, element, neighbors, lc);
+		AddConnections(base, element, neighbors, lm, lc);
 
 		ep = w;
 	}

src/hnswvacuum.c

@@ -2,7 +2,6 @@
 
 #include <math.h>
 
-#include "access/generic_xlog.h"
 #include "commands/vacuum.h"
 #include "hnsw.h"
 #include "storage/bufmgr.h"
@@ -60,7 +59,8 @@ RemoveHeapTids(HnswVacuumState * vacuumstate)
 		/* Iterate over nodes */
 		for (offno = FirstOffsetNumber; offno <= maxoffno; offno = OffsetNumberNext(offno))
 		{
-			HnswElementTuple etup = (HnswElementTuple) PageGetItem(page, PageGetItemId(page, offno));
+			ItemId		itemid = PageGetItemId(page, offno);
+			HnswElementTuple etup = (HnswElementTuple) PageGetItem(page, itemid);
 			int			idx = 0;
 			bool		itemUpdated = false;
 
@@ -91,10 +91,15 @@ RemoveHeapTids(HnswVacuumState * vacuumstate)
 
 				if (itemUpdated)
 				{
+					Size		etupSize = ItemIdGetLength(itemid);
+
 					/* Mark rest as invalid */
 					for (int i = idx; i < HNSW_HEAPTIDS; i++)
 						ItemPointerSetInvalid(&etup->heaptids[i]);
 
+					if (!PageIndexTupleOverwrite(page, offno, (Item) etup, etupSize))
+						elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
+
 					updated = true;
 				}
 			}
@@ -194,25 +199,21 @@ RepairGraphElement(HnswVacuumState * vacuumstate, HnswElement element, HnswEleme
 	BufferAccessStrategy bas = vacuumstate->bas;
 	HnswNeighborTuple ntup = vacuumstate->ntup;
 	Size		ntupSize = HNSW_NEIGHBOR_TUPLE_SIZE(element->level, m);
-	char	   *base = NULL;
 
 	/* Skip if element is entry point */
 	if (entryPoint != NULL && element->blkno == entryPoint->blkno && element->offno == entryPoint->offno)
 		return;
 
 	/* Init fields */
-	HnswInitNeighbors(base, element, m, NULL);
+	HnswInitNeighbors(NULL, element, m, NULL);
 	element->heaptidsLength = 0;
 
-	/* Find neighbors for element, skipping itself */
+	/* Add element to graph, skipping itself */
-	HnswFindElementNeighbors(base, element, entryPoint, index, procinfo, collation, m, efConstruction, true);
+	HnswInsertElement(NULL, element, entryPoint, index, procinfo, collation, m, efConstruction, true);
-
-	/* Zero memory for each element */
-	MemSet(ntup, 0, HNSW_TUPLE_ALLOC_SIZE);
 
 	/* Update neighbor tuple */
 	/* Do this before getting page to minimize locking */
-	HnswSetNeighborTuple(base, ntup, element, m);
+	HnswSetNeighborTuple(NULL, ntup, element, m);
 
 	/* Get neighbor page */
 	buf = ReadBufferExtended(index, MAIN_FORKNUM, element->neighborPage, RBM_NORMAL, bas);
@@ -229,7 +230,7 @@ RepairGraphElement(HnswVacuumState * vacuumstate, HnswElement element, HnswEleme
 	UnlockReleaseBuffer(buf);
 
 	/* Update neighbors */
-	HnswUpdateNeighborsOnDisk(index, procinfo, collation, element, m, true, false);
+	HnswUpdateNeighborPages(index, procinfo, collation, element, m, true, false);
 }
 
 /*
@@ -476,8 +477,11 @@ MarkDeleted(HnswVacuumState * vacuumstate)
 		/* Update element and neighbors together */
 		for (offno = FirstOffsetNumber; offno <= maxoffno; offno = OffsetNumberNext(offno))
 		{
-			HnswElementTuple etup = (HnswElementTuple) PageGetItem(page, PageGetItemId(page, offno));
+			ItemId		itemid = PageGetItemId(page, offno);
+			HnswElementTuple etup = (HnswElementTuple) PageGetItem(page, itemid);
 			HnswNeighborTuple ntup;
+			Size		etupSize;
+			Size		ntupSize;
 			Buffer		nbuf;
 			Page		npage;
 			BlockNumber neighborPage;
@@ -501,6 +505,10 @@ MarkDeleted(HnswVacuumState * vacuumstate)
 			if (ItemPointerIsValid(&etup->heaptids[0]))
 				continue;
 
+			/* Calculate sizes */
+			etupSize = ItemIdGetLength(itemid);
+			ntupSize = HNSW_NEIGHBOR_TUPLE_SIZE(etup->level, vacuumstate->m);
+
 			/* Get neighbor page */
 			neighborPage = ItemPointerGetBlockNumber(&etup->neighbortid);
 			neighborOffno = ItemPointerGetOffsetNumber(&etup->neighbortid);
@@ -527,10 +535,13 @@ MarkDeleted(HnswVacuumState * vacuumstate)
 			for (int i = 0; i < ntup->count; i++)
 				ItemPointerSetInvalid(&ntup->indextids[i]);
 
-			/*
+			/* Overwrite element tuple */
-			 * We modified the tuples in place, no need to call
+			if (!PageIndexTupleOverwrite(page, offno, (Item) etup, etupSize))
-			 * PageIndexTupleOverwrite
+				elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
-			 */
+
+			/* Overwrite neighbor tuple */
+			if (!PageIndexTupleOverwrite(npage, neighborOffno, (Item) ntup, ntupSize))
+				elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
 
 			/* Commit */
 			GenericXLogFinish(state);
@@ -575,7 +586,7 @@ InitVacuumState(HnswVacuumState * vacuumstate, IndexVacuumInfo *info, IndexBulkD
 	vacuumstate->bas = GetAccessStrategy(BAS_BULKREAD);
 	vacuumstate->procinfo = index_getprocinfo(index, 1, HNSW_DISTANCE_PROC);
 	vacuumstate->collation = index->rd_indcollation[0];
-	vacuumstate->ntup = palloc0(HNSW_TUPLE_ALLOC_SIZE);
+	vacuumstate->ntup = palloc0(BLCKSZ);
 	vacuumstate->tmpCtx = AllocSetContextCreate(CurrentMemoryContext,
 												"Hnsw vacuum temporary context",
 												ALLOCSET_DEFAULT_SIZES);

src/ivfbuild.c

@@ -2,38 +2,58 @@
 
 #include <float.h>
 
-#include "access/table.h"
-#include "access/tableam.h"
 #include "access/parallel.h"
 #include "access/xact.h"
 #include "catalog/index.h"
 #include "catalog/pg_operator_d.h"
 #include "catalog/pg_type_d.h"
-#include "commands/progress.h"
 #include "ivfflat.h"
 #include "miscadmin.h"
-#include "optimizer/optimizer.h"
 #include "storage/bufmgr.h"
 #include "tcop/tcopprot.h"
 #include "utils/memutils.h"
 
 #if PG_VERSION_NUM >= 140000
 #include "utils/backend_progress.h"
-#else
+#elif PG_VERSION_NUM >= 120000
 #include "pgstat.h"
 #endif
 
+#if PG_VERSION_NUM >= 120000
+#include "access/tableam.h"
+#include "commands/progress.h"
+#else
+#define PROGRESS_CREATEIDX_SUBPHASE 0
+#define PROGRESS_CREATEIDX_TUPLES_TOTAL 0
+#define PROGRESS_CREATEIDX_TUPLES_DONE 0
+#endif
+
 #if PG_VERSION_NUM >= 130000
 #define CALLBACK_ITEM_POINTER ItemPointer tid
 #else
 #define CALLBACK_ITEM_POINTER HeapTuple hup
 #endif
 
+#if PG_VERSION_NUM >= 120000
+#define UpdateProgress(index, val) pgstat_progress_update_param(index, val)
+#else
+#define UpdateProgress(index, val) ((void)val)
+#endif
+
 #if PG_VERSION_NUM >= 140000
 #include "utils/backend_status.h"
 #include "utils/wait_event.h"
 #endif
 
+#if PG_VERSION_NUM >= 120000
+#include "access/table.h"
+#include "optimizer/optimizer.h"
+#else
+#include "access/heapam.h"
+#include "optimizer/planner.h"
+#include "pgstat.h"
+#endif
+
 #define PARALLEL_KEY_IVFFLAT_SHARED		UINT64CONST(0xA000000000000001)
 #define PARALLEL_KEY_TUPLESORT			UINT64CONST(0xA000000000000002)
 #define PARALLEL_KEY_IVFFLAT_CENTERS	UINT64CONST(0xA000000000000003)
@@ -130,8 +150,13 @@ SampleRows(IvfflatBuildState * buildstate)
 	{
 		BlockNumber targblock = BlockSampler_Next(&buildstate->bs);
 
+#if PG_VERSION_NUM >= 120000
 		table_index_build_range_scan(buildstate->heap, buildstate->index, buildstate->indexInfo,
 									 false, true, false, targblock, 1, SampleCallback, (void *) buildstate, NULL);
+#else
+		IndexBuildHeapRangeScan(buildstate->heap, buildstate->index, buildstate->indexInfo,
+								false, true, targblock, 1, SampleCallback, (void *) buildstate, NULL);
+#endif
 	}
 }
 
@@ -257,12 +282,16 @@ InsertTuples(Relation index, IvfflatBuildState * buildstate, ForkNumber forkNum)
 	IndexTuple	itup = NULL;	/* silence compiler warning */
 	int64		inserted = 0;
 
+#if PG_VERSION_NUM >= 120000
 	TupleTableSlot *slot = MakeSingleTupleTableSlot(buildstate->tupdesc, &TTSOpsMinimalTuple);
+#else
+	TupleTableSlot *slot = MakeSingleTupleTableSlot(buildstate->tupdesc);
+#endif
 	TupleDesc	tupdesc = RelationGetDescr(index);
 
-	pgstat_progress_update_param(PROGRESS_CREATEIDX_SUBPHASE, PROGRESS_IVFFLAT_PHASE_LOAD);
+	UpdateProgress(PROGRESS_CREATEIDX_SUBPHASE, PROGRESS_IVFFLAT_PHASE_LOAD);
 
-	pgstat_progress_update_param(PROGRESS_CREATEIDX_TUPLES_TOTAL, buildstate->indtuples);
+	UpdateProgress(PROGRESS_CREATEIDX_TUPLES_TOTAL, buildstate->indtuples);
 
 	GetNextTuple(buildstate->sortstate, tupdesc, slot, &itup, &list);
 
@@ -298,7 +327,7 @@ InsertTuples(Relation index, IvfflatBuildState * buildstate, ForkNumber forkNum)
 
 			pfree(itup);
 
-			pgstat_progress_update_param(PROGRESS_CREATEIDX_TUPLES_DONE, ++inserted);
+			UpdateProgress(PROGRESS_CREATEIDX_TUPLES_DONE, ++inserted);
 
 			GetNextTuple(buildstate->sortstate, tupdesc, slot, &itup, &list);
 		}
@@ -346,12 +375,20 @@ InitBuildState(IvfflatBuildState * buildstate, Relation heap, Relation index, In
 		elog(ERROR, "dimensions must be greater than one for this opclass");
 
 	/* Create tuple description for sorting */
+#if PG_VERSION_NUM >= 120000
 	buildstate->tupdesc = CreateTemplateTupleDesc(3);
+#else
+	buildstate->tupdesc = CreateTemplateTupleDesc(3, false);
+#endif
 	TupleDescInitEntry(buildstate->tupdesc, (AttrNumber) 1, "list", INT4OID, -1, 0);
 	TupleDescInitEntry(buildstate->tupdesc, (AttrNumber) 2, "tid", TIDOID, -1, 0);
 	TupleDescInitEntry(buildstate->tupdesc, (AttrNumber) 3, "vector", RelationGetDescr(index)->attrs[0].atttypid, -1, 0);
 
+#if PG_VERSION_NUM >= 120000
 	buildstate->slot = MakeSingleTupleTableSlot(buildstate->tupdesc, &TTSOpsVirtual);
+#else
+	buildstate->slot = MakeSingleTupleTableSlot(buildstate->tupdesc);
+#endif
 
 	buildstate->centers = VectorArrayInit(buildstate->lists, buildstate->dimensions);
 	buildstate->listInfo = palloc(sizeof(ListInfo) * buildstate->lists);
@@ -398,7 +435,7 @@ ComputeCenters(IvfflatBuildState * buildstate)
 {
 	int			numSamples;
 
-	pgstat_progress_update_param(PROGRESS_CREATEIDX_SUBPHASE, PROGRESS_IVFFLAT_PHASE_KMEANS);
+	UpdateProgress(PROGRESS_CREATEIDX_SUBPHASE, PROGRESS_IVFFLAT_PHASE_KMEANS);
 
 	/* Target 50 samples per list, with at least 10000 samples */
 	/* The number of samples has a large effect on index build time */
@@ -473,7 +510,7 @@ CreateListPages(Relation index, VectorArray centers, int dimensions,
 	IvfflatList list;
 
 	listSize = MAXALIGN(IVFFLAT_LIST_SIZE(dimensions));
-	list = palloc0(listSize);
+	list = palloc(listSize);
 
 	buf = IvfflatNewBuffer(index, forkNum);
 	IvfflatInitRegisterPage(index, &buf, &page, &state);
@@ -594,7 +631,11 @@ IvfflatParallelScanAndSort(IvfflatSpool * ivfspool, IvfflatShared * ivfshared, S
 {
 	SortCoordinate coordinate;
 	IvfflatBuildState buildstate;
+#if PG_VERSION_NUM >= 120000
 	TableScanDesc scan;
+#else
+	HeapScanDesc scan;
+#endif
 	double		reltuples;
 	IndexInfo  *indexInfo;
 
@@ -618,11 +659,18 @@ IvfflatParallelScanAndSort(IvfflatSpool * ivfspool, IvfflatShared * ivfshared, S
 	buildstate.centers->length = buildstate.centers->maxlen;
 	ivfspool->sortstate = tuplesort_begin_heap(buildstate.tupdesc, 1, attNums, sortOperators, sortCollations, nullsFirstFlags, sortmem, coordinate, false);
 	buildstate.sortstate = ivfspool->sortstate;
+#if PG_VERSION_NUM >= 120000
 	scan = table_beginscan_parallel(ivfspool->heap,
 									ParallelTableScanFromIvfflatShared(ivfshared));
 	reltuples = table_index_build_scan(ivfspool->heap, ivfspool->index, indexInfo,
 									   true, progress, BuildCallback,
 									   (void *) &buildstate, scan);
+#else
+	scan = heap_beginscan_parallel(ivfspool->heap, &ivfshared->heapdesc);
+	reltuples = IndexBuildHeapScan(ivfspool->heap, ivfspool->index, indexInfo,
+								   true, BuildCallback,
+								   (void *) &buildstate, scan);
+#endif
 
 	/* Execute this worker's part of the sort */
 	tuplesort_performsort(ivfspool->sortstate);
@@ -692,7 +740,11 @@ IvfflatParallelBuildMain(dsm_segment *seg, shm_toc *toc)
 	}
 
 	/* Open relations within worker */
+#if PG_VERSION_NUM >= 120000
 	heapRel = table_open(ivfshared->heaprelid, heapLockmode);
+#else
+	heapRel = heap_open(ivfshared->heaprelid, heapLockmode);
+#endif
 	indexRel = index_open(ivfshared->indexrelid, indexLockmode);
 
 	/* Initialize worker's own spool */
@@ -712,7 +764,11 @@ IvfflatParallelBuildMain(dsm_segment *seg, shm_toc *toc)
 
 	/* Close relations within worker */
 	index_close(indexRel, indexLockmode);
+#if PG_VERSION_NUM >= 120000
 	table_close(heapRel, heapLockmode);
+#else
+	heap_close(heapRel, heapLockmode);
+#endif
 }
 
 /*
@@ -737,7 +793,19 @@ IvfflatEndParallel(IvfflatLeader * ivfleader)
 static Size
 ParallelEstimateShared(Relation heap, Snapshot snapshot)
 {
+#if PG_VERSION_NUM >= 120000
 	return add_size(BUFFERALIGN(sizeof(IvfflatShared)), table_parallelscan_estimate(heap, snapshot));
+#else
+	if (!IsMVCCSnapshot(snapshot))
+	{
+		Assert(snapshot == SnapshotAny);
+		return sizeof(IvfflatShared);
+	}
+
+	return add_size(offsetof(IvfflatShared, heapdesc) +
+					offsetof(ParallelHeapScanDescData, phs_snapshot_data),
+					EstimateSnapshotSpace(snapshot));
+#endif
 }
 
 /*
@@ -788,7 +856,11 @@ IvfflatBeginParallel(IvfflatBuildState * buildstate, bool isconcurrent, int requ
 	/* Enter parallel mode and create context */
 	EnterParallelMode();
 	Assert(request > 0);
+#if PG_VERSION_NUM >= 120000
 	pcxt = CreateParallelContext("vector", "IvfflatParallelBuildMain", request);
+#else
+	pcxt = CreateParallelContext("vector", "IvfflatParallelBuildMain", request, true);
+#endif
 
 	scantuplesortstates = leaderparticipates ? request + 1 : request;
 
@@ -846,9 +918,13 @@ IvfflatBeginParallel(IvfflatBuildState * buildstate, bool isconcurrent, int requ
 #ifdef IVFFLAT_KMEANS_DEBUG
 	ivfshared->inertia = 0;
 #endif
+#if PG_VERSION_NUM >= 120000
 	table_parallelscan_initialize(buildstate->heap,
 								  ParallelTableScanFromIvfflatShared(ivfshared),
 								  snapshot);
+#else
+	heap_parallelscan_initialize(&ivfshared->heapdesc, buildstate->heap, snapshot);
+#endif
 
 	/* Store shared tuplesort-private state, for which we reserved space */
 	sharedsort = (Sharedsort *) shm_toc_allocate(pcxt->toc, estsort);
@@ -919,7 +995,7 @@ AssignTuples(IvfflatBuildState * buildstate)
 	Oid			sortCollations[] = {InvalidOid};
 	bool		nullsFirstFlags[] = {false};
 
-	pgstat_progress_update_param(PROGRESS_CREATEIDX_SUBPHASE, PROGRESS_IVFFLAT_PHASE_ASSIGN);
+	UpdateProgress(PROGRESS_CREATEIDX_SUBPHASE, PROGRESS_IVFFLAT_PHASE_ASSIGN);
 
 	/* Calculate parallel workers */
 	if (buildstate->heap != NULL)
@@ -947,8 +1023,15 @@ AssignTuples(IvfflatBuildState * buildstate)
 		if (buildstate->ivfleader)
 			buildstate->reltuples = ParallelHeapScan(buildstate);
 		else
+		{
+#if PG_VERSION_NUM >= 120000
 			buildstate->reltuples = table_index_build_scan(buildstate->heap, buildstate->index, buildstate->indexInfo,
 														   true, true, BuildCallback, (void *) buildstate, NULL);
+#else
+			buildstate->reltuples = IndexBuildHeapScan(buildstate->heap, buildstate->index, buildstate->indexInfo,
+													   true, BuildCallback, (void *) buildstate, NULL);
+#endif
+		}
 
 #ifdef IVFFLAT_KMEANS_DEBUG
 		PrintKmeansMetrics(buildstate);

src/ivfflat.c

@@ -3,16 +3,14 @@
 #include <float.h>
 
 #include "access/amapi.h"
-#include "access/reloptions.h"
-#include "commands/progress.h"
 #include "commands/vacuum.h"
 #include "ivfflat.h"
 #include "utils/guc.h"
 #include "utils/selfuncs.h"
 #include "utils/spccache.h"
 
-#if PG_VERSION_NUM < 150000
+#if PG_VERSION_NUM >= 120000
-#define MarkGUCPrefixReserved(x) EmitWarningsOnPlaceholders(x)
+#include "commands/progress.h"
 #endif
 
 int			ivfflat_probes;
@@ -35,13 +33,12 @@ IvfflatInit(void)
 	DefineCustomIntVariable("ivfflat.probes", "Sets the number of probes",
 							"Valid range is 1..lists.", &ivfflat_probes,
 							IVFFLAT_DEFAULT_PROBES, IVFFLAT_MIN_LISTS, IVFFLAT_MAX_LISTS, PGC_USERSET, 0, NULL, NULL, NULL);
-
-	MarkGUCPrefixReserved("ivfflat");
 }
 
 /*
  * Get the name of index build phase
  */
+#if PG_VERSION_NUM >= 120000
 static char *
 ivfflatbuildphasename(int64 phasenum)
 {
@@ -59,6 +56,7 @@ ivfflatbuildphasename(int64 phasenum)
 			return NULL;
 	}
 }
+#endif
 
 /*
  * Estimate the cost of an index scan
@@ -74,6 +72,9 @@ ivfflatcostestimate(PlannerInfo *root, IndexPath *path, double loop_count,
 	double		ratio;
 	double		spc_seq_page_cost;
 	Relation	index;
+#if PG_VERSION_NUM < 120000
+	List	   *qinfos;
+#endif
 
 	/* Never use index without order */
 	if (path->indexorderbys == NULL)
@@ -104,7 +105,12 @@ ivfflatcostestimate(PlannerInfo *root, IndexPath *path, double loop_count,
 	 */
 	costs.numIndexTuples = path->indexinfo->tuples * ratio;
 
+#if PG_VERSION_NUM >= 120000
 	genericcostestimate(root, path, loop_count, &costs);
+#else
+	qinfos = deconstruct_indexquals(path);
+	genericcostestimate(root, path, loop_count, qinfos, &costs);
+#endif
 
 	get_tablespace_page_costs(path->indexinfo->reltablespace, NULL, &spc_seq_page_cost);
 
@@ -221,7 +227,9 @@ ivfflathandler(PG_FUNCTION_ARGS)
 	amroutine->amcostestimate = ivfflatcostestimate;
 	amroutine->amoptions = ivfflatoptions;
 	amroutine->amproperty = NULL;	/* TODO AMPROP_DISTANCE_ORDERABLE */
+#if PG_VERSION_NUM >= 120000
 	amroutine->ambuildphasename = ivfflatbuildphasename;
+#endif
 	amroutine->amvalidate = ivfflatvalidate;
 #if PG_VERSION_NUM >= 140000
 	amroutine->amadjustmembers = NULL;

src/ivfflat.h

@@ -3,10 +3,9 @@
 
 #include "postgres.h"
 
-#include "access/genam.h"
 #include "access/generic_xlog.h"
 #include "access/parallel.h"
-#include "lib/pairingheap.h"
+#include "access/reloptions.h"
 #include "nodes/execnodes.h"
 #include "port.h"				/* for random() */
 #include "utils/sampling.h"
@@ -17,6 +16,10 @@
 #include "common/pg_prng.h"
 #endif
 
+#if PG_VERSION_NUM < 120000
+#include "access/relscan.h"
+#endif
+
 #ifdef IVFFLAT_BENCH
 #include "portability/instr_time.h"
 #endif
@@ -132,10 +135,16 @@ typedef struct IvfflatShared
 #ifdef IVFFLAT_KMEANS_DEBUG
 	double		inertia;
 #endif
+
+#if PG_VERSION_NUM < 120000
+	ParallelHeapScanDescData heapdesc;	/* must come last */
+#endif
 }			IvfflatShared;
 
+#if PG_VERSION_NUM >= 120000
 #define ParallelTableScanFromIvfflatShared(shared) \
 	(ParallelTableScanDesc) ((char *) (shared) + BUFFERALIGN(sizeof(IvfflatShared)))
+#endif
 
 typedef struct IvfflatLeader
 {

src/ivfinsert.c

@@ -2,7 +2,6 @@
 
 #include <float.h>
 
-#include "access/generic_xlog.h"
 #include "ivfflat.h"
 #include "storage/bufmgr.h"
 #include "storage/lmgr.h"

src/ivfscan.c

@@ -5,7 +5,6 @@
 #include "access/relscan.h"
 #include "catalog/pg_operator_d.h"
 #include "catalog/pg_type_d.h"
-#include "lib/pairingheap.h"
 #include "ivfflat.h"
 #include "miscadmin.h"
 #include "pgstat.h"
@@ -106,7 +105,12 @@ GetScanItems(IndexScanDesc scan, Datum value)
 	IvfflatScanOpaque so = (IvfflatScanOpaque) scan->opaque;
 	TupleDesc	tupdesc = RelationGetDescr(scan->indexRelation);
 	double		tuples = 0;
+
+#if PG_VERSION_NUM >= 120000
 	TupleTableSlot *slot = MakeSingleTupleTableSlot(so->tupdesc, &TTSOpsVirtual);
+#else
+	TupleTableSlot *slot = MakeSingleTupleTableSlot(so->tupdesc);
+#endif
 
 	/*
 	 * Reuse same set of shared buffers for scan
@@ -212,14 +216,22 @@ ivfflatbeginscan(Relation index, int nkeys, int norderbys)
 	so->collation = index->rd_indcollation[0];
 
 	/* Create tuple description for sorting */
+#if PG_VERSION_NUM >= 120000
 	so->tupdesc = CreateTemplateTupleDesc(2);
+#else
+	so->tupdesc = CreateTemplateTupleDesc(2, false);
+#endif
 	TupleDescInitEntry(so->tupdesc, (AttrNumber) 1, "distance", FLOAT8OID, -1, 0);
 	TupleDescInitEntry(so->tupdesc, (AttrNumber) 2, "heaptid", TIDOID, -1, 0);
 
 	/* Prep sort */
 	so->sortstate = tuplesort_begin_heap(so->tupdesc, 1, attNums, sortOperators, sortCollations, nullsFirstFlags, work_mem, NULL, false);
 
+#if PG_VERSION_NUM >= 120000
 	so->slot = MakeSingleTupleTableSlot(so->tupdesc, &TTSOpsMinimalTuple);
+#else
+	so->slot = MakeSingleTupleTableSlot(so->tupdesc);
+#endif
 
 	so->listQueue = pairingheap_allocate(CompareLists, scan);
 
@@ -309,8 +321,12 @@ ivfflatgettuple(IndexScanDesc scan, ScanDirection dir)
 	{
 		ItemPointer heaptid = (ItemPointer) DatumGetPointer(slot_getattr(so->slot, 2, &so->isnull));
 
+#if PG_VERSION_NUM >= 120000
 		scan->xs_heaptid = *heaptid;
-		scan->xs_recheck = false;
+#else
+		scan->xs_ctup.t_self = *heaptid;
+#endif
+
 		scan->xs_recheckorderby = false;
 		return true;
 	}

src/ivfutils.c

@@ -1,6 +1,5 @@
 #include "postgres.h"
 
-#include "access/generic_xlog.h"
 #include "ivfflat.h"
 #include "storage/bufmgr.h"
 #include "vector.h"

src/ivfvacuum.c

@@ -1,6 +1,5 @@
 #include "postgres.h"
 
-#include "access/generic_xlog.h"
 #include "commands/vacuum.h"
 #include "ivfflat.h"
 #include "storage/bufmgr.h"

src/vector.c

@@ -2,9 +2,7 @@
 
 #include <math.h>
 
-#include "bitvector.h"
 #include "catalog/pg_type.h"
-#include "common/shortest_dec.h"
 #include "fmgr.h"
 #include "hnsw.h"
 #include "ivfflat.h"
@@ -13,7 +11,6 @@
 #include "port.h"				/* for strtof() */
 #include "utils/array.h"
 #include "utils/builtins.h"
-#include "utils/float.h"
 #include "utils/lsyscache.h"
 #include "utils/numeric.h"
 #include "vector.h"
@@ -22,6 +19,13 @@
 #include "varatt.h"
 #endif
 
+#if PG_VERSION_NUM >= 120000
+#include "common/shortest_dec.h"
+#include "utils/float.h"
+#else
+#include <float.h>
+#endif
+
 #if PG_VERSION_NUM < 130000
 #define TYPALIGN_DOUBLE 'd'
 #define TYPALIGN_INT 'i'
@@ -173,15 +177,14 @@ PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_in);
 Datum
 vector_in(PG_FUNCTION_ARGS)
 {
-	char	   *lit = PG_GETARG_CSTRING(0);
+	char	   *str = PG_GETARG_CSTRING(0);
 	int32		typmod = PG_GETARG_INT32(2);
 	float		x[VECTOR_MAX_DIM];
 	int			dim = 0;
 	char	   *pt;
 	char	   *stringEnd;
 	Vector	   *result;
-	char	   *litcopy = pstrdup(lit);
+	char	   *lit = pstrdup(str);
-	char	   *str = litcopy;
 
 	while (vector_isspace(*str))
 		str++;
@@ -265,7 +268,7 @@ vector_in(PG_FUNCTION_ARGS)
 				(errcode(ERRCODE_DATA_EXCEPTION),
 				 errmsg("vector must have at least 1 dimension")));
 
-	pfree(litcopy);
+	pfree(lit);
 
 	CheckExpectedDim(typmod, dim);
 
@@ -289,6 +292,15 @@ vector_out(PG_FUNCTION_ARGS)
 	char	   *ptr;
 	int			n;
 
+#if PG_VERSION_NUM < 120000
+	int			ndig = FLT_DIG + extra_float_digits;
+
+	if (ndig < 1)
+		ndig = 1;
+
+#define FLOAT_SHORTEST_DECIMAL_LEN (ndig + 10)
+#endif
+
 	/*
 	 * Need:
 	 *
@@ -312,7 +324,11 @@ vector_out(PG_FUNCTION_ARGS)
 			ptr++;
 		}
 
+#if PG_VERSION_NUM >= 120000
 		n = float_to_shortest_decimal_bufn(vector->x[i], ptr);
+#else
+		n = sprintf(ptr, "%.*g", ndig, vector->x[i]);
+#endif
 		ptr += n;
 	}
 	*ptr = ']';
@@ -861,36 +877,15 @@ vector_mul(PG_FUNCTION_ARGS)
 	PG_RETURN_POINTER(result);
 }
 
-/*
- * Quantize a vector
- */
-PGDLLEXPORT PG_FUNCTION_INFO_V1(quantize_binary);
-Datum
-quantize_binary(PG_FUNCTION_ARGS)
-{
-	Vector	   *a = PG_GETARG_VECTOR_P(0);
-	float	   *ax = a->x;
-	VarBit	   *result = InitBitVector(a->dim);
-	unsigned char *rx = VARBITS(result);
-
-	/* TODO Improve */
-	for (int i = 0; i < a->dim; i++)
-		rx[i / 8] |= (ax[i] > 0) << (7 - (i % 8));
-
-	PG_RETURN_VARBIT_P(result);
-}
-
-
 /*
  * Internal helper to compare vectors
  */
 int
 vector_cmp_internal(Vector * a, Vector * b)
 {
-	int			dim = Min(a->dim, b->dim);
+	CheckDims(a, b);
 
-	/* Check values before dimensions to be consistent with Postgres arrays */
+	for (int i = 0; i < a->dim; i++)
-	for (int i = 0; i < dim; i++)
 	{
 		if (a->x[i] < b->x[i])
 			return -1;
@@ -898,13 +893,6 @@ vector_cmp_internal(Vector * a, Vector * b)
 		if (a->x[i] > b->x[i])
 			return 1;
 	}
-
-	if (a->dim < b->dim)
-		return -1;
-
-	if (a->dim > b->dim)
-		return 1;
-
 	return 0;
 }
 
@@ -918,9 +906,6 @@ vector_lt(PG_FUNCTION_ARGS)
 	Vector	   *a = PG_GETARG_VECTOR_P(0);
 	Vector	   *b = PG_GETARG_VECTOR_P(1);
 
-	/* TODO Remove in 0.7.0 */
-	CheckDims(a, b);
-
 	PG_RETURN_BOOL(vector_cmp_internal(a, b) < 0);
 }
 
@@ -934,9 +919,6 @@ vector_le(PG_FUNCTION_ARGS)
 	Vector	   *a = PG_GETARG_VECTOR_P(0);
 	Vector	   *b = PG_GETARG_VECTOR_P(1);
 
-	/* TODO Remove in 0.7.0 */
-	CheckDims(a, b);
-
 	PG_RETURN_BOOL(vector_cmp_internal(a, b) <= 0);
 }
 
@@ -950,9 +932,6 @@ vector_eq(PG_FUNCTION_ARGS)
 	Vector	   *a = PG_GETARG_VECTOR_P(0);
 	Vector	   *b = PG_GETARG_VECTOR_P(1);
 
-	/* TODO Remove in 0.7.0 */
-	CheckDims(a, b);
-
 	PG_RETURN_BOOL(vector_cmp_internal(a, b) == 0);
 }
 
@@ -966,9 +945,6 @@ vector_ne(PG_FUNCTION_ARGS)
 	Vector	   *a = PG_GETARG_VECTOR_P(0);
 	Vector	   *b = PG_GETARG_VECTOR_P(1);
 
-	/* TODO Remove in 0.7.0 */
-	CheckDims(a, b);
-
 	PG_RETURN_BOOL(vector_cmp_internal(a, b) != 0);
 }
 
@@ -982,9 +958,6 @@ vector_ge(PG_FUNCTION_ARGS)
 	Vector	   *a = PG_GETARG_VECTOR_P(0);
 	Vector	   *b = PG_GETARG_VECTOR_P(1);
 
-	/* TODO Remove in 0.7.0 */
-	CheckDims(a, b);
-
 	PG_RETURN_BOOL(vector_cmp_internal(a, b) >= 0);
 }
 
@@ -998,9 +971,6 @@ vector_gt(PG_FUNCTION_ARGS)
 	Vector	   *a = PG_GETARG_VECTOR_P(0);
 	Vector	   *b = PG_GETARG_VECTOR_P(1);
 
-	/* TODO Remove in 0.7.0 */
-	CheckDims(a, b);
-
 	PG_RETURN_BOOL(vector_cmp_internal(a, b) > 0);
 }
 

test/expected/bit_functions.out

@@ -1,64 +0,0 @@
-SELECT hamming_distance(B'111', B'111');
- hamming_distance 
-------------------
-                0
-(1 row)
-
-SELECT hamming_distance(B'111', B'110');
- hamming_distance 
-------------------
-                1
-(1 row)
-
-SELECT hamming_distance(B'111', B'100');
- hamming_distance 
-------------------
-                2
-(1 row)
-
-SELECT hamming_distance(B'111', B'000');
- hamming_distance 
-------------------
-                3
-(1 row)
-
-SELECT hamming_distance(B'111', B'00');
-ERROR:  different bit lengths 3 and 2
-SELECT jaccard_distance(B'1111', B'1111');
- jaccard_distance 
-------------------
-                0
-(1 row)
-
-SELECT jaccard_distance(B'1111', B'1110');
- jaccard_distance 
-------------------
-             0.25
-(1 row)
-
-SELECT jaccard_distance(B'1111', B'1100');
- jaccard_distance 
-------------------
-              0.5
-(1 row)
-
-SELECT jaccard_distance(B'1111', B'1000');
- jaccard_distance 
-------------------
-             0.75
-(1 row)
-
-SELECT jaccard_distance(B'1111', B'0000');
- jaccard_distance 
-------------------
-                1
-(1 row)
-
-SELECT jaccard_distance(B'1100', B'1000');
- jaccard_distance 
-------------------
-              0.5
-(1 row)
-
-SELECT jaccard_distance(B'1111', B'000');
-ERROR:  different bit lengths 4 and 3

test/expected/functions.out

@@ -24,56 +24,6 @@ SELECT '[1e37]'::vector * '[1e37]';
 ERROR:  value out of range: overflow
 SELECT '[1e-37]'::vector * '[1e-37]';
 ERROR:  value out of range: underflow
-SELECT '[1,2,3]'::vector = '[1,2,3]';
- ?column? 
-----------
- t
-(1 row)
-
-SELECT '[1,2,3]'::vector = '[1,2]';
-ERROR:  different vector dimensions 3 and 2
-SELECT vector_cmp('[1,2,3]', '[1,2,3]');
- vector_cmp 
-------------
-          0
-(1 row)
-
-SELECT vector_cmp('[1,2,3]', '[0,0,0]');
- vector_cmp 
-------------
-          1
-(1 row)
-
-SELECT vector_cmp('[0,0,0]', '[1,2,3]');
- vector_cmp 
-------------
-         -1
-(1 row)
-
-SELECT vector_cmp('[1,2]', '[1,2,3]');
- vector_cmp 
-------------
-         -1
-(1 row)
-
-SELECT vector_cmp('[1,2,3]', '[1,2]');
- vector_cmp 
-------------
-          1
-(1 row)
-
-SELECT vector_cmp('[1,2]', '[2,3,4]');
- vector_cmp 
-------------
-         -1
-(1 row)
-
-SELECT vector_cmp('[2,3]', '[1,2,3]');
- vector_cmp 
-------------
-          1
-(1 row)
-
 SELECT vector_dims('[1,2,3]');
  vector_dims 
 -------------
@@ -208,18 +158,6 @@ SELECT l1_distance('[3e38]', '[-3e38]');
     Infinity
 (1 row)
 
-SELECT quantize_binary('[1,0,-1]');
- quantize_binary 
------------------
- 100
-(1 row)
-
-SELECT quantize_binary('[0,0.1,-0.2,-0.3,0.4,0.5,0.6,-0.7,0.8,-0.9,1]');
- quantize_binary 
------------------
- 01001110101
-(1 row)
-
 SELECT avg(v) FROM unnest(ARRAY['[1,2,3]'::vector, '[3,5,7]']) v;
     avg    
 -----------

test/expected/hnsw_hamming.out

@@ -1,21 +0,0 @@
-SET enable_seqscan = off;
-CREATE TABLE t (val bit(3));
-INSERT INTO t (val) VALUES (B'000'), (B'100'), (B'111'), (NULL);
-CREATE INDEX ON t USING hnsw (val bit_hamming_ops);
-INSERT INTO t (val) VALUES (B'110');
-SELECT * FROM t ORDER BY val <~> B'111';
- val 
------
- 111
- 110
- 100
- 000
-(4 rows)
-
-SELECT COUNT(*) FROM (SELECT * FROM t ORDER BY val <~> (SELECT NULL::bit)) t2;
- count 
--------
-     4
-(1 row)
-
-DROP TABLE t;

test/expected/hnsw_jaccard.out

@@ -1,21 +0,0 @@
-SET enable_seqscan = off;
-CREATE TABLE t (val bit(4));
-INSERT INTO t (val) VALUES (B'0000'), (B'1100'), (B'1111'), (NULL);
-CREATE INDEX ON t USING hnsw (val bit_jaccard_ops);
-INSERT INTO t (val) VALUES (B'1110');
-SELECT * FROM t ORDER BY val <%> B'1111';
- val  
-------
- 1111
- 1110
- 1100
- 0000
-(4 rows)
-
-SELECT COUNT(*) FROM (SELECT * FROM t ORDER BY val <%> (SELECT NULL::bit)) t2;
- count 
--------
-     4
-(1 row)
-
-DROP TABLE t;

test/expected/input.out

@@ -116,30 +116,8 @@ SELECT '[1, ,3]'::vector;
 ERROR:  invalid input syntax for type vector: "[1, ,3]"
 LINE 1: SELECT '[1, ,3]'::vector;
                ^
-SELECT '[1,2,3]'::vector(3);
- vector  
----------
- [1,2,3]
-(1 row)
-
 SELECT '[1,2,3]'::vector(2);
 ERROR:  expected 2 dimensions, not 3
-SELECT '[1,2,3]'::vector(3, 2);
-ERROR:  invalid type modifier
-LINE 1: SELECT '[1,2,3]'::vector(3, 2);
-                          ^
-SELECT '[1,2,3]'::vector('a');
-ERROR:  invalid input syntax for type integer: "a"
-LINE 1: SELECT '[1,2,3]'::vector('a');
-                          ^
-SELECT '[1,2,3]'::vector(0);
-ERROR:  dimensions for type vector must be at least 1
-LINE 1: SELECT '[1,2,3]'::vector(0);
-                          ^
-SELECT '[1,2,3]'::vector(16001);
-ERROR:  dimensions for type vector cannot exceed 16000
-LINE 1: SELECT '[1,2,3]'::vector(16001);
-                          ^
 SELECT unnest('{"[1,2,3]", "[4,5,6]"}'::vector[]);
  unnest  
 ---------

test/sql/bit_functions.sql

@@ -1,13 +0,0 @@
-SELECT hamming_distance(B'111', B'111');
-SELECT hamming_distance(B'111', B'110');
-SELECT hamming_distance(B'111', B'100');
-SELECT hamming_distance(B'111', B'000');
-SELECT hamming_distance(B'111', B'00');
-
-SELECT jaccard_distance(B'1111', B'1111');
-SELECT jaccard_distance(B'1111', B'1110');
-SELECT jaccard_distance(B'1111', B'1100');
-SELECT jaccard_distance(B'1111', B'1000');
-SELECT jaccard_distance(B'1111', B'0000');
-SELECT jaccard_distance(B'1100', B'1000');
-SELECT jaccard_distance(B'1111', B'000');

test/sql/functions.sql

@@ -6,17 +6,6 @@ SELECT '[1,2,3]'::vector * '[4,5,6]';
 SELECT '[1e37]'::vector * '[1e37]';
 SELECT '[1e-37]'::vector * '[1e-37]';
 
-SELECT '[1,2,3]'::vector = '[1,2,3]';
-SELECT '[1,2,3]'::vector = '[1,2]';
-
-SELECT vector_cmp('[1,2,3]', '[1,2,3]');
-SELECT vector_cmp('[1,2,3]', '[0,0,0]');
-SELECT vector_cmp('[0,0,0]', '[1,2,3]');
-SELECT vector_cmp('[1,2]', '[1,2,3]');
-SELECT vector_cmp('[1,2,3]', '[1,2]');
-SELECT vector_cmp('[1,2]', '[2,3,4]');
-SELECT vector_cmp('[2,3]', '[1,2,3]');
-
 SELECT vector_dims('[1,2,3]');
 
 SELECT round(vector_norm('[1,1]')::numeric, 5);
@@ -48,9 +37,6 @@ SELECT l1_distance('[0,0]', '[0,1]');
 SELECT l1_distance('[1,2]', '[3]');
 SELECT l1_distance('[3e38]', '[-3e38]');
 
-SELECT quantize_binary('[1,0,-1]');
-SELECT quantize_binary('[0,0.1,-0.2,-0.3,0.4,0.5,0.6,-0.7,0.8,-0.9,1]');
-
 SELECT avg(v) FROM unnest(ARRAY['[1,2,3]'::vector, '[3,5,7]']) v;
 SELECT avg(v) FROM unnest(ARRAY['[1,2,3]'::vector, '[3,5,7]', NULL]) v;
 SELECT avg(v) FROM unnest(ARRAY[]::vector[]) v;

test/sql/hnsw_hamming.sql

@@ -1,12 +0,0 @@
-SET enable_seqscan = off;
-
-CREATE TABLE t (val bit(3));
-INSERT INTO t (val) VALUES (B'000'), (B'100'), (B'111'), (NULL);
-CREATE INDEX ON t USING hnsw (val bit_hamming_ops);
-
-INSERT INTO t (val) VALUES (B'110');
-
-SELECT * FROM t ORDER BY val <~> B'111';
-SELECT COUNT(*) FROM (SELECT * FROM t ORDER BY val <~> (SELECT NULL::bit)) t2;
-
-DROP TABLE t;

test/sql/hnsw_jaccard.sql

@@ -1,12 +0,0 @@
-SET enable_seqscan = off;
-
-CREATE TABLE t (val bit(4));
-INSERT INTO t (val) VALUES (B'0000'), (B'1100'), (B'1111'), (NULL);
-CREATE INDEX ON t USING hnsw (val bit_jaccard_ops);
-
-INSERT INTO t (val) VALUES (B'1110');
-
-SELECT * FROM t ORDER BY val <%> B'1111';
-SELECT COUNT(*) FROM (SELECT * FROM t ORDER BY val <%> (SELECT NULL::bit)) t2;
-
-DROP TABLE t;

test/sql/input.sql

@@ -22,13 +22,7 @@ SELECT '[1,]'::vector;
 SELECT '[1a]'::vector;
 SELECT '[1,,3]'::vector;
 SELECT '[1, ,3]'::vector;
-
-SELECT '[1,2,3]'::vector(3);
 SELECT '[1,2,3]'::vector(2);
-SELECT '[1,2,3]'::vector(3, 2);
-SELECT '[1,2,3]'::vector('a');
-SELECT '[1,2,3]'::vector(0);
-SELECT '[1,2,3]'::vector(16001);
 
 SELECT unnest('{"[1,2,3]", "[4,5,6]"}'::vector[]);
 SELECT '{"[1,2,3]"}'::vector(2)[];

test/t/018_hnsw_filtering.pl

@@ -1,114 +0,0 @@
-use strict;
-use warnings;
-use PostgresNode;
-use TestLib;
-use Test::More;
-
-my $dim = 3;
-my $nc = 50;
-my $limit = 20;
-
-my $array_sql = join(",", ('random()') x $dim);
-
-# Initialize node
-my $node = get_new_node('node');
-$node->init;
-$node->start;
-
-# Create table and index
-$node->safe_psql("postgres", "CREATE EXTENSION vector;");
-$node->safe_psql("postgres", "CREATE TABLE tst (i int4, v vector($dim), c int4, t text);");
-$node->safe_psql("postgres",
-	"INSERT INTO tst SELECT i, ARRAY[$array_sql], i % $nc, 'test ' || i FROM generate_series(1, 10000) i;"
-);
-$node->safe_psql("postgres", "CREATE INDEX idx ON tst USING hnsw (v vector_l2_ops);");
-$node->safe_psql("postgres", "ANALYZE tst;");
-
-# Generate query
-my @r = ();
-for (1 .. $dim)
-{
-	push(@r, rand());
-}
-my $query = "[" . join(",", @r) . "]";
-my $c = int(rand() * $nc);
-
-# Test attribute filtering
-my $explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE c = $c ORDER BY v <-> '$query' LIMIT $limit;
-));
-# TODO Do not use index
-like($explain, qr/Index Scan using idx/);
-
-# Test attribute filtering with few rows removed
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE c != $c ORDER BY v <-> '$query' LIMIT $limit;
-));
-like($explain, qr/Index Scan using idx/);
-
-# Test attribute filtering with few rows removed comparison
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE c >= 1 ORDER BY v <-> '$query' LIMIT $limit;
-));
-like($explain, qr/Index Scan using idx/);
-
-# Test attribute filtering with many rows removed comparison
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE c < 1 ORDER BY v <-> '$query' LIMIT $limit;
-));
-# TODO Do not use index
-like($explain, qr/Index Scan using idx/);
-
-# Test attribute filtering with few rows removed like
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE t LIKE '%%test%%' ORDER BY v <-> '$query' LIMIT $limit;
-));
-like($explain, qr/Index Scan using idx/);
-
-# Test attribute filtering with many rows removed like
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE t LIKE '%%other%%' ORDER BY v <-> '$query' LIMIT $limit;
-));
-like($explain, qr/Seq Scan/);
-
-# Test distance filtering
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE v <-> '$query' < 1 ORDER BY v <-> '$query' LIMIT $limit;
-));
-like($explain, qr/Index Scan using idx/);
-
-# Test distance filtering greater than distance
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE v <-> '$query' > 1 ORDER BY v <-> '$query' LIMIT $limit;
-));
-# TODO Do not use index
-like($explain, qr/Index Scan using idx/);
-
-# Test distance filtering without order
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE v <-> '$query' < 1;
-));
-like($explain, qr/Seq Scan/);
-
-# Test distance filtering without limit
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE v <-> '$query' < 1 ORDER BY v <-> '$query';
-));
-like($explain, qr/Seq Scan/);
-
-# Test attribute index
-$node->safe_psql("postgres", "CREATE INDEX attribute_idx ON tst (c);");
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE c = $c ORDER BY v <-> '$query' LIMIT $limit;
-));
-# TODO Use attribute index
-like($explain, qr/Index Scan using idx/);
-
-# Test partial index
-$node->safe_psql("postgres", "CREATE INDEX partial_idx ON tst USING hnsw (v vector_l2_ops) WHERE (c = $c);");
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE c = $c ORDER BY v <-> '$query' LIMIT $limit;
-));
-like($explain, qr/Index Scan using partial_idx/);
-
-done_testing();

test/t/018_ivfflat_deletes.pl

@@ -0,0 +1,43 @@
+use strict;
+use warnings;
+use PostgresNode;
+use TestLib;
+use Test::More;
+
+my $dim = 3;
+
+my $array_sql = join(",", ('random()') x $dim);
+
+# Initialize node
+my $node = get_new_node('node');
+$node->init;
+$node->start;
+
+# Create table and index
+$node->safe_psql("postgres", "CREATE EXTENSION vector;");
+$node->safe_psql("postgres", "CREATE TABLE tst (i serial, v vector($dim));");
+$node->safe_psql("postgres",
+	"INSERT INTO tst (v) SELECT ARRAY[$array_sql] FROM generate_series(1, 10000) i;"
+);
+$node->safe_psql("postgres", "CREATE INDEX ON tst USING ivfflat (v vector_l2_ops);");
+
+# Delete data
+$node->safe_psql("postgres", "DELETE FROM tst WHERE i % 100 != 0;");
+
+my $exp = $node->safe_psql("postgres", qq(
+	SET enable_indexscan = off;
+	SELECT i FROM tst ORDER BY v <-> '[0,0,0]';
+));
+
+# Run twice to make sure correct tuples marked as dead
+for (1 .. 2)
+{
+	my $res = $node->safe_psql("postgres", qq(
+		SET enable_seqscan = off;
+		SET ivfflat.probes = 100;
+		SELECT i FROM tst ORDER BY v <-> '[0,0,0]';
+	));
+	is($res, $exp);
+}
+
+done_testing();

test/t/019_ivfflat_filtering.pl

@@ -1,116 +0,0 @@
-use strict;
-use warnings;
-use PostgresNode;
-use TestLib;
-use Test::More;
-
-my $dim = 3;
-my $nc = 50;
-my $limit = 20;
-
-my $array_sql = join(",", ('random()') x $dim);
-
-# Initialize node
-my $node = get_new_node('node');
-$node->init;
-$node->start;
-
-# Create table and index
-$node->safe_psql("postgres", "CREATE EXTENSION vector;");
-$node->safe_psql("postgres", "CREATE TABLE tst (i int4, v vector($dim), c int4, t text);");
-$node->safe_psql("postgres",
-	"INSERT INTO tst SELECT i, ARRAY[$array_sql], i % $nc, 'test ' || i FROM generate_series(1, 10000) i;"
-);
-$node->safe_psql("postgres", "CREATE INDEX idx ON tst USING ivfflat (v vector_l2_ops) WITH (lists = 100);");
-$node->safe_psql("postgres", "ANALYZE tst;");
-
-# Generate query
-my @r = ();
-for (1 .. $dim)
-{
-	push(@r, rand());
-}
-my $query = "[" . join(",", @r) . "]";
-my $c = int(rand() * $nc);
-
-# Test attribute filtering
-my $explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE c = $c ORDER BY v <-> '$query' LIMIT $limit;
-));
-# TODO Do not use index
-like($explain, qr/Index Scan using idx/);
-
-# Test attribute filtering with few rows removed
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE c != $c ORDER BY v <-> '$query' LIMIT $limit;
-));
-like($explain, qr/Index Scan using idx/);
-
-# Test attribute filtering with few rows removed comparison
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE c >= 1 ORDER BY v <-> '$query' LIMIT $limit;
-));
-like($explain, qr/Index Scan using idx/);
-
-# Test attribute filtering with many rows removed comparison
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE c < 1 ORDER BY v <-> '$query' LIMIT $limit;
-));
-# TODO Do not use index
-like($explain, qr/Index Scan using idx/);
-
-# Test attribute filtering with few rows removed like
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE t LIKE '%%test%%' ORDER BY v <-> '$query' LIMIT $limit;
-));
-like($explain, qr/Index Scan using idx/);
-
-# Test attribute filtering with many rows removed like
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE t LIKE '%%other%%' ORDER BY v <-> '$query' LIMIT $limit;
-));
-like($explain, qr/Seq Scan/);
-
-# Test distance filtering
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE v <-> '$query' < 1 ORDER BY v <-> '$query' LIMIT $limit;
-));
-like($explain, qr/Index Scan using idx/);
-
-# Test distance filtering greater than distance
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE v <-> '$query' > 1 ORDER BY v <-> '$query' LIMIT $limit;
-));
-# TODO Do not use index
-like($explain, qr/Index Scan using idx/);
-
-# Test distance filtering without order
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE v <-> '$query' < 1;
-));
-like($explain, qr/Seq Scan/);
-
-# Test distance filtering without limit
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE v <-> '$query' < 1 ORDER BY v <-> '$query';
-));
-# TODO Do not use index
-like($explain, qr/Index Scan using idx/);
-
-# Test attribute index
-$node->safe_psql("postgres", "CREATE INDEX attribute_idx ON tst (c);");
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE c = $c ORDER BY v <-> '$query' LIMIT $limit;
-));
-# TODO Use attribute index
-like($explain, qr/Index Scan using idx/);
-
-# Test partial index
-$node->safe_psql("postgres", "CREATE INDEX partial_idx ON tst USING ivfflat (v vector_l2_ops) WITH (lists = 5) WHERE (c = $c);");
-$explain = $node->safe_psql("postgres", qq(
-	EXPLAIN ANALYZE SELECT i FROM tst WHERE c = $c ORDER BY v <-> '$query' LIMIT $limit;
-));
-# TODO Use partial index
-like($explain, qr/Index Scan using idx/);
-
-done_testing();

test/t/020_hnsw_bit_build_recall.pl

@@ -1,137 +0,0 @@
-use strict;
-use warnings;
-use PostgresNode;
-use TestLib;
-use Test::More;
-
-my $node;
-my @queries = ();
-my @expected;
-my $limit = 20;
-my $dim = 52;
-my $max = 2**$dim;
-
-sub test_recall
-{
-	my ($min, $operator) = @_;
-	my $correct = 0;
-	my $total = 0;
-
-	my $explain = $node->safe_psql("postgres", qq(
-		SET enable_seqscan = off;
-		SET hnsw.ef_search = 100;
-		EXPLAIN ANALYZE SELECT i FROM tst ORDER BY v $operator $queries[0] LIMIT $limit;
-	));
-	like($explain, qr/Index Scan/);
-
-	for my $i (0 .. $#queries)
-	{
-		my $actual = $node->safe_psql("postgres", qq(
-			SET enable_seqscan = off;
-			SET hnsw.ef_search = 100;
-			SELECT i FROM tst ORDER BY v $operator $queries[$i] LIMIT $limit;
-		));
-		my @actual_ids = split("\n", $actual);
-
-		my @expected_ids = split("\n", $expected[$i]);
-		my %expected_set = map { $_ => 1 } @expected_ids;
-
-		foreach (@actual_ids)
-		{
-			if (exists($expected_set{$_}))
-			{
-				$correct++;
-			}
-		}
-
-		$total += $limit;
-	}
-
-	cmp_ok($correct / $total, ">=", $min, $operator);
-}
-
-# Initialize node
-$node = get_new_node('node');
-$node->init;
-$node->start;
-
-# Create table
-$node->safe_psql("postgres", "CREATE EXTENSION vector;");
-$node->safe_psql("postgres", "CREATE TABLE tst (i int4, v bit($dim));");
-$node->safe_psql("postgres",
-	"INSERT INTO tst SELECT i, (random() * $max)::bigint::bit($dim) FROM generate_series(1, 10000) i;"
-);
-
-# Generate queries
-for (1 .. 20)
-{
-	my $r = int(rand() * $max);
-	push(@queries, "${r}::bigint::bit($dim)");
-}
-
-# Check each index type
-my @operators = ("<~>", "<\%>");
-my @opclasses = ("bit_hamming_ops", "bit_jaccard_ops");
-
-for my $i (0 .. $#operators)
-{
-	my $operator = $operators[$i];
-	my $opclass = $opclasses[$i];
-
-	# Get exact results
-	@expected = ();
-	foreach (@queries)
-	{
-		# Handle ties
-		my $res = $node->safe_psql("postgres", qq(
-			WITH top AS (
-				SELECT v $operator $_ AS distance FROM tst ORDER BY v $operator $_ LIMIT $limit
-			)
-			SELECT i FROM tst WHERE (v $operator $_) <= (SELECT MAX(distance) FROM top)
-		));
-		push(@expected, $res);
-	}
-
-	# Build index serially
-	$node->safe_psql("postgres", qq(
-		SET max_parallel_maintenance_workers = 0;
-		CREATE INDEX idx ON tst USING hnsw (v $opclass);
-	));
-
-	# Test approximate results
-	my $min = $operator eq "<\%>" ? 0.96 : 0.99;
-	test_recall($min, $operator);
-
-	$node->safe_psql("postgres", "DROP INDEX idx;");
-
-	# Build index in parallel in memory
-	my ($ret, $stdout, $stderr) = $node->psql("postgres", qq(
-		SET client_min_messages = DEBUG;
-		SET min_parallel_table_scan_size = 1;
-		CREATE INDEX idx ON tst USING hnsw (v $opclass);
-	));
-	is($ret, 0, $stderr);
-	like($stderr, qr/using \d+ parallel workers/);
-
-	# Test approximate results
-	test_recall($min, $operator);
-
-	$node->safe_psql("postgres", "DROP INDEX idx;");
-
-	# Build index in parallel on disk
-	# Set parallel_workers on table to use workers with low maintenance_work_mem
-	($ret, $stdout, $stderr) = $node->psql("postgres", qq(
-		ALTER TABLE tst SET (parallel_workers = 2);
-		SET client_min_messages = DEBUG;
-		SET maintenance_work_mem = '4MB';
-		CREATE INDEX idx ON tst USING hnsw (v $opclass);
-		ALTER TABLE tst RESET (parallel_workers);
-	));
-	is($ret, 0, $stderr);
-	like($stderr, qr/using \d+ parallel workers/);
-	like($stderr, qr/hnsw graph no longer fits into maintenance_work_mem/);
-
-	$node->safe_psql("postgres", "DROP INDEX idx;");
-}
-
-done_testing();

vector.control

@@ -1,4 +1,4 @@
 comment = 'vector data type and ivfflat and hnsw access methods'
-default_version = '0.6.2'
+default_version = '0.5.1'
 module_pathname = '$libdir/vector'
 relocatable = true