Improved code [skip ci]

This reverts commit 32e95a8598.

.github/workflows/build.yml

@@ -13,13 +13,13 @@ jobs:
           - postgres: 17
             os: ubuntu-24.04
           - postgres: 16
-            os: ubuntu-22.04
+            os: ubuntu-24.04-arm
           - postgres: 15
             os: ubuntu-22.04
           - postgres: 14
-            os: ubuntu-20.04
+            os: ubuntu-22.04-arm
           - postgres: 13
-            os: ubuntu-20.04
+            os: ubuntu-22.04
     steps:
       - uses: actions/checkout@v4
       - uses: ankane/setup-postgres@v1
@@ -28,7 +28,7 @@ jobs:
           dev-files: true
       - run: make
         env:
-          PG_CFLAGS: -DUSE_ASSERT_CHECKING -Wall -Wextra -Werror -Wno-unused-parameter -Wno-sign-compare
+          PG_CFLAGS: ${{ matrix.postgres == 18 && '-Wno-missing-field-initializers' || '' }} -DUSE_ASSERT_CHECKING -Wall -Wextra -Werror -Wno-unused-parameter -Wno-sign-compare
       - run: |
           export PG_CONFIG=`which pg_config`
           sudo --preserve-env=PG_CONFIG make install
@@ -46,8 +46,8 @@ jobs:
       fail-fast: false
       matrix:
         include:
-          - postgres: 16
-            os: macos-14
+          - postgres: 17
+            os: macos-15
           - postgres: 14
             os: macos-13
     steps:
@@ -70,12 +70,13 @@ jobs:
           tar xf $TAG.tar.gz
           mv postgres-$TAG postgres
         env:
-          TAG: ${{ matrix.postgres == 16 && 'REL_16_2' || 'REL_14_11' }}
+          TAG: ${{ matrix.postgres == 17 && 'REL_17_2' || 'REL_14_15' }}
       - run: make prove_installcheck PROVE_FLAGS="-I ./postgres/src/test/perl -I ./test/perl"
         env:
           PERL5LIB: /Users/runner/perl5/lib/perl5
-      - run: make clean && $(brew --prefix llvm@15)/bin/scan-build --status-bugs make
+      - run: make clean && $(brew --prefix llvm@$LLVM_VERSION)/bin/scan-build --status-bugs make
         env:
+          LLVM_VERSION: ${{ matrix.os == 'macos-15' && 18 || 15 }}
           PG_CFLAGS: -DUSE_ASSERT_CHECKING
   windows:
     runs-on: windows-latest
@@ -125,7 +126,7 @@ jobs:
       - uses: actions/checkout@v4
       - uses: ankane/setup-postgres-valgrind@v1
         with:
-          postgres-version: 16
+          postgres-version: 17
           check-ub: yes
       - run: make OPTFLAGS=""
       - run: sudo --preserve-env=PG_CONFIG make install

CHANGELOG.md

@@ -1,4 +1,4 @@
-## 0.8.0 (unreleased)
+## 0.8.0 (2024-10-30)
 
 - Added support for iterative index scans
 - Added casts for arrays to `sparsevec`

LICENSE

@@ -1,4 +1,4 @@
-Portions Copyright (c) 1996-2024, PostgreSQL Global Development Group
+Portions Copyright (c) 1996-2025, 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": "0.7.4",
+	"version": "0.8.0",
 	"maintainer": [
 		"Andrew Kane <andrew@ankane.org>"
 	],
@@ -12,7 +12,7 @@
 	"prereqs": {
 		"runtime": {
 			"requires": {
-				"PostgreSQL": "12.0.0"
+				"PostgreSQL": "13.0.0"
 			}
 		}
 	},
@@ -20,7 +20,7 @@
 		"vector": {
 			"file": "sql/vector.sql",
 			"docfile": "README.md",
-			"version": "0.7.4",
+			"version": "0.8.0",
 			"abstract": "Open-source vector similarity search for Postgres"
 		}
 	},

Makefile

@@ -1,5 +1,5 @@
 EXTENSION = vector
-EXTVERSION = 0.7.4
+EXTVERSION = 0.8.0
 
 MODULE_big = vector
 DATA = $(wildcard sql/*--*--*.sql)

Makefile.win

@@ -1,5 +1,5 @@
 EXTENSION = vector
-EXTVERSION = 0.7.4
+EXTVERSION = 0.8.0
 
 DATA_built = sql\$(EXTENSION)--$(EXTVERSION).sql
 OBJS = src\bitutils.obj src\bitvec.obj src\halfutils.obj src\halfvec.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\sparsevec.obj src\vector.obj

README.md

@@ -17,11 +17,11 @@ Plus [ACID](https://en.wikipedia.org/wiki/ACID) compliance, point-in-time recove
 
 ### Linux and Mac
 
-Compile and install the extension (supports Postgres 12+)
+Compile and install the extension (supports Postgres 13+)
 
 ```sh
 cd /tmp
-git clone --branch v0.7.4 https://github.com/pgvector/pgvector.git
+git clone --branch v0.8.0 https://github.com/pgvector/pgvector.git
 cd pgvector
 make
 make install # may need sudo
@@ -46,14 +46,12 @@ Then use `nmake` to build:
 ```cmd
 set "PGROOT=C:\Program Files\PostgreSQL\16"
 cd %TEMP%
-git clone --branch v0.7.4 https://github.com/pgvector/pgvector.git
+git clone --branch v0.8.0 https://github.com/pgvector/pgvector.git
 cd pgvector
 nmake /F Makefile.win
 nmake /F Makefile.win install
 ```
 
-Note: Postgres 17 is not supported yet due to an upstream issue
-
 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).
@@ -84,7 +82,7 @@ Get the nearest neighbors by L2 distance
 SELECT * FROM items ORDER BY embedding <-> '[3,1,2]' LIMIT 5;
 ```
 
-Also supports inner product (`<#>`), cosine distance (`<=>`), and L1 distance (`<+>`, added in 0.7.0)
+Also supports inner product (`<#>`), cosine distance (`<=>`), and L1 distance (`<+>`)
 
 Note: `<#>` returns the negative inner product since Postgres only supports `ASC` order index scans on operators
 
@@ -148,9 +146,9 @@ Supported distance functions are:
 - `<->` - L2 distance
 - `<#>` - (negative) inner product
 - `<=>` - cosine distance
-- `<+>` - L1 distance (added in 0.7.0)
-- `<~>` - Hamming distance (binary vectors, added in 0.7.0)
-- `<%>` - Jaccard distance (binary vectors, added in 0.7.0)
+- `<+>` - L1 distance
+- `<~>` - Hamming distance (binary vectors)
+- `<%>` - Jaccard distance (binary vectors)
 
 Get the nearest neighbors to a row
 
@@ -237,19 +235,19 @@ Cosine distance
 CREATE INDEX ON items USING hnsw (embedding vector_cosine_ops);
 ```
 
-L1 distance - added in 0.7.0
+L1 distance
 
 ```sql
 CREATE INDEX ON items USING hnsw (embedding vector_l1_ops);
 ```
 
-Hamming distance - added in 0.7.0
+Hamming distance
 
 ```sql
 CREATE INDEX ON items USING hnsw (embedding bit_hamming_ops);
 ```
 
-Jaccard distance - added in 0.7.0
+Jaccard distance
 
 ```sql
 CREATE INDEX ON items USING hnsw (embedding bit_jaccard_ops);
@@ -258,9 +256,9 @@ CREATE INDEX ON items USING hnsw (embedding bit_jaccard_ops);
 Supported types are:
 
 - `vector` - up to 2,000 dimensions
-- `halfvec` - up to 4,000 dimensions (added in 0.7.0)
-- `bit` - up to 64,000 dimensions (added in 0.7.0)
-- `sparsevec` - up to 1,000 non-zero elements (added in 0.7.0)
+- `halfvec` - up to 4,000 dimensions
+- `bit` - up to 64,000 dimensions
+- `sparsevec` - up to 1,000 non-zero elements
 
 ### Index Options
 
@@ -314,17 +312,19 @@ Note: Do not set `maintenance_work_mem` so high that it exhausts the memory on t
 
 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)
+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)
+For a large number of workers, you may need to increase `max_parallel_workers` (8 by default)
+
+The [index options](#index-options) also have a significant impact on build time (use the defaults unless seeing low recall)
 
 ### Indexing Progress
 
-Check [indexing progress](https://www.postgresql.org/docs/current/progress-reporting.html#CREATE-INDEX-PROGRESS-REPORTING) with Postgres 12+
+Check [indexing progress](https://www.postgresql.org/docs/current/progress-reporting.html#CREATE-INDEX-PROGRESS-REPORTING)
 
 ```sql
 SELECT phase, round(100.0 * blocks_done / nullif(blocks_total, 0), 1) AS "%" FROM pg_stat_progress_create_index;
@@ -367,7 +367,7 @@ Cosine distance
 CREATE INDEX ON items USING ivfflat (embedding vector_cosine_ops) WITH (lists = 100);
 ```
 
-Hamming distance - added in 0.7.0
+Hamming distance
 
 ```sql
 CREATE INDEX ON items USING ivfflat (embedding bit_hamming_ops) WITH (lists = 100);
@@ -376,8 +376,8 @@ CREATE INDEX ON items USING ivfflat (embedding bit_hamming_ops) WITH (lists = 10
 Supported types are:
 
 - `vector` - up to 2,000 dimensions
-- `halfvec` - up to 4,000 dimensions (added in 0.7.0)
-- `bit` - up to 64,000 dimensions (added in 0.7.0)
+- `halfvec` - up to 4,000 dimensions
+- `bit` - up to 64,000 dimensions
 
 ### Query Options
 
@@ -410,7 +410,7 @@ For a large number of workers, you may also need to increase `max_parallel_worke
 
 ### Indexing Progress
 
-Check [indexing progress](https://www.postgresql.org/docs/current/progress-reporting.html#CREATE-INDEX-PROGRESS-REPORTING) with Postgres 12+
+Check [indexing progress](https://www.postgresql.org/docs/current/progress-reporting.html#CREATE-INDEX-PROGRESS-REPORTING)
 
 ```sql
 SELECT phase, round(100.0 * tuples_done / nullif(tuples_total, 0), 1) AS "%" FROM pg_stat_progress_create_index;
@@ -477,8 +477,6 @@ CREATE TABLE items (embedding vector(3), category_id int) PARTITION BY LIST(cate
 
 ## Iterative Index Scans
 
-*Unreleased*
-
 With approximate indexes, queries with filtering can return less results since filtering is applied *after* the index is scanned. Starting with 0.8.0, you can enable iterative index scans, which will automatically scan more of the index until enough results are found (or it reaches `hnsw.max_scan_tuples` or `ivfflat.max_probes`).
 
 Iterative scans can use strict or relaxed ordering.
@@ -549,8 +547,6 @@ Note: If this is lower than `ivfflat.probes`, `ivfflat.probes` will be used
 
 ## Half-Precision Vectors
 
-*Added in 0.7.0*
-
 Use the `halfvec` type to store half-precision vectors
 
 ```sql
@@ -559,8 +555,6 @@ CREATE TABLE items (id bigserial PRIMARY KEY, embedding halfvec(3));
 
 ## Half-Precision Indexing
 
-*Added in 0.7.0*
-
 Index vectors at half precision for smaller indexes
 
 ```sql
@@ -582,24 +576,16 @@ CREATE TABLE items (id bigserial PRIMARY KEY, embedding bit(3));
 INSERT INTO items (embedding) VALUES ('000'), ('111');
 ```
 
-Get the nearest neighbors by Hamming distance (added in 0.7.0)
+Get the nearest neighbors by Hamming distance
 
 ```sql
 SELECT * FROM items ORDER BY embedding <~> '101' LIMIT 5;
 ```
 
-Or (before 0.7.0)
-
-```sql
-SELECT * FROM items ORDER BY bit_count(embedding # '101') LIMIT 5;
-```
-
 Also supports Jaccard distance (`<%>`)
 
 ## Binary Quantization
 
-*Added in 0.7.0*
-
 Use expression indexing for binary quantization
 
 ```sql
@@ -622,8 +608,6 @@ SELECT * FROM (
 
 ## Sparse Vectors
 
-*Added in 0.7.0*
-
 Use the `sparsevec` type to store sparse vectors
 
 ```sql
@@ -657,8 +641,6 @@ You can use [Reciprocal Rank Fusion](https://github.com/pgvector/pgvector-python
 
 ## Indexing Subvectors
 
-*Added in 0.7.0*
-
 Use expression indexing to index subvectors
 
 ```sql
@@ -772,8 +754,6 @@ SELECT query, calls, ROUND((total_plan_time + total_exec_time) / calls) AS avg_t
     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
@@ -801,8 +781,12 @@ C | [pgvector-c](https://github.com/pgvector/pgvector-c)
 C++ | [pgvector-cpp](https://github.com/pgvector/pgvector-cpp)
 C#, F#, Visual Basic | [pgvector-dotnet](https://github.com/pgvector/pgvector-dotnet)
 Crystal | [pgvector-crystal](https://github.com/pgvector/pgvector-crystal)
+D | [pgvector-d](https://github.com/pgvector/pgvector-d)
 Dart | [pgvector-dart](https://github.com/pgvector/pgvector-dart)
 Elixir | [pgvector-elixir](https://github.com/pgvector/pgvector-elixir)
+Erlang | [pgvector-erlang](https://github.com/pgvector/pgvector-erlang)
+Fortran | [pgvector-fortran](https://github.com/pgvector/pgvector-fortran)
+Gleam | [pgvector-gleam](https://github.com/pgvector/pgvector-gleam)
 Go | [pgvector-go](https://github.com/pgvector/pgvector-go)
 Haskell | [pgvector-haskell](https://github.com/pgvector/pgvector-haskell)
 Java, Kotlin, Groovy, Scala | [pgvector-java](https://github.com/pgvector/pgvector-java)
@@ -816,6 +800,7 @@ Perl | [pgvector-perl](https://github.com/pgvector/pgvector-perl)
 PHP | [pgvector-php](https://github.com/pgvector/pgvector-php)
 Python | [pgvector-python](https://github.com/pgvector/pgvector-python)
 R | [pgvector-r](https://github.com/pgvector/pgvector-r)
+Raku | [pgvector-raku](https://github.com/pgvector/pgvector-raku)
 Ruby | [pgvector-ruby](https://github.com/pgvector/pgvector-ruby), [Neighbor](https://github.com/ankane/neighbor)
 Rust | [pgvector-rust](https://github.com/pgvector/pgvector-rust)
 Swift | [pgvector-swift](https://github.com/pgvector/pgvector-swift)
@@ -837,7 +822,7 @@ You can use [half-precision indexing](#half-precision-indexing) to index up to 4
 
 #### Can I store vectors with different dimensions in the same column?
 
-You can use `vector` as the type (instead of `vector(3)`).
+You can use `vector` as the type (instead of `vector(n)`).
 
 ```sql
 CREATE TABLE embeddings (model_id bigint, item_id bigint, embedding vector, PRIMARY KEY (model_id, item_id));
@@ -937,7 +922,7 @@ 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.
+Results are limited by the size of the dynamic candidate list (`hnsw.ef_search`), which is 40 by default. There may be even less results due to dead tuples or filtering conditions in the query. Enabling [iterative index scans](#iterative-index-scans) can help address this.
 
 Also, note that `NULL` vectors are not indexed (as well as zero vectors for cosine distance).
 
@@ -949,7 +934,7 @@ 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`).
+Results can also be limited by the number of probes (`ivfflat.probes`). Enabling [iterative index scans](#iterative-index-scans) can address this.
 
 Also, note that `NULL` vectors are not indexed (as well as zero vectors for cosine distance).
 
@@ -1116,7 +1101,13 @@ Note: Replace `17` with your Postgres server version
 
 ### Missing SDK
 
-If compilation fails and the output includes `warning: no such sysroot directory` on Mac, reinstall Xcode Command Line Tools.
+If compilation fails and the output includes `warning: no such sysroot directory` on Mac, your Postgres installation points to a path that no longer exists.
+
+```sh
+pg_config --cppflags
+```
+
+Reinstall Postgres to fix this.
 
 ### Portability
 
@@ -1134,6 +1125,14 @@ make OPTFLAGS=""
 
 If compilation fails with `Cannot open include file: 'postgres.h': No such file or directory`, make sure `PGROOT` is correct.
 
+### Mismatched Architecture
+
+If compilation fails with `error C2196: case value '4' already used`, make sure `vcvars64.bat` was called. Then run `nmake /F Makefile.win clean` and re-run the installation instructions.
+
+### Missing Symbol
+
+If linking fails with `unresolved external symbol float_to_shortest_decimal_bufn` with Postgres 17.0-17.2, upgrade to Postgres 17.3+.
+
 ### Permissions
 
 If installation fails with `Access is denied`, re-run the installation instructions as an administrator.
@@ -1153,11 +1152,17 @@ This adds pgvector to the [Postgres image](https://hub.docker.com/_/postgres) (r
 You can also build the image manually:
 
 ```sh
-git clone --branch v0.7.4 https://github.com/pgvector/pgvector.git
+git clone --branch v0.8.0 https://github.com/pgvector/pgvector.git
 cd pgvector
 docker build --pull --build-arg PG_MAJOR=17 -t myuser/pgvector .
 ```
 
+If you increase `maintenance_work_mem`, make sure `--shm-size` is at least that size to avoid an error with parallel HNSW index builds.
+
+```sh
+docker run --shm-size=1g ...
+```
+
 ### Homebrew
 
 With Homebrew Postgres, you can use:
@@ -1203,7 +1208,7 @@ Note: Replace `17` with your Postgres server version
 Install the FreeBSD package with:
 
 ```sh
-pkg install postgresql15-pgvector
+pkg install postgresql16-pgvector
 ```
 
 or the port with:
@@ -1245,36 +1250,6 @@ You can check the version in the current database with:
 SELECT extversion FROM pg_extension WHERE extname = 'vector';
 ```
 
-## Upgrade Notes
-
-### 0.6.0
-
-#### Postgres 12
-
-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);
-```
-
-Then run `make install` and `ALTER EXTENSION vector UPDATE;`.
-
-#### Docker
-
-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).
-
-```sh
-docker pull pgvector/pgvector:pg16
-# or
-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.
-
-```sh
-docker run --shm-size=1g ...
-```
-
 ## Thanks
 
 Thanks to:

src/hnsw.c

@@ -259,6 +259,11 @@ hnswhandler(PG_FUNCTION_ARGS)
 	amroutine->amoptsprocnum = 0;
 	amroutine->amcanorder = false;
 	amroutine->amcanorderbyop = true;
+#if PG_VERSION_NUM >= 180000
+	amroutine->amcanhash = false;
+	amroutine->amconsistentequality = false;
+	amroutine->amconsistentordering = false;
+#endif
 	amroutine->amcanbackward = false;	/* can change direction mid-scan */
 	amroutine->amcanunique = false;
 	amroutine->amcanmulticol = false;
@@ -291,6 +296,9 @@ hnswhandler(PG_FUNCTION_ARGS)
 	amroutine->amvacuumcleanup = hnswvacuumcleanup;
 	amroutine->amcanreturn = NULL;
 	amroutine->amcostestimate = hnswcostestimate;
+#if PG_VERSION_NUM >= 180000
+	amroutine->amgettreeheight = NULL;
+#endif
 	amroutine->amoptions = hnswoptions;
 	amroutine->amproperty = NULL;	/* TODO AMPROP_DISTANCE_ORDERABLE */
 	amroutine->ambuildphasename = hnswbuildphasename;
@@ -311,5 +319,10 @@ hnswhandler(PG_FUNCTION_ARGS)
 	amroutine->aminitparallelscan = NULL;
 	amroutine->amparallelrescan = NULL;
 
+#if PG_VERSION_NUM >= 180000
+	amroutine->amtranslatestrategy = NULL;
+	amroutine->amtranslatecmptype = NULL;
+#endif
+
 	PG_RETURN_POINTER(amroutine);
 }

src/hnsw.h

@@ -126,7 +126,7 @@ typedef struct HnswNeighborArray HnswNeighborArray;
 
 #define HnswPtrDeclare(type, relptrtype, ptrtype) \
 	relptr_declare(type, relptrtype); \
-	typedef union { type *ptr; relptrtype relptr; } ptrtype;
+	typedef union { type *ptr; relptrtype relptr; } ptrtype
 
 /* Pointers that can be absolute or relative */
 /* Use char for DatumPtr so works with Pointer */

src/hnswutils.c

@@ -722,8 +722,6 @@ CountElement(HnswElement skipElement, HnswElement e)
 static void
 HnswLoadUnvisitedFromMemory(char *base, HnswElement element, HnswUnvisited * unvisited, int *unvisitedLength, visited_hash * v, int lc, HnswNeighborArray * localNeighborhood, Size neighborhoodSize)
 {
-	uint32		hashes[HNSW_MAX_M * 2];
-
 	/* Get the neighborhood at layer lc */
 	HnswNeighborArray *neighborhood = HnswGetNeighbors(base, element, lc);
 
@@ -735,33 +733,14 @@ HnswLoadUnvisitedFromMemory(char *base, HnswElement element, HnswUnvisited * unv
 	*unvisitedLength = 0;
 
 	for (int i = 0; i < localNeighborhood->length; i++)
-		hashes[i] = HnswPtrAccess(base, localNeighborhood->items[i].element)->hash;
-
-	if (base != NULL)
 	{
-		for (int i = 0; i < localNeighborhood->length; i++)
-		{
-			HnswCandidate *hc = &localNeighborhood->items[i];
-			bool		found;
+		HnswCandidate *hc = &localNeighborhood->items[i];
+		bool		found;
 
-			offsethash_insert_hash(v->offsets, HnswPtrOffset(hc->element), hashes[i], &found);
+		AddToVisited(base, v, hc->element, true, &found);
 
-			if (!found)
-				unvisited[(*unvisitedLength)++].element = HnswPtrAccess(base, hc->element);
-		}
-	}
-	else
-	{
-		for (int i = 0; i < localNeighborhood->length; i++)
-		{
-			HnswCandidate *hc = &localNeighborhood->items[i];
-			bool		found;
-
-			pointerhash_insert_hash(v->pointers, (uintptr_t) HnswPtrPointer(hc->element), hashes[i], &found);
-
-			if (!found)
-				unvisited[(*unvisitedLength)++].element = HnswPtrAccess(base, hc->element);
-		}
+		if (!found)
+			unvisited[(*unvisitedLength)++].element = HnswPtrAccess(base, hc->element);
 	}
 }
 
@@ -826,6 +805,12 @@ HnswLoadUnvisitedFromDisk(HnswElement element, HnswUnvisited * unvisited, int *u
 		if (!found)
 			unvisited[(*unvisitedLength)++].indextid = *indextid;
 	}
+
+#ifdef USE_PREFETCH
+	/* TODO limit by get_tablespace_io_concurrency */
+	for (int i = 0; i < *unvisitedLength; i++)
+		PrefetchBuffer(index, MAIN_FORKNUM, ItemPointerGetBlockNumber(&unvisited[i].indextid));
+#endif
 }
 
 /*
@@ -1414,7 +1399,7 @@ hnsw_halfvec_support(PG_FUNCTION_ARGS)
 	};
 
 	PG_RETURN_POINTER(&typeInfo);
-};
+}
 
 FUNCTION_PREFIX PG_FUNCTION_INFO_V1(hnsw_bit_support);
 Datum
@@ -1427,7 +1412,7 @@ hnsw_bit_support(PG_FUNCTION_ARGS)
 	};
 
 	PG_RETURN_POINTER(&typeInfo);
-};
+}
 
 FUNCTION_PREFIX PG_FUNCTION_INFO_V1(hnsw_sparsevec_support);
 Datum
@@ -1440,4 +1425,4 @@ hnsw_sparsevec_support(PG_FUNCTION_ARGS)
 	};
 
 	PG_RETURN_POINTER(&typeInfo);
-};
+}

src/hnswvacuum.c

@@ -9,6 +9,10 @@
 #include "storage/lmgr.h"
 #include "utils/memutils.h"
 
+#if PG_VERSION_NUM >= 180000
+#define vacuum_delay_point() vacuum_delay_point(false)
+#endif
+
 /*
  * Check if deleted list contains an index TID
  */

src/ivfbuild.c

@@ -360,7 +360,7 @@ InitBuildState(IvfflatBuildState * buildstate, Relation heap, Relation index, In
 	buildstate->sortdesc = CreateTemplateTupleDesc(3);
 	TupleDescInitEntry(buildstate->sortdesc, (AttrNumber) 1, "list", INT4OID, -1, 0);
 	TupleDescInitEntry(buildstate->sortdesc, (AttrNumber) 2, "tid", TIDOID, -1, 0);
-	TupleDescInitEntry(buildstate->sortdesc, (AttrNumber) 3, "vector", buildstate->tupdesc->attrs[0].atttypid, -1, 0);
+	TupleDescInitEntry(buildstate->sortdesc, (AttrNumber) 3, "vector", TupleDescAttr(buildstate->tupdesc, 0)->atttypid, -1, 0);
 
 	buildstate->slot = MakeSingleTupleTableSlot(buildstate->sortdesc, &TTSOpsVirtual);
 

src/ivfflat.c

@@ -186,6 +186,11 @@ ivfflathandler(PG_FUNCTION_ARGS)
 	amroutine->amoptsprocnum = 0;
 	amroutine->amcanorder = false;
 	amroutine->amcanorderbyop = true;
+#if PG_VERSION_NUM >= 180000
+	amroutine->amcanhash = false;
+	amroutine->amconsistentequality = false;
+	amroutine->amconsistentordering = false;
+#endif
 	amroutine->amcanbackward = false;	/* can change direction mid-scan */
 	amroutine->amcanunique = false;
 	amroutine->amcanmulticol = false;
@@ -218,6 +223,9 @@ ivfflathandler(PG_FUNCTION_ARGS)
 	amroutine->amvacuumcleanup = ivfflatvacuumcleanup;
 	amroutine->amcanreturn = NULL;	/* tuple not included in heapsort */
 	amroutine->amcostestimate = ivfflatcostestimate;
+#if PG_VERSION_NUM >= 180000
+	amroutine->amgettreeheight = NULL;
+#endif
 	amroutine->amoptions = ivfflatoptions;
 	amroutine->amproperty = NULL;	/* TODO AMPROP_DISTANCE_ORDERABLE */
 	amroutine->ambuildphasename = ivfflatbuildphasename;
@@ -238,5 +246,10 @@ ivfflathandler(PG_FUNCTION_ARGS)
 	amroutine->aminitparallelscan = NULL;
 	amroutine->amparallelrescan = NULL;
 
+#if PG_VERSION_NUM >= 180000
+	amroutine->amtranslatestrategy = NULL;
+	amroutine->amtranslatecmptype = NULL;
+#endif
+
 	PG_RETURN_POINTER(amroutine);
 }

src/ivfutils.c

@@ -355,7 +355,7 @@ ivfflat_halfvec_support(PG_FUNCTION_ARGS)
 	};
 
 	PG_RETURN_POINTER(&typeInfo);
-};
+}
 
 FUNCTION_PREFIX PG_FUNCTION_INFO_V1(ivfflat_bit_support);
 Datum
@@ -370,4 +370,4 @@ ivfflat_bit_support(PG_FUNCTION_ARGS)
 	};
 
 	PG_RETURN_POINTER(&typeInfo);
-};
+}

src/ivfvacuum.c

@@ -5,6 +5,10 @@
 #include "ivfflat.h"
 #include "storage/bufmgr.h"
 
+#if PG_VERSION_NUM >= 180000
+#define vacuum_delay_point() vacuum_delay_point(false)
+#endif
+
 /*
  * Bulk delete tuples from the index
  */

src/sparsevec.c

@@ -4,6 +4,7 @@
 #include <math.h>
 
 #include "catalog/pg_type.h"
+#include "common/shortest_dec.h"
 #include "common/string.h"
 #include "fmgr.h"
 #include "halfutils.h"
@@ -12,17 +13,10 @@
 #include "sparsevec.h"
 #include "utils/array.h"
 #include "utils/builtins.h"
+#include "utils/float.h"
 #include "utils/lsyscache.h"
 #include "vector.h"
 
-#if PG_VERSION_NUM >= 120000
-#include "common/shortest_dec.h"
-#include "utils/float.h"
-#else
-#include <float.h>
-#include "utils/builtins.h"
-#endif
-
 typedef struct SparseInputElement
 {
 	int32		index;

test/expected/hnsw_vector.out

@@ -123,6 +123,12 @@ SELECT * FROM t ORDER BY val <-> '[3,3,3]';
  [0,0,0]
 (3 rows)
 
+TRUNCATE t;
+SELECT * FROM t ORDER BY val <-> '[3,3,3]';
+ val 
+-----
+(0 rows)
+
 RESET hnsw.iterative_scan;
 RESET hnsw.ef_search;
 DROP TABLE t;

test/expected/ivfflat_vector.out

@@ -110,6 +110,15 @@ SELECT * FROM t ORDER BY val <-> '[3,3,3]';
  [1,1,1]
 (2 rows)
 
+TRUNCATE t;
+NOTICE:  ivfflat index created with little data
+DETAIL:  This will cause low recall.
+HINT:  Drop the index until the table has more data.
+SELECT * FROM t ORDER BY val <-> '[3,3,3]';
+ val 
+-----
+(0 rows)
+
 RESET ivfflat.iterative_scan;
 RESET ivfflat.max_probes;
 DROP TABLE t;

test/sql/hnsw_vector.sql

@@ -70,6 +70,9 @@ SELECT * FROM t ORDER BY val <-> '[3,3,3]';
 SET hnsw.iterative_scan = relaxed_order;
 SELECT * FROM t ORDER BY val <-> '[3,3,3]';
 
+TRUNCATE t;
+SELECT * FROM t ORDER BY val <-> '[3,3,3]';
+
 RESET hnsw.iterative_scan;
 RESET hnsw.ef_search;
 DROP TABLE t;

test/sql/ivfflat_vector.sql

@@ -59,6 +59,9 @@ SELECT * FROM t ORDER BY val <-> '[3,3,3]';
 SET ivfflat.max_probes = 2;
 SELECT * FROM t ORDER BY val <-> '[3,3,3]';
 
+TRUNCATE t;
+SELECT * FROM t ORDER BY val <-> '[3,3,3]';
+
 RESET ivfflat.iterative_scan;
 RESET ivfflat.max_probes;
 DROP TABLE t;

vector.control

@@ -1,4 +1,4 @@
 comment = 'vector data type and ivfflat and hnsw access methods'
-default_version = '0.7.4'
+default_version = '0.8.0'
 module_pathname = '$libdir/vector'
 relocatable = true