Added search parameters to EXPLAIN [skip ci]

.github/workflows/build.yml

@@ -46,8 +46,8 @@ jobs:
       fail-fast: false
       matrix:
         include:
-          - postgres: 17
-            os: macos-15
+          - postgres: 16
+            os: macos-14
           - postgres: 14
             os: macos-13
     steps:
@@ -70,13 +70,12 @@ jobs:
           tar xf $TAG.tar.gz
           mv postgres-$TAG postgres
         env:
-          TAG: ${{ matrix.postgres == 17 && 'REL_17_2' || 'REL_14_15' }}
+          TAG: ${{ matrix.postgres == 16 && 'REL_16_2' || 'REL_14_11' }}
       - run: make prove_installcheck PROVE_FLAGS="-I ./postgres/src/test/perl -I ./test/perl"
         env:
           PERL5LIB: /Users/runner/perl5/lib/perl5
-      - run: make clean && $(brew --prefix llvm@$LLVM_VERSION)/bin/scan-build --status-bugs make
+      - run: make clean && $(brew --prefix llvm@15)/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
@@ -87,7 +86,7 @@ jobs:
         with:
           postgres-version: 14
       - run: |
-          call "C:\Program Files\Microsoft Visual Studio\2022\Enterprise\VC\Auxiliary\Build\vcvars32.bat" && ^
+          call "C:\Program Files\Microsoft Visual Studio\2022\Enterprise\VC\Auxiliary\Build\vcvars64.bat" && ^
           nmake /NOLOGO /F Makefile.win && ^
           nmake /NOLOGO /F Makefile.win install && ^
           nmake /NOLOGO /F Makefile.win installcheck && ^
@@ -126,7 +125,7 @@ jobs:
       - uses: actions/checkout@v4
       - uses: ankane/setup-postgres-valgrind@v1
         with:
-          postgres-version: 17
+          postgres-version: 16
           check-ub: yes
       - run: make OPTFLAGS=""
       - run: sudo --preserve-env=PG_CONFIG make install

CHANGELOG.md

@@ -1,10 +1,10 @@
-## 0.8.0 (2024-10-30)
+## 0.8.0 (unreleased)
 
 - Added support for iterative index scans
 - Added casts for arrays to `sparsevec`
-- Improved cost estimation for better index selection when filtering
-- Improved performance of HNSW index scans
+- Improved cost estimation
 - Improved performance of HNSW inserts and on-disk index builds
+- Reduced memory usage for HNSW index scans
 - Dropped support for Postgres 12
 
 ## 0.7.4 (2024-08-05)

LICENSE

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

Makefile

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

Makefile.win

@@ -1,5 +1,5 @@
 EXTENSION = vector
-EXTVERSION = 0.8.0
+EXTVERSION = 0.7.4
 
 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 13+)
+Compile and install the extension (supports Postgres 12+)
 
 ```sh
 cd /tmp
-git clone --branch v0.8.0 https://github.com/pgvector/pgvector.git
+git clone --branch v0.7.4 https://github.com/pgvector/pgvector.git
 cd pgvector
 make
 make install # may need sudo
@@ -46,12 +46,14 @@ Then use `nmake` to build:
 ```cmd
 set "PGROOT=C:\Program Files\PostgreSQL\16"
 cd %TEMP%
-git clone --branch v0.8.0 https://github.com/pgvector/pgvector.git
+git clone --branch v0.7.4 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).
@@ -82,7 +84,7 @@ 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 (`<+>`)
+Also supports inner product (`<#>`), cosine distance (`<=>`), and L1 distance (`<+>`, added in 0.7.0)
 
 Note: `<#>` returns the negative inner product since Postgres only supports `ASC` order index scans on operators
 
@@ -146,9 +148,9 @@ Supported distance functions are:
 - `<->` - L2 distance
 - `<#>` - (negative) inner product
 - `<=>` - cosine distance
-- `<+>` - L1 distance
-- `<~>` - Hamming distance (binary vectors)
-- `<%>` - Jaccard distance (binary vectors)
+- `<+>` - 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)
 
 Get the nearest neighbors to a row
 
@@ -235,19 +237,19 @@ Cosine distance
 CREATE INDEX ON items USING hnsw (embedding vector_cosine_ops);
 ```
 
-L1 distance
+L1 distance - added in 0.7.0
 
 ```sql
 CREATE INDEX ON items USING hnsw (embedding vector_l1_ops);
 ```
 
-Hamming distance
+Hamming distance - added in 0.7.0
 
 ```sql
 CREATE INDEX ON items USING hnsw (embedding bit_hamming_ops);
 ```
 
-Jaccard distance
+Jaccard distance - added in 0.7.0
 
 ```sql
 CREATE INDEX ON items USING hnsw (embedding bit_jaccard_ops);
@@ -256,9 +258,9 @@ 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
-- `bit` - up to 64,000 dimensions
-- `sparsevec` - up to 1,000 non-zero elements
+- `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)
 
 ### Index Options
 
@@ -312,7 +314,7 @@ 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
 
-You can also speed up index creation by increasing the number of parallel workers (2 by default)
+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
@@ -322,7 +324,7 @@ 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)
+Check [indexing progress](https://www.postgresql.org/docs/current/progress-reporting.html#CREATE-INDEX-PROGRESS-REPORTING) with Postgres 12+
 
 ```sql
 SELECT phase, round(100.0 * blocks_done / nullif(blocks_total, 0), 1) AS "%" FROM pg_stat_progress_create_index;
@@ -365,7 +367,7 @@ Cosine distance
 CREATE INDEX ON items USING ivfflat (embedding vector_cosine_ops) WITH (lists = 100);
 ```
 
-Hamming distance
+Hamming distance - added in 0.7.0
 
 ```sql
 CREATE INDEX ON items USING ivfflat (embedding bit_hamming_ops) WITH (lists = 100);
@@ -374,8 +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
-- `bit` - up to 64,000 dimensions
+- `halfvec` - up to 4,000 dimensions (added in 0.7.0)
+- `bit` - up to 64,000 dimensions (added in 0.7.0)
 
 ### Query Options
 
@@ -408,7 +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)
+Check [indexing progress](https://www.postgresql.org/docs/current/progress-reporting.html#CREATE-INDEX-PROGRESS-REPORTING) with Postgres 12+
 
 ```sql
 SELECT phase, round(100.0 * tuples_done / nullif(tuples_total, 0), 1) AS "%" FROM pg_stat_progress_create_index;
@@ -425,49 +427,25 @@ Note: `%` is only populated during the `loading tuples` phase
 
 ## Filtering
 
-There are a few ways to index nearest neighbor queries with a `WHERE` clause.
+There are a few ways to index nearest neighbor queries with a `WHERE` clause
 
 ```sql
 SELECT * FROM items WHERE category_id = 123 ORDER BY embedding <-> '[3,1,2]' LIMIT 5;
 ```
 
-A good place to start is creating an index on the filter column. This can provide fast, exact nearest neighbor search in many cases. Postgres has a number of [index types](https://www.postgresql.org/docs/current/indexes-types.html) for this: B-tree (default), hash, GiST, SP-GiST, GIN, and BRIN.
+Create an index on one [or more](https://www.postgresql.org/docs/current/indexes-multicolumn.html) of the `WHERE` columns for exact search
 
 ```sql
 CREATE INDEX ON items (category_id);
 ```
 
-For multiple columns, consider a [multicolumn index](https://www.postgresql.org/docs/current/indexes-multicolumn.html).
-
-```sql
-CREATE INDEX ON items (location_id, category_id);
-```
-
-Exact indexes work well for conditions that match a low percentage of rows. Otherwise, [approximate indexes](#indexing) can work better.
-
-```sql
-CREATE INDEX ON items USING hnsw (embedding vector_l2_ops);
-```
-
-With approximate indexes, filtering is applied *after* the index is scanned. If a condition matches 10% of rows, with HNSW and the default `hnsw.ef_search` of 40, only 4 rows will match on average. For more rows, increase `hnsw.ef_search`.
-
-```sql
-SET hnsw.ef_search = 200;
-```
-
-Starting with 0.8.0, you can enable [iterative index scans](#iterative-index-scans), which will automatically scan more of the index when needed.
-
-```sql
-SET hnsw.iterative_scan = strict_order;
-```
-
-If filtering by only a few distinct values, consider [partial indexing](https://www.postgresql.org/docs/current/indexes-partial.html).
+Or a [partial index](https://www.postgresql.org/docs/current/indexes-partial.html) on the vector column for approximate search
 
 ```sql
 CREATE INDEX ON items USING hnsw (embedding vector_l2_ops) WHERE (category_id = 123);
 ```
 
-If filtering by many different values, consider [partitioning](https://www.postgresql.org/docs/current/ddl-partitioning.html).
+Use [partitioning](https://www.postgresql.org/docs/current/ddl-partitioning.html) for approximate search on many different values of the `WHERE` columns
 
 ```sql
 CREATE TABLE items (embedding vector(3), category_id int) PARTITION BY LIST(category_id);
@@ -475,11 +453,11 @@ CREATE TABLE items (embedding vector(3), category_id int) PARTITION BY LIST(cate
 
 ## Iterative Index Scans
 
-*Added in 0.8.0*
+*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`).
+With approximate indexes, queries with filtering can return less results (due to post-filtering). Starting with 0.8.0, you can enable iterative index scans. If too few results from the initial scan match the filters, the scan will resume until enough results are found (or it reaches `hnsw.max_scan_tuples` or `ivfflat.max_probes`). This can significantly improve recall.
 
-Iterative scans can use strict or relaxed ordering.
+There are two modes for iterative scans: strict and relaxed.
 
 Strict ensures results are in the exact order by distance
 
@@ -515,7 +493,7 @@ Note: Place any other filters inside the CTE
 
 ### Iterative Scan Options
 
-Since scanning a large portion of an approximate index is expensive, there are options to control when a scan ends.
+Since scanning a large portion of an approximate index is expensive, there are options to control when a scan ends
 
 #### HNSW
 
@@ -533,7 +511,18 @@ Specify the max amount of memory to use, as a multiple of `work_mem` (1 by defau
 SET hnsw.scan_mem_multiplier = 2;
 ```
 
-Note: Try increasing this if increasing `hnsw.max_scan_tuples` does not improve recall
+You can see when increasing this is needed by enabling debug messages
+
+```sql
+SET client_min_messages = debug1;
+```
+
+which will show when a scan reaches the memory limit
+
+```text
+DEBUG:  hnsw index scan reached memory limit after 20000 tuples
+HINT:  Increase hnsw.scan_mem_multiplier to scan more tuples.
+```
 
 #### IVFFlat
 
@@ -547,6 +536,8 @@ 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
@@ -555,6 +546,8 @@ 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
@@ -576,16 +569,24 @@ CREATE TABLE items (id bigserial PRIMARY KEY, embedding bit(3));
 INSERT INTO items (embedding) VALUES ('000'), ('111');
 ```
 
-Get the nearest neighbors by Hamming distance
+Get the nearest neighbors by Hamming distance (added in 0.7.0)
 
 ```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
@@ -608,6 +609,8 @@ SELECT * FROM (
 
 ## Sparse Vectors
 
+*Added in 0.7.0*
+
 Use the `sparsevec` type to store sparse vectors
 
 ```sql
@@ -641,6 +644,8 @@ 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
@@ -783,12 +788,8 @@ 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)
@@ -802,7 +803,6 @@ 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)
@@ -924,7 +924,7 @@ 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`), 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.
+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.
 
 Also, note that `NULL` vectors are not indexed (as well as zero vectors for cosine distance).
 
@@ -936,7 +936,7 @@ 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`). Enabling [iterative index scans](#iterative-index-scans) can address this.
+Results can also be limited by the number of probes (`ivfflat.probes`).
 
 Also, note that `NULL` vectors are not indexed (as well as zero vectors for cosine distance).
 
@@ -1103,13 +1103,7 @@ Note: Replace `17` with your Postgres server version
 
 ### Missing SDK
 
-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.
+If compilation fails and the output includes `warning: no such sysroot directory` on Mac, reinstall Xcode Command Line Tools.
 
 ### Portability
 
@@ -1127,14 +1121,6 @@ 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.
-
-### 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.
@@ -1154,17 +1140,11 @@ This adds pgvector to the [Postgres image](https://hub.docker.com/_/postgres) (r
 You can also build the image manually:
 
 ```sh
-git clone --branch v0.8.0 https://github.com/pgvector/pgvector.git
+git clone --branch v0.7.4 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:
@@ -1210,7 +1190,7 @@ Note: Replace `17` with your Postgres server version
 Install the FreeBSD package with:
 
 ```sh
-pkg install postgresql16-pgvector
+pkg install postgresql15-pgvector
 ```
 
 or the port with:
@@ -1252,6 +1232,36 @@ 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/bitutils.h

@@ -8,9 +8,6 @@
 #error "Requires PostgreSQL 13+"
 #endif
 
-/* Check architecture in first header */
-StaticAssertDecl(sizeof(Datum) == SIZEOF_DATUM, "Architecture mismatch");
-
 extern uint64 (*BitHammingDistance) (uint32 bytes, unsigned char *ax, unsigned char *bx, uint64 distance);
 extern double (*BitJaccardDistance) (uint32 bytes, unsigned char *ax, unsigned char *bx, uint64 ab, uint64 aa, uint64 bb);
 

src/hnsw.c

@@ -77,21 +77,21 @@ 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);
+							HNSW_DEFAULT_EF_SEARCH, HNSW_MIN_EF_SEARCH, HNSW_MAX_EF_SEARCH, PGC_USERSET, GUC_EXPLAIN, NULL, NULL, NULL);
 
 	DefineCustomEnumVariable("hnsw.iterative_scan", "Sets the mode for iterative scans",
 							 NULL, &hnsw_iterative_scan,
-							 HNSW_ITERATIVE_SCAN_OFF, hnsw_iterative_scan_options, PGC_USERSET, 0, NULL, NULL, NULL);
+							 HNSW_ITERATIVE_SCAN_OFF, hnsw_iterative_scan_options, PGC_USERSET, GUC_EXPLAIN, NULL, NULL, NULL);
 
 	/* This is approximate and does not affect the initial scan */
 	DefineCustomIntVariable("hnsw.max_scan_tuples", "Sets the max number of tuples to visit for iterative scans",
 							NULL, &hnsw_max_scan_tuples,
-							20000, 1, INT_MAX, PGC_USERSET, 0, NULL, NULL, NULL);
+							20000, 1, INT_MAX, PGC_USERSET, GUC_EXPLAIN, NULL, NULL, NULL);
 
 	/* Same range as hash_mem_multiplier */
 	DefineCustomRealVariable("hnsw.scan_mem_multiplier", "Sets the multiple of work_mem to use for iterative scans",
 							 NULL, &hnsw_scan_mem_multiplier,
-							 1, 1, 1000, PGC_USERSET, 0, NULL, NULL, NULL);
+							 1, 1, 1000, PGC_USERSET, GUC_EXPLAIN, NULL, NULL, NULL);
 
 	MarkGUCPrefixReserved("hnsw");
 }

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/hnswscan.c

@@ -240,8 +240,8 @@ hnswgettuple(IndexScanDesc scan, ScanDirection dir)
 			if (so->discarded == NULL)
 				break;
 
-			/* Reached max number of tuples or memory limit */
-			if (so->tuples >= hnsw_max_scan_tuples || MemoryContextMemAllocated(so->tmpCtx, false) > so->maxMemory)
+			/* Reached max number of tuples */
+			if (so->tuples >= hnsw_max_scan_tuples)
 			{
 				if (pairingheap_is_empty(so->discarded))
 					break;
@@ -249,6 +249,21 @@ hnswgettuple(IndexScanDesc scan, ScanDirection dir)
 				/* Return remaining tuples */
 				so->w = lappend(so->w, HnswGetSearchCandidate(w_node, pairingheap_remove_first(so->discarded)));
 			}
+			/* Prevent scans from consuming too much memory */
+			else if (MemoryContextMemAllocated(so->tmpCtx, false) > so->maxMemory)
+			{
+				if (pairingheap_is_empty(so->discarded))
+				{
+					ereport(DEBUG1,
+							(errmsg("hnsw index scan reached memory limit after " INT64_FORMAT " tuples", so->tuples),
+							 errhint("Increase hnsw.scan_mem_multiplier to scan more tuples.")));
+
+					break;
+				}
+
+				/* Return remaining tuples */
+				so->w = lappend(so->w, HnswGetSearchCandidate(w_node, pairingheap_remove_first(so->discarded)));
+			}
 			else
 			{
 				/*

src/hnswutils.c

@@ -1393,7 +1393,7 @@ hnsw_halfvec_support(PG_FUNCTION_ARGS)
 	};
 
 	PG_RETURN_POINTER(&typeInfo);
-}
+};
 
 FUNCTION_PREFIX PG_FUNCTION_INFO_V1(hnsw_bit_support);
 Datum
@@ -1406,7 +1406,7 @@ hnsw_bit_support(PG_FUNCTION_ARGS)
 	};
 
 	PG_RETURN_POINTER(&typeInfo);
-}
+};
 
 FUNCTION_PREFIX PG_FUNCTION_INFO_V1(hnsw_sparsevec_support);
 Datum
@@ -1419,4 +1419,4 @@ hnsw_sparsevec_support(PG_FUNCTION_ARGS)
 	};
 
 	PG_RETURN_POINTER(&typeInfo);
-}
+};

src/hnswvacuum.c

@@ -9,10 +9,6 @@
 #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", TupleDescAttr(buildstate->tupdesc, 0)->atttypid, -1, 0);
+	TupleDescInitEntry(buildstate->sortdesc, (AttrNumber) 3, "vector", buildstate->tupdesc->attrs[0].atttypid, -1, 0);
 
 	buildstate->slot = MakeSingleTupleTableSlot(buildstate->sortdesc, &TTSOpsVirtual);
 

src/ivfflat.c

@@ -39,16 +39,16 @@ 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);
+							IVFFLAT_DEFAULT_PROBES, IVFFLAT_MIN_LISTS, IVFFLAT_MAX_LISTS, PGC_USERSET, GUC_EXPLAIN, NULL, NULL, NULL);
 
 	DefineCustomEnumVariable("ivfflat.iterative_scan", "Sets the mode for iterative scans",
 							 NULL, &ivfflat_iterative_scan,
-							 IVFFLAT_ITERATIVE_SCAN_OFF, ivfflat_iterative_scan_options, PGC_USERSET, 0, NULL, NULL, NULL);
+							 IVFFLAT_ITERATIVE_SCAN_OFF, ivfflat_iterative_scan_options, PGC_USERSET, GUC_EXPLAIN, NULL, NULL, NULL);
 
 	/* If this is less than probes, probes is used */
 	DefineCustomIntVariable("ivfflat.max_probes", "Sets the max number of probes for iterative scans",
 							NULL, &ivfflat_max_probes,
-							IVFFLAT_MAX_LISTS, IVFFLAT_MIN_LISTS, IVFFLAT_MAX_LISTS, PGC_USERSET, 0, NULL, NULL, NULL);
+							IVFFLAT_MAX_LISTS, IVFFLAT_MIN_LISTS, IVFFLAT_MAX_LISTS, PGC_USERSET, GUC_EXPLAIN, NULL, NULL, NULL);
 
 	MarkGUCPrefixReserved("ivfflat");
 }

src/ivfscan.c

@@ -114,6 +114,7 @@ GetScanItems(IndexScanDesc scan, Datum value)
 {
 	IvfflatScanOpaque so = (IvfflatScanOpaque) scan->opaque;
 	TupleDesc	tupdesc = RelationGetDescr(scan->indexRelation);
+	double		tuples = 0;
 	TupleTableSlot *slot = so->vslot;
 	int			batchProbes = 0;
 
@@ -160,6 +161,8 @@ GetScanItems(IndexScanDesc scan, Datum value)
 				ExecStoreVirtualTuple(slot);
 
 				tuplesort_puttupleslot(so->sortstate, slot);
+
+				tuples++;
 			}
 
 			searchPage = IvfflatPageGetOpaque(page)->nextblkno;
@@ -168,6 +171,12 @@ GetScanItems(IndexScanDesc scan, Datum value)
 		}
 	}
 
+	if (tuples < 100 && ivfflat_iterative_scan == IVFFLAT_ITERATIVE_SCAN_OFF)
+		ereport(DEBUG1,
+				(errmsg("index scan found few tuples"),
+				 errdetail("Index may have been created with little data."),
+				 errhint("Recreate the index and possibly decrease lists.")));
+
 	tuplesort_performsort(so->sortstate);
 
 #if defined(IVFFLAT_MEMORY)

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

@@ -5,10 +5,6 @@
 #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

@@ -4,7 +4,6 @@
 #include <math.h>
 
 #include "catalog/pg_type.h"
-#include "common/shortest_dec.h"
 #include "common/string.h"
 #include "fmgr.h"
 #include "halfutils.h"
@@ -13,10 +12,17 @@
 #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

@@ -123,12 +123,6 @@ 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;
@@ -196,6 +190,4 @@ SHOW hnsw.scan_mem_multiplier;
 
 SET hnsw.scan_mem_multiplier = 0;
 ERROR:  0 is outside the valid range for parameter "hnsw.scan_mem_multiplier" (1 .. 1000)
-SET hnsw.scan_mem_multiplier = 1001;
-ERROR:  1001 is outside the valid range for parameter "hnsw.scan_mem_multiplier" (1 .. 1000)
 DROP TABLE t;

test/expected/ivfflat_vector.out

@@ -110,15 +110,6 @@ 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

@@ -70,9 +70,6 @@ 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;
@@ -112,6 +109,5 @@ SET hnsw.max_scan_tuples = 0;
 SHOW hnsw.scan_mem_multiplier;
 
 SET hnsw.scan_mem_multiplier = 0;
-SET hnsw.scan_mem_multiplier = 1001;
 
 DROP TABLE t;

test/sql/ivfflat_vector.sql

@@ -59,9 +59,6 @@ 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;

test/t/043_hnsw_iterative_scan.pl

@@ -56,4 +56,13 @@ foreach ((30000, 50000, 70000))
 	cmp_ok($avg, '<', $expected + 2);
 }
 
+my ($ret, $stdout, $stderr) = $node->psql("postgres", qq(
+	SET enable_seqscan = off;
+	SET hnsw.iterative_scan = relaxed_order;
+	SET client_min_messages = debug1;
+	SET work_mem = '1MB';
+	SELECT COUNT(*) FROM (SELECT v FROM tst WHERE i % 10000 = 0 ORDER BY v <-> (SELECT v FROM tst LIMIT 1) LIMIT 11) t;
+));
+like($stderr, qr/hnsw index scan reached memory limit after \d+ tuples/);
+
 done_testing();

vector.control

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