mirror of
https://github.com/pgvector/pgvector.git
synced 2026-07-22 03:57:34 +08:00
Compare commits
1 Commits
amgettuple
...
hnsw-read-
| Author | SHA1 | Date | |
|---|---|---|---|
|
67c72ac8c3 |
10
.github/workflows/build.yml
vendored
10
.github/workflows/build.yml
vendored
@@ -23,7 +23,7 @@ jobs:
|
||||
- postgres: 13
|
||||
os: ubuntu-22.04
|
||||
steps:
|
||||
- uses: actions/checkout@v6
|
||||
- uses: actions/checkout@v5
|
||||
- uses: ankane/setup-postgres@v1
|
||||
with:
|
||||
postgres-version: ${{ matrix.postgres }}
|
||||
@@ -53,7 +53,7 @@ jobs:
|
||||
- postgres: 14
|
||||
os: macos-15-intel
|
||||
steps:
|
||||
- uses: actions/checkout@v6
|
||||
- uses: actions/checkout@v5
|
||||
- uses: ankane/setup-postgres@v1
|
||||
with:
|
||||
postgres-version: ${{ matrix.postgres }}
|
||||
@@ -72,7 +72,7 @@ jobs:
|
||||
tar xf $TAG.tar.gz
|
||||
mv postgres-$TAG postgres
|
||||
env:
|
||||
TAG: ${{ matrix.postgres == 18 && 'REL_18_2' || 'REL_14_21' }}
|
||||
TAG: ${{ matrix.postgres == 18 && 'REL_18_0' || 'REL_14_19' }}
|
||||
- run: make prove_installcheck PROVE_FLAGS="-I ./postgres/src/test/perl -I ./test/perl"
|
||||
env:
|
||||
PERL5LIB: /Users/runner/perl5/lib/perl5
|
||||
@@ -92,7 +92,7 @@ jobs:
|
||||
- postgres: 14
|
||||
os: windows-2022
|
||||
steps:
|
||||
- uses: actions/checkout@v6
|
||||
- uses: actions/checkout@v5
|
||||
- uses: ankane/setup-postgres@v1
|
||||
with:
|
||||
postgres-version: ${{ matrix.postgres }}
|
||||
@@ -133,7 +133,7 @@ jobs:
|
||||
if: ${{ !startsWith(github.ref_name, 'mac') && !startsWith(github.ref_name, 'windows') }}
|
||||
runs-on: ubuntu-latest
|
||||
steps:
|
||||
- uses: actions/checkout@v6
|
||||
- uses: actions/checkout@v5
|
||||
- uses: ankane/setup-postgres-valgrind@v1
|
||||
with:
|
||||
postgres-version: 18
|
||||
|
||||
@@ -1,6 +1,5 @@
|
||||
## 0.8.2 (2026-02-25)
|
||||
## 0.8.2 (unreleased)
|
||||
|
||||
- Fixed buffer overflow with parallel HNSW index build - [more info](https://github.com/pgvector/pgvector/issues/959)
|
||||
- Improved `install` target on Windows
|
||||
- Fixed `Index Searches` in `EXPLAIN` output for Postgres 18
|
||||
|
||||
|
||||
@@ -5,7 +5,7 @@ ARG DEBIAN_CODENAME=bookworm
|
||||
FROM postgres:$PG_MAJOR-$DEBIAN_CODENAME
|
||||
ARG PG_MAJOR
|
||||
|
||||
ADD https://github.com/pgvector/pgvector.git#v0.8.2 /tmp/pgvector
|
||||
ADD https://github.com/pgvector/pgvector.git#v0.8.1 /tmp/pgvector
|
||||
|
||||
RUN apt-get update && \
|
||||
apt-mark hold locales && \
|
||||
|
||||
2
LICENSE
2
LICENSE
@@ -1,4 +1,4 @@
|
||||
Portions Copyright (c) 1996-2026, PostgreSQL Global Development Group
|
||||
Portions Copyright (c) 1996-2025, PostgreSQL Global Development Group
|
||||
|
||||
Portions Copyright (c) 1994, The Regents of the University of California
|
||||
|
||||
|
||||
@@ -2,7 +2,7 @@
|
||||
"name": "vector",
|
||||
"abstract": "Open-source vector similarity search for Postgres",
|
||||
"description": "Supports L2 distance, inner product, and cosine distance",
|
||||
"version": "0.8.2",
|
||||
"version": "0.8.1",
|
||||
"maintainer": [
|
||||
"Andrew Kane <andrew@ankane.org>"
|
||||
],
|
||||
@@ -20,7 +20,7 @@
|
||||
"vector": {
|
||||
"file": "sql/vector.sql",
|
||||
"docfile": "README.md",
|
||||
"version": "0.8.2",
|
||||
"version": "0.8.1",
|
||||
"abstract": "Open-source vector similarity search for Postgres"
|
||||
}
|
||||
},
|
||||
|
||||
7
Makefile
7
Makefile
@@ -1,5 +1,5 @@
|
||||
EXTENSION = vector
|
||||
EXTVERSION = 0.8.2
|
||||
EXTVERSION = 0.8.1
|
||||
|
||||
MODULE_big = vector
|
||||
DATA = $(wildcard sql/*--*--*.sql)
|
||||
@@ -27,11 +27,6 @@ ifneq ($(filter ppc64%, $(shell uname -m)), )
|
||||
OPTFLAGS =
|
||||
endif
|
||||
|
||||
# RISC-V64 doesn't support -march=native
|
||||
ifeq ($(shell uname -m), riscv64)
|
||||
OPTFLAGS =
|
||||
endif
|
||||
|
||||
# For auto-vectorization:
|
||||
# - GCC (needs -ftree-vectorize OR -O3) - https://gcc.gnu.org/projects/tree-ssa/vectorization.html
|
||||
# - Clang (could use pragma instead) - https://llvm.org/docs/Vectorizers.html
|
||||
|
||||
@@ -1,5 +1,5 @@
|
||||
EXTENSION = vector
|
||||
EXTVERSION = 0.8.2
|
||||
EXTVERSION = 0.8.1
|
||||
|
||||
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
|
||||
|
||||
100
README.md
100
README.md
@@ -11,8 +11,6 @@ Store your vectors with the rest of your data. Supports:
|
||||
|
||||
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
|
||||
|
||||
Have a lot of vectors? Use [quantization](#scaling) to scale
|
||||
|
||||
[](https://github.com/pgvector/pgvector/actions)
|
||||
|
||||
## Installation
|
||||
@@ -23,7 +21,7 @@ Compile and install the extension (supports Postgres 13+)
|
||||
|
||||
```sh
|
||||
cd /tmp
|
||||
git clone --branch v0.8.2 https://github.com/pgvector/pgvector.git
|
||||
git clone --branch v0.8.1 https://github.com/pgvector/pgvector.git
|
||||
cd pgvector
|
||||
make
|
||||
make install # may need sudo
|
||||
@@ -31,7 +29,7 @@ make install # may need sudo
|
||||
|
||||
See the [installation notes](#installation-notes---linux-and-mac) if you run into issues
|
||||
|
||||
You can also install it with [Docker](#docker), [Homebrew](#homebrew), [PGXN](#pgxn), [APT](#apt), [Yum](#yum), [pkg](#pkg), [APK](#apk), 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), [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).
|
||||
|
||||
### Windows
|
||||
|
||||
@@ -40,7 +38,7 @@ Ensure [C++ support in Visual Studio](https://learn.microsoft.com/en-us/cpp/buil
|
||||
```cmd
|
||||
set "PGROOT=C:\Program Files\PostgreSQL\18"
|
||||
cd %TEMP%
|
||||
git clone --branch v0.8.2 https://github.com/pgvector/pgvector.git
|
||||
git clone --branch v0.8.1 https://github.com/pgvector/pgvector.git
|
||||
cd pgvector
|
||||
nmake /F Makefile.win
|
||||
nmake /F Makefile.win install
|
||||
@@ -316,8 +314,6 @@ For a large number of workers, you may need to increase `max_parallel_workers` (
|
||||
|
||||
The [index options](#index-options) also have a significant impact on build time (use the defaults unless seeing low recall)
|
||||
|
||||
Use [binary quantization](#binary-quantization) for faster build times at scale
|
||||
|
||||
### Indexing Progress
|
||||
|
||||
Check [indexing progress](https://www.postgresql.org/docs/current/progress-reporting.html#CREATE-INDEX-PROGRESS-REPORTING)
|
||||
@@ -447,7 +443,13 @@ Exact indexes work well for conditions that match a low percentage of rows. Othe
|
||||
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, enable [iterative index scans](#iterative-index-scans), which will automatically scan more of the index when needed.
|
||||
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;
|
||||
@@ -671,10 +673,6 @@ SHOW shared_buffers;
|
||||
|
||||
Be sure to restart Postgres for changes to take effect.
|
||||
|
||||
### Storing
|
||||
|
||||
Use the `halfvec` type instead of `vector` for a smaller working set.
|
||||
|
||||
### Loading
|
||||
|
||||
Use `COPY` for bulk loading data ([example](https://github.com/pgvector/pgvector-python/blob/master/examples/loading/example.py)).
|
||||
@@ -689,8 +687,6 @@ Add any indexes *after* loading the initial data for best performance.
|
||||
|
||||
See index build time for [HNSW](#index-build-time) and [IVFFlat](#index-build-time-1).
|
||||
|
||||
Use [binary quantization](#binary-quantization) for smaller indexes and faster build times at scale.
|
||||
|
||||
In production environments, create indexes concurrently to avoid blocking writes.
|
||||
|
||||
```sql
|
||||
@@ -721,8 +717,6 @@ SELECT * FROM items ORDER BY embedding <#> '[3,1,2]' LIMIT 5;
|
||||
|
||||
#### Approximate Search
|
||||
|
||||
Use [binary quantization](#binary-quantization) with re-ranking to keep indexes in-memory at scale.
|
||||
|
||||
To speed up queries with an IVFFlat index, increase the number of inverted lists (at the expense of recall).
|
||||
|
||||
```sql
|
||||
@@ -738,20 +732,21 @@ REINDEX INDEX CONCURRENTLY index_name;
|
||||
VACUUM table_name;
|
||||
```
|
||||
|
||||
## Scaling
|
||||
|
||||
For a smaller working set:
|
||||
|
||||
1. Use the `halfvec` type instead of `vector` for tables
|
||||
2. Use [binary quantization](#binary-quantization) for indexes (with re-ranking for search)
|
||||
|
||||
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), [PgDog](https://github.com/pgdogdev/pgdog), or another approach for sharding ([example](https://github.com/pgvector/pgvector-python/blob/master/examples/citus/example.py)).
|
||||
|
||||
## Monitoring
|
||||
|
||||
Use existing tools like [pg_stat_statements](https://www.postgresql.org/docs/current/pgstatstatements.html) or [PgHero](https://github.com/ankane/pghero) to monitor performance.
|
||||
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;
|
||||
```
|
||||
|
||||
Monitor recall by comparing results from approximate search with exact search.
|
||||
|
||||
@@ -762,6 +757,14 @@ 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/example.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.
|
||||
@@ -769,11 +772,9 @@ Use pgvector from any language with a Postgres client. You can even generate and
|
||||
Language | Libraries / Examples
|
||||
--- | ---
|
||||
Ada | [pgvector-ada](https://github.com/pgvector/pgvector-ada)
|
||||
Algol | [pgvector-algol](https://github.com/pgvector/pgvector-algol)
|
||||
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)
|
||||
COBOL | [pgvector-cobol](https://github.com/pgvector/pgvector-cobol)
|
||||
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)
|
||||
@@ -793,15 +794,12 @@ OCaml | [pgvector-ocaml](https://github.com/pgvector/pgvector-ocaml)
|
||||
Pascal | [pgvector-pascal](https://github.com/pgvector/pgvector-pascal)
|
||||
Perl | [pgvector-perl](https://github.com/pgvector/pgvector-perl)
|
||||
PHP | [pgvector-php](https://github.com/pgvector/pgvector-php)
|
||||
Prolog | [pgvector-prolog](https://github.com/pgvector/pgvector-prolog)
|
||||
Python | [pgvector-python](https://github.com/pgvector/pgvector-python)
|
||||
R | [pgvector-r](https://github.com/pgvector/pgvector-r)
|
||||
Racket | [pgvector-racket](https://github.com/pgvector/pgvector-racket)
|
||||
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)
|
||||
Tcl | [pgvector-tcl](https://github.com/pgvector/pgvector-tcl)
|
||||
Zig | [pgvector-zig](https://github.com/pgvector/pgvector-zig)
|
||||
|
||||
## Frequently Asked Questions
|
||||
@@ -875,8 +873,6 @@ No, but like other index types, you’ll likely see better performance if they d
|
||||
SELECT pg_size_pretty(pg_relation_size('index_name'));
|
||||
```
|
||||
|
||||
Use [half-precision indexing](#half-precision-indexing) or [binary quantization](#binary-quantization) for smaller indexes.
|
||||
|
||||
## Troubleshooting
|
||||
|
||||
#### Why isn’t a query using an index?
|
||||
@@ -1151,23 +1147,23 @@ This adds pgvector to the [Postgres image](https://hub.docker.com/_/postgres) (r
|
||||
|
||||
Supported tags are:
|
||||
|
||||
- `pg18-trixie`, `0.8.2-pg18-trixie`
|
||||
- `pg18-bookworm`, `0.8.2-pg18-bookworm`, `pg18`, `0.8.2-pg18`
|
||||
- `pg17-trixie`, `0.8.2-pg17-trixie`
|
||||
- `pg17-bookworm`, `0.8.2-pg17-bookworm`, `pg17`, `0.8.2-pg17`
|
||||
- `pg16-trixie`, `0.8.2-pg16-trixie`
|
||||
- `pg16-bookworm`, `0.8.2-pg16-bookworm`, `pg16`, `0.8.2-pg16`
|
||||
- `pg15-trixie`, `0.8.2-pg15-trixie`
|
||||
- `pg15-bookworm`, `0.8.2-pg15-bookworm`, `pg15`, `0.8.2-pg15`
|
||||
- `pg14-trixie`, `0.8.2-pg14-trixie`
|
||||
- `pg14-bookworm`, `0.8.2-pg14-bookworm`, `pg14`, `0.8.2-pg14`
|
||||
- `pg13-trixie`, `0.8.2-pg13-trixie`
|
||||
- `pg13-bookworm`, `0.8.2-pg13-bookworm`, `pg13`, `0.8.2-pg13`
|
||||
- `pg18-trixie`, `0.8.1-pg18-trixie`
|
||||
- `pg18-bookworm`, `0.8.1-pg18-bookworm`, `pg18`, `0.8.1-pg18`
|
||||
- `pg17-trixie`, `0.8.1-pg17-trixie`
|
||||
- `pg17-bookworm`, `0.8.1-pg17-bookworm`, `pg17`, `0.8.1-pg17`
|
||||
- `pg16-trixie`, `0.8.1-pg16-trixie`
|
||||
- `pg16-bookworm`, `0.8.1-pg16-bookworm`, `pg16`, `0.8.1-pg16`
|
||||
- `pg15-trixie`, `0.8.1-pg15-trixie`
|
||||
- `pg15-bookworm`, `0.8.1-pg15-bookworm`, `pg15`, `0.8.1-pg15`
|
||||
- `pg14-trixie`, `0.8.1-pg14-trixie`
|
||||
- `pg14-bookworm`, `0.8.1-pg14-bookworm`, `pg14`, `0.8.1-pg14`
|
||||
- `pg13-trixie`, `0.8.1-pg13-trixie`
|
||||
- `pg13-bookworm`, `0.8.1-pg13-bookworm`, `pg13`, `0.8.1-pg13`
|
||||
|
||||
You can also build the image manually:
|
||||
|
||||
```sh
|
||||
git clone --branch v0.8.2 https://github.com/pgvector/pgvector.git
|
||||
git clone --branch v0.8.1 https://github.com/pgvector/pgvector.git
|
||||
cd pgvector
|
||||
docker build --pull --build-arg PG_MAJOR=18 -t myuser/pgvector .
|
||||
```
|
||||
@@ -1233,14 +1229,6 @@ cd /usr/ports/databases/pgvector
|
||||
make install
|
||||
```
|
||||
|
||||
### APK
|
||||
|
||||
Install the Alpine package with:
|
||||
|
||||
```sh
|
||||
apk add postgresql-pgvector
|
||||
```
|
||||
|
||||
### conda-forge
|
||||
|
||||
With Conda Postgres, install from [conda-forge](https://anaconda.org/conda-forge/pgvector) with:
|
||||
|
||||
@@ -1,2 +0,0 @@
|
||||
-- complain if script is sourced in psql, rather than via CREATE EXTENSION
|
||||
\echo Use "ALTER EXTENSION vector UPDATE TO '0.8.2'" to load this file. \quit
|
||||
@@ -169,7 +169,7 @@ BitJaccardDistanceAvx512Popcount(uint32 bytes, unsigned char *ax, unsigned char
|
||||
#endif
|
||||
|
||||
TARGET_XSAVE static bool
|
||||
SupportsAvx512Popcount(void)
|
||||
SupportsAvx512Popcount()
|
||||
{
|
||||
unsigned int exx[4] = {0, 0, 0, 0};
|
||||
|
||||
|
||||
@@ -2,7 +2,6 @@
|
||||
|
||||
#include "bitutils.h"
|
||||
#include "bitvec.h"
|
||||
#include "fmgr.h"
|
||||
#include "utils/varbit.h"
|
||||
#include "vector.h"
|
||||
|
||||
|
||||
@@ -1,7 +1,5 @@
|
||||
#include "postgres.h"
|
||||
|
||||
#include <math.h>
|
||||
|
||||
#include "halfutils.h"
|
||||
#include "halfvec.h"
|
||||
|
||||
|
||||
@@ -13,20 +13,12 @@
|
||||
#include "port.h" /* for strtof() */
|
||||
#include "sparsevec.h"
|
||||
#include "utils/array.h"
|
||||
#include "utils/builtins.h"
|
||||
#include "utils/float.h"
|
||||
#include "utils/fmgrprotos.h"
|
||||
#include "utils/lsyscache.h"
|
||||
#include "utils/varbit.h"
|
||||
#include "utils/numeric.h"
|
||||
#include "vector.h"
|
||||
|
||||
#if PG_VERSION_NUM >= 160000
|
||||
#include "varatt.h"
|
||||
#endif
|
||||
|
||||
#if PG_VERSION_NUM >= 170000
|
||||
#include "parser/scansup.h"
|
||||
#endif
|
||||
|
||||
#define STATE_DIMS(x) (ARR_DIMS(x)[0] - 1)
|
||||
#define CreateStateDatums(dim) palloc(sizeof(Datum) * (dim + 1))
|
||||
|
||||
@@ -137,9 +129,9 @@ InitHalfVector(int dim)
|
||||
return result;
|
||||
}
|
||||
|
||||
#if PG_VERSION_NUM >= 170000
|
||||
#define halfvec_isspace(ch) scanner_isspace(ch)
|
||||
#else
|
||||
/*
|
||||
* Check for whitespace, since array_isspace() is static
|
||||
*/
|
||||
static inline bool
|
||||
halfvec_isspace(char ch)
|
||||
{
|
||||
@@ -152,7 +144,6 @@ halfvec_isspace(char ch)
|
||||
return true;
|
||||
return false;
|
||||
}
|
||||
#endif
|
||||
|
||||
/*
|
||||
* Check state array
|
||||
|
||||
70
src/hnsw.c
70
src/hnsw.c
@@ -1,25 +1,18 @@
|
||||
#include "postgres.h"
|
||||
|
||||
#include <float.h>
|
||||
#include <limits.h>
|
||||
#include <math.h>
|
||||
|
||||
#include "access/amapi.h"
|
||||
#include "access/genam.h"
|
||||
#include "access/reloptions.h"
|
||||
#include "commands/progress.h"
|
||||
#include "commands/vacuum.h"
|
||||
#include "fmgr.h"
|
||||
#include "hnsw.h"
|
||||
#include "miscadmin.h"
|
||||
#include "nodes/pg_list.h"
|
||||
#include "storage/lwlock.h"
|
||||
#include "utils/float.h"
|
||||
#include "utils/guc.h"
|
||||
#include "utils/relcache.h"
|
||||
#include "utils/selfuncs.h"
|
||||
#include "utils/spccache.h"
|
||||
#include "vector.h"
|
||||
|
||||
#if PG_VERSION_NUM < 150000
|
||||
#define MarkGUCPrefixReserved(x) EmitWarningsOnPlaceholders(x)
|
||||
@@ -267,68 +260,6 @@ FUNCTION_PREFIX PG_FUNCTION_INFO_V1(hnswhandler);
|
||||
Datum
|
||||
hnswhandler(PG_FUNCTION_ARGS)
|
||||
{
|
||||
#if PG_VERSION_NUM >= 190000
|
||||
static const IndexAmRoutine amroutine = {
|
||||
.type = T_IndexAmRoutine,
|
||||
.amstrategies = 0,
|
||||
.amsupport = 3,
|
||||
.amoptsprocnum = 0,
|
||||
.amcanorder = false,
|
||||
.amcanorderbyop = true,
|
||||
.amcanhash = false,
|
||||
.amconsistentequality = false,
|
||||
.amconsistentordering = false,
|
||||
.amcanbackward = false,
|
||||
.amcanmarkpos = false,
|
||||
.amcanunique = false,
|
||||
.amcanmulticol = false,
|
||||
.amoptionalkey = true,
|
||||
.amsearcharray = false,
|
||||
.amsearchnulls = false,
|
||||
.amstorage = false,
|
||||
.amclusterable = false,
|
||||
.ampredlocks = false,
|
||||
.amcanparallel = false,
|
||||
.amcanbuildparallel = true,
|
||||
.amcaninclude = false,
|
||||
.amusemaintenanceworkmem = false,
|
||||
.amsummarizing = false,
|
||||
.amparallelvacuumoptions = VACUUM_OPTION_PARALLEL_BULKDEL,
|
||||
.amkeytype = InvalidOid,
|
||||
|
||||
.ambuild = hnswbuild,
|
||||
.ambuildempty = hnswbuildempty,
|
||||
.aminsert = hnswinsert,
|
||||
.aminsertcleanup = NULL,
|
||||
.ambulkdelete = hnswbulkdelete,
|
||||
.amvacuumcleanup = hnswvacuumcleanup,
|
||||
.amcanreturn = NULL,
|
||||
.amcostestimate = hnswcostestimate,
|
||||
.amgettreeheight = NULL,
|
||||
.amoptions = hnswoptions,
|
||||
.amproperty = NULL,
|
||||
.ambuildphasename = hnswbuildphasename,
|
||||
.amvalidate = hnswvalidate,
|
||||
.amadjustmembers = NULL,
|
||||
.ambeginscan = hnswbeginscan,
|
||||
.amrescan = hnswrescan,
|
||||
.amgettuple = hnswgettuple,
|
||||
.amgetbatch = NULL,
|
||||
.amunguardbatch = NULL,
|
||||
.amkillitemsbatch = NULL,
|
||||
.amgettransform = NULL,
|
||||
.amgetbitmap = NULL,
|
||||
.amendscan = hnswendscan,
|
||||
.amposreset = NULL,
|
||||
.amestimateparallelscan = NULL,
|
||||
.aminitparallelscan = NULL,
|
||||
.amparallelrescan = NULL,
|
||||
.amtranslatestrategy = NULL,
|
||||
.amtranslatecmptype = NULL,
|
||||
};
|
||||
|
||||
PG_RETURN_POINTER(&amroutine);
|
||||
#else
|
||||
IndexAmRoutine *amroutine = makeNode(IndexAmRoutine);
|
||||
|
||||
amroutine->amstrategies = 0;
|
||||
@@ -402,5 +333,4 @@ hnswhandler(PG_FUNCTION_ARGS)
|
||||
#endif
|
||||
|
||||
PG_RETURN_POINTER(amroutine);
|
||||
#endif
|
||||
}
|
||||
|
||||
23
src/hnsw.h
23
src/hnsw.h
@@ -3,25 +3,15 @@
|
||||
|
||||
#include "postgres.h"
|
||||
|
||||
#include <math.h>
|
||||
|
||||
#include "access/genam.h"
|
||||
#include "access/parallel.h"
|
||||
#include "lib/pairingheap.h"
|
||||
#include "nodes/execnodes.h"
|
||||
#include "port.h" /* for random() */
|
||||
#include "storage/bufpage.h"
|
||||
#include "storage/condition_variable.h"
|
||||
#include "storage/lwlock.h"
|
||||
#include "storage/s_lock.h"
|
||||
#include "utils/relptr.h"
|
||||
#include "utils/sampling.h"
|
||||
#include "vector.h"
|
||||
|
||||
#if PG_VERSION_NUM >= 190000
|
||||
typedef Pointer Item;
|
||||
#endif
|
||||
|
||||
#define HNSW_MAX_DIM 2000
|
||||
#define HNSW_MAX_NNZ 1000
|
||||
|
||||
@@ -372,6 +362,13 @@ typedef union
|
||||
ItemPointerData indextid;
|
||||
} HnswUnvisited;
|
||||
|
||||
typedef struct HnswReadStreamData
|
||||
{
|
||||
HnswUnvisited *unvisited;
|
||||
int unvisitedLength;
|
||||
int visited;
|
||||
} HnswReadStreamData;
|
||||
|
||||
typedef struct HnswScanOpaqueData
|
||||
{
|
||||
const HnswTypeInfo *typeInfo;
|
||||
@@ -427,14 +424,14 @@ bool HnswCheckNorm(HnswSupport * support, Datum value);
|
||||
Buffer HnswNewBuffer(Relation index, ForkNumber forkNum);
|
||||
void HnswInitPage(Buffer buf, Page page);
|
||||
void HnswInit(void);
|
||||
List *HnswSearchLayer(char *base, HnswQuery * q, List *ep, int ef, int lc, Relation index, HnswSupport * support, int m, bool inserting, HnswElement skipElement, visited_hash * v, pairingheap **discarded, bool initVisited, int64 *tuples);
|
||||
List *HnswSearchLayer(char *base, HnswQuery * q, List *ep, int ef, int lc, Relation index, HnswSupport * support, int m, bool inserting, HnswElement skipElement, visited_hash * v, pairingheap **discarded, bool initVisited, int64 *tuples, bool maintenance);
|
||||
HnswElement HnswGetEntryPoint(Relation index);
|
||||
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, HnswSupport * support, int m, int efConstruction, bool existing);
|
||||
HnswSearchCandidate *HnswEntryCandidate(char *base, HnswElement entryPoint, HnswQuery * q, Relation index, HnswSupport * support, bool loadVec);
|
||||
void HnswFindElementNeighbors(char *base, HnswElement element, HnswElement entryPoint, Relation index, HnswSupport * support, int m, int efConstruction, bool existing, bool maintenance);
|
||||
HnswSearchCandidate *HnswEntryCandidate(char *base, HnswElement em, HnswQuery * q, Relation rel, HnswSupport * support, 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);
|
||||
|
||||
@@ -36,14 +36,11 @@
|
||||
*/
|
||||
#include "postgres.h"
|
||||
|
||||
#include <limits.h>
|
||||
#include <math.h>
|
||||
|
||||
#include "access/genam.h"
|
||||
#include "access/parallel.h"
|
||||
#include "access/relscan.h"
|
||||
#include "access/table.h"
|
||||
#include "access/tableam.h"
|
||||
#include "access/tupdesc.h"
|
||||
#include "access/xact.h"
|
||||
#include "access/xloginsert.h"
|
||||
#include "catalog/index.h"
|
||||
@@ -51,15 +48,11 @@
|
||||
#include "commands/progress.h"
|
||||
#include "hnsw.h"
|
||||
#include "miscadmin.h"
|
||||
#include "nodes/execnodes.h"
|
||||
#include "optimizer/optimizer.h"
|
||||
#include "storage/bufmgr.h"
|
||||
#include "storage/condition_variable.h"
|
||||
#include "tcop/tcopprot.h"
|
||||
#include "utils/datum.h"
|
||||
#include "utils/memutils.h"
|
||||
#include "utils/rel.h"
|
||||
#include "utils/snapmgr.h"
|
||||
|
||||
#if PG_VERSION_NUM >= 160000
|
||||
#include "varatt.h"
|
||||
@@ -80,8 +73,6 @@
|
||||
#define PARALLEL_KEY_HNSW_AREA UINT64CONST(0xA000000000000002)
|
||||
#define PARALLEL_KEY_QUERY_TEXT UINT64CONST(0xA000000000000003)
|
||||
|
||||
#define HNSW_MAX_GRAPH_MEMORY (SIZE_MAX / 2)
|
||||
|
||||
/*
|
||||
* Create the metapage
|
||||
*/
|
||||
@@ -470,7 +461,7 @@ InsertTupleInMemory(HnswBuildState * buildstate, HnswElement element)
|
||||
}
|
||||
|
||||
/* Find neighbors for element */
|
||||
HnswFindElementNeighbors(base, element, entryPoint, NULL, support, m, efConstruction, false);
|
||||
HnswFindElementNeighbors(base, element, entryPoint, NULL, support, m, efConstruction, false, true);
|
||||
|
||||
/* Update graph in memory */
|
||||
UpdateGraphInMemory(support, element, m, entryPoint, buildstate);
|
||||
@@ -494,7 +485,6 @@ InsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heaptid, Hn
|
||||
LWLock *flushLock = &graph->flushLock;
|
||||
char *base = buildstate->hnswarea;
|
||||
Datum value;
|
||||
Size memoryMargin;
|
||||
|
||||
/* Form index value */
|
||||
if (!HnswFormIndexValue(&value, values, isnull, buildstate->typeInfo, support))
|
||||
@@ -503,9 +493,6 @@ InsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heaptid, Hn
|
||||
/* Get datum size */
|
||||
valueSize = VARSIZE_ANY(DatumGetPointer(value));
|
||||
|
||||
/* In a parallel build, add a margin so allocations never fail */
|
||||
memoryMargin = base == NULL ? 0 : 1024 * 1024;
|
||||
|
||||
/* Ensure graph not flushed when inserting */
|
||||
LWLockAcquire(flushLock, LW_SHARED);
|
||||
|
||||
@@ -527,7 +514,7 @@ InsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heaptid, Hn
|
||||
* 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 + memoryMargin >= graph->memoryTotal)
|
||||
if (graph->memoryUsed >= graph->memoryTotal)
|
||||
{
|
||||
LWLockRelease(&graph->allocatorLock);
|
||||
|
||||
@@ -562,7 +549,7 @@ InsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heaptid, Hn
|
||||
|
||||
/* Copy the datum */
|
||||
memcpy(valuePtr, DatumGetPointer(value), valueSize);
|
||||
HnswPtrStore(base, element->value, (char *) valuePtr);
|
||||
HnswPtrStore(base, element->value, valuePtr);
|
||||
|
||||
/* Create a lock for the element */
|
||||
LWLockInitialize(&element->lock, hnsw_lock_tranche_id);
|
||||
@@ -620,7 +607,7 @@ InitGraph(HnswGraph * graph, char *base, Size memoryTotal)
|
||||
HnswPtrStore(base, graph->head, (HnswElement) NULL);
|
||||
HnswPtrStore(base, graph->entryPoint, (HnswElement) NULL);
|
||||
graph->memoryUsed = 0;
|
||||
graph->memoryTotal = Min(memoryTotal, HNSW_MAX_GRAPH_MEMORY);
|
||||
graph->memoryTotal = memoryTotal;
|
||||
graph->flushed = false;
|
||||
graph->indtuples = 0;
|
||||
SpinLockInit(&graph->lock);
|
||||
@@ -661,17 +648,9 @@ static void *
|
||||
HnswSharedMemoryAlloc(Size size, void *state)
|
||||
{
|
||||
HnswBuildState *buildstate = (HnswBuildState *) state;
|
||||
Size alignedSize = MAXALIGN(size);
|
||||
void *chunk;
|
||||
void *chunk = buildstate->hnswarea + buildstate->graph->memoryUsed;
|
||||
|
||||
if (alignedSize > 1024 * 1024)
|
||||
elog(ERROR, "hnsw allocation too large");
|
||||
|
||||
if (buildstate->graph->memoryUsed + alignedSize > buildstate->graph->memoryTotal)
|
||||
elog(ERROR, "hnsw allocator out of memory");
|
||||
|
||||
chunk = buildstate->hnswarea + buildstate->graph->memoryUsed;
|
||||
buildstate->graph->memoryUsed += alignedSize;
|
||||
buildstate->graph->memoryUsed += MAXALIGN(size);
|
||||
return chunk;
|
||||
}
|
||||
|
||||
@@ -803,11 +782,7 @@ HnswParallelScanAndInsert(Relation heapRel, Relation indexRel, HnswShared * hnsw
|
||||
buildstate.hnswarea = hnswarea;
|
||||
InitAllocator(&buildstate.allocator, &HnswSharedMemoryAlloc, &buildstate);
|
||||
scan = table_beginscan_parallel(heapRel,
|
||||
ParallelTableScanFromHnswShared(hnswshared)
|
||||
#if PG_VERSION_NUM >= 190000
|
||||
,SO_NONE
|
||||
#endif
|
||||
);
|
||||
ParallelTableScanFromHnswShared(hnswshared));
|
||||
reltuples = table_index_build_scan(heapRel, indexRel, indexInfo,
|
||||
true, progress, BuildCallback,
|
||||
(void *) &buildstate, scan);
|
||||
@@ -961,8 +936,6 @@ HnswBeginParallel(HnswBuildState * buildstate, bool isconcurrent, int request)
|
||||
if (esthnswarea > estother)
|
||||
esthnswarea -= estother;
|
||||
|
||||
esthnswarea = Min(esthnswarea, HNSW_MAX_GRAPH_MEMORY);
|
||||
|
||||
shm_toc_estimate_chunk(&pcxt->estimator, esthnswarea);
|
||||
shm_toc_estimate_keys(&pcxt->estimator, 2);
|
||||
|
||||
@@ -1005,7 +978,8 @@ HnswBeginParallel(HnswBuildState * buildstate, bool isconcurrent, int request)
|
||||
snapshot);
|
||||
|
||||
hnswarea = (char *) shm_toc_allocate(pcxt->toc, esthnswarea);
|
||||
InitGraph(&hnswshared->graphData, hnswarea, esthnswarea);
|
||||
/* Report less than allocated so never fails */
|
||||
InitGraph(&hnswshared->graphData, hnswarea, esthnswarea - 1024 * 1024);
|
||||
|
||||
/*
|
||||
* Avoid base address for relptr for Postgres < 14.5
|
||||
|
||||
@@ -1,15 +1,13 @@
|
||||
#include "postgres.h"
|
||||
|
||||
#include "access/genam.h"
|
||||
#include <math.h>
|
||||
|
||||
#include "access/generic_xlog.h"
|
||||
#include "hnsw.h"
|
||||
#include "nodes/execnodes.h"
|
||||
#include "storage/bufmgr.h"
|
||||
#include "storage/lmgr.h"
|
||||
#include "storage/lwlock.h"
|
||||
#include "utils/datum.h"
|
||||
#include "utils/memutils.h"
|
||||
#include "utils/rel.h"
|
||||
|
||||
#if PG_VERSION_NUM >= 160000
|
||||
#include "varatt.h"
|
||||
@@ -714,7 +712,7 @@ HnswInsertTupleOnDisk(Relation index, HnswSupport * support, Datum value, ItemPo
|
||||
|
||||
/* Create an element */
|
||||
element = HnswInitElement(base, heaptid, m, HnswGetMl(m), HnswGetMaxLevel(m), NULL);
|
||||
HnswPtrStore(base, element->value, (char *) DatumGetPointer(value));
|
||||
HnswPtrStore(base, element->value, DatumGetPointer(value));
|
||||
|
||||
/* Prevent concurrent inserts when likely updating entry point */
|
||||
if (entryPoint == NULL || element->level > entryPoint->level)
|
||||
@@ -731,7 +729,7 @@ HnswInsertTupleOnDisk(Relation index, HnswSupport * support, Datum value, ItemPo
|
||||
}
|
||||
|
||||
/* Find neighbors for element */
|
||||
HnswFindElementNeighbors(base, element, entryPoint, index, support, m, efConstruction, false);
|
||||
HnswFindElementNeighbors(base, element, entryPoint, index, support, m, efConstruction, false, building);
|
||||
|
||||
/* Update graph on disk */
|
||||
UpdateGraphOnDisk(index, support, element, m, entryPoint, building);
|
||||
|
||||
@@ -1,23 +1,12 @@
|
||||
#include "postgres.h"
|
||||
|
||||
#include <limits.h>
|
||||
|
||||
#include "access/genam.h"
|
||||
#include "access/relscan.h"
|
||||
#include "hnsw.h"
|
||||
#include "lib/pairingheap.h"
|
||||
#include "miscadmin.h"
|
||||
#include "nodes/pg_list.h"
|
||||
#include "pgstat.h"
|
||||
#include "storage/bufmgr.h"
|
||||
#include "storage/lmgr.h"
|
||||
#include "utils/float.h"
|
||||
#include "utils/memutils.h"
|
||||
#include "utils/relcache.h"
|
||||
#include "utils/snapmgr.h"
|
||||
|
||||
#if PG_VERSION_NUM >= 160000
|
||||
#include "varatt.h"
|
||||
#endif
|
||||
|
||||
/*
|
||||
* Algorithm 5 from paper
|
||||
@@ -48,11 +37,11 @@ GetScanItems(IndexScanDesc scan, Datum value)
|
||||
|
||||
for (int lc = entryPoint->level; lc >= 1; lc--)
|
||||
{
|
||||
w = HnswSearchLayer(base, q, ep, 1, lc, index, support, m, false, NULL, NULL, NULL, true, NULL);
|
||||
w = HnswSearchLayer(base, q, ep, 1, lc, index, support, m, false, NULL, NULL, NULL, true, NULL, false);
|
||||
ep = w;
|
||||
}
|
||||
|
||||
return HnswSearchLayer(base, q, ep, hnsw_ef_search, 0, index, support, m, false, NULL, &so->v, hnsw_iterative_scan != HNSW_ITERATIVE_SCAN_OFF ? &so->discarded : NULL, true, &so->tuples);
|
||||
return HnswSearchLayer(base, q, ep, hnsw_ef_search, 0, index, support, m, false, NULL, &so->v, hnsw_iterative_scan != HNSW_ITERATIVE_SCAN_OFF ? &so->discarded : NULL, true, &so->tuples, false);
|
||||
}
|
||||
|
||||
/*
|
||||
@@ -83,7 +72,7 @@ ResumeScanItems(IndexScanDesc scan)
|
||||
ep = lappend(ep, sc);
|
||||
}
|
||||
|
||||
return HnswSearchLayer(base, &so->q, ep, batch_size, 0, index, &so->support, so->m, false, NULL, &so->v, &so->discarded, false, &so->tuples);
|
||||
return HnswSearchLayer(base, &so->q, ep, batch_size, 0, index, &so->support, so->m, false, NULL, &so->v, &so->discarded, false, &so->tuples, false);
|
||||
}
|
||||
|
||||
/*
|
||||
@@ -153,7 +142,7 @@ hnswbeginscan(Relation index, int nkeys, int norderbys)
|
||||
/* Calculate max memory */
|
||||
/* Add 256 extra bytes to fill last block when close */
|
||||
maxMemory = (double) work_mem * hnsw_scan_mem_multiplier * 1024.0 + 256;
|
||||
so->maxMemory = Min(maxMemory, (double) (SIZE_MAX / 2));
|
||||
so->maxMemory = Min(maxMemory, (double) SIZE_MAX);
|
||||
|
||||
scan->opaque = so;
|
||||
|
||||
|
||||
124
src/hnswutils.c
124
src/hnswutils.c
@@ -2,25 +2,27 @@
|
||||
|
||||
#include <math.h>
|
||||
|
||||
#include "access/genam.h"
|
||||
#include "access/generic_xlog.h"
|
||||
#include "catalog/pg_type.h"
|
||||
#include "catalog/pg_type_d.h"
|
||||
#include "common/hashfn.h"
|
||||
#include "fmgr.h"
|
||||
#include "hnsw.h"
|
||||
#include "lib/pairingheap.h"
|
||||
#include "nodes/pg_list.h"
|
||||
#include "port/atomics.h"
|
||||
#include "sparsevec.h"
|
||||
#include "storage/bufmgr.h"
|
||||
#include "utils/datum.h"
|
||||
#include "utils/memdebug.h"
|
||||
#include "utils/rel.h"
|
||||
#include "vector.h"
|
||||
|
||||
#if PG_VERSION_NUM >= 160000
|
||||
#include "varatt.h"
|
||||
#endif
|
||||
|
||||
#if PG_VERSION_NUM >= 170000
|
||||
#include "storage/read_stream.h"
|
||||
#endif
|
||||
|
||||
#if PG_VERSION_NUM < 170000
|
||||
static inline uint64
|
||||
murmurhash64(uint64 data)
|
||||
@@ -262,7 +264,7 @@ HnswInitElement(char *base, ItemPointer heaptid, int m, double ml, int maxLevel,
|
||||
|
||||
HnswInitNeighbors(base, element, m, allocator);
|
||||
|
||||
HnswPtrStore(base, element->value, (char *) NULL);
|
||||
HnswPtrStore(base, element->value, (Pointer) NULL);
|
||||
|
||||
return element;
|
||||
}
|
||||
@@ -288,7 +290,7 @@ HnswInitElementFromBlock(BlockNumber blkno, OffsetNumber offno)
|
||||
element->blkno = blkno;
|
||||
element->offno = offno;
|
||||
HnswPtrStore(base, element->neighbors, (HnswNeighborArrayPtr *) NULL);
|
||||
HnswPtrStore(base, element->value, (char *) NULL);
|
||||
HnswPtrStore(base, element->value, (Pointer) NULL);
|
||||
return element;
|
||||
}
|
||||
|
||||
@@ -514,7 +516,7 @@ HnswLoadElementFromTuple(HnswElement element, HnswElementTuple etup, bool loadHe
|
||||
char *base = NULL;
|
||||
Datum value = datumCopy(PointerGetDatum(&etup->data), false, -1);
|
||||
|
||||
HnswPtrStore(base, element->value, (char *) DatumGetPointer(value));
|
||||
HnswPtrStore(base, element->value, DatumGetPointer(value));
|
||||
}
|
||||
}
|
||||
|
||||
@@ -531,14 +533,12 @@ HnswGetDistance(Datum a, Datum b, HnswSupport * support)
|
||||
* Load an element and optionally get its distance from q
|
||||
*/
|
||||
static void
|
||||
HnswLoadElementImpl(BlockNumber blkno, OffsetNumber offno, double *distance, HnswQuery * q, Relation index, HnswSupport * support, bool loadVec, double *maxDistance, HnswElement * element)
|
||||
HnswLoadElementImpl(Buffer buf, OffsetNumber offno, double *distance, HnswQuery * q, Relation index, HnswSupport * support, bool loadVec, double *maxDistance, HnswElement * element)
|
||||
{
|
||||
Buffer buf;
|
||||
Page page;
|
||||
HnswElementTuple etup;
|
||||
|
||||
/* Read vector */
|
||||
buf = ReadBuffer(index, blkno);
|
||||
LockBuffer(buf, BUFFER_LOCK_SHARE);
|
||||
page = BufferGetPage(buf);
|
||||
|
||||
@@ -559,7 +559,7 @@ HnswLoadElementImpl(BlockNumber blkno, OffsetNumber offno, double *distance, Hns
|
||||
if (distance == NULL || maxDistance == NULL || *distance < *maxDistance)
|
||||
{
|
||||
if (*element == NULL)
|
||||
*element = HnswInitElementFromBlock(blkno, offno);
|
||||
*element = HnswInitElementFromBlock(BufferGetBlockNumber(buf), offno);
|
||||
|
||||
HnswLoadElementFromTuple(*element, etup, true, loadVec);
|
||||
}
|
||||
@@ -573,7 +573,9 @@ HnswLoadElementImpl(BlockNumber blkno, OffsetNumber offno, double *distance, Hns
|
||||
void
|
||||
HnswLoadElement(HnswElement element, double *distance, HnswQuery * q, Relation index, HnswSupport * support, bool loadVec, double *maxDistance)
|
||||
{
|
||||
HnswLoadElementImpl(element->blkno, element->offno, distance, q, index, support, loadVec, maxDistance, &element);
|
||||
Buffer buf = ReadBuffer(index, element->blkno);
|
||||
|
||||
HnswLoadElementImpl(buf, element->offno, distance, q, index, support, loadVec, maxDistance, &element);
|
||||
}
|
||||
|
||||
/*
|
||||
@@ -813,11 +815,31 @@ HnswLoadUnvisitedFromDisk(HnswElement element, HnswUnvisited * unvisited, int *u
|
||||
}
|
||||
}
|
||||
|
||||
#if PG_VERSION_NUM >= 170000
|
||||
/*
|
||||
* Get next block number for read stream
|
||||
*/
|
||||
static BlockNumber
|
||||
HnswReadStreamNextBlock(ReadStream *stream, void *callback_private_data, void *per_buffer_data)
|
||||
{
|
||||
HnswReadStreamData *streamData = callback_private_data;
|
||||
OffsetNumber *offno = per_buffer_data;
|
||||
HnswUnvisited *uv;
|
||||
|
||||
if (streamData->visited == streamData->unvisitedLength)
|
||||
return InvalidBlockNumber;
|
||||
|
||||
uv = &streamData->unvisited[streamData->visited++];
|
||||
*offno = ItemPointerGetOffsetNumber(&uv->indextid);
|
||||
return ItemPointerGetBlockNumber(&uv->indextid);
|
||||
}
|
||||
#endif
|
||||
|
||||
/*
|
||||
* Algorithm 2 from paper
|
||||
*/
|
||||
List *
|
||||
HnswSearchLayer(char *base, HnswQuery * q, List *ep, int ef, int lc, Relation index, HnswSupport * support, int m, bool inserting, HnswElement skipElement, visited_hash * v, pairingheap **discarded, bool initVisited, int64 *tuples)
|
||||
HnswSearchLayer(char *base, HnswQuery * q, List *ep, int ef, int lc, Relation index, HnswSupport * support, int m, bool inserting, HnswElement skipElement, visited_hash * v, pairingheap **discarded, bool initVisited, int64 *tuples, bool maintenance)
|
||||
{
|
||||
List *w = NIL;
|
||||
pairingheap *C = pairingheap_allocate(CompareNearestCandidates, NULL);
|
||||
@@ -832,6 +854,27 @@ HnswSearchLayer(char *base, HnswQuery * q, List *ep, int ef, int lc, Relation in
|
||||
int unvisitedLength;
|
||||
bool inMemory = index == NULL;
|
||||
|
||||
#if PG_VERSION_NUM >= 170000
|
||||
HnswReadStreamData streamData;
|
||||
ReadStream *stream = NULL;
|
||||
|
||||
if (!inMemory)
|
||||
{
|
||||
int flags = READ_STREAM_DEFAULT;
|
||||
|
||||
if (maintenance)
|
||||
{
|
||||
flags |= READ_STREAM_MAINTENANCE;
|
||||
}
|
||||
|
||||
#if PG_VERSION_NUM >= 180000
|
||||
flags |= READ_STREAM_USE_BATCHING;
|
||||
#endif
|
||||
|
||||
stream = read_stream_begin_relation(flags, NULL, index, MAIN_FORKNUM, HnswReadStreamNextBlock, &streamData, sizeof(OffsetNumber));
|
||||
}
|
||||
#endif
|
||||
|
||||
if (v == NULL)
|
||||
{
|
||||
v = &vh;
|
||||
@@ -894,13 +937,23 @@ HnswSearchLayer(char *base, HnswQuery * q, List *ep, int ef, int lc, Relation in
|
||||
if (inMemory)
|
||||
HnswLoadUnvisitedFromMemory(base, cElement, unvisited, &unvisitedLength, v, lc, localNeighborhood, neighborhoodSize);
|
||||
else
|
||||
{
|
||||
HnswLoadUnvisitedFromDisk(cElement, unvisited, &unvisitedLength, v, index, m, lm, lc);
|
||||
|
||||
#if PG_VERSION_NUM >= 170000
|
||||
read_stream_resume(stream);
|
||||
|
||||
streamData.unvisited = unvisited;
|
||||
streamData.unvisitedLength = unvisitedLength;
|
||||
streamData.visited = 0;
|
||||
#endif
|
||||
}
|
||||
|
||||
/* OK to count elements instead of tuples */
|
||||
if (tuples != NULL)
|
||||
(*tuples) += unvisitedLength;
|
||||
|
||||
for (int i = 0; i < unvisitedLength; i++)
|
||||
for (int i = 0;; i++)
|
||||
{
|
||||
HnswElement eElement;
|
||||
HnswSearchCandidate *e;
|
||||
@@ -911,24 +964,46 @@ HnswSearchLayer(char *base, HnswQuery * q, List *ep, int ef, int lc, Relation in
|
||||
|
||||
if (inMemory)
|
||||
{
|
||||
if (i == unvisitedLength)
|
||||
break;
|
||||
|
||||
eElement = unvisited[i].element;
|
||||
eDistance = GetElementDistance(base, eElement, q, support);
|
||||
}
|
||||
else
|
||||
{
|
||||
ItemPointer indextid = &unvisited[i].indextid;
|
||||
BlockNumber blkno = ItemPointerGetBlockNumber(indextid);
|
||||
OffsetNumber offno = ItemPointerGetOffsetNumber(indextid);
|
||||
Buffer buf;
|
||||
OffsetNumber offno;
|
||||
|
||||
#if PG_VERSION_NUM >= 170000
|
||||
void *offnoPtr;
|
||||
|
||||
buf = read_stream_next_buffer(stream, &offnoPtr);
|
||||
|
||||
if (!BufferIsValid(buf))
|
||||
break;
|
||||
|
||||
offno = *((OffsetNumber *) offnoPtr);
|
||||
#else
|
||||
ItemPointer indextid;
|
||||
|
||||
if (i == unvisitedLength)
|
||||
break;
|
||||
|
||||
indextid = &unvisited[i].indextid;
|
||||
buf = ReadBuffer(index, ItemPointerGetBlockNumber(indextid));
|
||||
offno = ItemPointerGetOffsetNumber(indextid);
|
||||
#endif
|
||||
|
||||
/* Avoid any allocations if not adding */
|
||||
eElement = NULL;
|
||||
HnswLoadElementImpl(blkno, offno, &eDistance, q, index, support, inserting, alwaysAdd || discarded != NULL ? NULL : &f->distance, &eElement);
|
||||
HnswLoadElementImpl(buf, offno, &eDistance, q, index, support, inserting, alwaysAdd || discarded != NULL ? NULL : &f->distance, &eElement);
|
||||
|
||||
if (eElement == NULL)
|
||||
continue;
|
||||
}
|
||||
|
||||
if (!(eDistance < f->distance || alwaysAdd))
|
||||
if (eElement == NULL || !(eDistance < f->distance || alwaysAdd))
|
||||
{
|
||||
if (discarded != NULL)
|
||||
{
|
||||
@@ -978,6 +1053,11 @@ HnswSearchLayer(char *base, HnswQuery * q, List *ep, int ef, int lc, Relation in
|
||||
w = lappend(w, sc);
|
||||
}
|
||||
|
||||
#if PG_VERSION_NUM >= 170000
|
||||
if (!inMemory)
|
||||
read_stream_end(stream);
|
||||
#endif
|
||||
|
||||
return w;
|
||||
}
|
||||
|
||||
@@ -1273,7 +1353,7 @@ PrecomputeHash(char *base, HnswElement element)
|
||||
* Algorithm 1 from paper
|
||||
*/
|
||||
void
|
||||
HnswFindElementNeighbors(char *base, HnswElement element, HnswElement entryPoint, Relation index, HnswSupport * support, int m, int efConstruction, bool existing)
|
||||
HnswFindElementNeighbors(char *base, HnswElement element, HnswElement entryPoint, Relation index, HnswSupport * support, int m, int efConstruction, bool existing, bool maintenance)
|
||||
{
|
||||
List *ep;
|
||||
List *w;
|
||||
@@ -1300,7 +1380,7 @@ HnswFindElementNeighbors(char *base, HnswElement element, HnswElement entryPoint
|
||||
/* 1st phase: greedy search to insert level */
|
||||
for (int lc = entryLevel; lc >= level + 1; lc--)
|
||||
{
|
||||
w = HnswSearchLayer(base, &q, ep, 1, lc, index, support, m, true, skipElement, NULL, NULL, true, NULL);
|
||||
w = HnswSearchLayer(base, &q, ep, 1, lc, index, support, m, true, skipElement, NULL, NULL, true, NULL, maintenance);
|
||||
ep = w;
|
||||
}
|
||||
|
||||
@@ -1319,7 +1399,7 @@ HnswFindElementNeighbors(char *base, HnswElement element, HnswElement entryPoint
|
||||
List *lw = NIL;
|
||||
ListCell *lc2;
|
||||
|
||||
w = HnswSearchLayer(base, &q, ep, efConstruction, lc, index, support, m, true, skipElement, NULL, NULL, true, NULL);
|
||||
w = HnswSearchLayer(base, &q, ep, efConstruction, lc, index, support, m, true, skipElement, NULL, NULL, true, NULL, maintenance);
|
||||
|
||||
/* Convert search candidates to candidates */
|
||||
foreach(lc2, w)
|
||||
|
||||
@@ -1,14 +1,13 @@
|
||||
#include "postgres.h"
|
||||
|
||||
#include "access/genam.h"
|
||||
#include <math.h>
|
||||
|
||||
#include "access/generic_xlog.h"
|
||||
#include "commands/vacuum.h"
|
||||
#include "hnsw.h"
|
||||
#include "nodes/pg_list.h"
|
||||
#include "storage/bufmgr.h"
|
||||
#include "storage/lmgr.h"
|
||||
#include "utils/memutils.h"
|
||||
#include "utils/rel.h"
|
||||
|
||||
#if PG_VERSION_NUM >= 160000
|
||||
#include "varatt.h"
|
||||
@@ -179,12 +178,7 @@ NeedsUpdated(HnswVacuumState * vacuumstate, HnswElement element)
|
||||
/* Also update if layer 0 is not full */
|
||||
/* This could indicate too many candidates being deleted during insert */
|
||||
if (!needsUpdated)
|
||||
{
|
||||
/* Keep clang-tidy happy */
|
||||
Assert(ntup->count > 0);
|
||||
|
||||
needsUpdated = !ItemPointerIsValid(&ntup->indextids[ntup->count - 1]);
|
||||
}
|
||||
|
||||
UnlockReleaseBuffer(buf);
|
||||
|
||||
@@ -218,7 +212,7 @@ RepairGraphElement(HnswVacuumState * vacuumstate, HnswElement element, HnswEleme
|
||||
element->heaptidsLength = 0;
|
||||
|
||||
/* Find neighbors for element, skipping itself */
|
||||
HnswFindElementNeighbors(base, element, entryPoint, index, support, m, efConstruction, true);
|
||||
HnswFindElementNeighbors(base, element, entryPoint, index, support, m, efConstruction, true, true);
|
||||
|
||||
/* Zero memory for each element */
|
||||
MemSet(ntup, 0, HNSW_TUPLE_ALLOC_SIZE);
|
||||
@@ -534,9 +528,8 @@ MarkDeleted(HnswVacuumState * vacuumstate)
|
||||
ntup = (HnswNeighborTuple) PageGetItem(npage, PageGetItemId(npage, neighborOffno));
|
||||
|
||||
/* Overwrite element */
|
||||
/* Use memset instead of MemSet to keep clang-tidy happy */
|
||||
etup->deleted = 1;
|
||||
memset(&etup->data, 0, VARSIZE_ANY(&etup->data));
|
||||
MemSet(&etup->data, 0, VARSIZE_ANY(&etup->data));
|
||||
|
||||
/* Overwrite neighbors */
|
||||
for (int i = 0; i < ntup->count; i++)
|
||||
|
||||
@@ -2,33 +2,23 @@
|
||||
|
||||
#include <float.h>
|
||||
|
||||
#include "access/genam.h"
|
||||
#include "access/generic_xlog.h"
|
||||
#include "access/itup.h"
|
||||
#include "access/relscan.h"
|
||||
#include "access/table.h"
|
||||
#include "access/tableam.h"
|
||||
#include "access/tupdesc.h"
|
||||
#include "access/parallel.h"
|
||||
#include "access/xact.h"
|
||||
#include "access/xloginsert.h"
|
||||
#include "bitvec.h"
|
||||
#include "catalog/index.h"
|
||||
#include "catalog/pg_operator_d.h"
|
||||
#include "catalog/pg_type_d.h"
|
||||
#include "commands/progress.h"
|
||||
#include "fmgr.h"
|
||||
#include "halfvec.h"
|
||||
#include "ivfflat.h"
|
||||
#include "miscadmin.h"
|
||||
#include "nodes/execnodes.h"
|
||||
#include "optimizer/optimizer.h"
|
||||
#include "storage/bufmgr.h"
|
||||
#include "storage/condition_variable.h"
|
||||
#include "tcop/tcopprot.h"
|
||||
#include "utils/memutils.h"
|
||||
#include "utils/rel.h"
|
||||
#include "utils/sampling.h"
|
||||
#include "utils/snapmgr.h"
|
||||
#include "utils/tuplesort.h"
|
||||
#include "vector.h"
|
||||
|
||||
#if PG_VERSION_NUM >= 160000
|
||||
#include "varatt.h"
|
||||
@@ -63,13 +53,15 @@ AddSample(Datum *values, IvfflatBuildState * buildstate)
|
||||
Datum value = PointerGetDatum(PG_DETOAST_DATUM(values[0]));
|
||||
|
||||
/*
|
||||
* Check with KMEANS_NORM_PROC that the value can be normalized since
|
||||
* spherical distance function expects unit vectors
|
||||
* Normalize with KMEANS_NORM_PROC since spherical distance function
|
||||
* expects unit vectors
|
||||
*/
|
||||
if (buildstate->kmeansnormprocinfo != NULL)
|
||||
{
|
||||
if (!IvfflatCheckNorm(buildstate->kmeansnormprocinfo, buildstate->collation, value))
|
||||
return;
|
||||
|
||||
value = IvfflatNormValue(buildstate->typeInfo, buildstate->collation, value);
|
||||
}
|
||||
|
||||
if (samples->length < targsamples)
|
||||
@@ -80,7 +72,7 @@ AddSample(Datum *values, IvfflatBuildState * buildstate)
|
||||
else
|
||||
{
|
||||
if (buildstate->rowstoskip < 0)
|
||||
buildstate->rowstoskip = reservoir_get_next_S(&buildstate->rstate, buildstate->samplerows, targsamples);
|
||||
buildstate->rowstoskip = reservoir_get_next_S(&buildstate->rstate, samples->length, targsamples);
|
||||
|
||||
if (buildstate->rowstoskip <= 0)
|
||||
{
|
||||
@@ -96,9 +88,6 @@ AddSample(Datum *values, IvfflatBuildState * buildstate)
|
||||
|
||||
buildstate->rowstoskip -= 1;
|
||||
}
|
||||
|
||||
/* Increment after reservoir_get_next_S */
|
||||
buildstate->samplerows += 1;
|
||||
}
|
||||
|
||||
/*
|
||||
@@ -135,7 +124,6 @@ SampleRows(IvfflatBuildState * buildstate)
|
||||
int targsamples = buildstate->samples->maxlen;
|
||||
BlockNumber totalblocks = RelationGetNumberOfBlocks(buildstate->heap);
|
||||
|
||||
buildstate->samplerows = 0;
|
||||
buildstate->rowstoskip = -1;
|
||||
|
||||
BlockSampler_Init(&buildstate->bs, totalblocks, targsamples, RandomInt());
|
||||
@@ -145,24 +133,8 @@ SampleRows(IvfflatBuildState * buildstate)
|
||||
{
|
||||
BlockNumber targblock = BlockSampler_Next(&buildstate->bs);
|
||||
|
||||
/* Set anyvisible to false like table_index_build_scan */
|
||||
table_index_build_range_scan(buildstate->heap, buildstate->index, buildstate->indexInfo,
|
||||
false, false, false, targblock, 1, SampleCallback, (void *) buildstate, NULL);
|
||||
}
|
||||
|
||||
/* Normalize if needed */
|
||||
if (buildstate->kmeansnormprocinfo != NULL)
|
||||
{
|
||||
VectorArray samples = buildstate->samples;
|
||||
|
||||
for (int i = 0; i < samples->length; i++)
|
||||
{
|
||||
Datum value = PointerGetDatum(VectorArrayGet(samples, i));
|
||||
Datum normValue = IvfflatNormValue(buildstate->typeInfo, buildstate->collation, value);
|
||||
|
||||
VectorArraySet(samples, i, DatumGetPointer(normValue));
|
||||
pfree(DatumGetPointer(normValue));
|
||||
}
|
||||
false, true, false, targblock, 1, SampleCallback, (void *) buildstate, NULL);
|
||||
}
|
||||
}
|
||||
|
||||
@@ -393,9 +365,6 @@ InitBuildState(IvfflatBuildState * buildstate, Relation heap, Relation index, In
|
||||
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);
|
||||
#if PG_VERSION_NUM >= 190000
|
||||
TupleDescFinalize(buildstate->sortdesc);
|
||||
#endif
|
||||
|
||||
buildstate->slot = MakeSingleTupleTableSlot(buildstate->sortdesc, &TTSOpsVirtual);
|
||||
|
||||
@@ -457,7 +426,7 @@ ComputeCenters(IvfflatBuildState * buildstate)
|
||||
buildstate->samples = VectorArrayInit(numSamples, buildstate->dimensions, buildstate->centers->itemsize);
|
||||
if (buildstate->heap != NULL)
|
||||
{
|
||||
IvfflatBench("sample rows", SampleRows(buildstate));
|
||||
SampleRows(buildstate);
|
||||
|
||||
if (buildstate->samples->length < buildstate->lists)
|
||||
{
|
||||
@@ -672,11 +641,7 @@ IvfflatParallelScanAndSort(IvfflatSpool * ivfspool, IvfflatShared * ivfshared, S
|
||||
ivfspool->sortstate = InitBuildSortState(buildstate.sortdesc, sortmem, coordinate);
|
||||
buildstate.sortstate = ivfspool->sortstate;
|
||||
scan = table_beginscan_parallel(ivfspool->heap,
|
||||
ParallelTableScanFromIvfflatShared(ivfshared)
|
||||
#if PG_VERSION_NUM >= 190000
|
||||
,SO_NONE
|
||||
#endif
|
||||
);
|
||||
ParallelTableScanFromIvfflatShared(ivfshared));
|
||||
reltuples = table_index_build_scan(ivfspool->heap, ivfspool->index, indexInfo,
|
||||
true, progress, BuildCallback,
|
||||
(void *) &buildstate, scan);
|
||||
|
||||
@@ -3,19 +3,14 @@
|
||||
#include <float.h>
|
||||
|
||||
#include "access/amapi.h"
|
||||
#include "access/genam.h"
|
||||
#include "access/reloptions.h"
|
||||
#include "commands/progress.h"
|
||||
#include "commands/vacuum.h"
|
||||
#include "fmgr.h"
|
||||
#include "ivfflat.h"
|
||||
#include "nodes/pg_list.h"
|
||||
#include "utils/float.h"
|
||||
#include "utils/guc.h"
|
||||
#include "utils/relcache.h"
|
||||
#include "utils/selfuncs.h"
|
||||
#include "utils/spccache.h"
|
||||
#include "vector.h"
|
||||
|
||||
#if PG_VERSION_NUM < 150000
|
||||
#define MarkGUCPrefixReserved(x) EmitWarningsOnPlaceholders(x)
|
||||
@@ -184,68 +179,6 @@ FUNCTION_PREFIX PG_FUNCTION_INFO_V1(ivfflathandler);
|
||||
Datum
|
||||
ivfflathandler(PG_FUNCTION_ARGS)
|
||||
{
|
||||
#if PG_VERSION_NUM >= 190000
|
||||
static const IndexAmRoutine amroutine = {
|
||||
.type = T_IndexAmRoutine,
|
||||
.amstrategies = 0,
|
||||
.amsupport = 5,
|
||||
.amoptsprocnum = 0,
|
||||
.amcanorder = false,
|
||||
.amcanorderbyop = true,
|
||||
.amcanhash = false,
|
||||
.amconsistentequality = false,
|
||||
.amconsistentordering = false,
|
||||
.amcanbackward = false,
|
||||
.amcanmarkpos = false,
|
||||
.amcanunique = false,
|
||||
.amcanmulticol = false,
|
||||
.amoptionalkey = true,
|
||||
.amsearcharray = false,
|
||||
.amsearchnulls = false,
|
||||
.amstorage = false,
|
||||
.amclusterable = false,
|
||||
.ampredlocks = false,
|
||||
.amcanparallel = false,
|
||||
.amcanbuildparallel = true,
|
||||
.amcaninclude = false,
|
||||
.amusemaintenanceworkmem = false,
|
||||
.amsummarizing = false,
|
||||
.amparallelvacuumoptions = VACUUM_OPTION_PARALLEL_BULKDEL,
|
||||
.amkeytype = InvalidOid,
|
||||
|
||||
.ambuild = ivfflatbuild,
|
||||
.ambuildempty = ivfflatbuildempty,
|
||||
.aminsert = ivfflatinsert,
|
||||
.aminsertcleanup = NULL,
|
||||
.ambulkdelete = ivfflatbulkdelete,
|
||||
.amvacuumcleanup = ivfflatvacuumcleanup,
|
||||
.amcanreturn = NULL,
|
||||
.amcostestimate = ivfflatcostestimate,
|
||||
.amgettreeheight = NULL,
|
||||
.amoptions = ivfflatoptions,
|
||||
.amproperty = NULL,
|
||||
.ambuildphasename = ivfflatbuildphasename,
|
||||
.amvalidate = ivfflatvalidate,
|
||||
.amadjustmembers = NULL,
|
||||
.ambeginscan = ivfflatbeginscan,
|
||||
.amrescan = ivfflatrescan,
|
||||
.amgettuple = ivfflatgettuple,
|
||||
.amgetbatch = NULL,
|
||||
.amunguardbatch = NULL,
|
||||
.amkillitemsbatch = NULL,
|
||||
.amgettransform = NULL,
|
||||
.amgetbitmap = NULL,
|
||||
.amendscan = ivfflatendscan,
|
||||
.amposreset = NULL,
|
||||
.amestimateparallelscan = NULL,
|
||||
.aminitparallelscan = NULL,
|
||||
.amparallelrescan = NULL,
|
||||
.amtranslatestrategy = NULL,
|
||||
.amtranslatecmptype = NULL,
|
||||
};
|
||||
|
||||
PG_RETURN_POINTER(&amroutine);
|
||||
#else
|
||||
IndexAmRoutine *amroutine = makeNode(IndexAmRoutine);
|
||||
|
||||
amroutine->amstrategies = 0;
|
||||
@@ -319,5 +252,4 @@ ivfflathandler(PG_FUNCTION_ARGS)
|
||||
#endif
|
||||
|
||||
PG_RETURN_POINTER(amroutine);
|
||||
#endif
|
||||
}
|
||||
|
||||
@@ -9,7 +9,6 @@
|
||||
#include "lib/pairingheap.h"
|
||||
#include "nodes/execnodes.h"
|
||||
#include "port.h" /* for random() */
|
||||
#include "storage/condition_variable.h"
|
||||
#include "utils/sampling.h"
|
||||
#include "utils/tuplesort.h"
|
||||
#include "vector.h"
|
||||
@@ -26,10 +25,6 @@
|
||||
#include "portability/instr_time.h"
|
||||
#endif
|
||||
|
||||
#if PG_VERSION_NUM >= 190000
|
||||
typedef Pointer Item;
|
||||
#endif
|
||||
|
||||
#define IVFFLAT_MAX_DIM 2000
|
||||
|
||||
/* Support functions */
|
||||
@@ -214,8 +209,7 @@ typedef struct IvfflatBuildState
|
||||
/* Sampling */
|
||||
BlockSamplerData bs;
|
||||
ReservoirStateData rstate;
|
||||
double samplerows;
|
||||
double rowstoskip;
|
||||
int rowstoskip;
|
||||
|
||||
/* Sorting */
|
||||
Tuplesortstate *sortstate;
|
||||
|
||||
@@ -2,16 +2,11 @@
|
||||
|
||||
#include <float.h>
|
||||
|
||||
#include "access/genam.h"
|
||||
#include "access/generic_xlog.h"
|
||||
#include "access/itup.h"
|
||||
#include "fmgr.h"
|
||||
#include "ivfflat.h"
|
||||
#include "nodes/execnodes.h"
|
||||
#include "storage/bufmgr.h"
|
||||
#include "storage/lmgr.h"
|
||||
#include "utils/memutils.h"
|
||||
#include "utils/rel.h"
|
||||
|
||||
/*
|
||||
* Find the list that minimizes the distance function
|
||||
|
||||
@@ -1,15 +1,17 @@
|
||||
#include "postgres.h"
|
||||
|
||||
#include <float.h>
|
||||
#include <limits.h>
|
||||
#include <math.h>
|
||||
|
||||
#include "access/genam.h"
|
||||
#include "fmgr.h"
|
||||
#include "bitvec.h"
|
||||
#include "halfutils.h"
|
||||
#include "halfvec.h"
|
||||
#include "ivfflat.h"
|
||||
#include "miscadmin.h"
|
||||
#include "utils/builtins.h"
|
||||
#include "utils/datum.h"
|
||||
#include "utils/memutils.h"
|
||||
#include "utils/relcache.h"
|
||||
#include "vector.h"
|
||||
|
||||
#if PG_VERSION_NUM >= 160000
|
||||
#include "varatt.h"
|
||||
|
||||
@@ -2,26 +2,15 @@
|
||||
|
||||
#include <float.h>
|
||||
|
||||
#include "access/genam.h"
|
||||
#include "access/itup.h"
|
||||
#include "access/relscan.h"
|
||||
#include "access/tupdesc.h"
|
||||
#include "catalog/pg_operator_d.h"
|
||||
#include "catalog/pg_type_d.h"
|
||||
#include "fmgr.h"
|
||||
#include "lib/pairingheap.h"
|
||||
#include "ivfflat.h"
|
||||
#include "miscadmin.h"
|
||||
#include "pgstat.h"
|
||||
#include "storage/bufmgr.h"
|
||||
#include "utils/memutils.h"
|
||||
#include "utils/rel.h"
|
||||
#include "utils/snapmgr.h"
|
||||
#include "utils/tuplesort.h"
|
||||
|
||||
#if PG_VERSION_NUM >= 160000
|
||||
#include "varatt.h"
|
||||
#endif
|
||||
|
||||
#define GetScanList(ptr) pairingheap_container(IvfflatScanList, ph_node, ptr)
|
||||
#define GetScanListConst(ptr) pairingheap_const_container(IvfflatScanList, ph_node, ptr)
|
||||
@@ -298,9 +287,6 @@ ivfflatbeginscan(Relation index, int nkeys, int norderbys)
|
||||
so->tupdesc = CreateTemplateTupleDesc(2);
|
||||
TupleDescInitEntry(so->tupdesc, (AttrNumber) 1, "distance", FLOAT8OID, -1, 0);
|
||||
TupleDescInitEntry(so->tupdesc, (AttrNumber) 2, "heaptid", TIDOID, -1, 0);
|
||||
#if PG_VERSION_NUM >= 190000
|
||||
TupleDescFinalize(so->tupdesc);
|
||||
#endif
|
||||
|
||||
/* Prep sort */
|
||||
so->sortstate = InitScanSortState(so->tupdesc);
|
||||
|
||||
@@ -1,19 +1,13 @@
|
||||
#include "postgres.h"
|
||||
|
||||
#include "access/genam.h"
|
||||
#include "access/generic_xlog.h"
|
||||
#include "bitvec.h"
|
||||
#include "catalog/pg_type.h"
|
||||
#include "fmgr.h"
|
||||
#include "halfutils.h"
|
||||
#include "halfvec.h"
|
||||
#include "ivfflat.h"
|
||||
#include "storage/bufmgr.h"
|
||||
#include "utils/relcache.h"
|
||||
#include "utils/varbit.h"
|
||||
#include "vector.h"
|
||||
|
||||
#if PG_VERSION_NUM >= 160000
|
||||
#include "varatt.h"
|
||||
#endif
|
||||
|
||||
/*
|
||||
* Allocate a vector array
|
||||
|
||||
@@ -1,12 +1,9 @@
|
||||
#include "postgres.h"
|
||||
|
||||
#include "access/genam.h"
|
||||
#include "access/generic_xlog.h"
|
||||
#include "access/itup.h"
|
||||
#include "commands/vacuum.h"
|
||||
#include "ivfflat.h"
|
||||
#include "storage/bufmgr.h"
|
||||
#include "utils/relcache.h"
|
||||
|
||||
#if PG_VERSION_NUM >= 180000
|
||||
#define vacuum_delay_point() vacuum_delay_point(false)
|
||||
|
||||
@@ -5,27 +5,18 @@
|
||||
|
||||
#include "catalog/pg_type.h"
|
||||
#include "common/shortest_dec.h"
|
||||
#include "common/string.h"
|
||||
#include "fmgr.h"
|
||||
#include "halfutils.h"
|
||||
#include "halfvec.h"
|
||||
#include "lib/stringinfo.h"
|
||||
#include "libpq/pqformat.h"
|
||||
#include "sparsevec.h"
|
||||
#include "utils/array.h"
|
||||
#include "utils/builtins.h"
|
||||
#include "utils/float.h"
|
||||
#include "utils/fmgrprotos.h"
|
||||
#include "utils/lsyscache.h"
|
||||
#include "vector.h"
|
||||
|
||||
#if PG_VERSION_NUM >= 160000
|
||||
#include "varatt.h"
|
||||
#endif
|
||||
|
||||
#if PG_VERSION_NUM >= 170000
|
||||
#include "parser/scansup.h"
|
||||
#endif
|
||||
|
||||
typedef struct SparseInputElement
|
||||
{
|
||||
int32 index;
|
||||
@@ -159,9 +150,9 @@ InitSparseVector(int dim, int nnz)
|
||||
return result;
|
||||
}
|
||||
|
||||
#if PG_VERSION_NUM >= 170000
|
||||
#define sparsevec_isspace(ch) scanner_isspace(ch)
|
||||
#else
|
||||
/*
|
||||
* Check for whitespace, since array_isspace() is static
|
||||
*/
|
||||
static inline bool
|
||||
sparsevec_isspace(char ch)
|
||||
{
|
||||
@@ -174,7 +165,6 @@ sparsevec_isspace(char ch)
|
||||
return true;
|
||||
return false;
|
||||
}
|
||||
#endif
|
||||
|
||||
/*
|
||||
* Compare indices
|
||||
@@ -182,10 +172,10 @@ sparsevec_isspace(char ch)
|
||||
static int
|
||||
CompareIndices(const void *a, const void *b)
|
||||
{
|
||||
if (((const SparseInputElement *) a)->index < ((const SparseInputElement *) b)->index)
|
||||
if (((SparseInputElement *) a)->index < ((SparseInputElement *) b)->index)
|
||||
return -1;
|
||||
|
||||
if (((const SparseInputElement *) a)->index > ((const SparseInputElement *) b)->index)
|
||||
if (((SparseInputElement *) a)->index > ((SparseInputElement *) b)->index)
|
||||
return 1;
|
||||
|
||||
return 0;
|
||||
|
||||
17
src/vector.c
17
src/vector.c
@@ -16,20 +16,16 @@
|
||||
#include "port.h" /* for strtof() */
|
||||
#include "sparsevec.h"
|
||||
#include "utils/array.h"
|
||||
#include "utils/builtins.h"
|
||||
#include "utils/float.h"
|
||||
#include "utils/fmgrprotos.h"
|
||||
#include "utils/lsyscache.h"
|
||||
#include "utils/varbit.h"
|
||||
#include "utils/numeric.h"
|
||||
#include "vector.h"
|
||||
|
||||
#if PG_VERSION_NUM >= 160000
|
||||
#include "varatt.h"
|
||||
#endif
|
||||
|
||||
#if PG_VERSION_NUM >= 170000
|
||||
#include "parser/scansup.h"
|
||||
#endif
|
||||
|
||||
#define STATE_DIMS(x) (ARR_DIMS(x)[0] - 1)
|
||||
#define CreateStateDatums(dim) palloc(sizeof(Datum) * (dim + 1))
|
||||
|
||||
@@ -40,7 +36,7 @@
|
||||
#endif
|
||||
|
||||
#if PG_VERSION_NUM >= 180000
|
||||
PG_MODULE_MAGIC_EXT(.name = "vector", .version = "0.8.2");
|
||||
PG_MODULE_MAGIC_EXT(.name = "vector",.version = "0.8.1");
|
||||
#else
|
||||
PG_MODULE_MAGIC;
|
||||
#endif
|
||||
@@ -133,9 +129,9 @@ InitVector(int dim)
|
||||
return result;
|
||||
}
|
||||
|
||||
#if PG_VERSION_NUM >= 170000
|
||||
#define vector_isspace(ch) scanner_isspace(ch)
|
||||
#else
|
||||
/*
|
||||
* Check for whitespace, since array_isspace() is static
|
||||
*/
|
||||
static inline bool
|
||||
vector_isspace(char ch)
|
||||
{
|
||||
@@ -148,7 +144,6 @@ vector_isspace(char ch)
|
||||
return true;
|
||||
return false;
|
||||
}
|
||||
#endif
|
||||
|
||||
/*
|
||||
* Check state array
|
||||
|
||||
@@ -16,19 +16,29 @@ $node->safe_psql("postgres",
|
||||
"INSERT INTO tst SELECT i, ARRAY[random(), random(), random()] FROM generate_series(1, 100000) i;"
|
||||
);
|
||||
|
||||
# Add index
|
||||
$node->safe_psql("postgres", "CREATE INDEX ON tst USING ivfflat (v vector_l2_ops);");
|
||||
# Check each index type
|
||||
my @operators = ("<->", "<#>", "<=>");
|
||||
my @opclasses = ("vector_l2_ops", "vector_ip_ops", "vector_cosine_ops");
|
||||
|
||||
# Test 100% recall
|
||||
for (1 .. 20)
|
||||
for my $i (0 .. $#operators)
|
||||
{
|
||||
my $id = int(rand() * 100000);
|
||||
my $query = $node->safe_psql("postgres", "SELECT v FROM tst WHERE i = $id;");
|
||||
my $res = $node->safe_psql("postgres", qq(
|
||||
SET enable_seqscan = off;
|
||||
SELECT v FROM tst ORDER BY v <-> '$query' LIMIT 1;
|
||||
));
|
||||
is($res, $query);
|
||||
my $operator = $operators[$i];
|
||||
my $opclass = $opclasses[$i];
|
||||
|
||||
# Add index
|
||||
$node->safe_psql("postgres", "CREATE INDEX ON tst USING ivfflat (v $opclass);");
|
||||
|
||||
# Test 100% recall
|
||||
for (1 .. 20)
|
||||
{
|
||||
my $id = int(rand() * 100000);
|
||||
my $query = $node->safe_psql("postgres", "SELECT v FROM tst WHERE i = $id;");
|
||||
my $res = $node->safe_psql("postgres", qq(
|
||||
SET enable_seqscan = off;
|
||||
SELECT v FROM tst ORDER BY v <-> '$query' LIMIT 1;
|
||||
));
|
||||
is($res, $query);
|
||||
}
|
||||
}
|
||||
|
||||
done_testing();
|
||||
|
||||
@@ -1,29 +0,0 @@
|
||||
use strict;
|
||||
use warnings FATAL => 'all';
|
||||
use PostgreSQL::Test::Cluster;
|
||||
use PostgreSQL::Test::Utils;
|
||||
use Test::More;
|
||||
|
||||
# Initialize node
|
||||
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||
$node->init;
|
||||
$node->start;
|
||||
|
||||
# Create table
|
||||
$node->safe_psql("postgres", "CREATE EXTENSION vector;");
|
||||
$node->safe_psql("postgres", "CREATE TABLE tst (v vector(3));");
|
||||
$node->safe_psql("postgres",
|
||||
"INSERT INTO tst SELECT ARRAY[random(), random(), random()] FROM generate_series(1, 1000) i;"
|
||||
);
|
||||
|
||||
my ($ret, $stdout, $stderr) = $node->psql("postgres", qq(
|
||||
SET client_min_messages = DEBUG;
|
||||
SET maintenance_work_mem = '3073kB';
|
||||
ALTER TABLE tst SET (parallel_workers = 1);
|
||||
CREATE INDEX ON tst USING hnsw (v vector_l2_ops);
|
||||
));
|
||||
is($ret, 0, $stderr);
|
||||
like($stderr, qr/using \d+ parallel workers/);
|
||||
like($stderr, qr/hnsw graph no longer fits into maintenance_work_mem after 0 tuples/);
|
||||
|
||||
done_testing();
|
||||
@@ -1,4 +1,4 @@
|
||||
comment = 'vector data type and ivfflat and hnsw access methods'
|
||||
default_version = '0.8.2'
|
||||
default_version = '0.8.1'
|
||||
module_pathname = '$libdir/vector'
|
||||
relocatable = true
|
||||
|
||||
Reference in New Issue
Block a user