liuyu/pgvector
1
0
Fork 0
mirror of https://github.com/pgvector/pgvector.git synced 2026-07-15 09:06:56 +08:00
Code Issues Packages Projects Releases Wiki Activity

Added experimental support for Windows (including auto-vectorization) - closes #49

Browse Source
This commit is contained in:
Andrew Kane
2022-12-08 13:27:26 -08:00
parent 4fbe55f97e
commit 2621f9f947
6 changed files with 139 additions and 41 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
.github/workflows/build.yml vendored
View File

@@ -42,3 +42,16 @@ jobs:
wget -q https://github.com/postgres/postgres/archive/refs/tags/REL_14_5.tar.gz wget -q https://github.com/postgres/postgres/archive/refs/tags/REL_14_5.tar.gz
tar xf REL_14_5.tar.gz tar xf REL_14_5.tar.gz
make prove_installcheck PROVE=prove PROVE_FLAGS="-I ./postgres-REL_14_5/src/test/perl" PERL5LIB="/Users/runner/perl5/lib/perl5" make prove_installcheck PROVE=prove PROVE_FLAGS="-I ./postgres-REL_14_5/src/test/perl" PERL5LIB="/Users/runner/perl5/lib/perl5"
windows:
runs-on: windows-latest
steps:
- uses: actions/checkout@v3
- uses: ankane/setup-postgres@v1
with:
postgres-version: 14
- run: |
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
shell: cmd

1
CHANGELOG.md
View File

@@ -2,6 +2,7 @@
- Changed text representation for vector elements to match `real` - Changed text representation for vector elements to match `real`
- Improved accuracy of text parsing for certain inputs - Improved accuracy of text parsing for certain inputs
- Added experimental support for Windows
## 0.3.2 (2022-11-22) ## 0.3.2 (2022-11-22)

65
Makefile.win Normal file
View File

@@ -0,0 +1,65 @@
EXTENSION = vector
EXTVERSION = 0.3.2
OBJS = src\ivfbuild.obj src\ivfflat.obj src\ivfinsert.obj src\ivfkmeans.obj src\ivfscan.obj src\ivfutils.obj src\ivfvacuum.obj src\vector.obj
# For /arch flags
# https://learn.microsoft.com/en-us/cpp/build/reference/arch-minimum-cpu-architecture
OPTFLAGS =
# For auto-vectorization:
# - MSVC (needs /O2 /fp:fast) - https://learn.microsoft.com/en-us/cpp/parallel/auto-parallelization-and-auto-vectorization?#auto-vectorizer
PG_CFLAGS = $(PG_CFLAGS) $(OPTFLAGS) /O2 /fp:fast
# Debug MSVC auto-vectorization
# https://learn.microsoft.com/en-us/cpp/error-messages/tool-errors/vectorizer-and-parallelizer-messages
# PG_CFLAGS = $(PG_CFLAGS) /Qvec-report:2
all: sql\$(EXTENSION)--$(EXTVERSION).sql
sql\$(EXTENSION)--$(EXTVERSION).sql: sql\$(EXTENSION).sql
copy sql\$(EXTENSION).sql $@
# TODO use pg_config
INCLUDEDIR = $(PGROOT)\include
INCLUDEDIR_SERVER = $(PGROOT)\include\server
LIBDIR = $(PGROOT)\lib
PKGLIBDIR = $(PGROOT)\lib
SHAREDIR = $(PGROOT)\share
CFLAGS = /nologo /I"$(INCLUDEDIR_SERVER)\port\win32_msvc" /I"$(INCLUDEDIR_SERVER)\port\win32" /I"$(INCLUDEDIR_SERVER)" /I"$(INCLUDEDIR)"
CFLAGS = $(CFLAGS) $(PG_CFLAGS)
SHLIB = src\$(EXTENSION).dll
LIBS = "$(LIBDIR)\postgres.lib"
.c.obj:
$(CC) $(CFLAGS) /c $< /Fo$@
$(SHLIB): $(OBJS)
$(CC) $(CFLAGS) $(OBJS) $(LIBS) /link /DLL /OUT:$(SHLIB)
all: $(SHLIB)
install:
copy $(SHLIB) "$(PKGLIBDIR)"
copy $(EXTENSION).control "$(SHAREDIR)\extension"
copy sql\$(EXTENSION)--*.sql "$(SHAREDIR)\extension"
# TODO improve
installcheck:
createdb contrib_regression
mkdir -p results
psql -d contrib_regression -a -q -f test\sql\btree.sql > results\btree.out 2>&1
psql -d contrib_regression -a -q -f test\sql\cast.sql > results\cast.out 2>&1
psql -d contrib_regression -a -q -f test\sql\copy.sql > results\copy.out 2>&1
psql -d contrib_regression -a -q -f test\sql\functions.sql > results\functions.out 2>&1
psql -d contrib_regression -a -q -f test\sql\input.sql > results\input.out 2>&1
psql -d contrib_regression -a -q -f test\sql\ivfflat_cosine.sql > results\ivfflat_cosine.out 2>&1
psql -d contrib_regression -a -q -f test\sql\ivfflat_ip.sql > results\ivfflat_ip.out 2>&1
psql -d contrib_regression -a -q -f test\sql\ivfflat_l2.sql > results\ivfflat_l2.out 2>&1
psql -d contrib_regression -a -q -f test\sql\ivfflat_options.sql > results\ivfflat_options.out 2>&1
psql -d contrib_regression -a -q -f test\sql\ivfflat_unlogged.sql > results\ivfflat_unlogged.out 2>&1
cat results/*

2
src/ivfflat.c
View File

@@ -164,7 +164,7 @@ ivfflatvalidate(Oid opclassoid)
* *
* See https://www.postgresql.org/docs/current/index-api.html * See https://www.postgresql.org/docs/current/index-api.html
*/ */
PG_FUNCTION_INFO_V1(ivfflathandler); PGDLLEXPORT PG_FUNCTION_INFO_V1(ivfflathandler);
Datum Datum
ivfflathandler(PG_FUNCTION_ARGS) ivfflathandler(PG_FUNCTION_ARGS)
{ {

4
src/ivfflat.h
View File

@@ -78,6 +78,9 @@
/* Variables */ /* Variables */
extern int ivfflat_probes; extern int ivfflat_probes;
/* Exported functions */
PGDLLEXPORT void _PG_init(void);
typedef struct VectorArrayData typedef struct VectorArrayData
{ {
int length; int length;
@@ -210,7 +213,6 @@ typedef IvfflatScanOpaqueData * IvfflatScanOpaque;
#define VectorArraySet(_arr, _offset, _val) (memcpy(VECTOR_ARRAY_OFFSET(_arr, _offset), _val, VECTOR_SIZE(_arr->dim))) #define VectorArraySet(_arr, _offset, _val) (memcpy(VECTOR_ARRAY_OFFSET(_arr, _offset), _val, VECTOR_SIZE(_arr->dim)))
/* Methods */ /* Methods */
void _PG_init(void);
VectorArray VectorArrayInit(int maxlen, int dimensions); VectorArray VectorArrayInit(int maxlen, int dimensions);
void VectorArrayFree(VectorArray arr); void VectorArrayFree(VectorArray arr);
void PrintVectorArray(char *msg, VectorArray arr); void PrintVectorArray(char *msg, VectorArray arr);

95
src/vector.c
View File

@@ -109,7 +109,7 @@ PrintVector(char *msg, Vector * vector)
/* /*
* Convert textual representation to internal representation * Convert textual representation to internal representation
*/ */
PG_FUNCTION_INFO_V1(vector_in); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_in);
Datum Datum
vector_in(PG_FUNCTION_ARGS) vector_in(PG_FUNCTION_ARGS)
{ {
@@ -186,7 +186,7 @@ vector_in(PG_FUNCTION_ARGS)
/* /*
* Convert internal representation to textual representation * Convert internal representation to textual representation
*/ */
PG_FUNCTION_INFO_V1(vector_out); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_out);
Datum Datum
vector_out(PG_FUNCTION_ARGS) vector_out(PG_FUNCTION_ARGS)
{ {
@@ -247,7 +247,7 @@ vector_out(PG_FUNCTION_ARGS)
/* /*
* Convert type modifier * Convert type modifier
*/ */
PG_FUNCTION_INFO_V1(vector_typmod_in); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_typmod_in);
Datum Datum
vector_typmod_in(PG_FUNCTION_ARGS) vector_typmod_in(PG_FUNCTION_ARGS)
{ {
@@ -278,7 +278,7 @@ vector_typmod_in(PG_FUNCTION_ARGS)
/* /*
* Convert external binary representation to internal representation * Convert external binary representation to internal representation
*/ */
PG_FUNCTION_INFO_V1(vector_recv); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_recv);
Datum Datum
vector_recv(PG_FUNCTION_ARGS) vector_recv(PG_FUNCTION_ARGS)
{ {
@@ -310,7 +310,7 @@ vector_recv(PG_FUNCTION_ARGS)
/* /*
* Convert internal representation to the external binary representation * Convert internal representation to the external binary representation
*/ */
PG_FUNCTION_INFO_V1(vector_send); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_send);
Datum Datum
vector_send(PG_FUNCTION_ARGS) vector_send(PG_FUNCTION_ARGS)
{ {
@@ -330,7 +330,7 @@ vector_send(PG_FUNCTION_ARGS)
/* /*
* Convert vector to vector * Convert vector to vector
*/ */
PG_FUNCTION_INFO_V1(vector); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector);
Datum Datum
vector(PG_FUNCTION_ARGS) vector(PG_FUNCTION_ARGS)
{ {
@@ -345,7 +345,7 @@ vector(PG_FUNCTION_ARGS)
/* /*
* Convert array to vector * Convert array to vector
*/ */
PG_FUNCTION_INFO_V1(array_to_vector); PGDLLEXPORT PG_FUNCTION_INFO_V1(array_to_vector);
Datum Datum
array_to_vector(PG_FUNCTION_ARGS) array_to_vector(PG_FUNCTION_ARGS)
{ {
@@ -403,7 +403,7 @@ array_to_vector(PG_FUNCTION_ARGS)
/* /*
* Convert vector to float4[] * Convert vector to float4[]
*/ */
PG_FUNCTION_INFO_V1(vector_to_float4); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_to_float4);
Datum Datum
vector_to_float4(PG_FUNCTION_ARGS) vector_to_float4(PG_FUNCTION_ARGS)
{ {
@@ -426,12 +426,14 @@ vector_to_float4(PG_FUNCTION_ARGS)
/* /*
* Get the L2 distance between vectors * Get the L2 distance between vectors
*/ */
PG_FUNCTION_INFO_V1(l2_distance); PGDLLEXPORT PG_FUNCTION_INFO_V1(l2_distance);
Datum Datum
l2_distance(PG_FUNCTION_ARGS) l2_distance(PG_FUNCTION_ARGS)
{ {
Vector *a = PG_GETARG_VECTOR_P(0); Vector *a = PG_GETARG_VECTOR_P(0);
Vector *b = PG_GETARG_VECTOR_P(1); Vector *b = PG_GETARG_VECTOR_P(1);
float *ax = a->x;
float *bx = b->x;
double distance = 0.0; double distance = 0.0;
double diff; double diff;
@@ -439,7 +441,7 @@ l2_distance(PG_FUNCTION_ARGS)
for (int i = 0; i < a->dim; i++) for (int i = 0; i < a->dim; i++)
{ {
diff = a->x[i] - b->x[i]; diff = ax[i] - bx[i];
distance += diff * diff; distance += diff * diff;
} }
@@ -450,12 +452,14 @@ l2_distance(PG_FUNCTION_ARGS)
* Get the L2 squared distance between vectors * Get the L2 squared distance between vectors
* This saves a sqrt calculation * This saves a sqrt calculation
*/ */
PG_FUNCTION_INFO_V1(vector_l2_squared_distance); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_l2_squared_distance);
Datum Datum
vector_l2_squared_distance(PG_FUNCTION_ARGS) vector_l2_squared_distance(PG_FUNCTION_ARGS)
{ {
Vector *a = PG_GETARG_VECTOR_P(0); Vector *a = PG_GETARG_VECTOR_P(0);
Vector *b = PG_GETARG_VECTOR_P(1); Vector *b = PG_GETARG_VECTOR_P(1);
float *ax = a->x;
float *bx = b->x;
double distance = 0.0; double distance = 0.0;
double diff; double diff;
@@ -463,7 +467,7 @@ vector_l2_squared_distance(PG_FUNCTION_ARGS)
for (int i = 0; i < a->dim; i++) for (int i = 0; i < a->dim; i++)
{ {
diff = a->x[i] - b->x[i]; diff = ax[i] - bx[i];
distance += diff * diff; distance += diff * diff;
} }
@@ -473,18 +477,20 @@ vector_l2_squared_distance(PG_FUNCTION_ARGS)
/* /*
* Get the inner product of two vectors * Get the inner product of two vectors
*/ */
PG_FUNCTION_INFO_V1(inner_product); PGDLLEXPORT PG_FUNCTION_INFO_V1(inner_product);
Datum Datum
inner_product(PG_FUNCTION_ARGS) inner_product(PG_FUNCTION_ARGS)
{ {
Vector *a = PG_GETARG_VECTOR_P(0); Vector *a = PG_GETARG_VECTOR_P(0);
Vector *b = PG_GETARG_VECTOR_P(1); Vector *b = PG_GETARG_VECTOR_P(1);
float *ax = a->x;
float *bx = b->x;
double distance = 0.0; double distance = 0.0;
CheckDims(a, b); CheckDims(a, b);
for (int i = 0; i < a->dim; i++) for (int i = 0; i < a->dim; i++)
distance += a->x[i] * b->x[i]; distance += ax[i] * bx[i];
PG_RETURN_FLOAT8(distance); PG_RETURN_FLOAT8(distance);
} }
@@ -492,18 +498,20 @@ inner_product(PG_FUNCTION_ARGS)
/* /*
* Get the negative inner product of two vectors * Get the negative inner product of two vectors
*/ */
PG_FUNCTION_INFO_V1(vector_negative_inner_product); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_negative_inner_product);
Datum Datum
vector_negative_inner_product(PG_FUNCTION_ARGS) vector_negative_inner_product(PG_FUNCTION_ARGS)
{ {
Vector *a = PG_GETARG_VECTOR_P(0); Vector *a = PG_GETARG_VECTOR_P(0);
Vector *b = PG_GETARG_VECTOR_P(1); Vector *b = PG_GETARG_VECTOR_P(1);
float *ax = a->x;
float *bx = b->x;
double distance = 0.0; double distance = 0.0;
CheckDims(a, b); CheckDims(a, b);
for (int i = 0; i < a->dim; i++) for (int i = 0; i < a->dim; i++)
distance += a->x[i] * b->x[i]; distance += ax[i] * bx[i];
PG_RETURN_FLOAT8(distance * -1); PG_RETURN_FLOAT8(distance * -1);
} }
@@ -511,12 +519,14 @@ vector_negative_inner_product(PG_FUNCTION_ARGS)
/* /*
* Get the cosine distance between two vectors * Get the cosine distance between two vectors
*/ */
PG_FUNCTION_INFO_V1(cosine_distance); PGDLLEXPORT PG_FUNCTION_INFO_V1(cosine_distance);
Datum Datum
cosine_distance(PG_FUNCTION_ARGS) cosine_distance(PG_FUNCTION_ARGS)
{ {
Vector *a = PG_GETARG_VECTOR_P(0); Vector *a = PG_GETARG_VECTOR_P(0);
Vector *b = PG_GETARG_VECTOR_P(1); Vector *b = PG_GETARG_VECTOR_P(1);
float *ax = a->x;
float *bx = b->x;
double distance = 0.0; double distance = 0.0;
double norma = 0.0; double norma = 0.0;
double normb = 0.0; double normb = 0.0;
@@ -525,9 +535,9 @@ cosine_distance(PG_FUNCTION_ARGS)
for (int i = 0; i < a->dim; i++) for (int i = 0; i < a->dim; i++)
{ {
distance += a->x[i] * b->x[i]; distance += ax[i] * bx[i];
norma += a->x[i] * a->x[i]; norma += ax[i] * ax[i];
normb += b->x[i] * b->x[i]; normb += bx[i] * bx[i];
} }
PG_RETURN_FLOAT8(1 - (distance / (sqrt(norma) * sqrt(normb)))); PG_RETURN_FLOAT8(1 - (distance / (sqrt(norma) * sqrt(normb))));
@@ -538,7 +548,7 @@ cosine_distance(PG_FUNCTION_ARGS)
* Currently uses angular distance since needs to satisfy triangle inequality * Currently uses angular distance since needs to satisfy triangle inequality
* Assumes inputs are unit vectors (skips norm) * Assumes inputs are unit vectors (skips norm)
*/ */
PG_FUNCTION_INFO_V1(vector_spherical_distance); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_spherical_distance);
Datum Datum
vector_spherical_distance(PG_FUNCTION_ARGS) vector_spherical_distance(PG_FUNCTION_ARGS)
{ {
@@ -563,7 +573,7 @@ vector_spherical_distance(PG_FUNCTION_ARGS)
/* /*
* Get the dimensions of a vector * Get the dimensions of a vector
*/ */
PG_FUNCTION_INFO_V1(vector_dims); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_dims);
Datum Datum
vector_dims(PG_FUNCTION_ARGS) vector_dims(PG_FUNCTION_ARGS)
{ {
@@ -575,15 +585,16 @@ vector_dims(PG_FUNCTION_ARGS)
/* /*
* Get the L2 norm of a vector * Get the L2 norm of a vector
*/ */
PG_FUNCTION_INFO_V1(vector_norm); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_norm);
Datum Datum
vector_norm(PG_FUNCTION_ARGS) vector_norm(PG_FUNCTION_ARGS)
{ {
Vector *a = PG_GETARG_VECTOR_P(0); Vector *a = PG_GETARG_VECTOR_P(0);
float *ax = a->x;
double norm = 0.0; double norm = 0.0;
for (int i = 0; i < a->dim; i++) for (int i = 0; i < a->dim; i++)
norm += a->x[i] * a->x[i]; norm += ax[i] * ax[i];
PG_RETURN_FLOAT8(sqrt(norm)); PG_RETURN_FLOAT8(sqrt(norm));
} }
@@ -591,20 +602,23 @@ vector_norm(PG_FUNCTION_ARGS)
/* /*
* Add vectors * Add vectors
*/ */
PG_FUNCTION_INFO_V1(vector_add); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_add);
Datum Datum
vector_add(PG_FUNCTION_ARGS) vector_add(PG_FUNCTION_ARGS)
{ {
Vector *a = PG_GETARG_VECTOR_P(0); Vector *a = PG_GETARG_VECTOR_P(0);
Vector *b = PG_GETARG_VECTOR_P(1); Vector *b = PG_GETARG_VECTOR_P(1);
float *ax = a->x;
float *bx = b->x;
Vector *result; Vector *result;
int i; float *rx;
CheckDims(a, b); CheckDims(a, b);
result = InitVector(a->dim); result = InitVector(a->dim);
for (i = 0; i < a->dim; i++) rx = result->x;
result->x[i] = a->x[i] + b->x[i]; for (int i = 0, imax = a->dim; i < imax; i++)
rx[i] = ax[i] + bx[i];
PG_RETURN_POINTER(result); PG_RETURN_POINTER(result);
} }
@@ -612,20 +626,23 @@ vector_add(PG_FUNCTION_ARGS)
/* /*
* Subtract vectors * Subtract vectors
*/ */
PG_FUNCTION_INFO_V1(vector_sub); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_sub);
Datum Datum
vector_sub(PG_FUNCTION_ARGS) vector_sub(PG_FUNCTION_ARGS)
{ {
Vector *a = PG_GETARG_VECTOR_P(0); Vector *a = PG_GETARG_VECTOR_P(0);
Vector *b = PG_GETARG_VECTOR_P(1); Vector *b = PG_GETARG_VECTOR_P(1);
float *ax = a->x;
float *bx = b->x;
Vector *result; Vector *result;
int i; float *rx;
CheckDims(a, b); CheckDims(a, b);
result = InitVector(a->dim); result = InitVector(a->dim);
for (i = 0; i < a->dim; i++) rx = result->x;
result->x[i] = a->x[i] - b->x[i]; for (int i = 0, imax = a->dim; i < imax; i++)
rx[i] = ax[i] - bx[i];
PG_RETURN_POINTER(result); PG_RETURN_POINTER(result);
} }
@@ -654,7 +671,7 @@ vector_cmp_internal(Vector * a, Vector * b)
/* /*
* Less than * Less than
*/ */
PG_FUNCTION_INFO_V1(vector_lt); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_lt);
Datum Datum
vector_lt(PG_FUNCTION_ARGS) vector_lt(PG_FUNCTION_ARGS)
{ {
@@ -667,7 +684,7 @@ vector_lt(PG_FUNCTION_ARGS)
/* /*
* Less than or equal * Less than or equal
*/ */
PG_FUNCTION_INFO_V1(vector_le); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_le);
Datum Datum
vector_le(PG_FUNCTION_ARGS) vector_le(PG_FUNCTION_ARGS)
{ {
@@ -680,7 +697,7 @@ vector_le(PG_FUNCTION_ARGS)
/* /*
* Equal * Equal
*/ */
PG_FUNCTION_INFO_V1(vector_eq); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_eq);
Datum Datum
vector_eq(PG_FUNCTION_ARGS) vector_eq(PG_FUNCTION_ARGS)
{ {
@@ -693,7 +710,7 @@ vector_eq(PG_FUNCTION_ARGS)
/* /*
* Not equal * Not equal
*/ */
PG_FUNCTION_INFO_V1(vector_ne); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_ne);
Datum Datum
vector_ne(PG_FUNCTION_ARGS) vector_ne(PG_FUNCTION_ARGS)
{ {
@@ -706,7 +723,7 @@ vector_ne(PG_FUNCTION_ARGS)
/* /*
* Greater than or equal * Greater than or equal
*/ */
PG_FUNCTION_INFO_V1(vector_ge); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_ge);
Datum Datum
vector_ge(PG_FUNCTION_ARGS) vector_ge(PG_FUNCTION_ARGS)
{ {
@@ -719,7 +736,7 @@ vector_ge(PG_FUNCTION_ARGS)
/* /*
* Greater than * Greater than
*/ */
PG_FUNCTION_INFO_V1(vector_gt); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_gt);
Datum Datum
vector_gt(PG_FUNCTION_ARGS) vector_gt(PG_FUNCTION_ARGS)
{ {
@@ -732,7 +749,7 @@ vector_gt(PG_FUNCTION_ARGS)
/* /*
* Compare vectors * Compare vectors
*/ */
PG_FUNCTION_INFO_V1(vector_cmp); PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_cmp);
Datum Datum
vector_cmp(PG_FUNCTION_ARGS) vector_cmp(PG_FUNCTION_ARGS)
{ {