postgresql/src/backend/access/brin/brin_minmax.c

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

317 lines
9.3 KiB
C
Raw Normal View History

BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
/*
* brin_minmax.c
* Implementation of Min/Max opclass for BRIN
*
* Portions Copyright (c) 1996-2023, PostgreSQL Global Development Group
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
* Portions Copyright (c) 1994, Regents of the University of California
*
* IDENTIFICATION
* src/backend/access/brin/brin_minmax.c
*/
#include "postgres.h"
#include "access/brin_internal.h"
#include "access/brin_tuple.h"
#include "access/genam.h"
#include "access/stratnum.h"
#include "catalog/pg_amop.h"
#include "catalog/pg_type.h"
#include "utils/builtins.h"
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
#include "utils/datum.h"
#include "utils/lsyscache.h"
#include "utils/rel.h"
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
#include "utils/syscache.h"
typedef struct MinmaxOpaque
{
Oid cached_subtype;
FmgrInfo strategy_procinfos[BTMaxStrategyNumber];
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
} MinmaxOpaque;
static FmgrInfo *minmax_get_strategy_procinfo(BrinDesc *bdesc, uint16 attno,
Oid subtype, uint16 strategynum);
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
Datum
brin_minmax_opcinfo(PG_FUNCTION_ARGS)
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
{
Oid typoid = PG_GETARG_OID(0);
BrinOpcInfo *result;
/*
* opaque->strategy_procinfos is initialized lazily; here it is set to
* all-uninitialized by palloc0 which sets fn_oid to InvalidOid.
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
*/
result = palloc0(MAXALIGN(SizeofBrinOpcInfo(2)) +
sizeof(MinmaxOpaque));
result->oi_nstored = 2;
result->oi_regular_nulls = true;
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
result->oi_opaque = (MinmaxOpaque *)
MAXALIGN((char *) result + SizeofBrinOpcInfo(2));
result->oi_typcache[0] = result->oi_typcache[1] =
lookup_type_cache(typoid, 0);
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
PG_RETURN_POINTER(result);
}
/*
* Examine the given index tuple (which contains partial status of a certain
* page range) by comparing it to the given value that comes from another heap
* tuple. If the new value is outside the min/max range specified by the
* existing tuple values, update the index tuple and return true. Otherwise,
* return false and do not modify in this case.
*/
Datum
brin_minmax_add_value(PG_FUNCTION_ARGS)
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
{
BrinDesc *bdesc = (BrinDesc *) PG_GETARG_POINTER(0);
BrinValues *column = (BrinValues *) PG_GETARG_POINTER(1);
Datum newval = PG_GETARG_DATUM(2);
bool isnull PG_USED_FOR_ASSERTS_ONLY = PG_GETARG_DATUM(3);
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
Oid colloid = PG_GET_COLLATION();
FmgrInfo *cmpFn;
Datum compar;
bool updated = false;
Form_pg_attribute attr;
AttrNumber attno;
Assert(!isnull);
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
attno = column->bv_attno;
attr = TupleDescAttr(bdesc->bd_tupdesc, attno - 1);
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
/*
* If the recorded value is null, store the new value (which we know to be
* not null) as both minimum and maximum, and we're done.
*/
if (column->bv_allnulls)
{
column->bv_values[0] = datumCopy(newval, attr->attbyval, attr->attlen);
column->bv_values[1] = datumCopy(newval, attr->attbyval, attr->attlen);
column->bv_allnulls = false;
PG_RETURN_BOOL(true);
}
/*
* Otherwise, need to compare the new value with the existing boundaries
* and update them accordingly. First check if it's less than the
* existing minimum.
*/
cmpFn = minmax_get_strategy_procinfo(bdesc, attno, attr->atttypid,
BTLessStrategyNumber);
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
compar = FunctionCall2Coll(cmpFn, colloid, newval, column->bv_values[0]);
if (DatumGetBool(compar))
{
if (!attr->attbyval)
pfree(DatumGetPointer(column->bv_values[0]));
column->bv_values[0] = datumCopy(newval, attr->attbyval, attr->attlen);
updated = true;
}
/*
* And now compare it to the existing maximum.
*/
cmpFn = minmax_get_strategy_procinfo(bdesc, attno, attr->atttypid,
BTGreaterStrategyNumber);
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
compar = FunctionCall2Coll(cmpFn, colloid, newval, column->bv_values[1]);
if (DatumGetBool(compar))
{
if (!attr->attbyval)
pfree(DatumGetPointer(column->bv_values[1]));
column->bv_values[1] = datumCopy(newval, attr->attbyval, attr->attlen);
updated = true;
}
PG_RETURN_BOOL(updated);
}
/*
* Given an index tuple corresponding to a certain page range and a scan key,
* return whether the scan key is consistent with the index tuple's min/max
* values. Return true if so, false otherwise.
Pass all scan keys to BRIN consistent function at once This commit changes how we pass scan keys to BRIN consistent function. Instead of passing them one by one, we now pass all scan keys for a given attribute at once. That makes the consistent function a bit more complex, as it has to loop through the keys, but it does allow more elaborate opclasses that can use multiple keys to eliminate ranges much more effectively. The existing BRIN opclasses (minmax, inclusion) don't really benefit from this change. The primary purpose is to allow future opclases to benefit from seeing all keys at once. This does change the BRIN API, because the signature of the consistent function changes (a new parameter with number of scan keys). So this breaks existing opclasses, and will require supporting two variants of the code for different PostgreSQL versions. We've considered supporting two variants of the consistent, but we've decided not to do that. Firstly, there's another patch that moves handling of NULL values from the opclass, which means the opclasses need to be updated anyway. Secondly, we're not aware of any out-of-core BRIN opclasses, so it does not seem worth the extra complexity. Bump catversion, because of pg_proc changes. Author: Tomas Vondra <tomas.vondra@postgresql.org> Reviewed-by: Alvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Mark Dilger <hornschnorter@gmail.com> Reviewed-by: Alexander Korotkov <aekorotkov@gmail.com> Reviewed-by: John Naylor <john.naylor@enterprisedb.com> Reviewed-by: Nikita Glukhov <n.gluhov@postgrespro.ru> Discussion: https://postgr.es/m/c1138ead-7668-f0e1-0638-c3be3237e812@2ndquadrant.com
2021-03-23 00:12:19 +01:00
*
Support the old signature of BRIN consistent function Commit a1c649d889 changed the signature of the BRIN consistent function by adding a new required parameter. Treating the parameter as optional, which would make the change backwards incompatibile, was rejected with the justification that there are few out-of-core extensions, so it's not worth adding making the code more complex, and it's better to deal with that in the extension. But after further thought, that would be rather problematic, because pg_upgrade simply dumps catalog contents and the same version of an extension needs to work on both PostgreSQL versions. Supporting both variants of the consistent function (with 3 or 4 arguments) makes that possible. The signature is not the only thing that changed, as commit 72ccf55cb9 moved handling of IS [NOT] NULL keys from the support procedures. But this change is backward compatible - handling the keys in exension is unnecessary, but harmless. The consistent function will do a bit of unnecessary work, but it should be very cheap. This also undoes most of the changes to the existing opclasses (minmax and inclusion), making them use the old signature again. This should make backpatching simpler. Catversion bump, because of changes in pg_amproc. Author: Tomas Vondra <tomas.vondra@postgresql.org> Author: Nikita Glukhov <n.gluhov@postgrespro.ru> Reviewed-by: Mark Dilger <hornschnorter@gmail.com> Reviewed-by: Alexander Korotkov <aekorotkov@gmail.com> Reviewed-by: Masahiko Sawada <masahiko.sawada@enterprisedb.com> Reviewed-by: John Naylor <john.naylor@enterprisedb.com> Discussion: https://postgr.es/m/c1138ead-7668-f0e1-0638-c3be3237e812@2ndquadrant.com
2021-03-26 13:17:56 +01:00
* We're no longer dealing with NULL keys in the consistent function, that is
* now handled by the AM code. That means we should not get any all-NULL ranges
* either, because those can't be consistent with regular (not [IS] NULL) keys.
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
*/
Datum
brin_minmax_consistent(PG_FUNCTION_ARGS)
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
{
BrinDesc *bdesc = (BrinDesc *) PG_GETARG_POINTER(0);
BrinValues *column = (BrinValues *) PG_GETARG_POINTER(1);
Support the old signature of BRIN consistent function Commit a1c649d889 changed the signature of the BRIN consistent function by adding a new required parameter. Treating the parameter as optional, which would make the change backwards incompatibile, was rejected with the justification that there are few out-of-core extensions, so it's not worth adding making the code more complex, and it's better to deal with that in the extension. But after further thought, that would be rather problematic, because pg_upgrade simply dumps catalog contents and the same version of an extension needs to work on both PostgreSQL versions. Supporting both variants of the consistent function (with 3 or 4 arguments) makes that possible. The signature is not the only thing that changed, as commit 72ccf55cb9 moved handling of IS [NOT] NULL keys from the support procedures. But this change is backward compatible - handling the keys in exension is unnecessary, but harmless. The consistent function will do a bit of unnecessary work, but it should be very cheap. This also undoes most of the changes to the existing opclasses (minmax and inclusion), making them use the old signature again. This should make backpatching simpler. Catversion bump, because of changes in pg_amproc. Author: Tomas Vondra <tomas.vondra@postgresql.org> Author: Nikita Glukhov <n.gluhov@postgrespro.ru> Reviewed-by: Mark Dilger <hornschnorter@gmail.com> Reviewed-by: Alexander Korotkov <aekorotkov@gmail.com> Reviewed-by: Masahiko Sawada <masahiko.sawada@enterprisedb.com> Reviewed-by: John Naylor <john.naylor@enterprisedb.com> Discussion: https://postgr.es/m/c1138ead-7668-f0e1-0638-c3be3237e812@2ndquadrant.com
2021-03-26 13:17:56 +01:00
ScanKey key = (ScanKey) PG_GETARG_POINTER(2);
Oid colloid = PG_GET_COLLATION(),
subtype;
AttrNumber attno;
Datum value;
Datum matches;
FmgrInfo *finfo;
Pass all scan keys to BRIN consistent function at once This commit changes how we pass scan keys to BRIN consistent function. Instead of passing them one by one, we now pass all scan keys for a given attribute at once. That makes the consistent function a bit more complex, as it has to loop through the keys, but it does allow more elaborate opclasses that can use multiple keys to eliminate ranges much more effectively. The existing BRIN opclasses (minmax, inclusion) don't really benefit from this change. The primary purpose is to allow future opclases to benefit from seeing all keys at once. This does change the BRIN API, because the signature of the consistent function changes (a new parameter with number of scan keys). So this breaks existing opclasses, and will require supporting two variants of the code for different PostgreSQL versions. We've considered supporting two variants of the consistent, but we've decided not to do that. Firstly, there's another patch that moves handling of NULL values from the opclass, which means the opclasses need to be updated anyway. Secondly, we're not aware of any out-of-core BRIN opclasses, so it does not seem worth the extra complexity. Bump catversion, because of pg_proc changes. Author: Tomas Vondra <tomas.vondra@postgresql.org> Reviewed-by: Alvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Mark Dilger <hornschnorter@gmail.com> Reviewed-by: Alexander Korotkov <aekorotkov@gmail.com> Reviewed-by: John Naylor <john.naylor@enterprisedb.com> Reviewed-by: Nikita Glukhov <n.gluhov@postgrespro.ru> Discussion: https://postgr.es/m/c1138ead-7668-f0e1-0638-c3be3237e812@2ndquadrant.com
2021-03-23 00:12:19 +01:00
Support the old signature of BRIN consistent function Commit a1c649d889 changed the signature of the BRIN consistent function by adding a new required parameter. Treating the parameter as optional, which would make the change backwards incompatibile, was rejected with the justification that there are few out-of-core extensions, so it's not worth adding making the code more complex, and it's better to deal with that in the extension. But after further thought, that would be rather problematic, because pg_upgrade simply dumps catalog contents and the same version of an extension needs to work on both PostgreSQL versions. Supporting both variants of the consistent function (with 3 or 4 arguments) makes that possible. The signature is not the only thing that changed, as commit 72ccf55cb9 moved handling of IS [NOT] NULL keys from the support procedures. But this change is backward compatible - handling the keys in exension is unnecessary, but harmless. The consistent function will do a bit of unnecessary work, but it should be very cheap. This also undoes most of the changes to the existing opclasses (minmax and inclusion), making them use the old signature again. This should make backpatching simpler. Catversion bump, because of changes in pg_amproc. Author: Tomas Vondra <tomas.vondra@postgresql.org> Author: Nikita Glukhov <n.gluhov@postgrespro.ru> Reviewed-by: Mark Dilger <hornschnorter@gmail.com> Reviewed-by: Alexander Korotkov <aekorotkov@gmail.com> Reviewed-by: Masahiko Sawada <masahiko.sawada@enterprisedb.com> Reviewed-by: John Naylor <john.naylor@enterprisedb.com> Discussion: https://postgr.es/m/c1138ead-7668-f0e1-0638-c3be3237e812@2ndquadrant.com
2021-03-26 13:17:56 +01:00
/* This opclass uses the old signature with only three arguments. */
Assert(PG_NARGS() == 3);
Pass all scan keys to BRIN consistent function at once This commit changes how we pass scan keys to BRIN consistent function. Instead of passing them one by one, we now pass all scan keys for a given attribute at once. That makes the consistent function a bit more complex, as it has to loop through the keys, but it does allow more elaborate opclasses that can use multiple keys to eliminate ranges much more effectively. The existing BRIN opclasses (minmax, inclusion) don't really benefit from this change. The primary purpose is to allow future opclases to benefit from seeing all keys at once. This does change the BRIN API, because the signature of the consistent function changes (a new parameter with number of scan keys). So this breaks existing opclasses, and will require supporting two variants of the code for different PostgreSQL versions. We've considered supporting two variants of the consistent, but we've decided not to do that. Firstly, there's another patch that moves handling of NULL values from the opclass, which means the opclasses need to be updated anyway. Secondly, we're not aware of any out-of-core BRIN opclasses, so it does not seem worth the extra complexity. Bump catversion, because of pg_proc changes. Author: Tomas Vondra <tomas.vondra@postgresql.org> Reviewed-by: Alvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Mark Dilger <hornschnorter@gmail.com> Reviewed-by: Alexander Korotkov <aekorotkov@gmail.com> Reviewed-by: John Naylor <john.naylor@enterprisedb.com> Reviewed-by: Nikita Glukhov <n.gluhov@postgrespro.ru> Discussion: https://postgr.es/m/c1138ead-7668-f0e1-0638-c3be3237e812@2ndquadrant.com
2021-03-23 00:12:19 +01:00
Support the old signature of BRIN consistent function Commit a1c649d889 changed the signature of the BRIN consistent function by adding a new required parameter. Treating the parameter as optional, which would make the change backwards incompatibile, was rejected with the justification that there are few out-of-core extensions, so it's not worth adding making the code more complex, and it's better to deal with that in the extension. But after further thought, that would be rather problematic, because pg_upgrade simply dumps catalog contents and the same version of an extension needs to work on both PostgreSQL versions. Supporting both variants of the consistent function (with 3 or 4 arguments) makes that possible. The signature is not the only thing that changed, as commit 72ccf55cb9 moved handling of IS [NOT] NULL keys from the support procedures. But this change is backward compatible - handling the keys in exension is unnecessary, but harmless. The consistent function will do a bit of unnecessary work, but it should be very cheap. This also undoes most of the changes to the existing opclasses (minmax and inclusion), making them use the old signature again. This should make backpatching simpler. Catversion bump, because of changes in pg_amproc. Author: Tomas Vondra <tomas.vondra@postgresql.org> Author: Nikita Glukhov <n.gluhov@postgrespro.ru> Reviewed-by: Mark Dilger <hornschnorter@gmail.com> Reviewed-by: Alexander Korotkov <aekorotkov@gmail.com> Reviewed-by: Masahiko Sawada <masahiko.sawada@enterprisedb.com> Reviewed-by: John Naylor <john.naylor@enterprisedb.com> Discussion: https://postgr.es/m/c1138ead-7668-f0e1-0638-c3be3237e812@2ndquadrant.com
2021-03-26 13:17:56 +01:00
/* Should not be dealing with all-NULL ranges. */
Assert(!column->bv_allnulls);
Pass all scan keys to BRIN consistent function at once This commit changes how we pass scan keys to BRIN consistent function. Instead of passing them one by one, we now pass all scan keys for a given attribute at once. That makes the consistent function a bit more complex, as it has to loop through the keys, but it does allow more elaborate opclasses that can use multiple keys to eliminate ranges much more effectively. The existing BRIN opclasses (minmax, inclusion) don't really benefit from this change. The primary purpose is to allow future opclases to benefit from seeing all keys at once. This does change the BRIN API, because the signature of the consistent function changes (a new parameter with number of scan keys). So this breaks existing opclasses, and will require supporting two variants of the code for different PostgreSQL versions. We've considered supporting two variants of the consistent, but we've decided not to do that. Firstly, there's another patch that moves handling of NULL values from the opclass, which means the opclasses need to be updated anyway. Secondly, we're not aware of any out-of-core BRIN opclasses, so it does not seem worth the extra complexity. Bump catversion, because of pg_proc changes. Author: Tomas Vondra <tomas.vondra@postgresql.org> Reviewed-by: Alvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Mark Dilger <hornschnorter@gmail.com> Reviewed-by: Alexander Korotkov <aekorotkov@gmail.com> Reviewed-by: John Naylor <john.naylor@enterprisedb.com> Reviewed-by: Nikita Glukhov <n.gluhov@postgrespro.ru> Discussion: https://postgr.es/m/c1138ead-7668-f0e1-0638-c3be3237e812@2ndquadrant.com
2021-03-23 00:12:19 +01:00
Support the old signature of BRIN consistent function Commit a1c649d889 changed the signature of the BRIN consistent function by adding a new required parameter. Treating the parameter as optional, which would make the change backwards incompatibile, was rejected with the justification that there are few out-of-core extensions, so it's not worth adding making the code more complex, and it's better to deal with that in the extension. But after further thought, that would be rather problematic, because pg_upgrade simply dumps catalog contents and the same version of an extension needs to work on both PostgreSQL versions. Supporting both variants of the consistent function (with 3 or 4 arguments) makes that possible. The signature is not the only thing that changed, as commit 72ccf55cb9 moved handling of IS [NOT] NULL keys from the support procedures. But this change is backward compatible - handling the keys in exension is unnecessary, but harmless. The consistent function will do a bit of unnecessary work, but it should be very cheap. This also undoes most of the changes to the existing opclasses (minmax and inclusion), making them use the old signature again. This should make backpatching simpler. Catversion bump, because of changes in pg_amproc. Author: Tomas Vondra <tomas.vondra@postgresql.org> Author: Nikita Glukhov <n.gluhov@postgrespro.ru> Reviewed-by: Mark Dilger <hornschnorter@gmail.com> Reviewed-by: Alexander Korotkov <aekorotkov@gmail.com> Reviewed-by: Masahiko Sawada <masahiko.sawada@enterprisedb.com> Reviewed-by: John Naylor <john.naylor@enterprisedb.com> Discussion: https://postgr.es/m/c1138ead-7668-f0e1-0638-c3be3237e812@2ndquadrant.com
2021-03-26 13:17:56 +01:00
attno = key->sk_attno;
subtype = key->sk_subtype;
value = key->sk_argument;
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
switch (key->sk_strategy)
{
case BTLessStrategyNumber:
case BTLessEqualStrategyNumber:
finfo = minmax_get_strategy_procinfo(bdesc, attno, subtype,
key->sk_strategy);
matches = FunctionCall2Coll(finfo, colloid, column->bv_values[0],
value);
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
break;
case BTEqualStrategyNumber:
/*
* In the equality case (WHERE col = someval), we want to return
* the current page range if the minimum value in the range <=
* scan key, and the maximum value >= scan key.
*/
finfo = minmax_get_strategy_procinfo(bdesc, attno, subtype,
BTLessEqualStrategyNumber);
matches = FunctionCall2Coll(finfo, colloid, column->bv_values[0],
value);
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
if (!DatumGetBool(matches))
break;
/* max() >= scankey */
finfo = minmax_get_strategy_procinfo(bdesc, attno, subtype,
BTGreaterEqualStrategyNumber);
matches = FunctionCall2Coll(finfo, colloid, column->bv_values[1],
value);
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
break;
case BTGreaterEqualStrategyNumber:
case BTGreaterStrategyNumber:
finfo = minmax_get_strategy_procinfo(bdesc, attno, subtype,
key->sk_strategy);
matches = FunctionCall2Coll(finfo, colloid, column->bv_values[1],
value);
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
break;
default:
/* shouldn't happen */
elog(ERROR, "invalid strategy number %d", key->sk_strategy);
matches = 0;
break;
}
Support the old signature of BRIN consistent function Commit a1c649d889 changed the signature of the BRIN consistent function by adding a new required parameter. Treating the parameter as optional, which would make the change backwards incompatibile, was rejected with the justification that there are few out-of-core extensions, so it's not worth adding making the code more complex, and it's better to deal with that in the extension. But after further thought, that would be rather problematic, because pg_upgrade simply dumps catalog contents and the same version of an extension needs to work on both PostgreSQL versions. Supporting both variants of the consistent function (with 3 or 4 arguments) makes that possible. The signature is not the only thing that changed, as commit 72ccf55cb9 moved handling of IS [NOT] NULL keys from the support procedures. But this change is backward compatible - handling the keys in exension is unnecessary, but harmless. The consistent function will do a bit of unnecessary work, but it should be very cheap. This also undoes most of the changes to the existing opclasses (minmax and inclusion), making them use the old signature again. This should make backpatching simpler. Catversion bump, because of changes in pg_amproc. Author: Tomas Vondra <tomas.vondra@postgresql.org> Author: Nikita Glukhov <n.gluhov@postgrespro.ru> Reviewed-by: Mark Dilger <hornschnorter@gmail.com> Reviewed-by: Alexander Korotkov <aekorotkov@gmail.com> Reviewed-by: Masahiko Sawada <masahiko.sawada@enterprisedb.com> Reviewed-by: John Naylor <john.naylor@enterprisedb.com> Discussion: https://postgr.es/m/c1138ead-7668-f0e1-0638-c3be3237e812@2ndquadrant.com
2021-03-26 13:17:56 +01:00
PG_RETURN_DATUM(matches);
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
}
/*
* Given two BrinValues, update the first of them as a union of the summary
* values contained in both. The second one is untouched.
*/
Datum
brin_minmax_union(PG_FUNCTION_ARGS)
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
{
BrinDesc *bdesc = (BrinDesc *) PG_GETARG_POINTER(0);
BrinValues *col_a = (BrinValues *) PG_GETARG_POINTER(1);
BrinValues *col_b = (BrinValues *) PG_GETARG_POINTER(2);
Oid colloid = PG_GET_COLLATION();
AttrNumber attno;
Form_pg_attribute attr;
FmgrInfo *finfo;
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
bool needsadj;
Assert(col_a->bv_attno == col_b->bv_attno);
Assert(!col_a->bv_allnulls && !col_b->bv_allnulls);
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
attno = col_a->bv_attno;
attr = TupleDescAttr(bdesc->bd_tupdesc, attno - 1);
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
/* Adjust minimum, if B's min is less than A's min */
finfo = minmax_get_strategy_procinfo(bdesc, attno, attr->atttypid,
BTLessStrategyNumber);
needsadj = FunctionCall2Coll(finfo, colloid, col_b->bv_values[0],
col_a->bv_values[0]);
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
if (needsadj)
{
if (!attr->attbyval)
pfree(DatumGetPointer(col_a->bv_values[0]));
col_a->bv_values[0] = datumCopy(col_b->bv_values[0],
attr->attbyval, attr->attlen);
}
/* Adjust maximum, if B's max is greater than A's max */
finfo = minmax_get_strategy_procinfo(bdesc, attno, attr->atttypid,
BTGreaterStrategyNumber);
needsadj = FunctionCall2Coll(finfo, colloid, col_b->bv_values[1],
col_a->bv_values[1]);
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
if (needsadj)
{
if (!attr->attbyval)
pfree(DatumGetPointer(col_a->bv_values[1]));
col_a->bv_values[1] = datumCopy(col_b->bv_values[1],
attr->attbyval, attr->attlen);
}
PG_RETURN_VOID();
}
/*
* Cache and return the procedure for the given strategy.
*
* Note: this function mirrors inclusion_get_strategy_procinfo; see notes
* there. If changes are made here, see that function too.
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
*/
static FmgrInfo *
minmax_get_strategy_procinfo(BrinDesc *bdesc, uint16 attno, Oid subtype,
uint16 strategynum)
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
{
MinmaxOpaque *opaque;
Assert(strategynum >= 1 &&
strategynum <= BTMaxStrategyNumber);
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
opaque = (MinmaxOpaque *) bdesc->bd_info[attno - 1]->oi_opaque;
/*
* We cache the procedures for the previous subtype in the opaque struct,
* to avoid repetitive syscache lookups. If the subtype changed,
* invalidate all the cached entries.
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
*/
if (opaque->cached_subtype != subtype)
{
uint16 i;
for (i = 1; i <= BTMaxStrategyNumber; i++)
opaque->strategy_procinfos[i - 1].fn_oid = InvalidOid;
opaque->cached_subtype = subtype;
}
if (opaque->strategy_procinfos[strategynum - 1].fn_oid == InvalidOid)
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
{
Form_pg_attribute attr;
HeapTuple tuple;
Oid opfamily,
oprid;
opfamily = bdesc->bd_index->rd_opfamily[attno - 1];
attr = TupleDescAttr(bdesc->bd_tupdesc, attno - 1);
tuple = SearchSysCache4(AMOPSTRATEGY, ObjectIdGetDatum(opfamily),
ObjectIdGetDatum(attr->atttypid),
ObjectIdGetDatum(subtype),
Int16GetDatum(strategynum));
if (!HeapTupleIsValid(tuple))
elog(ERROR, "missing operator %d(%u,%u) in opfamily %u",
strategynum, attr->atttypid, subtype, opfamily);
oprid = DatumGetObjectId(SysCacheGetAttrNotNull(AMOPSTRATEGY, tuple,
Anum_pg_amop_amopopr));
ReleaseSysCache(tuple);
Assert(RegProcedureIsValid(oprid));
fmgr_info_cxt(get_opcode(oprid),
&opaque->strategy_procinfos[strategynum - 1],
bdesc->bd_context);
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
}
return &opaque->strategy_procinfos[strategynum - 1];
BRIN: Block Range Indexes BRIN is a new index access method intended to accelerate scans of very large tables, without the maintenance overhead of btrees or other traditional indexes. They work by maintaining "summary" data about block ranges. Bitmap index scans work by reading each summary tuple and comparing them with the query quals; all pages in the range are returned in a lossy TID bitmap if the quals are consistent with the values in the summary tuple, otherwise not. Normal index scans are not supported because these indexes do not store TIDs. As new tuples are added into the index, the summary information is updated (if the block range in which the tuple is added is already summarized) or not; in the latter case, a subsequent pass of VACUUM or the brin_summarize_new_values() function will create the summary information. For data types with natural 1-D sort orders, the summary info consists of the maximum and the minimum values of each indexed column within each page range. This type of operator class we call "Minmax", and we supply a bunch of them for most data types with B-tree opclasses. Since the BRIN code is generalized, other approaches are possible for things such as arrays, geometric types, ranges, etc; even for things such as enum types we could do something different than minmax with better results. In this commit I only include minmax. Catalog version bumped due to new builtin catalog entries. There's more that could be done here, but this is a good step forwards. Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera, with contribution by Heikki Linnakangas. Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas. Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo. PS: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 318633.
2014-11-07 20:38:14 +01:00
}