database/postgres
1
0
Fork 0
mirror of https://github.com/postgres/postgres.git synced 2025-07-03 20:02:46 +03:00
Code Activity

Revise hash join code so that we can increase the number of batches

Browse Source 
on-the-fly, and thereby avoid blowing out memory when the planner has
underestimated the hash table size.  Hash join will now obey the
work_mem limit with some faithfulness.  Per my recent proposal
(hash aggregate part isn't done yet though).
This commit is contained in:
Tom Lane
2005-03-06 22:15:05 +00:00
parent 31b6d840f6
commit 849074f9ae
9 changed files with 654 additions and 450 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
src/backend/utils/adt/selfuncs.c
View File

@ -15,7 +15,7 @@
*
*
* IDENTIFICATION
* $PostgreSQL: pgsql/src/backend/utils/adt/selfuncs.c,v 1.171 2005/02/01 23:07:58 tgl Exp $
* $PostgreSQL: pgsql/src/backend/utils/adt/selfuncs.c,v 1.172 2005/03/06 22:15:04 tgl Exp $
*
*-------------------------------------------------------------------------
*/
@ -2154,7 +2154,7 @@ estimate_num_groups(Query *root, List *groupExprs, double input_rows)
* inner rel is well-dispersed (or the alternatives seem much worse).
*/
Selectivity
estimate_hash_bucketsize(Query *root, Node *hashkey, int nbuckets)
estimate_hash_bucketsize(Query *root, Node *hashkey, double nbuckets)
{
VariableStatData vardata;
double estfract,
@ -2212,8 +2212,8 @@ estimate_hash_bucketsize(Query *root, Node *hashkey, int nbuckets)
* the number of buckets is less than the expected number of distinct
* values; otherwise it is 1/ndistinct.
*/
if (ndistinct > (double) nbuckets)
estfract = 1.0 / (double) nbuckets;
if (ndistinct > nbuckets)
estfract = 1.0 / nbuckets;
else
estfract = 1.0 / ndistinct;