database/postgres
1
0
Fork 0
mirror of https://github.com/postgres/postgres.git synced 2025-08-19 23:22:23 +03:00
Code Activity

Remove 'additional' pointer from TupleHashEntryData.

Browse Source 
Reduces memory required for hash aggregation by avoiding an allocation
and a pointer in the TupleHashEntryData structure. That structure is
used for all buckets, whether occupied or not, so the savings is
substantial.

Discussion: https://postgr.es/m/AApHDvpN4v3t_sdz4dvrv1Fx_ZPw=twSnxuTEytRYP7LFz5K9A@mail.gmail.com
Reviewed-by: David Rowley <dgrowleyml@gmail.com>
This commit is contained in:
Jeff Davis
2025-03-24 22:06:02 -07:00
parent a0942f441e
commit 626df47ad9
3 changed files with 16 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

17
src/backend/executor/execGrouping.c
View File

@@ -485,11 +485,18 @@ LookupTupleHashEntry_internal(TupleHashTable hashtable, TupleTableSlot *slot,
MemoryContextSwitchTo(hashtable->tablecxt);
entry->firstTuple = ExecCopySlotMinimalTuple(slot);
if (hashtable->additionalsize > 0)
entry->additional = palloc0(hashtable->additionalsize);
else
entry->additional = NULL;
/*
* Copy the first tuple into the table context, and request
* additionalsize extra bytes before the allocation.
*
* The caller can get a pointer to the additional data with
* TupleHashEntryGetAdditional(), and store arbitrary data there.
* Placing both the tuple and additional data in the same
* allocation avoids the need to store an extra pointer in
* TupleHashEntryData or allocate an additional chunk.
*/
entry->firstTuple = ExecCopySlotMinimalTupleExtra(slot,
hashtable->additionalsize);
}
}
else