Function ExecHashTableInsert

src/backend/executor/nodeHash.c:1876–1971

* ExecHashTableInsert * insert a tuple into the hash table depending on the hash value * it may just go to a temp file for later batches * * Note: the passed TupleTableSlot may contain a regular, minimal, or virtual * tuple; the minimal case in particular is certain to happen while reloading * tuples from batch files. We could save some cycles in the regular-tuple * case by not forcing t

 * was pushed to a temp file.
 */
bool
ExecHashTableInsert(HashState *hashState, HashJoinTable hashtable,
					TupleTableSlot *slot,
					uint32 hashvalue)
{
	bool		shouldFree;
	MinimalTuple tuple = ExecFetchSlotMinimalTuple(slot, &shouldFree);
	int			bucketno;
	int			batchno;
	PlanState *ps = &hashState->ps;

	ExecHashGetBucketAndBatch(hashtable, hashvalue,
							  &bucketno, &batchno);

	/*
	 * decide whether to put the tuple in the hash table or a temp file
	 */
	if (batchno == hashtable->curbatch)
	{
		/*
		 * put the tuple in hash table
		 */
		HashJoinTuple hashTuple;
		int			hashTupleSize;
		double		ntuples = (hashtable->totalTuples - hashtable->skewTuples);

		/* Create the HashJoinTuple */
		hashTupleSize = HJTUPLE_OVERHEAD + tuple->t_len;
		hashTuple = (HashJoinTuple) dense_alloc(hashtable, hashTupleSize);

		hashTuple->hashvalue = hashvalue;
		memcpy(HJTUPLE_MINTUPLE(hashTuple), tuple, tuple->t_len);

		/*
		 * We always reset the tuple-matched flag on insertion.  This is okay
		 * even when reloading a tuple from a batch file, since the tuple
		 * could not possibly have been matched to an outer tuple before it
		 * went into the batch file.
		 */
		HeapTupleHeaderClearMatch(HJTUPLE_MINTUPLE(hashTuple));

		/* Push it onto the front of the bucket's list */
		hashTuple->next.unshared = hashtable->buckets.unshared[bucketno];
		hashtable->buckets.unshared[bucketno] = hashTuple;

		/*
		 * Increase the (optimal) number of buckets if we just exceeded the
		 * NTUP_PER_BUCKET threshold, but only when there's still a single
		 * batch.
		 */
		if (hashtable->nbatch == 1 &&
			ntuples > (hashtable->nbuckets_optimal * gp_hashjoin_tuples_per_bucket))
		{
			/* Guard against integer overflow and alloc size overflow */
			if (hashtable->nbuckets_optimal <= INT_MAX / 2 &&
				hashtable->nbuckets_optimal * 2 <= MaxAllocSize / sizeof(HashJoinTuple))
			{
				hashtable->nbuckets_optimal *= 2;