pg.ddx.io pgsql-performance@postgresql.org mailing list archive help / color / mirror / Atom feed From: Tomas Vondra <tomas@2ndquadrant.com>
Subject: [PATCH 1/2] move BufFile stuff into separate context
Date: Sat, 20 Apr 2019 21:44:37 +0200
---
src/backend/executor/nodeHash.c | 48 +++++++++++++++++++++++++++++
src/backend/executor/nodeHashjoin.c | 7 +++++
src/include/executor/hashjoin.h | 1 +
3 files changed, 56 insertions(+)
diff --git a/src/backend/executor/nodeHash.c b/src/backend/executor/nodeHash.c
index 6ffaa751f2..7d36c22d61 100644
--- a/src/backend/executor/nodeHash.c
+++ b/src/backend/executor/nodeHash.c
@@ -498,6 +498,7 @@ ExecHashTableCreate(HashState *state, List *hashOperators, bool keepNulls)
hashtable->skewTuples = 0;
hashtable->innerBatchFile = NULL;
hashtable->outerBatchFile = NULL;
+ hashtable->fileCtx = NULL;
hashtable->spaceUsed = 0;
hashtable->spacePeak = 0;
hashtable->spaceAllowed = space_allowed;
@@ -900,6 +901,11 @@ ExecHashIncreaseNumBatches(HashJoinTable hashtable)
nbatch = oldnbatch * 2;
Assert(nbatch > 1);
+ elog(WARNING, "ExecHashIncreaseNumBatches: increasing number of batches from %d to %d", oldnbatch, nbatch);
+
+ elog(LOG, "ExecHashIncreaseNumBatches ======= context stats start =======");
+ MemoryContextStats(TopMemoryContext);
+
#ifdef HJDEBUG
printf("Hashjoin %p: increasing nbatch to %d because space = %zu\n",
hashtable, nbatch, hashtable->spaceUsed);
@@ -909,6 +915,14 @@ ExecHashIncreaseNumBatches(HashJoinTable hashtable)
if (hashtable->innerBatchFile == NULL)
{
+ MemoryContext oldctx;
+
+ hashtable->fileCtx = AllocSetContextCreate(CurrentMemoryContext,
+ "hash batch files",
+ ALLOCSET_DEFAULT_SIZES);
+
+ oldctx = MemoryContextSwitchTo(hashtable->fileCtx);
+
/* we had no file arrays before */
hashtable->innerBatchFile = (BufFile **)
palloc0(nbatch * sizeof(BufFile *));
@@ -916,9 +930,15 @@ ExecHashIncreaseNumBatches(HashJoinTable hashtable)
palloc0(nbatch * sizeof(BufFile *));
/* time to establish the temp tablespaces, too */
PrepareTempTablespaces();
+
+ MemoryContextSwitchTo(oldctx);
}
else
{
+ MemoryContext oldctx;
+
+ oldctx = MemoryContextSwitchTo(hashtable->fileCtx);
+
/* enlarge arrays and zero out added entries */
hashtable->innerBatchFile = (BufFile **)
repalloc(hashtable->innerBatchFile, nbatch * sizeof(BufFile *));
@@ -928,6 +948,9 @@ ExecHashIncreaseNumBatches(HashJoinTable hashtable)
(nbatch - oldnbatch) * sizeof(BufFile *));
MemSet(hashtable->outerBatchFile + oldnbatch, 0,
(nbatch - oldnbatch) * sizeof(BufFile *));
+
+ MemoryContextSwitchTo(oldctx);
+
}
MemoryContextSwitchTo(oldcxt);
@@ -999,12 +1022,19 @@ ExecHashIncreaseNumBatches(HashJoinTable hashtable)
}
else
{
+ MemoryContext oldctx;
+
/* dump it out */
Assert(batchno > curbatch);
+
+ oldctx = MemoryContextSwitchTo(hashtable->fileCtx);
+
ExecHashJoinSaveTuple(HJTUPLE_MINTUPLE(hashTuple),
hashTuple->hashvalue,
&hashtable->innerBatchFile[batchno]);
+ MemoryContextSwitchTo(oldctx);
+
hashtable->spaceUsed -= hashTupleSize;
nfreed++;
}
@@ -1042,6 +1072,9 @@ ExecHashIncreaseNumBatches(HashJoinTable hashtable)
hashtable);
#endif
}
+
+ elog(LOG, "ExecHashIncreaseNumBatches ======= context stats end =======");
+ MemoryContextStats(TopMemoryContext);
}
/*
@@ -1656,13 +1689,20 @@ ExecHashTableInsert(HashJoinTable hashtable,
}
else
{
+ MemoryContext oldctx;
+
/*
* put the tuple into a temp file for later batches
*/
Assert(batchno > hashtable->curbatch);
+
+ oldctx = MemoryContextSwitchTo(hashtable->fileCtx);
+
ExecHashJoinSaveTuple(tuple,
hashvalue,
&hashtable->innerBatchFile[batchno]);
+
+ MemoryContextSwitchTo(oldctx);
}
}
@@ -2488,10 +2528,18 @@ ExecHashRemoveNextSkewBucket(HashJoinTable hashtable)
}
else
{
+ MemoryContext oldctx;
+
/* Put the tuple into a temp file for later batches */
Assert(batchno > hashtable->curbatch);
+
+ oldctx = MemoryContextSwitchTo(hashtable->fileCtx);
+
ExecHashJoinSaveTuple(tuple, hashvalue,
&hashtable->innerBatchFile[batchno]);
+
+ MemoryContextSwitchTo(oldctx);
+
pfree(hashTuple);
hashtable->spaceUsed -= tupleSize;
hashtable->spaceUsedSkew -= tupleSize;
diff --git a/src/backend/executor/nodeHashjoin.c b/src/backend/executor/nodeHashjoin.c
index 5922e60eed..6a546344ac 100644
--- a/src/backend/executor/nodeHashjoin.c
+++ b/src/backend/executor/nodeHashjoin.c
@@ -389,16 +389,23 @@ ExecHashJoinImpl(PlanState *pstate, bool parallel)
if (batchno != hashtable->curbatch &&
node->hj_CurSkewBucketNo == INVALID_SKEW_BUCKET_NO)
{
+ MemoryContext oldctx;
+
/*
* Need to postpone this outer tuple to a later batch.
* Save it in the corresponding outer-batch file.
*/
Assert(parallel_state == NULL);
Assert(batchno > hashtable->curbatch);
+
+ oldctx = MemoryContextSwitchTo(hashtable->fileCtx);
+
ExecHashJoinSaveTuple(ExecFetchSlotMinimalTuple(outerTupleSlot),
hashvalue,
&hashtable->outerBatchFile[batchno]);
+ MemoryContextSwitchTo(oldctx);
+
/* Loop around, staying in HJ_NEED_NEW_OUTER state */
continue;
}
diff --git a/src/include/executor/hashjoin.h b/src/include/executor/hashjoin.h
index a9f9872a78..ef8e4475e8 100644
--- a/src/include/executor/hashjoin.h
+++ b/src/include/executor/hashjoin.h
@@ -328,6 +328,7 @@ typedef struct HashJoinTableData
*/
BufFile **innerBatchFile; /* buffered virtual temp file per batch */
BufFile **outerBatchFile; /* buffered virtual temp file per batch */
+ MemoryContext fileCtx;
/*
* Info about the datatype-specific hash functions for the datatypes being
--
2.20.1
--73g6wg24zkedq45b
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="0002-allocate-BufFile-eagerly.patch"
Message-ID: <no-message-id-5651@localhost>
Permalink: ../no-message-id-5651@localhost/
Also on: postgresql.org/message-id/no-message-id-5651@localhost
copy link · copy postgr.es
reply Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Reply to all the recipients using the --to and --cc options:
reply via email
To: pgsql-performance@postgresql.org
Cc: tomas@2ndquadrant.com
Subject: Re: [PATCH 1/2] move BufFile stuff into separate context
In-Reply-To: <no-message-id-5651@localhost>
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox