postgresql/src/backend/executor/nodeCtescan.c

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

340 lines
9.6 KiB
C
Raw Normal View History

/*-------------------------------------------------------------------------
*
* nodeCtescan.c
* routines to handle CteScan nodes.
*
* Portions Copyright (c) 1996-2026, PostgreSQL Global Development Group
* Portions Copyright (c) 1994, Regents of the University of California
*
*
* IDENTIFICATION
2010-09-20 16:08:53 -04:00
* src/backend/executor/nodeCtescan.c
*
*-------------------------------------------------------------------------
*/
#include "postgres.h"
#include "executor/executor.h"
#include "executor/nodeCtescan.h"
#include "miscadmin.h"
static TupleTableSlot *CteScanNext(CteScanState *node);
/* ----------------------------------------------------------------
* CteScanNext
*
* This is a workhorse for ExecCteScan
* ----------------------------------------------------------------
*/
static TupleTableSlot *
CteScanNext(CteScanState *node)
{
EState *estate;
ScanDirection dir;
bool forward;
Tuplestorestate *tuplestorestate;
bool eof_tuplestore;
TupleTableSlot *slot;
/*
* get state info from node
*/
estate = node->ss.ps.state;
dir = estate->es_direction;
forward = ScanDirectionIsForward(dir);
tuplestorestate = node->leader->cte_table;
tuplestore_select_read_pointer(tuplestorestate, node->readptr);
slot = node->ss.ss_ScanTupleSlot;
/*
* If we are not at the end of the tuplestore, or are going backwards, try
* to fetch a tuple from tuplestore.
*/
eof_tuplestore = tuplestore_ateof(tuplestorestate);
if (!forward && eof_tuplestore)
{
if (!node->leader->eof_cte)
{
/*
* When reversing direction at tuplestore EOF, the first
* gettupleslot call will fetch the last-added tuple; but we want
* to return the one before that, if possible. So do an extra
* fetch.
*/
if (!tuplestore_advance(tuplestorestate, forward))
return NULL; /* the tuplestore must be empty */
}
eof_tuplestore = false;
}
/*
* If we can fetch another tuple from the tuplestore, return it.
*
* Note: we have to use copy=true in the tuplestore_gettupleslot call,
* because we are sharing the tuplestore with other nodes that might write
* into the tuplestore before we get called again.
*/
if (!eof_tuplestore)
{
if (tuplestore_gettupleslot(tuplestorestate, forward, true, slot))
return slot;
if (forward)
eof_tuplestore = true;
}
/*
* If necessary, try to fetch another row from the CTE query.
*
* Note: the eof_cte state variable exists to short-circuit further calls
* of the CTE plan. It's not optional, unfortunately, because some plan
* node types are not robust about being called again when they've already
* returned NULL.
*/
if (eof_tuplestore && !node->leader->eof_cte)
{
TupleTableSlot *cteslot;
/*
* We can only get here with forward==true, so no need to worry about
* which direction the subplan will go.
*/
cteslot = ExecProcNode(node->cteplanstate);
if (TupIsNull(cteslot))
{
node->leader->eof_cte = true;
return NULL;
}
/*
* There are corner cases where the subplan could change which
* tuplestore read pointer is active, so be sure to reselect ours
* before storing the tuple we got.
*/
tuplestore_select_read_pointer(tuplestorestate, node->readptr);
/*
* Append a copy of the returned tuple to tuplestore. NOTE: because
* our read pointer is certainly in EOF state, its read position will
* move forward over the added tuple. This is what we want. Also,
* any other readers will *not* move past the new tuple, which is what
* they want.
*/
tuplestore_puttupleslot(tuplestorestate, cteslot);
/*
* We MUST copy the CTE query's output tuple into our own slot. This
* is because other CteScan nodes might advance the CTE query before
* we are called again, and our output tuple must stay stable over
* that.
*/
return ExecCopySlot(slot, cteslot);
}
/*
* Nothing left ...
*/
return ExecClearTuple(slot);
}
/*
* CteScanRecheck -- access method routine to recheck a tuple in EvalPlanQual
*/
static bool
CteScanRecheck(CteScanState *node, TupleTableSlot *slot)
{
/* nothing to check */
return true;
}
/* ----------------------------------------------------------------
* ExecCteScan(node)
*
* Scans the CTE sequentially and returns the next qualifying tuple.
* We call the ExecScan() routine and pass it the appropriate
* access method functions.
* ----------------------------------------------------------------
*/
static TupleTableSlot *
ExecCteScan(PlanState *pstate)
{
CteScanState *node = castNode(CteScanState, pstate);
return ExecScan(&node->ss,
(ExecScanAccessMtd) CteScanNext,
(ExecScanRecheckMtd) CteScanRecheck);
}
/* ----------------------------------------------------------------
* ExecInitCteScan
* ----------------------------------------------------------------
*/
CteScanState *
ExecInitCteScan(CteScan *node, EState *estate, int eflags)
{
CteScanState *scanstate;
ParamExecData *prmdata;
/* check for unsupported flags */
Assert(!(eflags & EXEC_FLAG_MARK));
/*
* For the moment we have to force the tuplestore to allow REWIND, because
* we might be asked to rescan the CTE even though upper levels didn't
* tell us to be prepared to do it efficiently. Annoying, since this
* prevents truncation of the tuplestore. XXX FIXME
*
* Note: if we are in an EPQ recheck plan tree, it's likely that no access
* to the tuplestore is needed at all, making this even more annoying.
* It's not worth improving that as long as all the read pointers would
* have REWIND anyway, but if we ever improve this logic then that aspect
* should be considered too.
*/
eflags |= EXEC_FLAG_REWIND;
/*
* CteScan should not have any children.
*/
Assert(outerPlan(node) == NULL);
Assert(innerPlan(node) == NULL);
/*
* create new CteScanState for node
*/
scanstate = makeNode(CteScanState);
scanstate->ss.ps.plan = (Plan *) node;
scanstate->ss.ps.state = estate;
scanstate->ss.ps.ExecProcNode = ExecCteScan;
scanstate->eflags = eflags;
scanstate->cte_table = NULL;
scanstate->eof_cte = false;
/*
* Find the already-initialized plan for the CTE query.
*/
scanstate->cteplanstate = (PlanState *) list_nth(estate->es_subplanstates,
node->ctePlanId - 1);
/*
* The Param slot associated with the CTE query is used to hold a pointer
* to the CteState of the first CteScan node that initializes for this
* CTE. This node will be the one that holds the shared state for all the
* CTEs, particularly the shared tuplestore.
*/
prmdata = &(estate->es_param_exec_vals[node->cteParam]);
Assert(prmdata->execPlan == NULL);
Assert(!prmdata->isnull);
scanstate->leader = castNode(CteScanState, DatumGetPointer(prmdata->value));
if (scanstate->leader == NULL)
{
/* I am the leader */
prmdata->value = PointerGetDatum(scanstate);
scanstate->leader = scanstate;
scanstate->cte_table = tuplestore_begin_heap(true, false, work_mem);
tuplestore_set_eflags(scanstate->cte_table, scanstate->eflags);
scanstate->readptr = 0;
}
else
{
/* Not the leader */
/* Create my own read pointer, and ensure it is at start */
scanstate->readptr =
tuplestore_alloc_read_pointer(scanstate->leader->cte_table,
scanstate->eflags);
tuplestore_select_read_pointer(scanstate->leader->cte_table,
scanstate->readptr);
tuplestore_rescan(scanstate->leader->cte_table);
}
/*
* Miscellaneous initialization
*
* create expression context for node
*/
ExecAssignExprContext(estate, &scanstate->ss.ps);
/*
* The scan tuple type (ie, the rowtype we expect to find in the work
* table) is the same as the result rowtype of the CTE query.
*/
ExecInitScanTupleSlot(estate, &scanstate->ss,
Introduce notion of different types of slots (without implementing them). Upcoming work intends to allow pluggable ways to introduce new ways of storing table data. Accessing those table access methods from the executor requires TupleTableSlots to be carry tuples in the native format of such storage methods; otherwise there'll be a significant conversion overhead. Different access methods will require different data to store tuples efficiently (just like virtual, minimal, heap already require fields in TupleTableSlot). To allow that without requiring additional pointer indirections, we want to have different structs (embedding TupleTableSlot) for different types of slots. Thus different types of slots are needed, which requires adapting creators of slots. The slot that most efficiently can represent a type of tuple in an executor node will often depend on the type of slot a child node uses. Therefore we need to track the type of slot is returned by nodes, so parent slots can create slots based on that. Relatedly, JIT compilation of tuple deforming needs to know which type of slot a certain expression refers to, so it can create an appropriate deforming function for the type of tuple in the slot. But not all nodes will only return one type of slot, e.g. an append node will potentially return different types of slots for each of its subplans. Therefore add function that allows to query the type of a node's result slot, and whether it'll always be the same type (whether it's fixed). This can be queried using ExecGetResultSlotOps(). The scan, result, inner, outer type of slots are automatically inferred from ExecInitScanTupleSlot(), ExecInitResultSlot(), left/right subtrees respectively. If that's not correct for a node, that can be overwritten using new fields in PlanState. This commit does not introduce the actually abstracted implementation of different kind of TupleTableSlots, that will be left for a followup commit. The different types of slots introduced will, for now, still use the same backing implementation. While this already partially invalidates the big comment in tuptable.h, it seems to make more sense to update it later, when the different TupleTableSlot implementations actually exist. Author: Ashutosh Bapat and Andres Freund, with changes by Amit Khandekar Discussion: https://postgr.es/m/20181105210039.hh4vvi4vwoq5ba2q@alap3.anarazel.de
2018-11-16 01:00:30 -05:00
ExecGetResultType(scanstate->cteplanstate),
Optimize tuple deformation This commit includes various optimizations to improve the performance of tuple deformation. We now precalculate CompactAttribute's attcacheoff, which allows us to remove the code from the deform routines which was setting the attcacheoff. Setting the attcacheoff is now handled by TupleDescFinalize(), which must be called before the TupleDesc is used for anything. Having TupleDescFinalize() means we can store the first attribute in the TupleDesc which does not have an offset cached. That allows us to add a dedicated deforming loop to deform all attributes up to the final one with an attcacheoff set, or up to the first NULL attribute, whichever comes first. Here we also improve tuple deformation performance of tuples with NULLs. Previously, if the HEAP_HASNULL bit was set in the tuple's t_infomask, deforming would, one-by-one, check each and every bit in the NULL bitmap to see if it was zero. Now, we process the NULL bitmap 1 byte at a time rather than 1 bit at a time to find the attnum with the first NULL. We can now deform the tuple without checking for NULLs up to just before that attribute. We also record the maximum attribute number which is guaranteed to exist in the tuple, that is, has a NOT NULL constraint and isn't an atthasmissing attribute. When deforming only attributes prior to the guaranteed attnum, we've no need to access the tuple's natt count. As an additional optimization, we only count fixed-width columns when calculating the maximum guaranteed column, as this eliminates the need to emit code to fetch byref types in the deformation loop for guaranteed attributes. Some locations in the code deform tuples that have yet to go through NOT NULL constraint validation. We're unable to perform the guaranteed attribute optimization when that's the case. This optimization is opt-in via the TupleTableSlot using the TTS_FLAG_OBEYS_NOT_NULL_CONSTRAINTS flag. This commit also adds a more efficient way of populating the isnull array by using a bit-wise SWAR trick which performs multiplication on the inverse of the tuple's bitmap byte and masking out all but the lower bit of each of the boolean's byte. This results in much more optimal code when compared to determining the NULLness via att_isnull(). 8 isnull elements are processed at once using this method, which means we need to round the tts_isnull array size up to the next 8 bytes. The palloc code does this anyway, but the round-up needed to be formalized so as not to overwrite the sentinel byte in MEMORY_CONTEXT_CHECKING builds. Doing this also allows the NULL-checking deforming loop to more efficiently check the isnull array, rather than doing the bit-wise processing for each attribute that att_isnull() does. The level of performance improvement from these changes seems to vary depending on the CPU architecture. Apple's M chips seem particularly fond of the changes, with some of the tested deform-heavy queries going over twice as fast as before. With x86-64, the speedups aren't quite as large. With tables containing only a small number of columns, the speedups will be less. Author: David Rowley <dgrowleyml@gmail.com> Reviewed-by: Chao Li <li.evan.chao@gmail.com> Reviewed-by: Andres Freund <andres@anarazel.de> Reviewed-by: John Naylor <johncnaylorls@gmail.com> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Zsolt Parragi <zsolt.parragi@percona.com> Reviewed-by: Álvaro Herrera <alvherre@kurilemu.de> Reviewed-by: Junwang Zhao <zhjwpku@gmail.com> Discussion: https://postgr.es/m/CAApHDvpoFjaj3%2Bw_jD5uPnGazaw41A71tVJokLDJg2zfcigpMQ%40mail.gmail.com
2026-03-15 18:46:00 -04:00
&TTSOpsMinimalTuple, 0);
/*
Don't require return slots for nodes without projection. In a lot of nodes the return slot is not required. That can either be because the node doesn't do any projection (say an Append node), or because the node does perform projections but the projection is optimized away because the projection would yield an identical row. Slots aren't that small, especially for wide rows, so it's worthwhile to avoid creating them. It's not possible to just skip creating the slot - it's currently used to determine the tuple descriptor returned by ExecGetResultType(). So separate the determination of the result type from the slot creation. The work previously done internally ExecInitResultTupleSlotTL() can now also be done separately with ExecInitResultTypeTL() and ExecInitResultSlot(). That way nodes that aren't guaranteed to need a result slot, can use ExecInitResultTypeTL() to determine the result type of the node, and ExecAssignScanProjectionInfo() (via ExecConditionalAssignProjectionInfo()) determines that a result slot is needed, it is created with ExecInitResultSlot(). Besides the advantage of avoiding to create slots that then are unused, this is necessary preparation for later patches around tuple table slot abstraction. In particular separating the return descriptor and slot is a prerequisite to allow JITing of tuple deforming with knowledge of the underlying tuple format, and to avoid unnecessarily creating JITed tuple deforming for virtual slots. This commit removes a redundant argument from ExecInitResultTupleSlotTL(). While this commit touches a lot of the relevant lines anyway, it'd normally still not worthwhile to cause breakage, except that aforementioned later commits will touch *all* ExecInitResultTupleSlotTL() callers anyway (but fits worse thematically). Author: Andres Freund Discussion: https://postgr.es/m/20181105210039.hh4vvi4vwoq5ba2q@alap3.anarazel.de
2018-11-09 20:19:39 -05:00
* Initialize result type and projection.
*/
Don't require return slots for nodes without projection. In a lot of nodes the return slot is not required. That can either be because the node doesn't do any projection (say an Append node), or because the node does perform projections but the projection is optimized away because the projection would yield an identical row. Slots aren't that small, especially for wide rows, so it's worthwhile to avoid creating them. It's not possible to just skip creating the slot - it's currently used to determine the tuple descriptor returned by ExecGetResultType(). So separate the determination of the result type from the slot creation. The work previously done internally ExecInitResultTupleSlotTL() can now also be done separately with ExecInitResultTypeTL() and ExecInitResultSlot(). That way nodes that aren't guaranteed to need a result slot, can use ExecInitResultTypeTL() to determine the result type of the node, and ExecAssignScanProjectionInfo() (via ExecConditionalAssignProjectionInfo()) determines that a result slot is needed, it is created with ExecInitResultSlot(). Besides the advantage of avoiding to create slots that then are unused, this is necessary preparation for later patches around tuple table slot abstraction. In particular separating the return descriptor and slot is a prerequisite to allow JITing of tuple deforming with knowledge of the underlying tuple format, and to avoid unnecessarily creating JITed tuple deforming for virtual slots. This commit removes a redundant argument from ExecInitResultTupleSlotTL(). While this commit touches a lot of the relevant lines anyway, it'd normally still not worthwhile to cause breakage, except that aforementioned later commits will touch *all* ExecInitResultTupleSlotTL() callers anyway (but fits worse thematically). Author: Andres Freund Discussion: https://postgr.es/m/20181105210039.hh4vvi4vwoq5ba2q@alap3.anarazel.de
2018-11-09 20:19:39 -05:00
ExecInitResultTypeTL(&scanstate->ss.ps);
ExecAssignScanProjectionInfo(&scanstate->ss);
/*
* initialize child expressions
*/
scanstate->ss.ps.qual =
ExecInitQual(node->scan.plan.qual, (PlanState *) scanstate);
return scanstate;
}
/* ----------------------------------------------------------------
* ExecEndCteScan
*
* frees any storage allocated through C routines.
* ----------------------------------------------------------------
*/
void
ExecEndCteScan(CteScanState *node)
{
/*
* If I am the leader, free the tuplestore.
*/
if (node->leader == node)
{
tuplestore_end(node->cte_table);
node->cte_table = NULL;
}
}
/* ----------------------------------------------------------------
* ExecReScanCteScan
*
* Rescans the relation.
* ----------------------------------------------------------------
*/
void
ExecReScanCteScan(CteScanState *node)
{
Tuplestorestate *tuplestorestate = node->leader->cte_table;
Don't require return slots for nodes without projection. In a lot of nodes the return slot is not required. That can either be because the node doesn't do any projection (say an Append node), or because the node does perform projections but the projection is optimized away because the projection would yield an identical row. Slots aren't that small, especially for wide rows, so it's worthwhile to avoid creating them. It's not possible to just skip creating the slot - it's currently used to determine the tuple descriptor returned by ExecGetResultType(). So separate the determination of the result type from the slot creation. The work previously done internally ExecInitResultTupleSlotTL() can now also be done separately with ExecInitResultTypeTL() and ExecInitResultSlot(). That way nodes that aren't guaranteed to need a result slot, can use ExecInitResultTypeTL() to determine the result type of the node, and ExecAssignScanProjectionInfo() (via ExecConditionalAssignProjectionInfo()) determines that a result slot is needed, it is created with ExecInitResultSlot(). Besides the advantage of avoiding to create slots that then are unused, this is necessary preparation for later patches around tuple table slot abstraction. In particular separating the return descriptor and slot is a prerequisite to allow JITing of tuple deforming with knowledge of the underlying tuple format, and to avoid unnecessarily creating JITed tuple deforming for virtual slots. This commit removes a redundant argument from ExecInitResultTupleSlotTL(). While this commit touches a lot of the relevant lines anyway, it'd normally still not worthwhile to cause breakage, except that aforementioned later commits will touch *all* ExecInitResultTupleSlotTL() callers anyway (but fits worse thematically). Author: Andres Freund Discussion: https://postgr.es/m/20181105210039.hh4vvi4vwoq5ba2q@alap3.anarazel.de
2018-11-09 20:19:39 -05:00
if (node->ss.ps.ps_ResultTupleSlot)
ExecClearTuple(node->ss.ps.ps_ResultTupleSlot);
ExecScanReScan(&node->ss);
/*
* Clear the tuplestore if a new scan of the underlying CTE is required.
* This implicitly resets all the tuplestore's read pointers. Note that
* multiple CTE nodes might redundantly clear the tuplestore; that's OK,
* and not unduly expensive. We'll stop taking this path as soon as
* somebody has attempted to read something from the underlying CTE
* (thereby causing its chgParam to be cleared).
*/
if (node->leader->cteplanstate->chgParam != NULL)
{
tuplestore_clear(tuplestorestate);
node->leader->eof_cte = false;
}
else
{
/*
* Else, just rewind my own pointer. Either the underlying CTE
* doesn't need a rescan (and we can re-read what's in the tuplestore
* now), or somebody else already took care of it.
*/
tuplestore_select_read_pointer(tuplestorestate, node->readptr);
tuplestore_rescan(tuplestorestate);
}
}