PostgreSQL源码解读（182）-查询#98(聚合函数#3-ExecAgg)
本节简单介绍了PostgreSQL执行聚合函数的实现,主要实现函数是ExecAgg.这里先行介绍ExecAgg->agg_fill_hash_table函数,其他子函数后续再行介绍.
成都创新互联是一家专注于成都网站建设、成都网站设计与策划设计,昆玉网站建设哪家好?成都创新互联做网站,专注于网站建设十余年,网设计领域的专业建站公司;建站业务涵盖:昆玉等地区。昆玉做网站价格咨询:18982081108
通过设置log输出,可得到SQL的planTree:
",,,,,"select bh,avg(c1),min(c1),max(c2) from t_agg group by bh;",,,"psql"
2019-04-30 14:33:11.998 CST,"xdb","testdb",1387,"[local]",5cc7ec00.56b,3,"SELECT",2019-04-30 14:32:32 CST,3/3,0,LOG,00000,"plan:"," {PLANNEDSTMT
:commandType 1
:queryId 0
:hasReturning false
:hasModifyingCTE false
:canSetTag true
:transientPlan false
:dependsOnRole false
:parallelModeNeeded false
:jitFlags 0
:planTree
{AGG
:startup_cost 13677.00
:total_cost 13677.06
:plan_rows 5
:plan_width 45
:parallel_aware false
:parallel_safe false
:plan_node_id 0
:targetlist (...
)
:qual <>
:lefttree
{SEQSCAN
:startup_cost 0.00
:total_cost 8677.00
:plan_rows 500000
:plan_width 13
:parallel_aware false
:parallel_safe false
:plan_node_id 1
:targetlist (...
)
:qual <>
:lefttree <>
:righttree <>
:initPlan <>
:extParam (b)
:allParam (b)
:scanrelid 1
}
:righttree <>
:initPlan <>
:extParam (b)
:allParam (b)
:aggstrategy 2
:aggsplit 0
:numCols 1
:grpColIdx 1
:grpOperators 98
:numGroups 5
:aggParams (b)
:groupingSets <>
:chain <>
}
:rtable (...
)
:resultRelations <>
:nonleafResultRelations <>
:rootResultRelations <>
:subplans <>
:rewindPlanIDs (b)
:rowMarks <>
:relationOids (o 245801)
:invalItems <>
:paramExecTypes <>
:utilityStmt <>
:stmt_location 0
:stmt_len 56
}
",,,,,"select bh,avg(c1),min(c1),max(c2) from t_agg group by bh;",,,"psql"
第一个节点为AGG,相应的实现函数为ExecAgg.
一、数据结构
AggState
聚合函数执行时状态结构体,内含AggStatePerAgg等结构体
/* ---------------------
* AggState information
*
* ss.ss_ScanTupleSlot refers to output of underlying plan.
* ss.ss_ScanTupleSlot指的是基础计划的输出.
* (ss = ScanState,ps = PlanState)
*
* Note: ss.ps.ps_ExprContext contains ecxt_aggvalues and
* ecxt_aggnulls arrays, which hold the computed agg values for the current
* input group during evaluation of an Agg node's output tuple(s). We
* create a second ExprContext, tmpcontext, in which to evaluate input
* expressions and run the aggregate transition functions.
* 注意:ss.ps.ps_ExprContext包含了ecxt_aggvalues和ecxt_aggnulls数组,
* 这两个数组保存了在计算agg节点的输出元组时当前输入组已计算的agg值.
* ---------------------
*/
/* these structs are private in nodeAgg.c: */
//在nodeAgg.c中私有的结构体
typedef struct AggStatePerAggData *AggStatePerAgg;
typedef struct AggStatePerTransData *AggStatePerTrans;
typedef struct AggStatePerGroupData *AggStatePerGroup;
typedef struct AggStatePerPhaseData *AggStatePerPhase;
typedef struct AggStatePerHashData *AggStatePerHash;
typedef struct AggState
{
//第一个字段是NodeTag(继承自ScanState)
ScanState ss; /* its first field is NodeTag */
//targetlist和quals中所有的Aggref
List *aggs; /* all Aggref nodes in targetlist & quals */
//链表的大小(可以为0)
int numaggs; /* length of list (could be zero!) */
//pertrans条目大小
int numtrans; /* number of pertrans items */
//Agg策略模式
AggStrategy aggstrategy; /* strategy mode */
//agg-splitting模式,参见nodes.h
AggSplit aggsplit; /* agg-splitting mode, see nodes.h */
//指向当前步骤数据的指针
AggStatePerPhase phase; /* pointer to current phase data */
//步骤数(包括0)
int numphases; /* number of phases (including phase 0) */
//当前步骤
int current_phase; /* current phase number */
//per-Aggref信息
AggStatePerAgg peragg; /* per-Aggref information */
//per-Trans状态信息
AggStatePerTrans pertrans; /* per-Trans state information */
//长生命周期数据的ExprContexts(hashtable)
ExprContext *hashcontext; /* econtexts for long-lived data (hashtable) */
////长生命周期数据的ExprContexts(每一个GS使用)
ExprContext **aggcontexts; /* econtexts for long-lived data (per GS) */
//输入表达式的ExprContext
ExprContext *tmpcontext; /* econtext for input expressions */
#define FIELDNO_AGGSTATE_CURAGGCONTEXT 14
//当前活跃的aggcontext
ExprContext *curaggcontext; /* currently active aggcontext */
//当前活跃的aggregate(如存在)
AggStatePerAgg curperagg; /* currently active aggregate, if any */
#define FIELDNO_AGGSTATE_CURPERTRANS 16
//当前活跃的trans state
AggStatePerTrans curpertrans; /* currently active trans state, if any */
//输入结束?
bool input_done; /* indicates end of input */
//Agg扫描结束?
bool agg_done; /* indicates completion of Agg scan */
//最后一个grouping set
int projected_set; /* The last projected grouping set */
#define FIELDNO_AGGSTATE_CURRENT_SET 20
//将要解析的当前grouping set
int current_set; /* The current grouping set being evaluated */
//当前投影操作的分组列
Bitmapset *grouped_cols; /* grouped cols in current projection */
//倒序的分组列链表
List *all_grouped_cols; /* list of all grouped cols in DESC order */
/* These fields are for grouping set phase data */
//-------- 下面的列用于grouping set步骤数据
//所有步骤中最大的sets大小
int maxsets; /* The max number of sets in any phase */
//所有步骤的数组
AggStatePerPhase phases; /* array of all phases */
//对于phases > 1,已排序的输入信息
Tuplesortstate *sort_in; /* sorted input to phases > 1 */
//对于下一个步骤,输入已拷贝
Tuplesortstate *sort_out; /* input is copied here for next phase */
//排序结果的slot
TupleTableSlot *sort_slot; /* slot for sort results */
/* these fields are used in AGG_PLAIN and AGG_SORTED modes: */
//------- 下面的列用于AGG_PLAIN和AGG_SORTED模式:
//per-group指针的grouping set编号数组
AggStatePerGroup *pergroups; /* grouping set indexed array of per-group
* pointers */
//当前组的第一个元组拷贝
HeapTuple grp_firstTuple; /* copy of first tuple of current group */
/* these fields are used in AGG_HASHED and AGG_MIXED modes: */
//--------- 下面的列用于AGG_HASHED和AGG_MIXED模式:
//是否已填充hash表?
bool table_filled; /* hash table filled yet? */
//hash桶数?
int num_hashes;
//相应的哈希表数据数组
AggStatePerHash perhash; /* array of per-hashtable data */
//per-group指针的grouping set编号数组
AggStatePerGroup *hash_pergroup; /* grouping set indexed array of
* per-group pointers */
/* support for evaluation of agg input expressions: */
//---------- agg输入表达式解析支持
#define FIELDNO_AGGSTATE_ALL_PERGROUPS 34
//首先是->pergroups,然后是hash_pergroup
AggStatePerGroup *all_pergroups; /* array of first ->pergroups, than
* ->hash_pergroup */
//投影实现机制
ProjectionInfo *combinedproj; /* projection machinery */
} AggState;
/* Primitive options supported by nodeAgg.c: */
//nodeag .c支持的基本选项
#define AGGSPLITOP_COMBINE 0x01 /* substitute combinefn for transfn */
#define AGGSPLITOP_SKIPFINAL 0x02 /* skip finalfn, return state as-is */
#define AGGSPLITOP_SERIALIZE 0x04 /* apply serializefn to output */
#define AGGSPLITOP_DESERIALIZE 0x08 /* apply deserializefn to input */
/* Supported operating modes (i.e., useful combinations of these options): */
//支持的操作模式
typedef enum AggSplit
{
/* Basic, non-split aggregation: */
//基本 : 非split聚合
AGGSPLIT_SIMPLE = 0,
/* Initial phase of partial aggregation, with serialization: */
//部分聚合的初始步骤,序列化
AGGSPLIT_INITIAL_SERIAL = AGGSPLITOP_SKIPFINAL | AGGSPLITOP_SERIALIZE,
/* Final phase of partial aggregation, with deserialization: */
//部分聚合的最终步骤,反序列化
AGGSPLIT_FINAL_DESERIAL = AGGSPLITOP_COMBINE | AGGSPLITOP_DESERIALIZE
} AggSplit;
/* Test whether an AggSplit value selects each primitive option: */
//测试AggSplit选择了哪些基本选项
#define DO_AGGSPLIT_COMBINE(as) (((as) & AGGSPLITOP_COMBINE) != 0)
#define DO_AGGSPLIT_SKIPFINAL(as) (((as) & AGGSPLITOP_SKIPFINAL) != 0)
#define DO_AGGSPLIT_SERIALIZE(as) (((as) & AGGSPLITOP_SERIALIZE) != 0)
#define DO_AGGSPLIT_DESERIALIZE(as) (((as) & AGGSPLITOP_DESERIALIZE) != 0)
二、源码解读
ExecAgg接收从outer子计划返回的元组合适的属性上为每一个聚合函数(出现在投影列或节点表达式)执行聚合.需要聚合的元组数量依赖于是否已分组或者选择普通聚合.在已分组的聚合操作宏,为每一个组产生结果行;普通聚合,整个查询只有一个结果行.
不管哪种情况,每一个聚合结果值都会存储在表达式上下文中(ExecProject会解析结果元组)
/*
* ExecAgg -
*
* ExecAgg receives tuples from its outer subplan and aggregates over
* the appropriate attribute for each aggregate function use (Aggref
* node) appearing in the targetlist or qual of the node. The number
* of tuples to aggregate over depends on whether grouped or plain
* aggregation is selected. In grouped aggregation, we produce a result
* row for each group; in plain aggregation there's a single result row
* for the whole query. In either case, the value of each aggregate is
* stored in the expression context to be used when ExecProject evaluates
* the result tuple.
* ExecAgg接收从outer子计划返回的元组合适的属性上为每一个聚合函数(出现在投影列或节点表达式)执行聚合.
* 需要聚合的元组数量依赖于是否已分组或者选择普通聚合.
* 在已分组的聚合操作宏,为每一个组产生结果行;普通聚合,整个查询只有一个结果行.
* 不管哪种情况,每一个聚合结果值都会存储在表达式上下文中(ExecProject会解析结果元组)
*/
static TupleTableSlot *
ExecAgg(PlanState *pstate)
{
AggState *node = castNode(AggState, pstate);
TupleTableSlot *result = NULL;
CHECK_FOR_INTERRUPTS();
if (!node->agg_done)
{
/* Dispatch based on strategy */
//基于策略进行分发
switch (node->phase->aggstrategy)
{
case AGG_HASHED:
if (!node->table_filled)
agg_fill_hash_table(node);
/* FALLTHROUGH */
//填充后,执行MIXED
case AGG_MIXED:
result = agg_retrieve_hash_table(node);
break;
case AGG_PLAIN:
case AGG_SORTED:
result = agg_retrieve_direct(node);
break;
}
if (!TupIsNull(result))
return result;
}
return NULL;
}
agg_fill_hash_table
读取输入并构建哈希表,逻辑较为简单,详细参考下面源码
/*
* ExecAgg for hashed case: read input and build hash table
* 读取输入并构建哈希表
*/
static void
agg_fill_hash_table(AggState *aggstate)
{
TupleTableSlot *outerslot;
ExprContext *tmpcontext = aggstate->tmpcontext;
/*
* Process each outer-plan tuple, and then fetch the next one, until we
* exhaust the outer plan.
* 处理每一个outer-plan返回的元组,然后继续提取下一个,直至完成所有元组的处理.
*/
for (;;)
{
//--------- 循环直至完成所有元组的处理
//提取输入的元组
outerslot = fetch_input_tuple(aggstate);
if (TupIsNull(outerslot))
break;//已完成处理,退出循环
/* set up for lookup_hash_entries and advance_aggregates */
//配置lookup_hash_entries和advance_aggregates函数
//把元组放在临时内存上下文中
tmpcontext->ecxt_outertuple = outerslot;
/* Find or build hashtable entries */
//检索或构建哈希表条目
lookup_hash_entries(aggstate);
/* Advance the aggregates (or combine functions) */
//增加聚合(或组合函数)
advance_aggregates(aggstate);
/*
* Reset per-input-tuple context after each tuple, but note that the
* hash lookups do this too
* 重置per-input-tuple内存上下文,但需要注意hash检索也会做这个事情
*/
ResetExprContext(aggstate->tmpcontext);
}
aggstate->table_filled = true;
/* Initialize to walk the first hash table */
//初始化用于遍历第一个哈希表
select_current_set(aggstate, 0, true);
ResetTupleHashIterator(aggstate->perhash[0].hashtable,
&aggstate->perhash[0].hashiter);
}
/*
* Advance each aggregate transition state for one input tuple. The input
* tuple has been stored in tmpcontext->ecxt_outertuple, so that it is
* accessible to ExecEvalExpr.
*
* We have two sets of transition states to handle: one for sorted aggregation
* and one for hashed; we do them both here, to avoid multiple evaluation of
* the inputs.
*
* When called, CurrentMemoryContext should be the per-query context.
*/
static void
advance_aggregates(AggState *aggstate)
{
bool dummynull;
ExecEvalExprSwitchContext(aggstate->phase->evaltrans,
aggstate->tmpcontext,
&dummynull);
}
/*
* ExecEvalExprSwitchContext
*
* Same as ExecEvalExpr, but get into the right allocation context explicitly.
*/
#ifndef FRONTEND
static inline Datum
ExecEvalExprSwitchContext(ExprState *state,
ExprContext *econtext,
bool *isNull)
{
Datum retDatum;
MemoryContext oldContext;
oldContext = MemoryContextSwitchTo(econtext->ecxt_per_tuple_memory);
retDatum = state->evalfunc(state, econtext, isNull);
MemoryContextSwitchTo(oldContext);
return retDatum;
}
#endif
/*
* Look up hash entries for the current tuple in all hashed grouping sets,
* returning an array of pergroup pointers suitable for advance_aggregates.
* 为当前元组在所有已完成hash的grouping sets中检索hash条目,
* 为后续的advance_aggregates函数调用返回pergroup指针数组.
*
* Be aware that lookup_hash_entry can reset the tmpcontext.
* 需要提醒的是lookup_hash_entry可以重置tmpcontext
*/
static void
lookup_hash_entries(AggState *aggstate)
{
//hash个数
int numHashes = aggstate->num_hashes;
//获取pergroup
AggStatePerGroup *pergroup = aggstate->hash_pergroup;
int setno;
for (setno = 0; setno < numHashes; setno++)
{
//设置当前集合
select_current_set(aggstate, setno, true);
//检索哈希条目
pergroup[setno] = lookup_hash_entry(aggstate)->additional;
}
}
/*
* Find or create a hashtable entry for the tuple group containing the current
* tuple (already set in tmpcontext's outertuple slot), in the current grouping
* set (which the caller must have selected - note that initialize_aggregate
* depends on this).
* 为包含当前元组的组检索或创建哈希表条目(已在tmpcontext上下文中设置了outertuple slot),
* 在当前grouping set中设置(调用者已完成选择 - 注意initialize_aggregate依赖于此)
*
* When called, CurrentMemoryContext should be the per-query context.
* 一旦完成调用,CurrentMemoryContext应该是per-query上下文
*/
static TupleHashEntryData *
lookup_hash_entry(AggState *aggstate)
{
//输入的元组
TupleTableSlot *inputslot = aggstate->tmpcontext->ecxt_outertuple;
//perhash
AggStatePerHash perhash = &aggstate->perhash[aggstate->current_set];
//hashslot
TupleTableSlot *hashslot = perhash->hashslot;
//条目入口
TupleHashEntryData *entry;
//变量
bool isnew;
int i;
/* transfer just the needed columns into hashslot */
//转换需要的列到hashslot中
slot_getsomeattrs(inputslot, perhash->largestGrpColIdx);
ExecClearTuple(hashslot);
for (i = 0; i < perhash->numhashGrpCols; i++)
{
//遍历分组列
//列编号
int varNumber = perhash->hashGrpColIdxInput[i] - 1;
//赋值
hashslot->tts_values[i] = inputslot->tts_values[varNumber];
hashslot->tts_isnull[i] = inputslot->tts_isnull[varNumber];
}
//存储虚拟元组
ExecStoreVirtualTuple(hashslot);
/* find or create the hashtable entry using the filtered tuple */
//使用已过滤的元组检索或者创建哈希表条目
entry = LookupTupleHashEntry(perhash->hashtable, hashslot, &isnew);
if (isnew)
{
//新条目
AggStatePerGroup pergroup;
int transno;
//分配内存
pergroup = (AggStatePerGroup)
MemoryContextAlloc(perhash->hashtable->tablecxt,
sizeof(AggStatePerGroupData) * aggstate->numtrans);
entry->additional = pergroup;
/*
* Initialize aggregates for new tuple group, lookup_hash_entries()
* already has selected the relevant grouping set.
* 为新元组group初始化聚合操作, lookup_hash_entries()已选择了相应的grouping set
*/
for (transno = 0; transno < aggstate->numtrans; transno++)
{
//遍历转换函数
AggStatePerTrans pertrans = &aggstate->pertrans[transno];
AggStatePerGroup pergroupstate = &pergroup[transno];
//初始化聚合
initialize_aggregate(aggstate, pertrans, pergroupstate);
}
}
return entry;
}
/*
* Find or create a hashtable entry for the tuple group containing the
* given tuple. The tuple must be the same type as the hashtable entries.
* 为包含给定元组的元组group检索或创建哈希表条目.
* 元组必须与哈希表条目具有相同的类型.
*
* If isnew is NULL, we do not create new entries; we return NULL if no
* match is found.
* 如isnew为NULL,不需要创建新的条目;如无匹配则返回NULL.
*
* If isnew isn't NULL, then a new entry is created if no existing entry
* matches. On return, *isnew is true if the entry is newly created,
* false if it existed already. ->additional_data in the new entry has
* been zeroed.
* 如isnew不是NULL,如没有与之匹配的现存条目,则创建新的条目.
* 在返回的时候,如新创建了条目,则*isnew为T,如已存在条目则为F.
* 新条目中的->additional_data已初始化为0.
*/
TupleHashEntry
LookupTupleHashEntry(TupleHashTable hashtable, TupleTableSlot *slot,
bool *isnew)
{
//哈希条目
TupleHashEntryData *entry;
MemoryContext oldContext;
bool found;
MinimalTuple key;
/* Need to run the hash functions in short-lived context */
//在短生命周期中执行哈希函数
oldContext = MemoryContextSwitchTo(hashtable->tempcxt);
/* set up data needed by hash and match functions */
//设置哈希和匹配函数需要的数据
hashtable->inputslot = slot;
hashtable->in_hash_funcs = hashtable->tab_hash_funcs;
hashtable->cur_eq_func = hashtable->tab_eq_func;
//参考inputslot的flag
key = NULL; /* flag to reference inputslot */
if (isnew)
{
//新条目,插入到哈希表中
entry = tuplehash_insert(hashtable->hashtab, key, &found);
if (found)
{
/* found pre-existing entry */
//发现上一个已存在的条目
*isnew = false;
}
else
{
/* created new entry */
//创建新条目
*isnew = true;
/* zero caller data */
//初始化调用者的数据
entry->additional = NULL;
MemoryContextSwitchTo(hashtable->tablecxt);
/* Copy the first tuple into the table context */
//拷贝第一个条目到数据表上下文中
entry->firstTuple = ExecCopySlotMinimalTuple(slot);
}
}
else
{
//isnew为NULL,调用tuplehash_lookup
entry = tuplehash_lookup(hashtable->hashtab, key);
}
MemoryContextSwitchTo(oldContext);
return entry;
}
/*
* (Re)Initialize an individual aggregate.
* (重新)初始化单独的聚合函数.
*
* This function handles only one grouping set, already set in
* aggstate->current_set.
* 该函数只处理一个grouping set(已在aggstate->current_set设置)
*
* When called, CurrentMemoryContext should be the per-query context.
* 调用完毕,CurrentMemoryContext应为per-query上下文.
*/
static void
initialize_aggregate(AggState *aggstate, AggStatePerTrans pertrans,
AggStatePerGroup pergroupstate)
{
/*
* Start a fresh sort operation for each DISTINCT/ORDER BY aggregate.
* 为每一个DISTINCT/ORDER BY聚合启动刷新排序操作
*/
if (pertrans->numSortCols > 0)
{
/*
* In case of rescan, maybe there could be an uncompleted sort
* operation? Clean it up if so.
* 如为重新扫描,可能存在未完成的排序操作.如存在,则需清除.
*/
if (pertrans->sortstates[aggstate->current_set])
tuplesort_end(pertrans->sortstates[aggstate->current_set]);
/*
* We use a plain Datum sorter when there's a single input column;
* otherwise sort the full tuple. (See comments for
* process_ordered_aggregate_single.)
* 如存在一个独立的输入列,使用普通的Datum排序器即可.
* 否则的话,排序全部元组(参见process_ordered_aggregate_single中的注释)
*/
if (pertrans->numInputs == 1)
{
//属性信息
Form_pg_attribute attr = TupleDescAttr(pertrans->sortdesc, 0);
//Datum sorter
pertrans->sortstates[aggstate->current_set] =
tuplesort_begin_datum(attr->atttypid,
pertrans->sortOperators[0],
pertrans->sortCollations[0],
pertrans->sortNullsFirst[0],
work_mem, NULL, false);
}
else
//full tuple sorter
pertrans->sortstates[aggstate->current_set] =
tuplesort_begin_heap(pertrans->sortdesc,
pertrans->numSortCols,
pertrans->sortColIdx,
pertrans->sortOperators,
pertrans->sortCollations,
pertrans->sortNullsFirst,
work_mem, NULL, false);
}
/*
* (Re)set transValue to the initial value.
* (重新)设置transValue为初始值
*
* Note that when the initial value is pass-by-ref, we must copy it (into
* the aggcontext) since we will pfree the transValue later.
* 注意初始值为pass-by-ref(引用传递),必须拷贝该参数(到aggcontext中),因为在后续会用pfree释放transValue.
*/
if (pertrans->initValueIsNull)
pergroupstate->transValue = pertrans->initValue;
else
{
MemoryContext oldContext;
oldContext = MemoryContextSwitchTo(
aggstate->curaggcontext->ecxt_per_tuple_memory);
//拷贝
pergroupstate->transValue = datumCopy(pertrans->initValue,
pertrans->transtypeByVal,
pertrans->transtypeLen);
MemoryContextSwitchTo(oldContext);
}
pergroupstate->transValueIsNull = pertrans->initValueIsNull;
/*
* If the initial value for the transition state doesn't exist in the
* pg_aggregate table then we will let the first non-NULL value returned
* from the outer procNode become the initial value. (This is useful for
* aggregates like max() and min().) The noTransValue flag signals that we
* still need to do this.
* 如转换状态的初始值在pg_aggregate表中不存在,那么让outer procNode中的第一个非NULL值返回作为初始值.
* (这在max()和min()聚合时会非常有用).noTransValue标记提示需要执行该动作.
*/
pergroupstate->noTransValue = pertrans->initValueIsNull;
}
/*
* Select the current grouping set; affects current_set and
* curaggcontext.
* 选择当前的goruping set;影响的参数包括current_set和curaggcontext.
*/
static void
select_current_set(AggState *aggstate, int setno, bool is_hash)
{
/* when changing this, also adapt ExecInterpExpr() and friends */
//在修改的时候,会同时调整ExecInterpExpr()和友元
if (is_hash)
aggstate->curaggcontext = aggstate->hashcontext;
else
aggstate->curaggcontext = aggstate->aggcontexts[setno];
aggstate->current_set = setno;
}
三、跟踪分析
测试脚本
//禁用并行
testdb=# set max_parallel_workers_per_gather=0;
SET
testdb=# explain verbose select bh,avg(c1),min(c1),max(c2) from t_agg group by bh;
QUERY PLAN
---------------------------------------------------------------------------
HashAggregate (cost=13677.00..13677.06 rows=5 width=45)
Output: bh, avg(c1), min(c1), max(c2)
Group Key: t_agg.bh
-> Seq Scan on public.t_agg (cost=0.00..8677.00 rows=500000 width=13)
Output: bh, c1, c2, c3, c4, c5, c6
(5 rows)
跟踪分析
(gdb) b ExecAgg
Breakpoint 1 at 0x6ee444: file nodeAgg.c, line 1536.
(gdb) c
Continuing.
Breakpoint 1, ExecAgg (pstate=0x1f895a0) at nodeAgg.c:1536
1536 AggState *node = castNode(AggState, pstate);
(gdb)
输入参数,AggState,在ExecInitAgg函数中初始化
(gdb) p *pstate
$1 = {type = T_AggState, plan = 0x1f7b1e0, state = 0x1f89388, ExecProcNode = 0x6ee438 ,
ExecProcNodeReal = 0x6ee438 , instrument = 0x0, worker_instrument = 0x0, worker_jit_instrument = 0x0,
qual = 0x0, lefttree = 0x1f89b10, righttree = 0x0, initPlan = 0x0, subPlan = 0x0, chgParam = 0x0,
ps_ResultTupleSlot = 0x1f8a710, ps_ExprContext = 0x1f89a50, ps_ProjInfo = 0x1f8a850, scandesc = 0x1f89e60}
使用Hash实现
(gdb) n
1537 TupleTableSlot *result = NULL;
(gdb)
1539 CHECK_FOR_INTERRUPTS();
(gdb)
1541 if (!node->agg_done)
(gdb)
1544 switch (node->phase->aggstrategy)
(gdb) p node->phase->aggstrategy
$2 = AGG_HASHED
(gdb) n
1547 if (!node->table_filled)
(gdb)
1548 agg_fill_hash_table(node);
(gdb)
进入agg_fill_hash_table
(gdb) step
agg_fill_hash_table (aggstate=0x1f895a0) at nodeAgg.c:1915
1915 ExprContext *tmpcontext = aggstate->tmpcontext;
agg_fill_hash_table->提取输入的元组
(gdb) n
1923 outerslot = fetch_input_tuple(aggstate);
(gdb) step
fetch_input_tuple (aggstate=0x1f895a0) at nodeAgg.c:396
396 if (aggstate->sort_in)
(gdb) p aggstate->sort_in
$3 = (Tuplesortstate *) 0x0
(gdb) n
406 slot = ExecProcNode(outerPlanState(aggstate));
(gdb)
408 if (!TupIsNull(slot) && aggstate->sort_out)
(gdb) p *slot
$4 = {type = T_TupleTableSlot, tts_isempty = false, tts_shouldFree = false, tts_shouldFreeMin = false, tts_slow = false,
tts_tuple = 0x1fa5998, tts_tupleDescriptor = 0x7ff7dd2d1380, tts_mcxt = 0x1f89270, tts_buffer = 124, tts_nvalid = 0,
tts_values = 0x1f89d48, tts_isnull = 0x1f89d80, tts_mintuple = 0x0, tts_minhdr = {t_len = 0, t_self = {ip_blkid = {
bi_hi = 0, bi_lo = 0}, ip_posid = 0}, t_tableOid = 0, t_data = 0x0}, tts_off = 0, tts_fixedTupleDescriptor = true}
(gdb) n
411 return slot;
(gdb)
412 }
(gdb)
agg_fill_hash_table (aggstate=0x1f895a0) at nodeAgg.c:1924
1924 if (TupIsNull(outerslot))
(gdb)
lookup_hash_entries->进入lookup_hash_entries,为当前元组在所有已完成hash的grouping sets中检索hash条目,为后续的advance_aggregates函数调用返回pergroup指针数组.
(gdb) n
1928 tmpcontext->ecxt_outertuple = outerslot;
(gdb)
1931 lookup_hash_entries(aggstate);
(gdb)
(gdb) step
lookup_hash_entries (aggstate=0x1f895a0) at nodeAgg.c:1509
1509 int numHashes = aggstate->num_hashes;
(gdb) n
1510 AggStatePerGroup *pergroup = aggstate->hash_pergroup;
(gdb) p numHashes
$5 = 1
(gdb) n
1513 for (setno = 0; setno < numHashes; setno++)
(gdb) p *pergroup
$6 = (AggStatePerGroup) 0x0
(gdb) n
1515 select_current_set(aggstate, setno, true);
(gdb) step
select_current_set (aggstate=0x1f895a0, setno=0, is_hash=true) at nodeAgg.c:306
306 if (is_hash)
(gdb) n
307 aggstate->curaggcontext = aggstate->hashcontext;
(gdb)
311 aggstate->current_set = setno;
(gdb)
312 }
(gdb)
lookup_hash_entries (aggstate=0x1f895a0) at nodeAgg.c:1516
1516 pergroup[setno] = lookup_hash_entry(aggstate)->additional;
(gdb)
lookup_hash_entry->调用lookup_hash_entry,该函数为包含当前元组的组检索或创建哈希表条目.
(gdb) step
lookup_hash_entry (aggstate=0x1f895a0) at nodeAgg.c:1451
1451 TupleTableSlot *inputslot = aggstate->tmpcontext->ecxt_outertuple;
(gdb) n
1452 AggStatePerHash perhash = &aggstate->perhash[aggstate->current_set];
(gdb) p aggstate->current_set
$7 = 0
(gdb) n
1453 TupleTableSlot *hashslot = perhash->hashslot;
(gdb) p *perhash
$8 = {hashtable = 0x1f9fc98, hashiter = {cur = 0, end = 0, done = false}, hashslot = 0x1f8b198, hashfunctions = 0x1f8b230,
eqfuncoids = 0x1f9fc50, numCols = 1, numhashGrpCols = 1, largestGrpColIdx = 1, hashGrpColIdxInput = 0x1f9fbb0,
hashGrpColIdxHash = 0x1f9fbd0, aggnode = 0x1f7b1e0}
(gdb) n
1459 slot_getsomeattrs(inputslot, perhash->largestGrpColIdx);
(gdb)
1460 ExecClearTuple(hashslot);
(gdb) p *perhash
$9 = {hashtable = 0x1f9fc98, hashiter = {cur = 0, end = 0, done = false}, hashslot = 0x1f8b198, hashfunctions = 0x1f8b230,
eqfuncoids = 0x1f9fc50, numCols = 1, numhashGrpCols = 1, largestGrpColIdx = 1, hashGrpColIdxInput = 0x1f9fbb0,
hashGrpColIdxHash = 0x1f9fbd0, aggnode = 0x1f7b1e0}
(gdb) p *perhash->hashslot
$10 = {type = T_TupleTableSlot, tts_isempty = true, tts_shouldFree = false, tts_shouldFreeMin = false, tts_slow = false,
tts_tuple = 0x0, tts_tupleDescriptor = 0x1f8b080, tts_mcxt = 0x1f89270, tts_buffer = 0, tts_nvalid = 0,
tts_values = 0x1f8b1f8, tts_isnull = 0x1f8b200, tts_mintuple = 0x0, tts_minhdr = {t_len = 0, t_self = {ip_blkid = {
bi_hi = 0, bi_lo = 0}, ip_posid = 0}, t_tableOid = 0, t_data = 0x0}, tts_off = 0, tts_fixedTupleDescriptor = true}
(gdb) p *perhash->hashfunctions
$11 = {fn_addr = 0x4c8a31 , fn_oid = 400, fn_nargs = 1, fn_strict = true, fn_retset = false, fn_stats = 2 '\002',
fn_extra = 0x0, fn_mcxt = 0x1f89270, fn_expr = 0x0}
(gdb) p *perhash->eqfuncoids
$12 = 67
(gdb) p *perhash->hashGrpColIdxInput
$13 = 1
(gdb) p *perhash->hashGrpColIdxHash
$14 = 1
(gdb) p *perhash->aggnode
$15 = {plan = {type = T_Agg, startup_cost = 13677, total_cost = 13677.0625, plan_rows = 5, plan_width = 45,
parallel_aware = false, parallel_safe = false, plan_node_id = 0, targetlist = 0x1f84108, qual = 0x0,
lefttree = 0x1f83bc8, righttree = 0x0, initPlan = 0x0, extParam = 0x0, allParam = 0x0}, aggstrategy = AGG_HASHED,
aggsplit = AGGSPLIT_SIMPLE, numCols = 1, grpColIdx = 0x1f83eb8, grpOperators = 0x1f83e98, numGroups = 5, aggParams = 0x0,
groupingSets = 0x0, chain = 0x0}
(gdb)
lookup_hash_entry->遍历分组键(这里是bh列)
(gdb) n
1462 for (i = 0; i < perhash->numhashGrpCols; i++)
(gdb)
1464 int varNumber = perhash->hashGrpColIdxInput[i] - 1;
(gdb)
1466 hashslot->tts_values[i] = inputslot->tts_values[varNumber];
(gdb) p varNumber
$16 = 0
(gdb) n
1467 hashslot->tts_isnull[i] = inputslot->tts_isnull[varNumber];
(gdb)
1462 for (i = 0; i < perhash->numhashGrpCols; i++)
(gdb) p *hashslot
$17 = {type = T_TupleTableSlot, tts_isempty = true, tts_shouldFree = false, tts_shouldFreeMin = false, tts_slow = false,
tts_tuple = 0x0, tts_tupleDescriptor = 0x1f8b080, tts_mcxt = 0x1f89270, tts_buffer = 0, tts_nvalid = 0,
tts_values = 0x1f8b1f8, tts_isnull = 0x1f8b200, tts_mintuple = 0x0, tts_minhdr = {t_len = 0, t_self = {ip_blkid = {
bi_hi = 0, bi_lo = 0}, ip_posid = 0}, t_tableOid = 0, t_data = 0x0}, tts_off = 0, tts_fixedTupleDescriptor = true}
(gdb) p *hashslot->tts_values[0]
$18 = 811222795
(gdb)
lookup_hash_entry->调用LookupTupleHashEntry,该函数使用已过滤的元组检索或者创建哈希表条目
(gdb) step
LookupTupleHashEntry (hashtable=0x1f9fc98, slot=0x1f8b198, isnew=0x7fff7f065e17) at execGrouping.c:290
290 oldContext = MemoryContextSwitchTo(hashtable->tempcxt);
(gdb) p *hashtable
$19 = {hashtab = 0x1f9fd30, numCols = 1, keyColIdx = 0x1f9fbd0, tab_hash_funcs = 0x1f8b230, tab_eq_func = 0x1fa0050,
tablecxt = 0x1f91370, tempcxt = 0x1f9d8e0, entrysize = 24, tableslot = 0x1f9ffb8, inputslot = 0x0, in_hash_funcs = 0x0,
cur_eq_func = 0x0, hash_iv = 0, exprcontext = 0x1fa0970}
(gdb) p *hashtable->hashtab
$20 = {size = 8, members = 0, sizemask = 7, grow_threshold = 7, data = 0x1f9fd88, ctx = 0x1f89270, private_data = 0x1f9fc98}
(gdb) p *hashtable->keyColIdx
$21 = 1
(gdb) p *hashtable->tab_hash_funcs
$22 = {fn_addr = 0x4c8a31 , fn_oid = 400, fn_nargs = 1, fn_strict = true, fn_retset = false, fn_stats = 2 '\002',
fn_extra = 0x0, fn_mcxt = 0x1f89270, fn_expr = 0x0}
(gdb) n
293 hashtable->inputslot = slot;
(gdb)
294 hashtable->in_hash_funcs = hashtable->tab_hash_funcs;
(gdb)
295 hashtable->cur_eq_func = hashtable->tab_eq_func;
(gdb)
297 key = NULL; /* flag to reference inputslot */
(gdb)
299 if (isnew)
(gdb)
301 entry = tuplehash_insert(hashtable->hashtab, key, &found);
(gdb) step
tuplehash_insert (tb=0x1f9fd30, key=0x0, found=0x7fff7f065dd7) at ../../../src/include/lib/simplehash.h:490
490 uint32 hash = SH_HASH_KEY(tb, key);
(gdb) finish
Run till exit from #0 tuplehash_insert (tb=0x1f9fd30, key=0x0, found=0x7fff7f065dd7)
at ../../../src/include/lib/simplehash.h:490
0x00000000006d3a1e in LookupTupleHashEntry (hashtable=0x1f9fc98, slot=0x1f8b198, isnew=0x7fff7f065e17) at execGrouping.c:301
301 entry = tuplehash_insert(hashtable->hashtab, key, &found);
Value returned is $23 = (TupleHashEntryData *) 0x1f9fdb8
(gdb) n
303 if (found)
(gdb) p found
$24 = false
(gdb)
LookupTupleHashEntry->插入新条目,返回entry
(gdb) n
311 *isnew = true;
(gdb)
313 entry->additional = NULL;
(gdb)
314 MemoryContextSwitchTo(hashtable->tablecxt);
(gdb)
316 entry->firstTuple = ExecCopySlotMinimalTuple(slot);
(gdb)
324 MemoryContextSwitchTo(oldContext);
(gdb)
326 return entry;
(gdb)
327 }
lookup_hash_entry->回到lookup_hash_entry
(gdb)
lookup_hash_entry (aggstate=0x1f895a0) at nodeAgg.c:1474
1474 if (isnew)
(gdb)
lookup_hash_entry->分配内存,设置条目的额外信息
(gdb) n
1481 sizeof(AggStatePerGroupData) * aggstate->numtrans);
(gdb) p *entry
$25 = {firstTuple = 0x1f91488, additional = 0x0, status = 1, hash = 443809650}
(gdb) n
1480 MemoryContextAlloc(perhash->hashtable->tablecxt,
(gdb)
1479 pergroup = (AggStatePerGroup)
(gdb)
1482 entry->additional = pergroup;
(gdb)
lookup_hash_entry->为新元组group初始化聚合操作, lookup_hash_entries()已选择了相应的grouping set(这里有3个聚合列)
1488 for (transno = 0; transno < aggstate->numtrans; transno++)
(gdb) p aggstate->numtrans
$26 = 3
(gdb)
(gdb) n
1490 AggStatePerTrans pertrans = &aggstate->pertrans[transno];
(gdb)
1491 AggStatePerGroup pergroupstate = &pergroup[transno];
(gdb) p *pertrans
$27 = {aggref = 0x1f84650, aggshared = false, numInputs = 1, numTransInputs = 1, transfn_oid = 768, serialfn_oid = 0,
deserialfn_oid = 0, aggtranstype = 23, transfn = {fn_addr = 0x93e877 , fn_oid = 768, fn_nargs = 2,
fn_strict = true, fn_retset = false, fn_stats = 2 '\002', fn_extra = 0x0, fn_mcxt = 0x1f89270, fn_expr = 0x1fa0b00},
serialfn = {fn_addr = 0x0, fn_oid = 0, fn_nargs = 0, fn_strict = false, fn_retset = false, fn_stats = 0 '\000',
fn_extra = 0x0, fn_mcxt = 0x0, fn_expr = 0x0}, deserialfn = {fn_addr = 0x0, fn_oid = 0, fn_nargs = 0,
fn_strict = false, fn_retset = false, fn_stats = 0 '\000', fn_extra = 0x0, fn_mcxt = 0x0, fn_expr = 0x0},
aggCollation = 0, numSortCols = 0, numDistinctCols = 0, sortColIdx = 0x0, sortOperators = 0x0, sortCollations = 0x0,
sortNullsFirst = 0x0, equalfnOne = {fn_addr = 0x0, fn_oid = 0, fn_nargs = 0, fn_strict = false, fn_retset = false,
fn_stats = 0 '\000', fn_extra = 0x0, fn_mcxt = 0x0, fn_expr = 0x0}, equalfnMulti = 0x0, initValue = 0,
initValueIsNull = true, inputtypeLen = 0, transtypeLen = 4, inputtypeByVal = false, transtypeByVal = true,
sortslot = 0x0, uniqslot = 0x0, sortdesc = 0x0, sortstates = 0x1f9fb70, transfn_fcinfo = {flinfo = 0x1f99418,
context = 0x1f895a0, resultinfo = 0x0, fncollation = 0, isnull = false, nargs = 2, arg = {0 },
argnull = {false }}, serialfn_fcinfo = {flinfo = 0x0, context = 0x0, resultinfo = 0x0,
fncollation = 0, isnull = false, nargs = 0, arg = {0 }, argnull = {false }},
deserialfn_fcinfo = {flinfo = 0x0, context = 0x0, resultinfo = 0x0, fncollation = 0, isnull = false, nargs = 0, arg = {
0 }, argnull = {false }}}
(gdb) n
1493 initialize_aggregate(aggstate, pertrans, pergroupstate);
(gdb) p *pergroupstate
$28 = {transValue = 9187201950435737471, transValueIsNull = 127, noTransValue = 127}
(gdb) n
1488 for (transno = 0; transno < aggstate->numtrans; transno++)
(gdb) p *aggstate
$29 = {ss = {ps = {type = T_AggState, plan = 0x1f7b1e0, state = 0x1f89388, ExecProcNode = 0x6ee438 ,
ExecProcNodeReal = 0x6ee438 , instrument = 0x0, worker_instrument = 0x0, worker_jit_instrument = 0x0,
qual = 0x0, lefttree = 0x1f89b10, righttree = 0x0, initPlan = 0x0, subPlan = 0x0, chgParam = 0x0,
ps_ResultTupleSlot = 0x1f8a710, ps_ExprContext = 0x1f89a50, ps_ProjInfo = 0x1f8a850, scandesc = 0x1f89e60},
ss_currentRelation = 0x0, ss_currentScanDesc = 0x0, ss_ScanTupleSlot = 0x1f8a3b8}, aggs = 0x1f8ad60, numaggs = 3,
numtrans = 3, aggstrategy = AGG_HASHED, aggsplit = AGGSPLIT_SIMPLE, phase = 0x1f8ae58, numphases = 1, current_phase = 0,
peragg = 0x1f9f930, pertrans = 0x1f993f0, hashcontext = 0x1f89990, aggcontexts = 0x1f897b8, tmpcontext = 0x1f897d8,
curaggcontext = 0x1f89990, curperagg = 0x0, curpertrans = 0x0, input_done = false, agg_done = false, projected_set = -1,
current_set = 0, grouped_cols = 0x0, all_grouped_cols = 0x1f8aff0, maxsets = 1, phases = 0x1f8ae58, sort_in = 0x0,
sort_out = 0x0, sort_slot = 0x0, pergroups = 0x0, grp_firstTuple = 0x0, table_filled = false, num_hashes = 1,
perhash = 0x1f8aeb0, hash_pergroup = 0x1f9fb48, all_pergroups = 0x1f9fb48, combinedproj = 0x0}
(gdb) n
1490 AggStatePerTrans pertrans = &aggstate->pertrans[transno];
(gdb)
1491 AggStatePerGroup pergroupstate = &pergroup[transno];
(gdb)
1493 initialize_aggregate(aggstate, pertrans, pergroupstate);
(gdb)
1488 for (transno = 0; transno < aggstate->numtrans; transno++)
(gdb)
1490 AggStatePerTrans pertrans = &aggstate->pertrans[transno];
(gdb)
1491 AggStatePerGroup pergroupstate = &pergroup[transno];
(gdb)
1493 initialize_aggregate(aggstate, pertrans, pergroupstate);
(gdb)
1488 for (transno = 0; transno < aggstate->numtrans; transno++)
(gdb)
1497 return entry;
(gdb)
1498 }
(gdb)
lookup_hash_entries->回到lookup_hash_entries
(gdb) n
lookup_hash_entries (aggstate=0x1f895a0) at nodeAgg.c:1513
1513 for (setno = 0; setno < numHashes; setno++)
agg_fill_hash_table->回到agg_fill_hash_table
(gdb) n
1518 }
(gdb)
agg_fill_hash_table (aggstate=0x1f895a0) at nodeAgg.c:1934
1934 advance_aggregates(aggstate);
(gdb)
advance_aggregates->进入advance_aggregates
(gdb) step
advance_aggregates (aggstate=0x1f895a0) at nodeAgg.c:680
680 ExecEvalExprSwitchContext(aggstate->phase->evaltrans,
(gdb) p *aggstate->phase->evaltrans
$30 = {tag = {type = T_ExprState}, flags = 6 '\006', resnull = false, resvalue = 0, resultslot = 0x0, steps = 0x1fa10d0,
evalfunc = 0x6cd882 , expr = 0x1f895a0, evalfunc_private = 0x6cb43e ,
steps_len = 16, steps_alloc = 16, parent = 0x1f895a0, ext_params = 0x0, innermost_caseval = 0x0,
innermost_casenull = 0x0, innermost_domainval = 0x0, innermost_domainnull = 0x0}
(gdb) step
ExecEvalExprSwitchContext (state=0x1fa1038, econtext=0x1f897d8, isNull=0x7fff7f065e9f)
at ../../../src/include/executor/executor.h:312
312 oldContext = MemoryContextSwitchTo(econtext->ecxt_per_tuple_memory);
(gdb) finish
Run till exit from #0 ExecEvalExprSwitchContext (state=0x1fa1038, econtext=0x1f897d8, isNull=0x7fff7f065e9f)
at ../../../src/include/executor/executor.h:312
advance_aggregates (aggstate=0x1f895a0) at nodeAgg.c:683
683 }
Value returned is $31 = 0
进入第2轮循环
(gdb) step
agg_fill_hash_table (aggstate=0x1f895a0) at nodeAgg.c:1940
1940 ResetExprContext(aggstate->tmpcontext);
(gdb) n
1941 }
查看相关信息
(gdb) n
1941 }
(gdb)
1923 outerslot = fetch_input_tuple(aggstate);
(gdb)
1924 if (TupIsNull(outerslot))
(gdb) n
1928 tmpcontext->ecxt_outertuple = outerslot;
(gdb)
1931 lookup_hash_entries(aggstate);
(gdb)
1934 advance_aggregates(aggstate);
(gdb)
1940 ResetExprContext(aggstate->tmpcontext);
(gdb) p *outerslot
$32 = {type = T_TupleTableSlot, tts_isempty = false, tts_shouldFree = false, tts_shouldFreeMin = false, tts_slow = true,
tts_tuple = 0x1fa5998, tts_tupleDescriptor = 0x7ff7dd2d1380, tts_mcxt = 0x1f89270, tts_buffer = 124, tts_nvalid = 3,
tts_values = 0x1f89d48, tts_isnull = 0x1f89d80, tts_mintuple = 0x0, tts_minhdr = {t_len = 0, t_self = {ip_blkid = {
bi_hi = 0, bi_lo = 0}, ip_posid = 0}, t_tableOid = 0, t_data = 0x0}, tts_off = 16, tts_fixedTupleDescriptor = true}
(gdb) x/32x outerslot->tts_values
0x1f89d48: 0x28 0xf6 0x0b 0xb1 0xf7 0x7f 0x00 0x00
0x1f89d50: 0x02 0x00 0x00 0x00 0x00 0x00 0x00 0x00
0x1f89d58: 0x02 0x00 0x00 0x00 0x00 0x00 0x00 0x00
0x1f89d60: 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
tuple数据
(gdb) x/56x outerslot->tts_tuple->t_data->t_bits
0x7ff7b2e1365f: 0x00 0x0b 0x47 0x5a 0x30 0x31 0x00 0x00
0x7ff7b2e13667: 0x00 0x01 0x00 0x00 0x00 0x01 0x00 0x00
0x7ff7b2e1366f: 0x00 0x01 0x00 0x00 0x00 0x01 0x00 0x00
0x7ff7b2e13677: 0x00 0x01 0x00 0x00 0x00 0x01 0x00 0x00
0x7ff7b2e1367f: 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
0x7ff7b2e13687: 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
0x7ff7b2e1368f: 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
DONE!
四、参考资料
N/A
名称栏目：PostgreSQL源码解读（182）-查询#98(聚合函数#3-ExecAgg)
URL地址：http://azwzsj.com/article/gpdeeg.html
