From 3343b439152083def34f7de064fee204bc8d0f4d Mon Sep 17 00:00:00 2001 From: drh Date: Thu, 5 Apr 2012 01:37:32 +0000 Subject: [PATCH] Performance improvements for the RowSet object when it undergoes many cycles between RowSetInsert and RowSetTest. FossilOrigin-Name: 49d20ede5f4c0895a165126d5cf7c95a0510ba35 --- manifest | 12 +-- manifest.uuid | 2 +- src/rowset.c | 212 +++++++++++++++++++++++++++++++++++--------------- 3 files changed, 156 insertions(+), 70 deletions(-) diff --git a/manifest b/manifest index 036172c168..dbc5928c3f 100644 --- a/manifest +++ b/manifest @@ -1,5 +1,5 @@ -C Add\sthe\s".trace"\soption\sto\sthe\scommand-line\sshell. -D 2012-04-04T16:56:23.330 +C Performance\simprovements\sfor\sthe\sRowSet\sobject\swhen\sit\sundergoes\smany\ncycles\sbetween\sRowSetInsert\sand\sRowSetTest. +D 2012-04-05T01:37:32.902 F Makefile.arm-wince-mingw32ce-gcc d6df77f1f48d690bd73162294bbba7f59507c72f F Makefile.in 2f37e468503dbe79d35c9f6dffcf3fae1ae9ec20 F Makefile.linux-gcc 91d710bdc4998cb015f39edf3cb314ec4f4d7e23 @@ -180,7 +180,7 @@ F src/prepare.c ec4989f7f480544bdc4192fe663470d2a2d7d61e F src/printf.c 7ffb4ebb8b341f67e049695ba031da717b3d2699 F src/random.c cd4a67b3953b88019f8cd4ccd81394a8ddfaba50 F src/resolve.c 3d3e80a98f203ac6b9329e9621e29eda85ddfd40 -F src/rowset.c 69afa95a97c524ba6faf3805e717b5b7ae85a697 +F src/rowset.c f6a49f3e9579428024662f6e2931832511f831a1 F src/select.c 5e0e481c7d215d3c7ca8ccae1e688aa30163c6c1 F src/shell.c ce4d41582182b8fad3be364e2fa295b70bc342ab F src/sqlite.h.in 11a883919b0baf4ffaa7550cfeef99be613ec2bf @@ -1000,7 +1000,7 @@ F tool/tostr.awk e75472c2f98dd76e06b8c9c1367f4ab07e122d06 F tool/vdbe-compress.tcl d70ea6d8a19e3571d7ab8c9b75cba86d1173ff0f F tool/warnings-clang.sh 9f406d66e750e8ac031c63a9ef3248aaa347ef2a F tool/warnings.sh fbc018d67fd7395f440c28f33ef0f94420226381 -P 9e1e2fe2950bb96784413eae934314d95bce08e7 -R 7ce64940a4af6b82ea18cf10b0da61e8 +P b9ac3d7e340eb616fd23cc7dbdef6fdd66a79fe4 +R a1d15ac4a2f200ebad8c15bdd80077ea U drh -Z 783b4beff01e2d5037d4d8570a1550e1 +Z 3eabbe65a412e07d9d0e849fd433908a diff --git a/manifest.uuid b/manifest.uuid index 18daab3bde..99e0c0b36c 100644 --- a/manifest.uuid +++ b/manifest.uuid @@ -1 +1 @@ -b9ac3d7e340eb616fd23cc7dbdef6fdd66a79fe4 \ No newline at end of file +49d20ede5f4c0895a165126d5cf7c95a0510ba35 \ No newline at end of file diff --git a/src/rowset.c b/src/rowset.c index d84bb93abf..58c18b78db 100644 --- a/src/rowset.c +++ b/src/rowset.c @@ -76,6 +76,11 @@ /* ** Each entry in a RowSet is an instance of the following object. +** +** This same object is reused to store a linked list of trees of RowSetEntry +** objects. In that alternative use, pRight points to the next entry +** in the list, pLeft points to the tree, and v is unused. The +** RowSet.pForest value points to the head of this forest list. */ struct RowSetEntry { i64 v; /* ROWID value for this entry */ @@ -105,12 +110,18 @@ struct RowSet { struct RowSetEntry *pEntry; /* List of entries using pRight */ struct RowSetEntry *pLast; /* Last entry on the pEntry list */ struct RowSetEntry *pFresh; /* Source of new entry objects */ - struct RowSetEntry *pTree; /* Binary tree of entries */ + struct RowSetEntry *pForest; /* List of binary trees of entries */ u16 nFresh; /* Number of objects on pFresh */ - u8 isSorted; /* True if pEntry is sorted */ + u8 rsFlags; /* Various flags */ u8 iBatch; /* Current insert batch */ }; +/* +** Allowed values for RowSet.rsFlags +*/ +#define ROWSET_SORTED 0x01 /* True if RowSet.pEntry is sorted */ +#define ROWSET_NEXT 0x02 /* True if sqlite3RowSetNext() has been called */ + /* ** Turn bulk memory into a RowSet object. N bytes of memory ** are available at pSpace. The db pointer is used as a memory context @@ -131,10 +142,10 @@ RowSet *sqlite3RowSetInit(sqlite3 *db, void *pSpace, unsigned int N){ p->db = db; p->pEntry = 0; p->pLast = 0; - p->pTree = 0; + p->pForest = 0; p->pFresh = (struct RowSetEntry*)(ROUND8(sizeof(*p)) + (char*)p); p->nFresh = (u16)((N - ROUND8(sizeof(*p)))/sizeof(struct RowSetEntry)); - p->isSorted = 1; + p->rsFlags = ROWSET_SORTED; p->iBatch = 0; return p; } @@ -154,8 +165,33 @@ void sqlite3RowSetClear(RowSet *p){ p->nFresh = 0; p->pEntry = 0; p->pLast = 0; - p->pTree = 0; - p->isSorted = 1; + p->pForest = 0; + p->rsFlags = ROWSET_SORTED; +} + +/* +** Allocate a new RowSetEntry object that is associated with the +** given RowSet. Return a pointer to the new and completely uninitialized +** objected. +** +** In an OOM situation, the RowSet.db->mallocFailed flag is set and this +** routine returns NULL. +*/ +static struct RowSetEntry *rowSetEntryAlloc(RowSet *p){ + assert( p!=0 ); + if( p->nFresh==0 ){ + struct RowSetChunk *pNew; + pNew = sqlite3DbMallocRaw(p->db, sizeof(*pNew)); + if( pNew==0 ){ + return 0; + } + pNew->pNextChunk = p->pChunk; + p->pChunk = pNew; + p->pFresh = pNew->aEntry; + p->nFresh = ROWSET_ENTRY_PER_CHUNK; + } + p->nFresh--; + return p->pFresh++; } /* @@ -167,30 +203,21 @@ void sqlite3RowSetClear(RowSet *p){ void sqlite3RowSetInsert(RowSet *p, i64 rowid){ struct RowSetEntry *pEntry; /* The new entry */ struct RowSetEntry *pLast; /* The last prior entry */ - assert( p!=0 ); - if( p->nFresh==0 ){ - struct RowSetChunk *pNew; - pNew = sqlite3DbMallocRaw(p->db, sizeof(*pNew)); - if( pNew==0 ){ - return; - } - pNew->pNextChunk = p->pChunk; - p->pChunk = pNew; - p->pFresh = pNew->aEntry; - p->nFresh = ROWSET_ENTRY_PER_CHUNK; - } - pEntry = p->pFresh++; - p->nFresh--; + + /* This routine is never called after sqlite3RowSetNext() */ + assert( p!=0 && (p->rsFlags & ROWSET_NEXT)==0 ); + + pEntry = rowSetEntryAlloc(p); + if( pEntry==0 ) return; pEntry->v = rowid; pEntry->pRight = 0; pLast = p->pLast; if( pLast ){ - if( p->isSorted && rowid<=pLast->v ){ - p->isSorted = 0; + if( (p->rsFlags & ROWSET_SORTED)!=0 && rowid<=pLast->v ){ + p->rsFlags &= ~ROWSET_SORTED; } pLast->pRight = pEntry; }else{ - assert( p->pEntry==0 ); /* Fires if INSERT after SMALLEST */ p->pEntry = pEntry; } p->pLast = pEntry; @@ -202,7 +229,7 @@ void sqlite3RowSetInsert(RowSet *p, i64 rowid){ ** The input lists are connected via pRight pointers and are ** assumed to each already be in sorted order. */ -static struct RowSetEntry *rowSetMerge( +static struct RowSetEntry *rowSetEntryMerge( struct RowSetEntry *pA, /* First sorted list to be merged */ struct RowSetEntry *pB /* Second sorted list to be merged */ ){ @@ -236,32 +263,29 @@ static struct RowSetEntry *rowSetMerge( } /* -** Sort all elements on the pEntry list of the RowSet into ascending order. +** Sort all elements on the list of RowSetEntry objects into order of +** increasing v. */ -static void rowSetSort(RowSet *p){ +static struct RowSetEntry *rowSetEntrySort(struct RowSetEntry *pIn){ unsigned int i; - struct RowSetEntry *pEntry; - struct RowSetEntry *aBucket[40]; + struct RowSetEntry *pNext, *aBucket[40]; - assert( p->isSorted==0 ); memset(aBucket, 0, sizeof(aBucket)); - while( p->pEntry ){ - pEntry = p->pEntry; - p->pEntry = pEntry->pRight; - pEntry->pRight = 0; + while( pIn ){ + pNext = pIn->pRight; + pIn->pRight = 0; for(i=0; aBucket[i]; i++){ - pEntry = rowSetMerge(aBucket[i], pEntry); + pIn = rowSetEntryMerge(aBucket[i], pIn); aBucket[i] = 0; } - aBucket[i] = pEntry; + aBucket[i] = pIn; + pIn = pNext; } - pEntry = 0; + pIn = 0; for(i=0; ipEntry = pEntry; - p->pLast = 0; - p->isSorted = 1; + return pIn; } @@ -355,20 +379,37 @@ static struct RowSetEntry *rowSetListToTree(struct RowSetEntry *pList){ } /* -** Convert the list in p->pEntry into a sorted list if it is not -** sorted already. If there is a binary tree on p->pTree, then -** convert it into a list too and merge it into the p->pEntry list. +** Take all the entries on p->pEntry and on the trees in p->pForest and +** sort them all together into one big ordered list on p->pEntry. +** +** This routine should only be called once in the life of a RowSet. */ static void rowSetToList(RowSet *p){ - if( !p->isSorted ){ - rowSetSort(p); + + /* This routine is called only once */ + assert( p!=0 && (p->rsFlags & ROWSET_NEXT)==0 ); + + if( (p->rsFlags & ROWSET_SORTED)==0 ){ + p->pEntry = rowSetEntrySort(p->pEntry); } - if( p->pTree ){ - struct RowSetEntry *pHead, *pTail; - rowSetTreeToList(p->pTree, &pHead, &pTail); - p->pTree = 0; - p->pEntry = rowSetMerge(p->pEntry, pHead); + + /* While this module could theoretically support it, sqlite3RowSetNext() + ** is never called after sqlite3RowSetText() for the same RowSet. So + ** there is never a forest to deal with. Should this change, simply + ** remove the assert() and the #if 0. */ + assert( p->pForest==0 ); +#if 0 + while( p->pForest ){ + struct RowSetEntry *pTree = p->pForest->pLeft; + if( pTree ){ + struct RowSetEntry *pHead, *pTail; + rowSetTreeToList(pTree, &pHead, &pTail); + p->pEntry = rowSetEntryMerge(p->pEntry, pHead); + } + p->pForest = p->pForest->pRight; } +#endif + p->rsFlags |= ROWSET_NEXT; /* Verify this routine is never called again */ } /* @@ -380,7 +421,12 @@ static void rowSetToList(RowSet *p){ ** routine may not be called again. */ int sqlite3RowSetNext(RowSet *p, i64 *pRowid){ - rowSetToList(p); + assert( p!=0 ); + + /* Merge the forest into a single sorted list on first call */ + if( (p->rsFlags & ROWSET_NEXT)==0 ) rowSetToList(p); + + /* Return the next entry on the list */ if( p->pEntry ){ *pRowid = p->pEntry->v; p->pEntry = p->pEntry->pRight; @@ -396,26 +442,66 @@ int sqlite3RowSetNext(RowSet *p, i64 *pRowid){ /* ** Check to see if element iRowid was inserted into the the rowset as ** part of any insert batch prior to iBatch. Return 1 or 0. +** +** If this is the first test of a new batch and if there exist entires +** on pRowSet->pEntry, then sort those entires into the forest at +** pRowSet->pForest so that they can be tested. */ int sqlite3RowSetTest(RowSet *pRowSet, u8 iBatch, sqlite3_int64 iRowid){ - struct RowSetEntry *p; + struct RowSetEntry *p, *pTree; + + /* This routine is never called after sqlite3RowSetNext() */ + assert( pRowSet!=0 && (pRowSet->rsFlags & ROWSET_NEXT)==0 ); + + /* Sort entries into the forest on the first test of a new batch + */ if( iBatch!=pRowSet->iBatch ){ - if( pRowSet->pEntry ){ - rowSetToList(pRowSet); - pRowSet->pTree = rowSetListToTree(pRowSet->pEntry); + p = pRowSet->pEntry; + if( p ){ + struct RowSetEntry **ppPrevTree = &pRowSet->pForest; + if( (pRowSet->rsFlags & ROWSET_SORTED)==0 ){ + p = rowSetEntrySort(p); + } + for(pTree = pRowSet->pForest; pTree; pTree=pTree->pRight){ + ppPrevTree = &pTree->pRight; + if( pTree->pLeft==0 ){ + pTree->pLeft = rowSetListToTree(p); + break; + }else{ + struct RowSetEntry *pAux, *pTail; + rowSetTreeToList(pTree->pLeft, &pAux, &pTail); + pTree->pLeft = 0; + p = rowSetEntryMerge(pAux, p); + } + } + if( pTree==0 ){ + *ppPrevTree = pTree = rowSetEntryAlloc(pRowSet); + if( pTree ){ + pTree->v = 0; + pTree->pRight = 0; + pTree->pLeft = rowSetListToTree(p); + } + } pRowSet->pEntry = 0; pRowSet->pLast = 0; + pRowSet->rsFlags |= ROWSET_SORTED; } pRowSet->iBatch = iBatch; } - p = pRowSet->pTree; - while( p ){ - if( p->vpRight; - }else if( p->v>iRowid ){ - p = p->pLeft; - }else{ - return 1; + + /* Test to see if the iRowid value appears anywhere in the forest. + ** Return 1 if it does and 0 if not. + */ + for(pTree = pRowSet->pForest; pTree; pTree=pTree->pRight){ + p = pTree->pLeft; + while( p ){ + if( p->vpRight; + }else if( p->v>iRowid ){ + p = p->pLeft; + }else{ + return 1; + } } } return 0;