SQLite

/*
** 2004 April 6
**
** The author disclaims copyright to this source code.  In place of
** a legal notice, here is a blessing:
**
**    May you do good and not evil.
**    May you find forgiveness for yourself and forgive others.
**    May you share freely, never taking more than you give.
**
*************************************************************************
** $Id: btree.c,v 1.414 2007/08/28 22:24:35 drh Exp $
**
** This file implements a external (disk-based) database using BTrees.
** See the header comment on "btreeInt.h" for additional information.
** Including a description of file format and an overview of operation.
*/
#include "btreeInt.h"

** SQLITE_OK if the lock may be obtained (by calling lockTable()), or
** SQLITE_LOCKED if not.
*/
static int queryTableLock(Btree *p, Pgno iTab, u8 eLock){
  BtShared *pBt = p->pBt;
  BtLock *pIter;

  assert( sqlite3BtreeMutexHeld(p->pSqlite->mutex) );
  assert( sqlite3BtreeMutexHeld(pBt->mutex) );
  
  /* This is a no-op if the shared-cache is not enabled */
  if( !p->sharable ){
    return SQLITE_OK;
  }

** SQLITE_NOMEM may also be returned.
*/
static int lockTable(Btree *p, Pgno iTable, u8 eLock){
  BtShared *pBt = p->pBt;
  BtLock *pLock = 0;
  BtLock *pIter;

  assert( sqlite3BtreeMutexHeld(p->pSqlite->mutex) );
  assert( sqlite3BtreeMutexHeld(pBt->mutex) );

  /* This is a no-op if the shared-cache is not enabled */
  if( !p->sharable ){
    return SQLITE_OK;
  }

/*
** Release all the table locks (locks obtained via calls to the lockTable()
** procedure) held by Btree handle p.
*/
static void unlockAllTables(Btree *p){
  BtLock **ppIter = &p->pBt->pLock;

  assert( sqlite3BtreeMutexHeld(p->pSqlite->mutex) );
  assert( sqlite3BtreeMutexHeld(p->pBt->mutex) );
  assert( p->sharable || 0==*ppIter );

  while( *ppIter ){
    BtLock *pLock = *ppIter;
    if( pLock->pBtree==p ){
      *ppIter = pLock->pNext;

static void releasePage(MemPage *pPage);  /* Forward reference */

#ifndef SQLITE_OMIT_INCRBLOB
/*
** Invalidate the overflow page-list cache for cursor pCur, if any.
*/
static void invalidateOverflowCache(BtCursor *pCur){
  assert( sqlite3BtreeMutexHeld(pCur->pBt->mutex) );
  sqlite3_free(pCur->aOverflow);
  pCur->aOverflow = 0;
}

/*
** Invalidate the overflow page-list cache for all cursors opened
** on the shared btree structure pBt.

** and BtCursor.pKey. The cursor's state is set to CURSOR_REQUIRESEEK.
*/
static int saveCursorPosition(BtCursor *pCur){
  int rc;

  assert( CURSOR_VALID==pCur->eState );
  assert( 0==pCur->pKey );
  assert( sqlite3BtreeMutexHeld(pCur->pBt->mutex) );

  rc = sqlite3BtreeKeySize(pCur, &pCur->nKey);

  /* If this is an intKey table, then the above call to BtreeKeySize()
  ** stores the integer key in pCur->nKey. In this case this value is
  ** all that is required. Otherwise, if pCur is not open on an intKey
  ** table, then malloc space for and store the pCur->nKey bytes of key 

** Save the positions of all cursors except pExcept open on the table 
** with root-page iRoot. Usually, this is called just before cursor
** pExcept is used to modify the table (BtreeDelete() or BtreeInsert()).
*/
static int saveAllCursors(BtShared *pBt, Pgno iRoot, BtCursor *pExcept){
  BtCursor *p;
  assert( sqlite3BtreeMutexHeld(pBt->mutex) );
  assert( pExcept==0 || pExcept->pBt==pBt );
  for(p=pBt->pCursor; p; p=p->pNext){
    if( p!=pExcept && (0==iRoot || p->pgnoRoot==iRoot) && 
        p->eState==CURSOR_VALID ){
      int rc = saveCursorPosition(p);
      if( SQLITE_OK!=rc ){
        return rc;
      }
    }
  }
  return SQLITE_OK;
}

/*
** Clear the current cursor position.
*/
static void clearCursorPosition(BtCursor *pCur){
  assert( sqlite3BtreeMutexHeld(pCur->pBt->mutex) );
  sqlite3_free(pCur->pKey);
  pCur->pKey = 0;
  pCur->eState = CURSOR_INVALID;
}

/*
** Restore the cursor to the position it was in (or as close to as possible)
** when saveCursorPosition() was called. Note that this call deletes the 
** saved position info stored by saveCursorPosition(), so there can be
** at most one effective restoreOrClearCursorPosition() call after each 
** saveCursorPosition().
**
** If the second argument argument - doSeek - is false, then instead of 
** returning the cursor to it's saved position, any saved position is deleted
** and the cursor state set to CURSOR_INVALID.
*/
int sqlite3BtreeRestoreOrClearCursorPosition(BtCursor *pCur){
  int rc;
  assert( sqlite3BtreeMutexHeld(pCur->pBt->mutex) );
  assert( pCur->eState==CURSOR_REQUIRESEEK );
#ifndef SQLITE_OMIT_INCRBLOB
  if( pCur->isIncrblobHandle ){
    return SQLITE_ABORT;
  }
#endif
  pCur->eState = CURSOR_INVALID;

** reaches zero.  We need to unref the pParent pointer when that
** happens.
*/
static void pageDestructor(DbPage *pData, int pageSize){
  MemPage *pPage;
  assert( (pageSize & 7)==0 );
  pPage = (MemPage *)sqlite3PagerGetExtra(pData);
  assert( pPage->isInit==0 || sqlite3BtreeMutexHeld(pPage->pBt->mutex) );
  if( pPage->pParent ){
    MemPage *pParent = pPage->pParent;
    assert( pPage->isInit==1 );
    assert( pParent->pBt==pPage->pBt );
    pPage->pParent = 0;
    releasePage(pParent);
  }
  pPage->isInit = 0;
}

/*

  #ifdef SQLITE_OMIT_MEMORYDB
    const int isMemdb = 0;
  #else
    const int isMemdb = zFilename && !strcmp(zFilename, ":memory:");
  #endif
#endif

  assert( pSqlite!=0 );
  assert( sqlite3_mutex_held(pSqlite->mutex) );




  pVfs = pSqlite->pVfs;
  p = sqlite3MallocZero(sizeof(Btree));
  if( !p ){
    return SQLITE_NOMEM;
  }
  p->inTrans = TRANS_NONE;
  p->pSqlite = pSqlite;

#if !defined(SQLITE_OMIT_SHARED_CACHE) && !defined(SQLITE_OMIT_DISKIO)
  /*
  ** If this Btree is a candidate for shared cache, try to find an
  ** existing BtShared object that we can share with
  */
  if( (flags & BTREE_PRIVATE)==0
   && isMemdb==0
   && (pSqlite->flags & SQLITE_Vtab)==0
   && zFilename && zFilename[0]
   && sqlite3SharedCacheEnabled
  ){
    char *zFullPathname = (char *)sqlite3_malloc(pVfs->mxPathname);
    sqlite3_mutex *mutexShared;
    p->sharable = 1;
    if( pSqlite ){
      pSqlite->flags |= SQLITE_SharedCache;
    }
    if( !zFullPathname ){
      sqlite3_free(p);
      return SQLITE_NOMEM;
    }
    sqlite3OsFullPathname(pVfs, zFilename, zFullPathname);
    mutexShared = sqlite3_mutex_alloc(SQLITE_MUTEX_STATIC_MASTER);
    sqlite3_mutex_enter(mutexShared);
    for(pBt=sqlite3SharedCacheList; pBt; pBt=pBt->pNext){
      assert( pBt->nRef>0 );
      if( 0==strcmp(zFullPathname, sqlite3PagerFilename(pBt->pPager))
               && sqlite3PagerVfs(pBt->pPager)==pVfs ){
        p->pBt = pBt;
        pBt->nRef++;
        break;
      }
    }
    sqlite3_mutex_leave(mutexShared);
    sqlite3_free(zFullPathname);

** Close an open database and invalidate all cursors.
*/
int sqlite3BtreeClose(Btree *p){
  BtShared *pBt = p->pBt;
  BtCursor *pCur;

  /* Close all cursors opened via this handle.  */
  assert( sqlite3_mutex_held(p->pSqlite->mutex) );
  sqlite3BtreeEnter(p);
  pCur = pBt->pCursor;
  while( pCur ){
    BtCursor *pTmp = pCur;
    pCur = pCur->pNext;
    if( pTmp->pBtree==p ){
      sqlite3BtreeCloseCursor(pTmp);

  return SQLITE_OK;
}

#if SQLITE_THREADSAFE && !defined(SQLITE_OMIT_SHARED_CACHE)
/*
** Short-cuts for entering and leaving mutexes on a cursor.
*/
# define cursorEnter(X) assert( sqlite3_mutex_held(X->pBt->mutex) )
# define cursorLeave(X)





#else
# define cursorEnter(X)
# define cursorLeave(X)
#endif /* !SQLITE_OMIT_SHARED_CACHE */

/*
** Change the busy handler callback function.
*/
int sqlite3BtreeSetBusyHandler(Btree *p, BusyHandler *pHandler){
  BtShared *pBt = p->pBt;
  assert( sqlite3_mutex_held(p->pSqlite->mutex) );
  sqlite3BtreeEnter(p);
  pBt->pBusyHandler = pHandler;
  sqlite3PagerSetBusyhandler(pBt->pPager, pHandler);
  sqlite3BtreeLeave(p);
  return SQLITE_OK;
}

** an abrupt power failure when synchronous is off, the database
** could be left in an inconsistent and unrecoverable state.
** Synchronous is on by default so database corruption is not
** normally a worry.
*/
int sqlite3BtreeSetCacheSize(Btree *p, int mxPage){
  BtShared *pBt = p->pBt;
  assert( sqlite3_mutex_held(p->pSqlite->mutex) );
  sqlite3BtreeEnter(p);
  sqlite3PagerSetCachesize(pBt->pPager, mxPage);
  sqlite3BtreeLeave(p);
  return SQLITE_OK;
}

/*
** Change the way data is synced to disk in order to increase or decrease
** how well the database resists damage due to OS crashes and power
** failures.  Level 1 is the same as asynchronous (no syncs() occur and
** there is a high probability of damage)  Level 2 is the default.  There
** is a very low but non-zero probability of damage.  Level 3 reduces the
** probability of damage to near zero but with a write performance reduction.
*/
#ifndef SQLITE_OMIT_PAGER_PRAGMAS
int sqlite3BtreeSetSafetyLevel(Btree *p, int level, int fullSync){
  BtShared *pBt = p->pBt;
  assert( sqlite3_mutex_held(p->pSqlite->mutex) );
  sqlite3BtreeEnter(p);
  sqlite3PagerSetSafetyLevel(pBt->pPager, level, fullSync);
  sqlite3BtreeLeave(p);
  return SQLITE_OK;
}
#endif

/*
** Return TRUE if the given btree is set to safety level 1.  In other
** words, return TRUE if no sync() occurs on the disk files.
*/
int sqlite3BtreeSyncDisabled(Btree *p){
  BtShared *pBt = p->pBt;
  int rc;
  assert( sqlite3_mutex_held(p->pSqlite->mutex) );  
  sqlite3BtreeEnter(p);
  assert( pBt && pBt->pPager );
  rc = sqlite3PagerNosync(pBt->pPager);
  sqlite3BtreeLeave(p);
  return rc;
}

#if !defined(SQLITE_OMIT_PAGER_PRAGMAS) || !defined(SQLITE_OMIT_VACUUM)
/*

  Pgno iDbPage = pDbPage->pgno;
  Pager *pPager = pBt->pPager;
  int rc;

  assert( eType==PTRMAP_OVERFLOW2 || eType==PTRMAP_OVERFLOW1 || 
      eType==PTRMAP_BTREE || eType==PTRMAP_ROOTPAGE );
  assert( sqlite3BtreeMutexHeld(pBt->mutex) );
  assert( pDbPage->pBt==pBt );

  /* Move page iDbPage from it's current location to page number iFreePage */
  TRACE(("AUTOVACUUM: Moving %d to free page %d (ptr page %d type %d)\n", 
      iDbPage, iFreePage, iPtrPage, eType));
  rc = sqlite3PagerMovepage(pPager, pDbPage->pDbPage, iFreePage);
  if( rc!=SQLITE_OK ){
    return rc;

  BtCursor **ppCur                            /* Write new cursor here */
){
  int rc;
  BtCursor *pCur;
  BtShared *pBt = p->pBt;

  assert( sqlite3BtreeMutexHeld(pBt->mutex) );
  assert( sqlite3BtreeMutexHeld(p->pSqlite->mutex) );
  *ppCur = 0;
  if( wrFlag ){
    if( pBt->readOnly ){
      return SQLITE_READONLY;
    }
    if( checkReadLocks(p, iTable, 0) ){
      return SQLITE_LOCKED;

  /* Now that no other errors can occur, finish filling in the BtCursor
  ** variables, link the cursor into the BtShared list and set *ppCur (the
  ** output argument to this function).
  */
  pCur->xCompare = xCmp ? xCmp : dfltCompare;
  pCur->pArg = pArg;
  pCur->pBtree = p;
  pCur->pBt = pBt;
  pCur->wrFlag = wrFlag;
  pCur->pNext = pBt->pCursor;
  if( pCur->pNext ){
    pCur->pNext->pPrev = pCur;
  }
  pBt->pCursor = pCur;
  pCur->eState = CURSOR_INVALID;

/*
** Close a cursor.  The read lock on the database file is released
** when the last cursor is closed.
*/
int sqlite3BtreeCloseCursor(BtCursor *pCur){
  BtShared *pBt = pCur->pBt;

  assert( sqlite3_mutex_held(pCur->pBt->mutex) );
  assert( sqlite3_mutex_held(pCur->pBtree->pSqlite->mutex) );
  clearCursorPosition(pCur);
  if( pCur->pPrev ){
    pCur->pPrev->pNext = pCur->pNext;
  }else{
    pBt->pCursor = pCur->pNext;
  }
  if( pCur->pNext ){
    pCur->pNext->pPrev = pCur->pPrev;
  }
  releasePage(pCur->pPage);
  unlockBtreeIfUnused(pBt);
  invalidateOverflowCache(pCur);

  sqlite3_free(pCur);
  return SQLITE_OK;
}

/*
** Make a temporary cursor by filling in the fields of pTempCur.
** The temporary cursor is not on the cursor list for the Btree.
*/
void sqlite3BtreeGetTempCursor(BtCursor *pCur, BtCursor *pTempCur){
  cursorEnter(pCur);
  memcpy(pTempCur, pCur, sizeof(*pCur));
  pTempCur->pNext = 0;
  pTempCur->pPrev = 0;
  if( pTempCur->pPage ){

    sqlite3PagerRef(pTempCur->pPage->pDbPage);

  }
  cursorLeave(pCur);
}

/*
** Delete a temporary cursor such as was made by the CreateTemporaryCursor()
** function above.
*/
void sqlite3BtreeReleaseTempCursor(BtCursor *pCur){

  cursorEnter(pCur);
  if( pCur->pPage ){
    sqlite3PagerUnref(pCur->pPage->pDbPage);

  }
  cursorLeave(pCur);
}

/*
** Make sure the BtCursor* given in the argument has a valid
** BtCursor.info structure.  If it is not already valid, call
** sqlite3BtreeParseCell() to fill it in.
**

**
** For a table with the INTKEY flag set, this routine returns the key
** itself, not the number of bytes in the key.
*/
int sqlite3BtreeKeySize(BtCursor *pCur, i64 *pSize){
  int rc;

  assert( sqlite3_mutex_held(pCur->pBt->mutex) );
  rc = restoreOrClearCursorPosition(pCur);
  if( rc==SQLITE_OK ){
    assert( pCur->eState==CURSOR_INVALID || pCur->eState==CURSOR_VALID );
    if( pCur->eState==CURSOR_INVALID ){
      *pSize = 0;
    }else{
      getCellInfo(pCur);
      *pSize = pCur->info.nKey;
    }
  }

  return rc;
}

/*
** Set *pSize to the number of bytes of data in the entry the
** cursor currently points to.  Always return SQLITE_OK.
** Failure is not possible.  If the cursor is not currently
** pointing to an entry (which can happen, for example, if
** the database is empty) then *pSize is set to 0.
*/
int sqlite3BtreeDataSize(BtCursor *pCur, u32 *pSize){
  int rc;

  assert( sqlite3_mutex_held(pCur->pBt->mutex) );
  rc = restoreOrClearCursorPosition(pCur);
  if( rc==SQLITE_OK ){
    assert( pCur->eState==CURSOR_INVALID || pCur->eState==CURSOR_VALID );
    if( pCur->eState==CURSOR_INVALID ){
      /* Not pointing at a valid entry - set *pSize to 0. */
      *pSize = 0;
    }else{
      getCellInfo(pCur);
      *pSize = pCur->info.nData;
    }
  }

  return rc;
}

/*
** Given the page number of an overflow page in the database (parameter
** ovfl), this function finds the page number of the next page in the 
** linked list of overflow pages. If possible, it uses the auto-vacuum

  int skipKey,         /* offset begins at data if this is true */
  int eOp              /* zero to read. non-zero to write. */
){
  unsigned char *aPayload;
  int rc = SQLITE_OK;
  u32 nKey;
  int iIdx = 0;
  MemPage *pPage = pCur->pPage;     /* Btree page of current cursor entry */
  BtShared *pBt = pCur->pBt;        /* Btree this cursor belongs to */

  assert( pPage );
  assert( pCur->eState==CURSOR_VALID );
  assert( pCur->idx>=0 && pCur->idx<pPage->nCell );
  assert( offset>=0 );
  assert( sqlite3BtreeMutexHeld(pCur->pBt->mutex) );

  getCellInfo(pCur);
  aPayload = pCur->info.pCell + pCur->info.nHeader;
  nKey = (pPage->intKey ? 0 : pCur->info.nKey);

  if( skipKey ){
    offset += nKey;

** Return SQLITE_OK on success or an error code if anything goes
** wrong.  An error is returned if "offset+amt" is larger than
** the available payload.
*/
int sqlite3BtreeKey(BtCursor *pCur, u32 offset, u32 amt, void *pBuf){
  int rc;

  cursorEnter(pCur);
  rc = restoreOrClearCursorPosition(pCur);
  if( rc==SQLITE_OK ){
    assert( pCur->eState==CURSOR_VALID );
    assert( pCur->pPage!=0 );
    if( pCur->pPage->intKey ){
      cursorLeave(pCur);
      return SQLITE_CORRUPT_BKPT;
    }
    assert( pCur->pPage->intKey==0 );
    assert( pCur->idx>=0 && pCur->idx<pCur->pPage->nCell );
    rc = accessPayload(pCur, offset, amt, (unsigned char*)pBuf, 0, 0);
  }
  cursorLeave(pCur);
  return rc;
}

/*
** Read part of the data associated with cursor pCur.  Exactly
** "amt" bytes will be transfered into pBuf[].  The transfer
** begins at "offset".
**
** Return SQLITE_OK on success or an error code if anything goes
** wrong.  An error is returned if "offset+amt" is larger than
** the available payload.
*/
int sqlite3BtreeData(BtCursor *pCur, u32 offset, u32 amt, void *pBuf){
  int rc;

  cursorEnter(pCur);
  rc = restoreOrClearCursorPosition(pCur);
  if( rc==SQLITE_OK ){
    assert( pCur->eState==CURSOR_VALID );
    assert( pCur->pPage!=0 );
    assert( pCur->idx>=0 && pCur->idx<pCur->pPage->nCell );
    rc = accessPayload(pCur, offset, amt, pBuf, 1, 0);
  }
  cursorLeave(pCur);
  return rc;
}

/*
** Return a pointer to payload information from the entry that the 
** pCur cursor is pointing to.  The pointer is to the beginning of
** the key if skipKey==0 and it points to the beginning of data if

  unsigned char *aPayload;
  MemPage *pPage;
  u32 nKey;
  int nLocal;

  assert( pCur!=0 && pCur->pPage!=0 );
  assert( pCur->eState==CURSOR_VALID );
  assert( sqlite3BtreeMutexHeld(pCur->pBt->mutex) );
  pPage = pCur->pPage;
  assert( pCur->idx>=0 && pCur->idx<pPage->nCell );
  getCellInfo(pCur);
  aPayload = pCur->info.pCell;
  aPayload += pCur->info.nHeader;
  if( pPage->intKey ){
    nKey = 0;

** Hence, a mutex on the BtShared should be held prior to calling
** this routine.
**
** These routines is used to get quick access to key and data
** in the common case where no overflow pages are used.
*/
const void *sqlite3BtreeKeyFetch(BtCursor *pCur, int *pAmt){
  assert( sqlite3BtreeMutexHeld(pCur->pBt->mutex) );
  if( pCur->eState==CURSOR_VALID ){
    return (const void*)fetchPayload(pCur, pAmt, 0);
  }
  return 0;
}
const void *sqlite3BtreeDataFetch(BtCursor *pCur, int *pAmt){
  assert( sqlite3BtreeMutexHeld(pCur->pBt->mutex) );
  if( pCur->eState==CURSOR_VALID ){
    return (const void*)fetchPayload(pCur, pAmt, 1);
  }
  return 0;
}


/*
** Move the cursor down to a new child page.  The newPgno argument is the
** page number of the child page to move to.
*/
static int moveToChild(BtCursor *pCur, u32 newPgno){
  int rc;
  MemPage *pNewPage;
  MemPage *pOldPage;
  BtShared *pBt = pCur->pBt;

  assert( sqlite3BtreeMutexHeld(pBt->mutex) );
  assert( pCur->eState==CURSOR_VALID );
  rc = getAndInitPage(pBt, newPgno, &pNewPage, pCur->pPage);
  if( rc ) return rc;
  pNewPage->idxParent = pCur->idx;
  pOldPage = pCur->pPage;

** the largest cell index.
*/
void sqlite3BtreeMoveToParent(BtCursor *pCur){
  MemPage *pParent;
  MemPage *pPage;
  int idxParent;

  cursorEnter(pCur);
  assert( pCur->eState==CURSOR_VALID );
  pPage = pCur->pPage;
  assert( pPage!=0 );
  assert( !sqlite3BtreeIsRootPage(pPage) );
  pParent = pPage->pParent;
  assert( pParent!=0 );
  idxParent = pPage->idxParent;
  sqlite3PagerRef(pParent->pDbPage);
  releasePage(pPage);
  pCur->pPage = pParent;
  pCur->info.nSize = 0;
  assert( pParent->idxShift==0 );
  pCur->idx = idxParent;
  cursorLeave(pCur);
}

/*
** Move the cursor to the root page
*/
static int moveToRoot(BtCursor *pCur){
  MemPage *pRoot;
  int rc = SQLITE_OK;
  Btree *p = pCur->pBtree;
  BtShared *pBt = p->pBt;


  assert( sqlite3BtreeMutexHeld(pBt->mutex) );
  if( pCur->eState==CURSOR_REQUIRESEEK ){
    clearCursorPosition(pCur);
  }
  pRoot = pCur->pPage;
  if( pRoot && pRoot->pgno==pCur->pgnoRoot ){
    assert( pRoot->isInit );

** in ascending order.
*/
static int moveToLeftmost(BtCursor *pCur){
  Pgno pgno;
  int rc = SQLITE_OK;
  MemPage *pPage;

  assert( sqlite3BtreeMutexHeld(pCur->pBt->mutex) );

  assert( pCur->eState==CURSOR_VALID );
  while( rc==SQLITE_OK && !(pPage = pCur->pPage)->leaf ){
    assert( pCur->idx>=0 && pCur->idx<pPage->nCell );
    pgno = get4byte(findCell(pPage, pCur->idx));
    rc = moveToChild(pCur, pgno);
  }
  return rc;

** key in ascending order.
*/
static int moveToRightmost(BtCursor *pCur){
  Pgno pgno;
  int rc = SQLITE_OK;
  MemPage *pPage;

  assert( sqlite3BtreeMutexHeld(pCur->pBt->mutex) );

  assert( pCur->eState==CURSOR_VALID );
  while( rc==SQLITE_OK && !(pPage = pCur->pPage)->leaf ){
    pgno = get4byte(&pPage->aData[pPage->hdrOffset+8]);
    pCur->idx = pPage->nCell;
    rc = moveToChild(pCur, pgno);
  }
  if( rc==SQLITE_OK ){
    pCur->idx = pPage->nCell - 1;
    pCur->info.nSize = 0;
  }
  return SQLITE_OK;
}

/* Move the cursor to the first entry in the table.  Return SQLITE_OK
** on success.  Set *pRes to 0 if the cursor actually points to something
** or set *pRes to 1 if the table is empty.
*/
int sqlite3BtreeFirst(BtCursor *pCur, int *pRes){
  int rc;

  assert( sqlite3BtreeMutexHeld(pCur->pBt->mutex) );
  assert( sqlite3BtreeMutexHeld(pCur->pBtree->pSqlite->mutex) );
  rc = moveToRoot(pCur);
  if( rc==SQLITE_OK ){
    if( pCur->eState==CURSOR_INVALID ){
      assert( pCur->pPage->nCell==0 );
      *pRes = 1;
      rc = SQLITE_OK;

/* Move the cursor to the last entry in the table.  Return SQLITE_OK
** on success.  Set *pRes to 0 if the cursor actually points to something
** or set *pRes to 1 if the table is empty.
*/
int sqlite3BtreeLast(BtCursor *pCur, int *pRes){
  int rc;
 
  assert( sqlite3BtreeMutexHeld(pCur->pBt->mutex) );
  assert( sqlite3BtreeMutexHeld(pCur->pBtree->pSqlite->mutex) );
  rc = moveToRoot(pCur);
  if( rc==SQLITE_OK ){
    if( CURSOR_INVALID==pCur->eState ){
      assert( pCur->pPage->nCell==0 );
      *pRes = 1;
    }else{

  const void *pKey,      /* The key content for indices.  Not used by tables */
  i64 nKey,              /* Size of pKey.  Or the key for tables */
  int biasRight,         /* If true, bias the search to the high end */
  int *pRes              /* Search result flag */
){
  int rc;

  assert( sqlite3BtreeMutexHeld(pCur->pBt->mutex) );
  assert( sqlite3BtreeMutexHeld(pCur->pBtree->pSqlite->mutex) );
  rc = moveToRoot(pCur);
  if( rc ){
    return rc;
  }
  assert( pCur->pPage );
  assert( pCur->pPage->isInit );

  return (CURSOR_VALID!=pCur->eState);
}

/*
** Return the database connection handle for a cursor.
*/
sqlite3 *sqlite3BtreeCursorDb(const BtCursor *pCur){
  assert( sqlite3_mutex_held(pCur->pBtree->pSqlite->mutex) );
  return pCur->pBtree->pSqlite;
}

/*
** Advance the cursor to the next entry in the database.  If
** successful then set *pRes=0.  If the cursor
** was already pointing to the last entry in the database before
** this routine was called, then set *pRes=1.
*/
static int btreeNext(BtCursor *pCur, int *pRes){
  int rc;
  MemPage *pPage;

  assert( sqlite3_mutex_held(pCur->pBt->mutex) );
  rc = restoreOrClearCursorPosition(pCur);
  if( rc!=SQLITE_OK ){
    return rc;
  }
  assert( pRes!=0 );
  pPage = pCur->pPage;
  if( CURSOR_INVALID==pCur->eState ){

** this routine was called, then set *pRes=1.
*/
static int btreePrevious(BtCursor *pCur, int *pRes){
  int rc;
  Pgno pgno;
  MemPage *pPage;

  assert( sqlite3_mutex_held(pCur->pBt->mutex) );
  rc = restoreOrClearCursorPosition(pCur);
  if( rc!=SQLITE_OK ){
    return rc;
  }
  if( CURSOR_INVALID==pCur->eState ){
    *pRes = 1;
    return SQLITE_OK;

  int szNew;
  MemPage *pPage;
  Btree *p = pCur->pBtree;
  BtShared *pBt = p->pBt;
  unsigned char *oldCell;
  unsigned char *newCell = 0;

  cursorEnter(pCur);
  if( pBt->inTransaction!=TRANS_WRITE ){
    /* Must start a transaction before doing an insert */
    rc = pBt->readOnly ? SQLITE_READONLY : SQLITE_ERROR;
    cursorLeave(pCur);
    return rc;
  }
  assert( !pBt->readOnly );
  if( !pCur->wrFlag ){
    cursorLeave(pCur);
    return SQLITE_PERM;   /* Cursor not open for writing */
  }
  if( checkReadLocks(pCur->pBtree, pCur->pgnoRoot, pCur) ){
    cursorLeave(pCur);
    return SQLITE_LOCKED; /* The table pCur points to has a read lock */
  }

  /* Save the positions of any other cursors open on this table */
  clearCursorPosition(pCur);
  if( 
    SQLITE_OK!=(rc = saveAllCursors(pBt, pCur->pgnoRoot, pCur)) ||
    SQLITE_OK!=(rc = sqlite3BtreeMoveto(pCur, pKey, nKey, appendBias, &loc))
  ){
    cursorLeave(pCur);
    return rc;
  }

  pPage = pCur->pPage;
  assert( pPage->intKey || nKey>=0 );
  assert( pPage->leaf || !pPage->leafData );
  TRACE(("INSERT: table=%d nkey=%lld ndata=%d page=%d %s\n",

  /* sqlite3BtreePageDump(pCur->pBt, pCur->pgnoRoot, 1); */
  /* fflush(stdout); */
  if( rc==SQLITE_OK ){
    moveToRoot(pCur);
  }
end_insert:
  sqlite3_free(newCell);
  cursorLeave(pCur);
  return rc;
}

/*
** Delete the entry that the cursor is pointing to.  The cursor
** is left pointing at a random location.
*/
int sqlite3BtreeDelete(BtCursor *pCur){
  MemPage *pPage = pCur->pPage;
  unsigned char *pCell;
  int rc;
  Pgno pgnoChild = 0;
  Btree *p = pCur->pBtree;
  BtShared *pBt = p->pBt;

  cursorEnter(pCur);
  assert( pPage->isInit );
  if( pBt->inTransaction!=TRANS_WRITE ){
    /* Must start a transaction before doing a delete */
    rc = pBt->readOnly ? SQLITE_READONLY : SQLITE_ERROR;
    cursorLeave(pCur);
    return rc;
  }
  assert( !pBt->readOnly );
  if( pCur->idx >= pPage->nCell ){
    cursorLeave(pCur);
    return SQLITE_ERROR;  /* The cursor is not pointing to anything */
  }
  if( !pCur->wrFlag ){
    cursorLeave(pCur);
    return SQLITE_PERM;   /* Did not open this cursor for writing */
  }
  if( checkReadLocks(pCur->pBtree, pCur->pgnoRoot, pCur) ){
    cursorLeave(pCur);
    return SQLITE_LOCKED; /* The table pCur points to has a read lock */
  }

  /* Restore the current cursor position (a no-op if the cursor is not in 
  ** CURSOR_REQUIRESEEK state) and save the positions of any other cursors 
  ** open on the same table. Then call sqlite3PagerWrite() on the page
  ** that the entry will be deleted from.
  */
  if( 
    (rc = restoreOrClearCursorPosition(pCur))!=0 ||
    (rc = saveAllCursors(pBt, pCur->pgnoRoot, pCur))!=0 ||
    (rc = sqlite3PagerWrite(pPage->pDbPage))!=0
  ){
    cursorLeave(pCur);
    return rc;
  }

  /* Locate the cell within it's page and leave pCell pointing to the
  ** data. The clearCell() call frees any overflow pages associated with the
  ** cell. The cell itself is still intact.
  */
  pCell = findCell(pPage, pCur->idx);
  if( !pPage->leaf ){
    pgnoChild = get4byte(pCell);
  }
  rc = clearCell(pPage, pCell);
  if( rc ){
    cursorLeave(pCur);
    return rc;
  }

  if( !pPage->leaf ){
    /*
    ** The entry we are about to delete is not a leaf so if we do not
    ** do something we will leave a hole on an internal page.

       pCur->pgnoRoot, pPage->pgno));
    dropCell(pPage, pCur->idx, cellSizePtr(pPage, pCell));
    rc = balance(pPage, 0);
  }
  if( rc==SQLITE_OK ){
    moveToRoot(pCur);
  }
  cursorLeave(pCur);
  return rc;
}

/*
** Create a new BTree table.  Write into *piTable the page
** number for the root page of the new table.
**
** The type of type is determined by the flags parameter.  Only the
** following values of flags are currently in use.  Other values for
** flags might not work:
**
**     BTREE_INTKEY|BTREE_LEAFDATA     Used for SQL tables with rowid keys
**     BTREE_ZERODATA                  Used for SQL indices
*/
static int btreeCreateTable(Btree *p, int *piTable, int flags){
  BtShared *pBt = p->pBt;
  MemPage *pRoot;
  Pgno pgnoRoot;
  int rc;

  assert( sqlite3_mutex_held(p->pSqlite->mutex) );
  assert( sqlite3_mutex_held(pBt->mutex) );
  if( pBt->inTransaction!=TRANS_WRITE ){
    /* Must start a transaction first */
    rc = pBt->readOnly ? SQLITE_READONLY : SQLITE_ERROR;
    return rc;
  }
  assert( !pBt->readOnly );

*/
int sqlite3BtreeFlags(BtCursor *pCur){
  /* TODO: What about CURSOR_REQUIRESEEK state? Probably need to call
  ** restoreOrClearCursorPosition() here.
  */
  MemPage *pPage = pCur->pPage;
  assert( sqlite3BtreeMutexHeld(pPage->pBt->mutex) );
  assert( pPage->pBt==pCur->pBt );
  return pPage ? pPage->aData[pPage->hdrOffset] : 0;
}


/*
** Return the pager associated with a BTree.  This routine is used for
** testing and debugging only.
*/
Pager *sqlite3BtreePager(Btree *p){


  return p->pBt->pPager;
}

#ifndef SQLITE_OMIT_INTEGRITY_CHECK
/*
** Append a message to the error message string.
*/

  *pnErr = sCheck.nErr;
  return sCheck.zErrMsg;
}
#endif /* SQLITE_OMIT_INTEGRITY_CHECK */

/*
** Return the full pathname of the underlying database file.
**
** The pager filename is invariant as long as the pager is
** open so it is safe to access without the BtShared mutex.
*/
const char *sqlite3BtreeGetFilename(Btree *p){
  assert( p->pBt->pPager!=0 );
  assert( sqlite3_mutex_held(p->pSqlite->mutex) );
  return sqlite3PagerFilename(p->pBt->pPager);
}

/*
** Return the pathname of the directory that contains the database file.
**
** The pager directory name is invariant as long as the pager is
** open so it is safe to access without the BtShared mutex.
*/
const char *sqlite3BtreeGetDirname(Btree *p){
  assert( p->pBt->pPager!=0 );
  assert( sqlite3_mutex_held(p->pSqlite->mutex) );
  return sqlite3PagerDirname(p->pBt->pPager);
}

/*
** Return the pathname of the journal file for this database. The return
** value of this routine is the same regardless of whether the journal file
** has been created or not.
**
** The pager journal filename is invariant as long as the pager is
** open so it is safe to access without the BtShared mutex.
*/
const char *sqlite3BtreeGetJournalname(Btree *p){
  assert( p->pBt->pPager!=0 );
  assert( sqlite3_mutex_held(p->pSqlite->mutex) );
  return sqlite3PagerJournalname(p->pBt->pPager);
}

#ifndef SQLITE_OMIT_VACUUM
/*
** Copy the complete content of pBtFrom into pBtTo.  A transaction
** must be active for both files.

#endif /* SQLITE_OMIT_VACUUM */

/*
** Return non-zero if a transaction is active.
*/
int sqlite3BtreeIsInTrans(Btree *p){
  assert( p==0 || sqlite3BtreeMutexHeld(p->pSqlite->mutex) );
  return (p && (p->inTrans==TRANS_WRITE));
}

/*
** Return non-zero if a statement transaction is active.
*/
int sqlite3BtreeIsInStmt(Btree *p){

** Just before the shared-btree is closed, the function passed as the 
** xFree argument when the memory allocation was made is invoked on the 
** blob of allocated memory. This function should not call sqlite3_free()
** on the memory, the btree layer does that.
*/
void *sqlite3BtreeSchema(Btree *p, int nBytes, void(*xFree)(void *)){
  BtShared *pBt = p->pBt;

  sqlite3BtreeEnter(p);
  if( !pBt->pSchema ){
    pBt->pSchema = sqlite3MallocZero(nBytes);
    pBt->xFreeSchema = xFree;
  }
  sqlite3BtreeLeave(p);
  return pBt->pSchema;

** Argument pCsr must be a cursor opened for writing on an 
** INTKEY table currently pointing at a valid table entry. 
** This function modifies the data stored as part of that entry.
** Only the data content may only be modified, it is not possible
** to change the length of the data stored.
*/
int sqlite3BtreePutData(BtCursor *pCsr, u32 offset, u32 amt, void *z){
  assert( sqlite3BtreeMutexHeld(pCsr->pBt->mutex) );
  assert( sqlite3BtreeMutexHeld(pCsr->pBtree->pSqlite->mutex) );
  assert(pCsr->isIncrblobHandle);
  if( pCsr->eState==CURSOR_REQUIRESEEK ){
    return SQLITE_ABORT;
  }

  /* Check some preconditions: 
  **   (a) the cursor is open for writing,
  **   (b) there is no read-lock on the table being modified and
  **   (c) the cursor points at a valid row of an intKey table.
  */
  if( !pCsr->wrFlag ){
    return SQLITE_READONLY;
  }
  assert( !pCsr->pBt->readOnly 
          && pCsr->pBt->inTransaction==TRANS_WRITE );
  if( checkReadLocks(pCsr->pBtree, pCsr->pgnoRoot, pCsr) ){
    return SQLITE_LOCKED; /* The table pCur points to has a read lock */
  }
  if( pCsr->eState==CURSOR_INVALID || !pCsr->pPage->intKey ){
    return SQLITE_ERROR;
  }

**
** This function sets a flag only. The actual page location cache
** (stored in BtCursor.aOverflow[]) is allocated and used by function
** accessPayload() (the worker function for sqlite3BtreeData() and
** sqlite3BtreePutData()).
*/
void sqlite3BtreeCacheOverflow(BtCursor *pCur){
  assert( sqlite3BtreeMutexHeld(pCur->pBt->mutex) );
  assert( sqlite3BtreeMutexHeld(pCur->pBtree->pSqlite->mutex) );
  assert(!pCur->isIncrblobHandle);
  assert(!pCur->aOverflow);
  pCur->isIncrblobHandle = 1;
}
#endif

/*
** 2004 April 6
**
** The author disclaims copyright to this source code.  In place of
** a legal notice, here is a blessing:
**
**    May you do good and not evil.
**    May you find forgiveness for yourself and forgive others.
**    May you share freely, never taking more than you give.
**
*************************************************************************
** $Id: btreeInt.h,v 1.11 2007/08/28 22:24:35 drh Exp $
**
** This file implements a external (disk-based) database using BTrees.
** For a detailed discussion of BTrees, refer to
**
**     Donald E. Knuth, THE ART OF COMPUTER PROGRAMMING, Volume 3:
**     "Sorting And Searching", pages 473-480. Addison-Wesley
**     Publishing Company, Reading, Massachusetts.

** For some database files, the same underlying database cache might be 
** shared between multiple connections.  In that case, each contection
** has it own pointer to this object.  But each instance of this object
** points to the same BtShared object.  The database cache and the
** schema associated with the database file are all contained within
** the BtShared object.
**
** All fields in this structure are accessed under sqlite3.mutex.
** The pBt pointer itself may not be changed while there exists cursors 
** in the referenced BtShared that point back to this Btree since those
** cursors have to do go through this Btree to find their BtShared and
** they often do so without holding sqlite3.mutex.
*/
struct Btree {
  sqlite3 *pSqlite;  /* The database connection holding this btree */
  BtShared *pBt;     /* Sharable content of this btree */
  u8 inTrans;        /* TRANS_NONE, TRANS_READ or TRANS_WRITE */
  u8 sharable;       /* True if we can share pBt with other pSqlite */
  u8 locked;         /* True if pSqlite currently has pBt locked */

** MemPage.aCell[] of the entry.
**
** When a single database file can shared by two more database connections,
** but cursors cannot be shared.  Each cursor is associated with a
** particular database connection identified BtCursor.pBtree.pSqlite.
**
** Fields in this structure are accessed under the BtShared.mutex
** found at self->pBt->mutex. 

*/
struct BtCursor {
  Btree *pBtree;            /* The Btree to which this cursor belongs */
  BtShared *pBt;            /* The BtShared this cursor points to */
  BtCursor *pNext, *pPrev;  /* Forms a linked list of all cursors */
  int (*xCompare)(void*,int,const void*,int,const void*); /* Key comp func */
  void *pArg;               /* First arg to xCompare() */
  Pgno pgnoRoot;            /* The root page of this tree */
  MemPage *pPage;           /* Page that contains the entry */
  int idx;                  /* Index of the entry in pPage->aCell[] */
  CellInfo info;            /* A parse of the cell we are pointing at */

**     CREATE INDEX
**     DROP INDEX
**     creating ID lists
**     BEGIN TRANSACTION
**     COMMIT
**     ROLLBACK
**
** $Id: build.c,v 1.440 2007/08/28 22:24:35 drh Exp $
*/
#include "sqliteInt.h"
#include <ctype.h>

/*
** This routine is called when a new SQL statement is beginning to
** be parsed.  Initialize the pParse structure as needed.

    */
    if( pParse->cookieGoto>0 ){
      u32 mask;
      int iDb;
      sqlite3VdbeJumpHere(v, pParse->cookieGoto-1);
      for(iDb=0, mask=1; iDb<db->nDb; mask<<=1, iDb++){
        if( (mask & pParse->cookieMask)==0 ) continue;
        sqlite3VdbeUsesBtree(v, iDb, db->aDb[iDb].pBt);
        sqlite3VdbeAddOp(v, OP_Transaction, iDb, (mask & pParse->writeMask)!=0);
        sqlite3VdbeAddOp(v, OP_VerifyCookie, iDb, pParse->cookieValue[iDb]);
      }
#ifndef SQLITE_OMIT_VIRTUALTABLE
      if( pParse->pVirtualLock ){
        char *vtab = (char *)pParse->pVirtualLock->pVtab;
        sqlite3VdbeOp3(v, OP_VBegin, 0, 0, vtab, P3_VTAB);

/*
** 2007 August 28
**
** The author disclaims copyright to this source code.  In place of
** a legal notice, here is a blessing:
**
**    May you do good and not evil.
**    May you find forgiveness for yourself and forgive others.
**    May you share freely, never taking more than you give.
**
*************************************************************************
** This file contains the C functions that implement mutexes for pthreads
**
** $Id: mutex_unix.c,v 1.2 2007/08/28 22:24:35 drh Exp $
*/
#include "sqliteInt.h"

/*
** The code in this file is only used if we are compiling threadsafe
** under unix with pthreads.
**

** Each recursive mutex is an instance of the following structure.
*/
struct sqlite3_mutex {
  pthread_mutex_t mutex;     /* Mutex controlling the lock */
  int id;                    /* Mutex type */
  int nRef;                  /* Number of entrances */
  pthread_t owner;           /* Thread that is within this mutex */
#ifdef SQLITE_DEBUG
  int trace;                 /* True to trace changes */
#endif
};

/*
** The sqlite3_mutex_alloc() routine allocates a new
** mutex and returns a pointer to it.  If it returns NULL
** that means that a mutex could not be allocated.  SQLite
** will unwind its stack and return an error.  The argument

*/
void sqlite3_mutex_enter(sqlite3_mutex *p){
  assert( p );
  assert( p->id==SQLITE_MUTEX_RECURSIVE || sqlite3_mutex_notheld(p) );
  pthread_mutex_lock(&p->mutex);
  p->owner = pthread_self();
  p->nRef++;
#ifdef SQLITE_DEBUG
  if( p->trace ){
    printf("enter mutex %p (%d) with nRef=%d\n", p, p->trace, p->nRef);
  }
#endif
}
int sqlite3_mutex_try(sqlite3_mutex *p){
  int rc;
  assert( p );
  assert( p->id==SQLITE_MUTEX_RECURSIVE || sqlite3_mutex_notheld(p) );
  if( pthread_mutex_trylock(&p->mutex)==0 ){
    p->owner = pthread_self();
    p->nRef++;
    rc = SQLITE_OK;
#ifdef SQLITE_DEBUG
    if( p->trace ){
      printf("enter mutex %p (%d) with nRef=%d\n", p, p->trace, p->nRef);
    }
#endif
  }else{
    rc = SQLITE_BUSY;
  }
  return rc;
}

/*
** The sqlite3_mutex_leave() routine exits a mutex that was
** previously entered by the same thread.  The behavior
** is undefined if the mutex is not currently entered or
** is not currently allocated.  SQLite will never do either.
*/
void sqlite3_mutex_leave(sqlite3_mutex *p){
  assert( p );
  assert( sqlite3_mutex_held(p) );
  p->nRef--;
  assert( p->nRef==0 || p->id==SQLITE_MUTEX_RECURSIVE );
#ifdef SQLITE_DEBUG
  if( p->trace ){
    printf("leave mutex %p (%d) with nRef=%d\n", p, p->trace, p->nRef);
  }
#endif
  pthread_mutex_unlock(&p->mutex);
}

/*
** The sqlite3_mutex_held() and sqlite3_mutex_notheld() routine are
** intended for use only inside assert() statements.  On some platforms,
** there might be race conditions that can cause these routines to

** The pager is used to access a database disk file.  It implements
** atomic commit and rollback through the use of a journal file that
** is separate from the database file.  The pager also implements file
** locking to prevent two processes from writing the same database
** file simultaneously, or one process from reading the database while
** another is writing.
**
** @(#) $Id: pager.c,v 1.379 2007/08/28 22:24:35 drh Exp $
*/
#ifndef SQLITE_OMIT_DISKIO
#include "sqliteInt.h"
#include <assert.h>
#include <string.h>

/*

/*
** Return the full pathname of the database file.
*/
const char *sqlite3PagerFilename(Pager *pPager){
  return pPager->zFilename;
}

/*
** Return the VFS structure for the pager.
*/
const sqlite3_vfs *sqlite3PagerVfs(Pager *pPager){
  return pPager->pVfs;
}

/*
** Return the directory of the database file.
*/
const char *sqlite3PagerDirname(Pager *pPager){
  return pPager->zDirectory;
}

**    May you share freely, never taking more than you give.
**
*************************************************************************
** This header file defines the interface that the sqlite page cache
** subsystem.  The page cache subsystem reads and writes a file a page
** at a time and provides a journal for rollback.
**
** @(#) $Id: pager.h,v 1.63 2007/08/28 22:24:35 drh Exp $
*/

#ifndef _PAGER_H_
#define _PAGER_H_

/*
** The type used to represent a page number.  The first page in a file

int sqlite3PagerStmtCommit(Pager*);
int sqlite3PagerStmtRollback(Pager*);
void sqlite3PagerDontRollback(DbPage*);
void sqlite3PagerDontWrite(DbPage*);
int sqlite3PagerRefcount(Pager*);
void sqlite3PagerSetSafetyLevel(Pager*,int,int);
const char *sqlite3PagerFilename(Pager*);
const sqlite3_vfs *sqlite3PagerVfs(Pager*);
const char *sqlite3PagerDirname(Pager*);
const char *sqlite3PagerJournalname(Pager*);
int sqlite3PagerNosync(Pager*);
int sqlite3PagerMovepage(Pager*,DbPage*,Pgno);
void *sqlite3PagerGetData(DbPage *); 
void *sqlite3PagerGetExtra(DbPage *); 
int sqlite3PagerLockingMode(Pager *, int);

*************************************************************************
** Header file for the Virtual DataBase Engine (VDBE)
**
** This header defines the interface to the virtual database engine
** or VDBE.  The VDBE implements an abstract machine that runs a
** simple program to access and modify the underlying database.
**
** $Id: vdbe.h,v 1.112 2007/08/28 22:24:35 drh Exp $
*/
#ifndef _SQLITE_VDBE_H_
#define _SQLITE_VDBE_H_
#include <stdio.h>

/*
** A single VDBE is an opaque structure named "Vdbe".  Only routines

int sqlite3VdbeOp3(Vdbe*,int,int,int,const char *zP3,int);
int sqlite3VdbeAddOpList(Vdbe*, int nOp, VdbeOpList const *aOp);
void sqlite3VdbeChangeP1(Vdbe*, int addr, int P1);
void sqlite3VdbeChangeP2(Vdbe*, int addr, int P2);
void sqlite3VdbeJumpHere(Vdbe*, int addr);
void sqlite3VdbeChangeToNoop(Vdbe*, int addr, int N);
void sqlite3VdbeChangeP3(Vdbe*, int addr, const char *zP1, int N);
void sqlite3VdbeUsesBtree(Vdbe*, int, Btree*);
VdbeOp *sqlite3VdbeGetOp(Vdbe*, int);
int sqlite3VdbeMakeLabel(Vdbe*);
void sqlite3VdbeDelete(Vdbe*);
void sqlite3VdbeMakeReady(Vdbe*,int,int,int,int);
int sqlite3VdbeFinalize(Vdbe*);
void sqlite3VdbeResolveLabel(Vdbe*, int);
int sqlite3VdbeCurrentAddr(Vdbe*);

  u8 changeCntOn;         /* True to update the change-counter */
  u8 aborted;             /* True if ROLLBACK in another VM causes an abort */
  u8 expired;             /* True if the VM needs to be recompiled */
  u8 minWriteFileFormat;  /* Minimum file format for writable database files */
  u8 inVtabMethod;        /* See comments above */
  int nChange;            /* Number of db changes made since last reset */
  i64 startTime;          /* Time when query started - used for profiling */
  int btreeMask;          /* Bitmask of db->aDb[] entries referenced */
  BtreeMutexSet mtxSet;   /* Set of Btree mutexes */
  int nSql;             /* Number of bytes in zSql */
  char *zSql;           /* Text of the SQL statement that generated this */
#ifdef SQLITE_DEBUG
  FILE *trace;        /* Write an execution trace here, if not NULL */
#endif
  int openedStatement;  /* True if this VM has opened a statement journal */

  }
  assert( zP3!=0 );
  return zP3;
}
#endif

/*
** Declare to the Vdbe that the BTree object at db->aDb[i] is used.
**
*/
void sqlite3VdbeUsesBtree(Vdbe *p, int i, Btree *pBtree){
  assert( i>=0 && i<p->db->nDb );
  assert( i<sizeof(p->btreeMask)*8 );
  assert( p->db->aDb[i].pBt==pBtree );
  p->btreeMask |= 1<<i;
  sqlite3BtreeMutexSetInsert(&p->mtxSet, pBtree);
}


#if defined(VDBE_PROFILE) || defined(SQLITE_DEBUG)
/*
** Print a single opcode.  This routine is used for debugging only.

  ** (b) how many database files have open write transactions, not 
  ** including the temp database. (b) is important because if more than 
  ** one database file has an open write transaction, a master journal
  ** file is required for an atomic commit.
  */ 
  for(i=0; i<db->nDb; i++){ 
    Btree *pBt = db->aDb[i].pBt;
    if( sqlite3BtreeIsInTrans(pBt) ){
      needXcommit = 1;
      if( i!=1 ) nTrans++;
    }
  }

  /* If there are any write-transactions at all, invoke the commit hook */
  if( needXcommit && db->xCommitCallback ){

    ** and delete the master journal file. All the individual journal files
    ** still have 'null' as the master journal pointer, so they will roll
    ** back independently if a failure occurs.
    */
    for(i=0; i<db->nDb; i++){
      Btree *pBt = db->aDb[i].pBt;
      if( i==1 ) continue;   /* Ignore the TEMP database */
      if( sqlite3BtreeIsInTrans(pBt) ){
        char const *zFile = sqlite3BtreeGetJournalname(pBt);
        if( zFile[0]==0 ) continue;  /* Ignore :memory: databases */
        if( !needSync && !sqlite3BtreeSyncDisabled(pBt) ){
          needSync = 1;
        }
        rc = sqlite3OsWrite(pMaster, zFile, strlen(zFile)+1, offset);
        offset += strlen(zFile)+1;

    ** sqlite3BtreeCommitPhaseOne(), then there is a chance that the
    ** master journal file will be orphaned. But we cannot delete it,
    ** in case the master journal file name was written into the journal
    ** file before the failure occured.
    */
    for(i=0; rc==SQLITE_OK && i<db->nDb; i++){ 
      Btree *pBt = db->aDb[i].pBt;
      if( pBt ){
        rc = sqlite3BtreeCommitPhaseOne(pBt, zMaster);
      }
    }
    sqlite3OsCloseFree(pMaster);
    if( rc!=SQLITE_OK ){
      sqlite3_free(zMaster);
      return rc;

#
#***********************************************************************
#
# The focus of the tests in this file are IO errors that occur in a shared
# cache context. What happens to connection B if one connection A encounters
# an IO-error whilst reading or writing the file-system?
#
# $Id: shared_err.test,v 1.13 2007/08/28 22:24:35 drh Exp $

proc skip {args} {}


set testdir [file dirname $argv0]
source $testdir/tester.tcl
source $testdir/malloc_common.tcl

  execsql {COMMIT} db2
  set ::DB2 [sqlite3_connection_pointer db2]
  set ::STMT [sqlite3_prepare $::DB2 "SELECT a FROM t1 ORDER BY a" -1 DUMMY]
  sqlite3_step $::STMT       ;# Cursor points at 000.000.000.000
  sqlite3_step $::STMT       ;# Cursor points at 001.001.001.001

} -tclbody {
btree_breakpoint
  execsql {
    BEGIN;
    INSERT INTO t1 VALUES('201.201.201.201.201', NULL);
    UPDATE t1 SET a = '202.202.202.202.202' WHERE a LIKE '201%';
    COMMIT;
  }
} -cleanup {