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 1.423 2007/09/06 22:19:15 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"

  int top;           /* First byte of the cell content area */

  pBt = pPage->pBt;
  assert( pBt!=0 );
  assert( pParent==0 || pParent->pBt==pBt );
  assert( sqlite3_mutex_held(pBt->mutex) );
  assert( pPage->pgno==sqlite3PagerPagenumber(pPage->pDbPage) );
  assert( pPage == sqlite3PagerGetExtra(pPage->pDbPage) );
  assert( pPage->aData == sqlite3PagerGetData(pPage->pDbPage) );
  if( pPage->pParent!=pParent && (pPage->pParent!=0 || pPage->isInit) ){
    /* The parent page should never change unless the file is corrupt */
    return SQLITE_CORRUPT_BKPT;
  }
  if( pPage->isInit ) return SQLITE_OK;
  if( pPage->pParent==0 && pParent!=0 ){
    pPage->pParent = pParent;

static void zeroPage(MemPage *pPage, int flags){
  unsigned char *data = pPage->aData;
  BtShared *pBt = pPage->pBt;
  int hdr = pPage->hdrOffset;
  int first;

  assert( sqlite3PagerPagenumber(pPage->pDbPage)==pPage->pgno );
  assert( sqlite3PagerGetExtra(pPage->pDbPage) == (void*)pPage );
  assert( sqlite3PagerGetData(pPage->pDbPage) == data );
  assert( sqlite3PagerIswriteable(pPage->pDbPage) );
  assert( sqlite3_mutex_held(pBt->mutex) );
  memset(&data[hdr], 0, pBt->usableSize - hdr);
  data[hdr] = flags;
  first = hdr + 8 + 4*((flags&PTF_LEAF)==0);
  memset(&data[hdr+1], 0, 4);
  data[hdr+7] = 0;

** Release a MemPage.  This should be called once for each prior
** call to sqlite3BtreeGetPage.
*/
static void releasePage(MemPage *pPage){
  if( pPage ){
    assert( pPage->aData );
    assert( pPage->pBt );
    assert( sqlite3PagerGetExtra(pPage->pDbPage) == (void*)pPage );
    assert( sqlite3PagerGetData(pPage->pDbPage)==pPage->aData );
    assert( sqlite3_mutex_held(pPage->pBt->mutex) );
    sqlite3PagerUnref(pPage->pDbPage);
  }
}

/*
** This routine is called when the reference count for a page

*/
static void unlockBtreeIfUnused(BtShared *pBt){
  assert( sqlite3_mutex_held(pBt->mutex) );
  if( pBt->inTransaction==TRANS_NONE && pBt->pCursor==0 && pBt->pPage1!=0 ){
    if( sqlite3PagerRefcount(pBt->pPager)>=1 ){
      if( pBt->pPage1->aData==0 ){
        MemPage *pPage = pBt->pPage1;
        pPage->aData = sqlite3PagerGetData(pPage->pDbPage);
        pPage->pBt = pBt;
        pPage->pgno = 1;
      }
      releasePage(pBt->pPage1);
    }
    pBt->pPage1 = 0;
    pBt->inStmt = 0;

  assert( pNewParent!=0 );
  if( pgno==0 ) return SQLITE_OK;
  assert( pBt->pPager!=0 );
  pDbPage = sqlite3PagerLookup(pBt->pPager, pgno);
  if( pDbPage ){
    pThis = (MemPage *)sqlite3PagerGetExtra(pDbPage);
    if( pThis->isInit ){
      assert( pThis->aData==sqlite3PagerGetData(pDbPage) );
      if( pThis->pParent!=pNewParent ){
        if( pThis->pParent ) sqlite3PagerUnref(pThis->pParent->pDbPage);
        pThis->pParent = pNewParent;
        sqlite3PagerRef(pNewParent->pDbPage);
      }
      pThis->idxParent = idx;
    }

  /*
  ** Make copies of the content of pPage and its siblings into aOld[].
  ** The rest of this function will use data from the copies rather
  ** that the original pages since the original pages will be in the
  ** process of being overwritten.
  */
  for(i=0; i<nOld; i++){
    MemPage *p = apCopy[i] = (MemPage*)aCopy[i];
    memcpy(p, apOld[i], sizeof(MemPage));
    p->aData = (void*)&p[1];
    memcpy(p->aData, apOld[i]->aData, pBt->pageSize);



  }

  /*
  ** Load pointers to all cells on sibling pages and the divider cells
  ** into the local apCell[] array.  Make copies of the divider cells
  ** into space obtained form aSpace[] and remove the the divider Cells
  ** from pParent.

** 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 1.386 2007/09/06 22:19:15 drh Exp $
*/
#ifndef SQLITE_OMIT_DISKIO
#include "sqliteInt.h"
#include <assert.h>
#include <string.h>

/*

  u8 alwaysRollback;             /* Disable DontRollback() for this page */
  u8 needRead;                   /* Read content if PagerWrite() is called */
  short int nRef;                /* Number of users of this page */
  PgHdr *pDirty, *pPrevDirty;    /* Dirty pages */
#ifdef SQLITE_ENABLE_MEMORY_MANAGEMENT
  PagerLruLink gfree;            /* Global list of nRef==0 pages */
#endif

#ifdef SQLITE_CHECK_PAGES
  u32 pageHash;
#endif
  void *pData;                   /* Page data */
  /* Pager.nExtra bytes of local data appended to this header */
};

/*
** For an in-memory only database, some extra information is recorded about
** each page so that changes can be rolled back.  (Journal files are not
** used for in-memory databases.)  The following information is added to
** the end of every EXTRA block for in-memory databases.

# define CODEC2(P,D,N,X) ((char*)D)
#endif

/*
** Convert a pointer to a PgHdr into a pointer to its data
** and back again.
*/
#define PGHDR_TO_DATA(P)    ((P)->pData)

#define PGHDR_TO_EXTRA(G,P) ((void*)&((G)[1]))
#define PGHDR_TO_HIST(P,PGR)  \
            ((PgHistory*)&((char*)(&(P)[1]))[(PGR)->nExtra])

/*
** A open page cache is an instance of the following structure.
**
** Pager.errCode may be set to SQLITE_IOERR, SQLITE_CORRUPT, or
** or SQLITE_FULL. Once one of the first three errors occurs, it persists
** and is returned as the result of every major pager API call.  The

  PgHdr *pPg, *pNext;
  if( pPager->errCode ) return;
  for(pPg=pPager->pAll; pPg; pPg=pNext){
    IOTRACE(("PGFREE %p %d\n", pPager, pPg->pgno));
    PAGER_INCR(sqlite3_pager_pgfree_count);
    pNext = pPg->pNextAll;
    lruListRemove(pPg);
    sqlite3_free(pPg->pData);
    sqlite3_free(pPg);
  }
  assert(pPager->lru.pFirst==0);
  assert(pPager->lru.pFirstSynced==0);
  assert(pPager->lru.pLast==0);
  pPager->pStmt = 0;
  pPager->pAll = 0;

      ppPg = &pPg->pNextAll;
    }else{
      *ppPg = pPg->pNextAll;
      IOTRACE(("PGFREE %p %d\n", pPager, pPg->pgno));
      PAGER_INCR(sqlite3_pager_pgfree_count);
      unlinkPage(pPg);
      makeClean(pPg);
      sqlite3_free(pPg->pData);
      sqlite3_free(pPg);
      pPager->nPage--;
    }
  }
}

/*

      nReleased += (
          sizeof(*pPg) + pPager->pageSize
          + sizeof(u32) + pPager->nExtra
          + MEMDB*sizeof(PgHistory) 
      );
      IOTRACE(("PGFREE %p %d *\n", pPager, pPg->pgno));
      PAGER_INCR(sqlite3_pager_pgfree_count);
      sqlite3_free(pPg->pData);
      sqlite3_free(pPg);
      pPager->nPage--;
    }else{
      /* An error occured whilst writing to the database file or 
      ** journal in pager_recycle(). The error is not returned to the 
      ** caller of this function. Instead, set the Pager.errCode variable.
      ** The error will be returned to the user (or users, in the case 

**     (4)  Either there is an available PgHdr that does not need
**          to be synced to disk or else disk syncing is currently
**          allowed.
*/
static int pagerAllocatePage(Pager *pPager, PgHdr **ppPg){
  int rc = SQLITE_OK;
  PgHdr *pPg;
  void *pData;

  /* Create a new PgHdr if any of the four conditions defined 
  ** above are met: */
  if( pPager->nPage<pPager->mxPage
   || pPager->lru.pFirst==0 
   || MEMDB
   || (pPager->lru.pFirstSynced==0 && pPager->doNotSync)
  ){
    if( pPager->nPage>=pPager->nHash ){
      pager_resize_hash_table(pPager,
         pPager->nHash<256 ? 256 : pPager->nHash*2);
      if( pPager->nHash==0 ){
        rc = SQLITE_NOMEM;
        goto pager_allocate_out;
      }
    }
    pagerLeave(pPager);
    pPg = sqlite3_malloc( sizeof(*pPg) + sizeof(u32) + pPager->nExtra

                            + MEMDB*sizeof(PgHistory) );
    if( pPg ){
      pData = sqlite3_malloc( pPager->pageSize );
      if( pData==0 ){
        sqlite3_free(pPg);
        pPg = 0;
      }
    }
    pagerEnter(pPager);
    if( pPg==0 ){
      rc = SQLITE_NOMEM;
      goto pager_allocate_out;
    }
    memset(pPg, 0, sizeof(*pPg));
    if( MEMDB ){
      memset(PGHDR_TO_HIST(pPg, pPager), 0, sizeof(PgHistory));
    }
    pPg->pData = pData;
    pPg->pPager = pPager;
    pPg->pNextAll = pPager->pAll;
    pPager->pAll = pPg;
    pPager->nPage++;
  }else{
    /* Recycle an existing page with a zero ref-count. */
    rc = pager_recycle(pPager, &pPg);

  
    /* The transaction journal now exists and we have a RESERVED or an
    ** EXCLUSIVE lock on the main database file.  Write the current page to
    ** the transaction journal if it is not there already.
    */
    if( !pPg->inJournal && (pPager->useJournal || MEMDB) ){
      if( (int)pPg->pgno <= pPager->origDbSize ){

        if( MEMDB ){
          PgHistory *pHist = PGHDR_TO_HIST(pPg, pPager);
          PAGERTRACE3("JOURNAL %d page %d\n", PAGERID(pPager), pPg->pgno);
          assert( pHist->pOrig==0 );
          pHist->pOrig = sqlite3_malloc( pPager->pageSize );
          if( pHist->pOrig ){
            memcpy(pHist->pOrig, PGHDR_TO_DATA(pPg), pPager->pageSize);
          }
        }else{
          u32 cksum;
          char *pData2;

          /* We should never write to the journal file the page that
          ** contains the database locks.  The following assert verifies
          ** that we do not. */
          assert( pPg->pgno!=PAGER_MJ_PGNO(pPager) );
          pData2 = CODEC2(pPager, pData, pPg->pgno, 7);
          cksum = pager_cksum(pPager, (u8*)pData2);
          rc = write32bits(pPager->jfd, pPager->journalOff, pPg->pgno);
          if( rc==SQLITE_OK ){
            rc = sqlite3OsWrite(pPager->jfd, pData2, pPager->pageSize,

                                pPager->journalOff + 4);
            pPager->journalOff += pPager->pageSize+4;
          }
          if( rc==SQLITE_OK ){
            rc = write32bits(pPager->jfd, pPager->journalOff, cksum);
            pPager->journalOff += 4;
          }
          IOTRACE(("JOUT %p %d %lld %d\n", pPager, pPg->pgno,
                   pPager->journalOff, szPg));
          PAGER_INCR(sqlite3_pager_writej_count);

          PAGERTRACE5("JOURNAL %d page %d needSync=%d hash(%08x)\n",
               PAGERID(pPager), pPg->pgno, pPg->needSync, pager_pagehash(pPg));


          /* An error has occured writing to the journal file. The 
          ** transaction will be rolled back by the layer above.
          */
          if( rc!=SQLITE_OK ){
            return rc;
          }

        if( pHist->pStmt ){
          memcpy(pHist->pStmt, PGHDR_TO_DATA(pPg), pPager->pageSize);
        }
        PAGERTRACE3("STMT-JOURNAL %d page %d\n", PAGERID(pPager), pPg->pgno);
        page_add_to_stmt_list(pPg);
      }else{
        i64 offset = pPager->stmtNRec*(4+pPager->pageSize);
        char *pData2 = CODEC2(pPager, pData, pPg->pgno, 7);
        rc = write32bits(pPager->stfd, offset, pPg->pgno);
        if( rc==SQLITE_OK ){
          rc = sqlite3OsWrite(pPager->stfd, pData2, pPager->pageSize, offset+4);
        }
        PAGERTRACE3("STMT-JOURNAL %d page %d\n", PAGERID(pPager), pPg->pgno);
        if( rc!=SQLITE_OK ){
          return rc;
        }
        pPager->stmtNRec++;
        assert( pPager->aInStmt!=0 );
        pPager->aInStmt[pPg->pgno/8] |= 1<<(pPg->pgno&7);

# 2001 September 15
#
# 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 implements regression tests for SQLite library.  The
# focus of this script is btree database backend
#
# $Id: btree.test,v 1.41 2007/09/06 22:19:15 drh Exp $


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

ifcapable default_autovacuum {
  finish_test

  lindex [btree_pager_stats $::b1] 1
} {1}
do_test btree-10.2 {
  set ::c1 [btree_cursor $::b1 2 1]
  lindex [btree_pager_stats $::b1] 1
} {2}
do_test btree-10.3 {
btree_breakpoint
  for {set i 1} {$i<=30} {incr i} {
    set key [format %03d $i]
    set data "*** $key *** $key *** $key *** $key ***"
    btree_insert $::c1 $key $data
  }
  select_keys $::c1
} {001 002 003 004 005 006 007 008 009 010 011 012 013 014 015 016 017 018 019 020 021 022 023 024 025 026 027 028 029 030}