From: drh <>
Date: Wed, 8 Dec 2021 19:50:45 +0000 (+0000)
Subject: Improvements on the decision of whether or not to use a Bloom filter.
X-Git-Tag: version-3.38.0~186^2~4
X-Git-Url: http://git.ipfire.org/?a=commitdiff_plain;h=fb82caf0bed2cd006cda440888939b87641b00c9;p=thirdparty%2Fsqlite.git

Improvements on the decision of whether or not to use a Bloom filter.

FossilOrigin-Name: 0fb2a4e08f518cb38ea3edc6a084d1e4874fd622ba3cf9101b49b3e7dc1a3f2b
---

diff --git a/manifest b/manifest
index ed505a6910..4c4a9d684d 100644
--- a/manifest
+++ b/manifest
@@ -1,5 +1,5 @@
-C Merge\sthe\sWhereClause\sfix\sfrom\strunk
-D 2021-12-08T16:15:41.698
+C Improvements\son\sthe\sdecision\sof\swhether\sor\snot\sto\suse\sa\sBloom\sfilter.
+D 2021-12-08T19:50:45.145
 F .fossil-settings/empty-dirs dbb81e8fc0401ac46a1491ab34a7f2c7c0452f2f06b54ebb845d024ca8283ef1
 F .fossil-settings/ignore-glob 35175cdfcf539b2318cb04a9901442804be81cd677d8b889fcc9149c21f239ea
 F LICENSE.md df5091916dbb40e6e9686186587125e1b2ff51f022cc334e886c19a0e9982724
@@ -638,8 +638,8 @@ F src/vxworks.h d2988f4e5a61a4dfe82c6524dd3d6e4f2ce3cdb9
 F src/wal.c ed0398a7adf02c31e34aada42cc86c58f413a7afe5f741a5d373ad087abde028
 F src/wal.h c3aa7825bfa2fe0d85bef2db94655f99870a285778baa36307c0a16da32b226a
 F src/walker.c f890a3298418d7cba3b69b8803594fdc484ea241206a8dfa99db6dd36f8cbb3b
-F src/where.c 74382d720f2a9a06f738467929ecf0d79fcdd6b9779184bddbb97352895a1b23
-F src/whereInt.h d55d5ce5c9de361e16cf5cd23da054cdecc1ae7735682ee55c2f40e58e960d22
+F src/where.c 89958d4fc7c45e916882ebc97481d98597f516ce3d778ace3271aacf34e24e91
+F src/whereInt.h c2cb535e755b25a7e152bdb407cbb2f62bdb8747c44bf2d984139f5cbebb8704
 F src/wherecode.c e2207f011b7e5bdef5722da5e8d95eb30ad01051b3526757901ecb19a9e98ff3
 F src/whereexpr.c 791544603b254cf11f8e84e3b50b0863c57322e9f213b828680f658e232ebc57
 F src/window.c 5d3b397b0c026d0ff5890244ac41359e524c01ae31e78782e1ff418c3e271a9e
@@ -1934,7 +1934,7 @@ F vsixtest/vsixtest.tcl 6a9a6ab600c25a91a7acc6293828957a386a8a93
 F vsixtest/vsixtest.vcxproj.data 2ed517e100c66dc455b492e1a33350c1b20fbcdc
 F vsixtest/vsixtest.vcxproj.filters 37e51ffedcdb064aad6ff33b6148725226cd608e
 F vsixtest/vsixtest_TemporaryKey.pfx e5b1b036facdb453873e7084e1cae9102ccc67a0
-P 2739ed5192058fbcc816ecbc252be687efc606e038bfcd6cf71194a3f4f5684e 6024682ca467fa4fe49608772b0bbfa2f8a419b32cebfa715941073c8b29da49
-R d3afb24f2c63fc5d59ade7856914b3ad
+P d3250256772e3348abe887c0ca3550a6647cce3804c9456a9d0112aea7ee1c46
+R a7fca9e9fc2528f3e921fdfbeaa2e361
 U drh
-Z 246d691185204c115a5a684ad14acaf7
+Z 815ca39cf2b788796cb5ebc9be23c44d
diff --git a/manifest.uuid b/manifest.uuid
index 1416a80b0a..a9d9e689b3 100644
--- a/manifest.uuid
+++ b/manifest.uuid
@@ -1 +1 @@
-d3250256772e3348abe887c0ca3550a6647cce3804c9456a9d0112aea7ee1c46
\ No newline at end of file
+0fb2a4e08f518cb38ea3edc6a084d1e4874fd622ba3cf9101b49b3e7dc1a3f2b
\ No newline at end of file
diff --git a/src/where.c b/src/where.c
index bf10262a76..065d14fd64 100644
--- a/src/where.c
+++ b/src/where.c
@@ -2028,9 +2028,9 @@ void sqlite3WhereLoopPrint(WhereLoop *p, WhereClause *pWC){
     sqlite3_free(z);
   }
   if( p->wsFlags & WHERE_SKIPSCAN ){
-    sqlite3DebugPrintf(" f %05x %d-%d", p->wsFlags, p->nLTerm,p->nSkip);
+    sqlite3DebugPrintf(" f %06x %d-%d", p->wsFlags, p->nLTerm,p->nSkip);
   }else{
-    sqlite3DebugPrintf(" f %05x N %d", p->wsFlags, p->nLTerm);
+    sqlite3DebugPrintf(" f %06x N %d", p->wsFlags, p->nLTerm);
   }
   sqlite3DebugPrintf(" cost %d,%d,%d\n", p->rSetup, p->rRun, p->nOut);
   if( p->nLTerm && (sqlite3WhereTrace & 0x100)!=0 ){
@@ -2502,6 +2502,9 @@ static void whereLoopOutputAdjust(
       if( pX->iParent>=0 && (&pWC->a[pX->iParent])==pTerm ) break;
     }
     if( j<0 ){
+      if( pLoop->maskSelf==pTerm->prereqAll ){
+        pLoop->wsFlags |= WHERE_CULLED;
+      }
       if( pTerm->truthProb<=0 ){
         /* If a truth probability is specified using the likelihood() hints,
         ** then use the probability provided by the application. */
@@ -2529,7 +2532,9 @@ static void whereLoopOutputAdjust(
       }
     }
   }
-  if( pLoop->nOut > nRow-iReduce )  pLoop->nOut = nRow - iReduce;
+  if( pLoop->nOut > nRow-iReduce ){
+    pLoop->nOut = nRow - iReduce;
+  }
 }
 
 /* 
@@ -4984,22 +4989,30 @@ static SQLITE_NOINLINE void whereCheckIfBloomFilterIsUseful(
 ){
   int i;
   LogEst nSearch;
-  SrcItem *pItem;
 
   assert( pWInfo->nLevel>=2 );
   assert( OptimizationEnabled(pWInfo->pParse->db, SQLITE_BloomFilter) );
   nSearch = pWInfo->a[0].pWLoop->nOut;
   for(i=1; i<pWInfo->nLevel; i++){
     WhereLoop *pLoop = pWInfo->a[i].pWLoop;
-    if( pLoop->nOut<0
+    const int reqFlags = (WHERE_CULLED|WHERE_COLUMN_EQ);
+    if( (pLoop->wsFlags & reqFlags)==reqFlags
      && (pLoop->wsFlags & (WHERE_IPK|WHERE_INDEXED))!=0
-     && (pLoop->wsFlags & WHERE_COLUMN_EQ)!=0
-     && nSearch > (pItem = &pWInfo->pTabList->a[pLoop->iTab])->pTab->nRowLogEst
-     && (pItem->fg.jointype & JT_LEFT)==0
     ){
-      pLoop->wsFlags |= WHERE_BLOOMFILTER;
-      pLoop->wsFlags &= ~WHERE_IDX_ONLY;
-      WHERETRACE(0xffff, ("-> use Bloom-filter on loop %c\n", pLoop->cId));
+      SrcItem *pItem = &pWInfo->pTabList->a[pLoop->iTab];
+      Table *pTab = pItem->pTab;
+      pTab->tabFlags |= TF_StatsUsed;
+      if( nSearch > pTab->nRowLogEst
+       && (pItem->fg.jointype & JT_LEFT)==0
+      ){
+        pLoop->wsFlags |= WHERE_BLOOMFILTER;
+        pLoop->wsFlags &= ~WHERE_IDX_ONLY;
+        WHERETRACE(0xffff, (
+           "-> use Bloom-filter on loop %c because there are ~%.1e "
+           "lookups into %s which has only ~%.1e rows\n",
+           pLoop->cId, (double)sqlite3LogEstToInt(nSearch), pTab->zName,
+           (double)sqlite3LogEstToInt(pTab->nRowLogEst)));
+      }
     }
     nSearch += pLoop->nOut;
   }
diff --git a/src/whereInt.h b/src/whereInt.h
index 558a1e3eb9..d790653a71 100644
--- a/src/whereInt.h
+++ b/src/whereInt.h
@@ -608,5 +608,6 @@ void sqlite3WhereTabFuncArgs(Parse*, SrcItem*, WhereClause*);
 #define WHERE_IN_SEEKSCAN  0x00100000  /* Seek-scan optimization for IN */
 #define WHERE_TRANSCONS    0x00200000  /* Uses a transitive constraint */
 #define WHERE_BLOOMFILTER  0x00400000  /* Consider using a Bloom-filter */
+#define WHERE_CULLED       0x00800000  /* nOut reduced by extra WHERE terms */
 
 #endif /* !defined(SQLITE_WHEREINT_H) */