Changes which attempt to address an obscure SQLITE_PROTOCOL error.

FossilOrigin-Name: 1725aa7501fa97283a397d4c7cf206f35c6e6bb5

manifest

@@ -1,8 +1,8 @@
 -----BEGIN PGP SIGNED MESSAGE-----
 Hash: SHA1
 
-C Pull\sover\sall\sthe\slatest\strunk\schanges.
-D 2011-01-18T17:34:39.027
+C Changes\swhich\sattempt\sto\saddress\san\sobscure\sSQLITE_PROTOCOL\serror.
+D 2011-02-19T14:19:17.314
 F Makefile.arm-wince-mingw32ce-gcc d6df77f1f48d690bd73162294bbba7f59507c72f
 F Makefile.in de6498556d536ae60bb8bb10e8c1ba011448658c
 F Makefile.linux-gcc 91d710bdc4998cb015f39edf3cb314ec4f4d7e23
@@ -240,7 +240,7 @@ F src/vdbeblob.c 18955f0ee6b133cd08e1592010cb9a6b11e9984c
 F src/vdbemem.c 411649a35686f54268ccabeda175322c4697f5a6
 F src/vdbetrace.c 864cef96919323482ebd9986f2132435115e9cc2
 F src/vtab.c b297e8fa656ab5e66244ab15680d68db0adbec30
-F src/wal.c 2db71704aad93c8af510b9d03db7ff83cb72d262
+F src/wal.c cb338d8f80dece4189ce62b435b53528f4b2d2a7
 F src/wal.h c1aac6593a0b02b15dc625987e619edeab39292e
 F src/walker.c 3112bb3afe1d85dc52317cb1d752055e9a781f8f
 F src/where.c af069e6b53234118014dabfece96a9515b69d76b
@@ -899,14 +899,14 @@ F tool/speedtest2.tcl ee2149167303ba8e95af97873c575c3e0fab58ff
 F tool/speedtest8.c 2902c46588c40b55661e471d7a86e4dd71a18224
 F tool/speedtest8inst1.c 293327bc76823f473684d589a8160bde1f52c14e
 F tool/vdbe-compress.tcl d70ea6d8a19e3571d7ab8c9b75cba86d1173ff0f
-P 2c2afdd0adad7d364915ac200c71603deddf148e e5ca59e63b18ac45a8c82ca39dc8cce1c4ce903c
-R 76d6e877b8871a8c47492488f4bcacce
+P ca86d04be158df89f474e5b82ce3418d282074d7
+R 8b3d99323c46204fea0ea724d430cb5f
 U drh
-Z 1d0fe9ba31d5ff2a091cca17e43bc7c2
+Z 9d0e6478a1d3388663538b3f4c4b6e71
 -----BEGIN PGP SIGNATURE-----
 Version: GnuPG v1.4.6 (GNU/Linux)
 
-iD8DBQFNNc8yoxKgR168RlERApvvAJ4wiRnEhv7J9Ra0GfvLu5YHjC50gQCfV0Qc
-1z7PgObQXL+khH0VurTRvgc=
-=Jmg/
+iD8DBQFNX9FooxKgR168RlERAvuxAJ48VDRP7OeedmVUn0DlNMsLqhjIzACfXrx1
+rn10uZfDJ4qrpkis/50ZxM4=
+=24hC
 -----END PGP SIGNATURE-----

manifest.uuid

@@ -1 +1 @@
-ca86d04be158df89f474e5b82ce3418d282074d7
+1725aa7501fa97283a397d4c7cf206f35c6e6bb5

src/wal.c

@@ -1972,11 +1972,29 @@ static int walTryBeginRead(Wal *pWal, int *pChanged, int useWal, int cnt){
 
   assert( pWal->readLock<0 );     /* Not currently locked */
 
-  /* Take steps to avoid spinning forever if there is a protocol error. */
+  /* Take steps to avoid spinning forever if there is a protocol error.
+  **
+  ** Circumstances that cause a RETRY should only last for the briefest
+  ** instances of time.  No I/O or other system calls are done while the
+  ** locks are held, so the locks should not be held for very long. But 
+  ** if we are unlucky, another process that is holding a lock might get
+  ** paged out or take a page-fault that is time-consuming to resolve, 
+  ** during the few nanoseconds that it is holding the lock.  In that case,
+  ** it might take longer than normal for the lock to free.
+  **
+  ** After 5 RETRYs, we begin calling sqlite3OsSleep().  The first few
+  ** calls to sqlite3OsSleep() have a delay of 1 microsecond.  Really this
+  ** is more of a scheduler yield than an actual delay.  But on the 10th
+  ** an subsequent retries, the delays start becoming longer and longer, 
+  ** so that on the 100th (and last) RETRY we delay for 21 milliseconds.
+  ** The total delay time before giving up is less than 1 second.
+  */
   if( cnt>5 ){
+    int nDelay = 1;                      /* Pause time in microseconds */
     walTrace("cnt=%d",cnt);
     if( cnt>100 ) return SQLITE_PROTOCOL;
-    sqlite3OsSleep(pWal->pVfs, 1);
+    if( cnt>=10 ) nDelay = (cnt-9)*238;  /* Max delay 21ms. Total delay 996ms */
+    sqlite3OsSleep(pWal->pVfs, nDelay);
   }
 
   if( !useWal ){
@@ -2060,24 +2078,9 @@ static int walTryBeginRead(Wal *pWal, int *pChanged, int useWal, int cnt){
       mxI = i;
     }
   }
-  if( mxI==0 ){
-    /* If we get here, it means that all of the aReadMark[] entries between
-    ** 1 and WAL_NREADER-1 are zero.  Try to initialize aReadMark[1] to
-    ** be mxFrame, then retry.
-    */
-    rc = walLockExclusive(pWal, WAL_READ_LOCK(1), 1);
-    if( rc==SQLITE_OK ){
-      pInfo->aReadMark[1] = pWal->hdr.mxFrame;
-      walUnlockExclusive(pWal, WAL_READ_LOCK(1), 1);
-      rc = WAL_RETRY;
-      walTrace("aReadMark[1] <- %d", pWal->hdr.mxFrame);
-    }else if( rc==SQLITE_BUSY ){
-      rc = WAL_RETRY;
-      walTrace("aReadMark[1] is busy");
-    }
-    return rc;
-  }else{
-    if( mxReadMark < pWal->hdr.mxFrame ){
+  /* There was once an "if" here. The extra "{" is to preserve indentation. */
+  {
+    if( mxReadMark < pWal->hdr.mxFrame || mxI==0 ){
       for(i=1; i<WAL_NREADER; i++){
         rc = walLockExclusive(pWal, WAL_READ_LOCK(i), 1);
         if( rc==SQLITE_OK ){
@@ -2091,6 +2094,11 @@ static int walTryBeginRead(Wal *pWal, int *pChanged, int useWal, int cnt){
         }
       }
     }
+    if( mxI==0 ){
+      assert( rc==SQLITE_BUSY );
+      walTrace("all readlocks busy: cannot set read mark");
+      return WAL_RETRY;
+    }
 
     rc = walLockShared(pWal, WAL_READ_LOCK(mxI));
     if( rc ){
@@ -2469,6 +2477,8 @@ static int walRestartLog(Wal *pWal){
     volatile WalCkptInfo *pInfo = walCkptInfo(pWal);
     assert( pInfo->nBackfill==pWal->hdr.mxFrame );
     if( pInfo->nBackfill>0 ){
+      u32 salt1;
+      sqlite3_randomness(4, &salt1);
       rc = walLockExclusive(pWal, WAL_READ_LOCK(1), WAL_NREADER-1);
       if( rc==SQLITE_OK ){
         /* If all readers are using WAL_READ_LOCK(0) (in other words if no
@@ -2486,7 +2496,7 @@ static int walRestartLog(Wal *pWal){
         pWal->nCkpt++;
         pWal->hdr.mxFrame = 0;
         sqlite3Put4byte((u8*)&aSalt[0], 1 + sqlite3Get4byte((u8*)&aSalt[0]));
-        sqlite3_randomness(4, &aSalt[1]);
+        aSalt[1] = salt1;
         walIndexWriteHdr(pWal);
         pInfo->nBackfill = 0;
         for(i=1; i<WAL_NREADER; i++) pInfo->aReadMark[i] = READMARK_NOT_USED;