aboutsummaryrefslogtreecommitdiff
path: root/src
diff options
context:
space:
mode:
Diffstat (limited to 'src')
-rw-r--r--src/wal.c66
1 files changed, 55 insertions, 11 deletions
diff --git a/src/wal.c b/src/wal.c
index eb97831dc..4e49e707b 100644
--- a/src/wal.c
+++ b/src/wal.c
@@ -2903,6 +2903,37 @@ static int walBeginShmUnreliable(Wal *pWal, int *pChanged){
}
/*
+** The final argument passed to walTryBeginRead() is of type (int*). The
+** caller should invoke walTryBeginRead as follows:
+**
+** int cnt = 0;
+** do {
+** rc = walTryBeginRead(..., &cnt);
+** }while( rc==WAL_RETRY );
+**
+** The final value of "cnt" is of no use to the caller. It is used by
+** the implementation of walTryBeginRead() as follows:
+**
+** + Each time walTryBeginRead() is called, it is incremented. Once
+** it reaches WAL_RETRY_PROTOCOL_LIMIT - indicating that walTryBeginRead()
+** has many times been invoked and failed with WAL_RETRY - walTryBeginRead()
+** returns SQLITE_PROTOCOL.
+**
+** + If SQLITE_ENABLE_SETLK_TIMEOUT is defined and walTryBeginRead() failed
+** because a blocking lock timed out (SQLITE_BUSY_TIMEOUT from the OS
+** layer), the WAL_RETRY_BLOCKED_MASK bit is set in "cnt". In this case
+** the next invocation of walTryBeginRead() may omit an expected call to
+** sqlite3OsSleep(). There has already been a delay when the previous call
+** waited on a lock.
+*/
+#define WAL_RETRY_PROTOCOL_LIMIT 100
+#ifdef SQLITE_ENABLE_SETLK_TIMEOUT
+# define WAL_RETRY_BLOCKED_MASK 0x10000000
+#else
+# define WAL_RETRY_BLOCKED_MASK 0
+#endif
+
+/*
** Attempt to start a read transaction. This might fail due to a race or
** other transient condition. When that happens, it returns WAL_RETRY to
** indicate to the caller that it is safe to retry immediately.
@@ -2952,7 +2983,7 @@ static int walBeginShmUnreliable(Wal *pWal, int *pChanged){
** so it takes care to hold an exclusive lock on the corresponding
** WAL_READ_LOCK() while changing values.
*/
-static int walTryBeginRead(Wal *pWal, int *pChanged, int useWal, int cnt){
+static int walTryBeginRead(Wal *pWal, int *pChanged, int useWal, int *pCnt){
volatile WalCkptInfo *pInfo; /* Checkpoint information in wal-index */
u32 mxReadMark; /* Largest aReadMark[] value */
int mxI; /* Index of largest aReadMark[] value */
@@ -2985,27 +3016,34 @@ static int walTryBeginRead(Wal *pWal, int *pChanged, int useWal, int cnt){
** so that on the 100th (and last) RETRY we delay for 323 milliseconds.
** The total delay time before giving up is less than 10 seconds.
*/
- if( cnt>5 ){
+ (*pCnt)++;
+ if( *pCnt>5 ){
int nDelay = 1; /* Pause time in microseconds */
- if( cnt>100 ){
+ int cnt = (*pCnt & ~WAL_RETRY_BLOCKED_MASK);
+ if( cnt>WAL_RETRY_PROTOCOL_LIMIT ){
VVA_ONLY( pWal->lockError = 1; )
return SQLITE_PROTOCOL;
}
- if( cnt>=10 ) nDelay = (cnt-9)*(cnt-9)*39;
+ if( *pCnt>=10 ) nDelay = (cnt-9)*(cnt-9)*39;
#ifdef SQLITE_ENABLE_SETLK_TIMEOUT
/* In SQLITE_ENABLE_SETLK_TIMEOUT builds, configure the file-descriptor
** to block for locks for approximately nDelay us. This affects three
** locks: (a) the shared lock taken on the DMS slot in os_unix.c (if
** using os_unix.c), (b) the WRITER lock taken in walIndexReadHdr() if the
- ** first attempted read fails, and (c) the shared lock taken on the DMS
- ** slot in os_unix.c. All three of these locks are attempted from within
- ** the call to walIndexReadHdr() below. */
+ ** first attempted read fails, and (c) the shared lock taken on the
+ ** read-mark.
+ **
+ ** If the previous call failed due to an SQLITE_BUSY_TIMEOUT error,
+ ** then sleep for the minimum of 1us. The previous call already provided
+ ** an extra delay while it was blocking on the lock.
+ */
nBlockTmout = (nDelay+998) / 1000;
if( !useWal && walEnableBlockingMs(pWal, nBlockTmout) ){
- nDelay = 1;
+ if( *pCnt & WAL_RETRY_BLOCKED_MASK ) nDelay = 1;
}
#endif
sqlite3OsSleep(pWal->pVfs, nDelay);
+ *pCnt &= ~WAL_RETRY_BLOCKED_MASK;
}
if( !useWal ){
@@ -3015,7 +3053,10 @@ static int walTryBeginRead(Wal *pWal, int *pChanged, int useWal, int cnt){
}
#ifdef SQLITE_ENABLE_SETLK_TIMEOUT
walDisableBlocking(pWal);
- if( rc==SQLITE_BUSY_TIMEOUT ) rc = SQLITE_BUSY;
+ if( rc==SQLITE_BUSY_TIMEOUT ){
+ rc = SQLITE_BUSY;
+ *pCnt |= WAL_RETRY_BLOCKED_MASK;
+ }
#endif
if( rc==SQLITE_BUSY ){
/* If there is not a recovery running in another thread or process
@@ -3135,6 +3176,9 @@ static int walTryBeginRead(Wal *pWal, int *pChanged, int useWal, int cnt){
rc = walLockShared(pWal, WAL_READ_LOCK(mxI));
walDisableBlocking(pWal);
if( rc ){
+ if( rc==SQLITE_BUSY_TIMEOUT ){
+ *pCnt |= WAL_RETRY_BLOCKED_MASK;
+ }
assert( (rc&0xFF)!=SQLITE_BUSY||rc==SQLITE_BUSY||rc==SQLITE_BUSY_TIMEOUT );
return (rc&0xFF)==SQLITE_BUSY ? WAL_RETRY : rc;
}
@@ -3324,7 +3368,7 @@ static int walBeginReadTransaction(Wal *pWal, int *pChanged){
#endif
do{
- rc = walTryBeginRead(pWal, pChanged, 0, ++cnt);
+ rc = walTryBeginRead(pWal, pChanged, 0, &cnt);
}while( rc==WAL_RETRY );
testcase( (rc&0xff)==SQLITE_BUSY );
testcase( (rc&0xff)==SQLITE_IOERR );
@@ -3809,7 +3853,7 @@ static int walRestartLog(Wal *pWal){
cnt = 0;
do{
int notUsed;
- rc = walTryBeginRead(pWal, &notUsed, 1, ++cnt);
+ rc = walTryBeginRead(pWal, &notUsed, 1, &cnt);
}while( rc==WAL_RETRY );
assert( (rc&0xff)!=SQLITE_BUSY ); /* BUSY not possible when useWal==1 */
testcase( (rc&0xff)==SQLITE_IOERR );