SQLITE_LOCKED is an immediate return (that is, it returns immediately, it is not subject to automatic retries or timeouts or the busy/wait handler -- it is an error indication, not necessarily a transient condition).
You may use the sqlite3_unlock_notify API (assuming that you have compiled your SQLite3 with the appropriate option set to enable unlock_notify) to get notified when the lock condition from another connection to the same shared cache clears. If the conflict is within the same connection, then unlock_notify will not work. https://sqlite.org/c3ref/unlock_notify.html Note that if the "extended return code" is SQLITE_LOCKED_SHAREDCACHE then you may use the unlock_notify so that you know when the other connection to the same shared cache has finished with its lock so that you can retry the failed operation (and you may need to do this multiple times since there may be multiple conflicting connections to the same shared cache, and where this is the case, unlock_notify is called when one of them, chosen at random, clears). On the other hand, if the extended error code is SQLITE_LOCKED then you should treat it as SQLITE_MISUSE and immediately longjump to the HEP instruction (Halt and Execute Programmer). --- The fact that there's a Highway to Hell but only a Stairway to Heaven says a lot about anticipated traffic volume. >-----Original Message----- >From: sqlite-users [mailto:sqlite-users- >[email protected]] On Behalf Of Simon Slavin >Sent: Tuesday, 19 March, 2019 21:36 >To: SQLite mailing list >Subject: [sqlite] Do SQLITE_LOCKED situations call the busy-handler ? > >(For those playing along at home, SQLITE_LOCKED is a special-case >version of SQLITE_BUSY. You get it only when the competing access >attempts come from the same connection (multi-tasking) or different >connections sharing the same cache.) > >Suppose I'm in a situation where I would get SQLITE_LOCKED. Does >SQLite attempt to resolve the problem using whatever busy-handler >I've set up before returning that result ? > >The documentation here > ><https://sqlite.org/c3ref/busy_timeout.html> > >seems to say otherwise but I'd like to be sure. > >Simon. >_______________________________________________ >sqlite-users mailing list >[email protected] >http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users _______________________________________________ sqlite-users mailing list [email protected] http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
