On Jul 26, 2008, at 2:12 PM, Ben Smith wrote: > Hi Folks - > > I have an issue (that others are reporting as well at Apple > discussions); when using the built in SQLite3 library it seems that > finalize and close doesn't actually release all of the memory of a > particular connection. It isn't reported as a leak or anything. But > if you use instruments you can see that the libsqlite3 is sitting > there talking up memory (not a lot). It doesn't seem to have anything > to do with what the db has in it or anything, it's just based on the > number connections you have ever established (even if they have all > been killed). > > > Is this a known issue? Or am I doing something wrong (_close and > _finalize are the only two de-constructors or is there something else > I should be doing as well).
SQLite is carefully tested to ensure that all memory is released after finalize & close. Such a memory leak would show up on all of our standard tests and would be very in-your-face. It is not the kind of thing we are likely to miss. I would be surprised to find that this is something in the SQLite core. Are you certain that all prepared statements are finalized prior to calling sqlite3_close()? I have no idea what kind of VFS layer is being used for the iPhone. It might be something custom. If so, there could be a leak in the VFS layer someplace. D. Richard Hipp [EMAIL PROTECTED] _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
