This sounds like bad RAM, a wild pointer, or a stack overflow. If it's one of those sorts of problems changing the code in any fashion moves things and the problem goes away/changes. An ICE or good debugger can usually find it. I've also seen a debugger than can run code backwards so you can back up after the problem occurs.
--- Chris Schirlinger <[EMAIL PROTECTED]> wrote: > We have been using SQLite3 for a while now without issues. We are > running a EXE calling a DLL (Which calls a DLL which eventually uses > the SQLite functions) > > These two DLL's are both written in Delphi - in Windows (with a > delphi wrapper that runs the SQLite DLL) > > This code has worked fine under fairly nasty conditions for a LONG > time. > > Recently we tried to tie in with a third party program that calls one > > of our DLL's (which access the SQLite3 DB and returns a data set) > [Note: We eventually started a fresh project for this DLL, so there > was less of a chance legacy code was the issue] > > We don't control the EXE that calls our DLL, basically we export a > couple of functions determined by the third party. The EXE that calls > > our DLL is written in C++ > > Now the issue is, randomly, but ALWAYS within a few dozen calls, the > function sqlite3_step access violates.The first 10 calls > (sqlite3_prepare -> sqlite3_step -> sqlite3_finalise basically) work > fine, then maybe the 11th sqlite3_step AV's > > It can be the first, or the 40th but is guarenteed. > > We tried using sqlite3_get_table and it does the same thing, same > sort of AV > > We tore out any SQLite3 code but kept every other line of code, > returning some made up garbage data. The code worked fine without any > > SQlite3 calls, it seems entirely located in sqlite3_step or any call > that uses it... > > Now having used SQLite3 for a while, I'm not convinced the issue is > in the SQLite3 dll, but maybe the pointer populated by > sqlite3_prepare is being trashed by one of the various parties... > > Has anyone had issues like this before? Or any ideas as to what we > can try next? > > __________________________________ Yahoo! Messenger Show us what our next emoticon should look like. Join the fun. http://www.advision.webevents.yahoo.com/emoticontest