Anush Elangovan wrote: > Bruce, > Is 5.8.5 also exposed to this on 64bit machines, even though the > make check doesn't fail ?
No. The earlier code recast a "char*" into whatever was convenient. Unfortunately, the semantics of "C" have changed over the years and one compiler implements code motion on the presumption that different pointer types point to different places in memory. Therfore, this: *(int*)&(od->pzLastArg) = int-expr; fails. Some of this code is about 15 years old and it was not a problem in those days. So, I now use a unionized bucket and not all elements erase the entire contents when set. That was the problem here. The code trying to determine zero-or-non-zero was looking at the whole "bucket". Oops. I'll go over the code looking for any more of that before I release this weekend. :) Thank you - Bruce ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys -- and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Autogen-users mailing list Autogen-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/autogen-users
