Bug Tracker item #2864278, was opened at 2009-09-22 14:26 Message generated for change (Comment added) made by ebischoff You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=1126467&aid=2864278&group_id=250683
Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: daemon Group: v3.9.0 Status: Open Resolution: None Priority: 5 Private: No Submitted By: Eric Bischoff (ebischoff) Assigned to: Nobody/Anonymous (nobody) Summary: floating point exception in cssclean Initial Comment: Command "cssclean filename.css" produces a "fFoating point exception (core dumped)" with my CSS file. Bug seen with dspam 3.9.0 (experimental kubuntu packages, or self-compiled from git version 2009-09-22) on a 64-bit Linux. I tracked the bug and it appears that, in the hash driver, _hash_drv_set_spamrecord() calls itself a great number of times without ever returning, and always seeks offset 0. Finally, _hash_drv_autoexetend() returns a nonzero value, and for the first time _hash_drv_set_spamrecord() returns with EFAILURE code. That's at this point that a floating point exception appears. I got the impression that the hash driver has some kind of protection against infinite recursion, but that it does not work properly. The funny thing is that dspam works very correctly with that CSS file. It's only cssclean that has a problem. The CSS file is too big (40 Mb) to be attached here. Please send a private email and I'll publish it on some web site. ---------------------------------------------------------------------- >Comment By: Eric Bischoff (ebischoff) Date: 2009-09-22 14:40 Message: Well, as a matter of fact, the core dump is not always reproductible: - when _hash_drv_set_spamrecord() returns with an error code, the program ends cleanly - otherwise, it at times tries to seek a non-zero offset (apparently very big and outside of the file) and then it generates a floating point exception. Sorry, I mixed up both cases. Still, how could we avoid having broken CSS files? Could cssclean fix them to some extend? Could the core dump be avoided by checking the value of the offset? ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=1126467&aid=2864278&group_id=250683 ------------------------------------------------------------------------------ Come build with us! The BlackBerry® Developer Conference in SF, CA is the only developer event you need to attend this year. Jumpstart your developing skills, take BlackBerry mobile applications to market and stay ahead of the curve. Join us from November 9-12, 2009. Register now! http://p.sf.net/sfu/devconf _______________________________________________ Dspam-devel mailing list Dspam-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dspam-devel
