Public bug reported: Glibc version 2.15 has a bug causing spurious FP crashes when calling exp(). It has been reported here:
http://sourceware.org/bugzilla/show_bug.cgi?id=13705 You can find all the details, including a testcase and the patch for libm in this report. I found that the 64-bit version of precise ships a version of glibc that still has this bug. I have not tested the 32-bit version, but I assume it has the same problem. I am one of the developers of the open source code Cloudy and this bug renders our code completely non-functional. I would therefore like to urge you to ship this patch as an update for the precise distribution, especially since this is a LTS release. ProblemType: Bug DistroRelease: Ubuntu 12.04 Package: libc6 2.15-0ubuntu10 ProcVersionSignature: Ubuntu 3.2.0-24.39-generic 3.2.16 Uname: Linux 3.2.0-24-generic x86_64 NonfreeKernelModules: nvidia ApportVersion: 2.0.1-0ubuntu8 Architecture: amd64 Date: Fri Jun 1 16:33:54 2012 InstallationMedia: Kubuntu 11.10 "Oneiric Ocelot" - Release amd64 (20111012) ProcEnviron: LC_TIME=en_DK.utf8 TERM=vt100 LANG=en_US.UTF-8 SHELL=/bin/tcsh SourcePackage: eglibc UpgradeStatus: Upgraded to precise on 2012-05-07 (25 days ago) ** Affects: eglibc (Ubuntu) Importance: Undecided Status: New ** Tags: amd64 apport-bug precise -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/1007457 Title: Bogus FPE on underflow for exp(double) To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/eglibc/+bug/1007457/+subscriptions -- ubuntu-bugs mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
