I said I wouldn't be fast and I wasn't. I have found that this bug is not in yorick. The problem lies in the ia64 math library. I am still unsure on how much detail I should have to submit a bug report and whether to send it to glibc maintainers or debian. I am quite sure this could not come from the kinds of changes the debian distro makes; this is deep in the internals of the exp function.
Richard Harke On Thursday 10 December 2009 03:12:46 am you wrote: > Le 10 déc. 09 à 00:25, Richard E. Harke a écrit : > > Has anyone picked this up? > > I have an ia64 machine and could take a look at it. It probably > > wouldn't be fast as I would need to get up to speed on yorick. > > > > Would I need to update my machine to current sid? > > and where would I get the source deb. (unless it is the > > current source for sid) > > Thank you, > > I guess the bug is there already in stable. It was uncovered because I > ran checks at build time with the last upload, but must have been > there for a long time. > > The bug report is there: > http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=560026 > > Once you install Yorick, just run > yorick -batch testfull.i > > If the bug is present, yorick should die with those last few lines: > > _____________ > > Shock tracker timing test: > Timing Category CPU sec System sec Wall sec > Time per pass 0.014 0.000 0.014 > Total time 0.280 0.000 0.283 > > Escape factor timing test: > ERROR (escout) mathlib function signals error > LINE: 251 FILE: /usr/lib/yorick/i/test2.i > yorick: quitting on error in batch mode > ______________ > > Can you start by testing that? > > Best regards, Thibaut. -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected]
