It was triggered by old non-conforming feature of ECL 16.1.2; ECL 16.1.3 is more standard-conforming, and so for our purposes we need to explicitly disable SIGFPE traps. Took a while to understand, sorry for noise.
On Tuesday, January 17, 2017 at 9:46:50 AM UTC, Dima Pasechnik wrote: > > On Tuesday, January 17, 2017 at 6:50:25 AM UTC, Ralf Stephan wrote: >> >> As you already noticed that the bug depends on the ECL/Maxima >> version, let me just add that there is no need for taylor() (Maxima) >> if you use series() (Pynac) which is way faster. Also there is >> >> https://trac.sagemath.org/ticket/6119 >> > > well, taylor() was there as it came from the doctest showing the problem > with ECL 16.1.3. > Anything that triggers loading ECL/Maxima library interface readies the > bug to go off, anyway. > > Then something like > > plot(log(x),(x,0.0,2.0)) > > does trigger the crash itself. And it appears that the latter does not use > ECL/Maxima, right? > (invoking this plot() before calling taylor() does not lead to a problem). > > -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at https://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
