This thread got a lot of replies on mpir-devel which didn't show up here for various random reasons.
The gist is that the problem only occurs with debian gcc 4.3.2, and only with fstrict-aliasing which is enabled by -O2. There's a discussion there about what best to do about it in Sage and MPIR. I refer you to mpir-devel for the follow up. Bill. On May 9, 6:28 pm, Nils Bruin <nbr...@sfu.ca> wrote: > Both GMP and MPIR documentation complain about gcc-4.3.2, but I > haven't seen an explicit code snippet that triggers faulty behaviour. > Closest I found was this: > > http://trac.mpir.org/mpir_trac/ticket/291 > > which indicates that at least something in "rootrem.c" goes wrong. I'd > assume the testsuites of both GMP and MPIR pick up on the problem, so > if you're finding everything is fine, it might be worth investigating. -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
