Yeah I don't know why the MPIR test suite didn't fail. It should have been segfaulting like crazy. I did set some fallback values in case tuning values are missing. But I obviously got one wrong or left one out. I'll fix that later.
In the mean time the solution is to tune for those architectures and put the output of the tuning code in mpn/x86/nehalem/gmp-mparam.h, etc. This will both prevent any algorithms being used in regions where they don't work and hopefully also set those two cutoffs I mentioned to 0, which will cut down any chances for a real bug to be exposed. Bill. On 13 September 2015 at 00:18, Volker Braun <vbraun.n...@gmail.com> wrote: > On Saturday, September 12, 2015 at 7:25:34 PM UTC+2, leif wrote: >> >> P.S.: With MPIR 2.7.0.alpha13 on nehalem-pc-linux-gnu ABI=32 (on Linux >> x86_64, Haswell) all tests pass for me with GCC 4.4.3 at least. > > > We do run the mpir testsuite on the Sage buildbots and it didn't fail. We > only noticed this via the arb testsuite. > -- You received this message because you are subscribed to the Google Groups "mpir-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to mpir-devel+unsubscr...@googlegroups.com. To post to this group, send email to mpir-devel@googlegroups.com. Visit this group at http://groups.google.com/group/mpir-devel. For more options, visit https://groups.google.com/d/optout.
