I was just looking at http://trac.sagemath.org/sage_trac/ticket/7095
"os x 10.6 port -- numerous mysterious errors caused by weird 'abort trap' issue" and see discussion about problems with different gcc versions. I've noticed that the MPFR test suite seems to have a habbit of picking up compiler bugs. * It was an MPFR test failure which found the memset() bug on Solaris http://trac.sagemath.org/sage_trac/ticket/6453 * When I got a test failure of MPFR on HP-UX, and reported it, I was told it was a known compiler bug of gcc 4.4.0, which an MPFR developer had reported, and has since been fixed. Reading the MPFR lists, you get the feeling they really do care about quality of code. Compiler warnings are rare, and any I have reported have been investigated and discounted. Although the tests take a bit of time (I think there are 148 of them last time I counted), it might be time well spent. Given the total time it takes to build Sage, I would estimate enabling the MPFR tests would add less than 1% to the build time of Sage Dave -- 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
