On 4-Jul-09, at 07:27 , Kevin Horton wrote:
> > On 2-Jul-09, at 13:29 , Robert Miller wrote: > >> Source tarball: >> >> http://sage.math.washington.edu/home/rlmill/release/ >> sage-4.1.alpha3.tar > > Sage-4.1.alpha3 built OK on 32 bit ubuntu 9.04 on an Atom N270. > > Other than two tests which timed out, the following tests failed: > > sage -t -long "devel/sage/doc/fr/tutorial/programming.rst" > sage -t -long "devel/sage/sage/misc/darwin_utilities.pyx" > > I note that the time to complete "make testlong" has increased > substantially since sage 4.0.2. Is that expected? > > On the same machine, with the same OS, on sage 4.0.2 make testlong > took 21342.4 seconds. With sage-4.1.alpha3 it took 28781.0 seconds. > In both cases the CPU was running at the highest selectable > performance level, and minimal other activity was happening during the > tests. > > Digging deeper into the logs, most tests took approximately the same > amount of time on both these versions. I see that on sage-4.1.alpha3, > sage -t -long "devel/sage/sage/groups/perm_gps/partn_ref/ > refinement_matrices.pyx" timed out after 1800 seconds, while on > sage-4.0.2 the same test completed in 759.8 seconds. > refinement_matrices.pyx does not seem to have changed, so there has > been a big performance regression somewhere. > > sage -t -long "devel/sage/sage/rings/qqbar.py" 578.9 s vs 420.7 s > > sage -t -long "devel/sage/sage/graphs/graph.py" 1007.7 s vs 620.7 s > > sage -t -long "devel/sage/sage/geometry/polyhedra.py" 636.3 s vs > 403.4 s > > sage -t -long "devel/sage/sage/schemes/elliptic_curves/padics.py" > 369.3 s vs 267.1 s > > sage -t -long "devel/sage/sage/schemes/elliptic_curves/ell_egros.py" > 598.6 s vs 383.8 s > > This is not an exhaustive list. I looked in the logs from another machine running 32 bit ubuntu 8.10, and see no significant difference in the total time for make testlong between sage-4.0.2 and sage-4.1.alpha3. Perhaps I have something strange going on with my Eee PC rather than performance regressions with sage. Sorry about the noise. -- Kevin Horton Ottawa, Canada --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to sage-devel-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
