On Tuesday, June 27, 2017 at 10:14:30 PM UTC+12, François Bissey wrote: > > Well after realising that in its current state #12426/#22646/#23046 > would at least let me try to build sage with the intel compiler, > I just did go ahead and got myself an open source license. > > On top of the above tickets I had to patch _2_ packages to finish the > build. > * sqlite [icc has all the pragma to pass itself for gcc 6.3.0 (the > installed version of gcc) but miss some builtin of said version] > * openblas [assembly problem, fixed upstream but not in a stable release] > > The build had a hiccup while building the documentation. Several > backtraces in ecl and then it hung. But after restarting the build > everything finished without trouble. > > doctesting results: > ---------------------------------------------------------------------- > sage -t --long src/sage/schemes/elliptic_curves/ell_rational_field.py # > 17 doctests failed > sage -t --long src/sage/rings/qqbar.py # 1 doctest failed > sage -t --long src/sage/symbolic/expression.pyx # 1 doctest failed > sage -t --long src/sage/modular/abvar/abvar.py # 1 doctest failed > sage -t --long src/sage/groups/perm_gps/cubegroup.py # 3 doctests failed > sage -t --long src/sage/modular/hecke/submodule.py # 1 doctest failed > sage -t --long src/sage/lfunctions/sympow.py # 13 doctests failed > sage -t --long src/sage/libs/pari/tests.py # 2 doctests failed > sage -t --long src/sage/libs/lcalc/lcalc_Lfunction.pyx # 1 doctest failed > sage -t --long src/sage/rings/real_double.pyx # 7 doctests failed > sage -t --long src/sage/tests/french_book/float_doctest.py # 1 doctest > failed > ——————————————————————————————————— > * sympow is broken [at least 30 doctests above] not really surprising. > * some problems in pari/cypari > * quite a bit of numerical noise > > Overall surprisingly good results. > > icc over-optimise pari and sympow. pari fails its test suite if compiled with -O2 or -O3. At -O1 it pass it test suite and sage doctests failure related to pari go away. sympow needs to be compiled at -O0. The only level, where as far as I can tell, something like -ffp-contract=on is enforced. With these two adjustments all the remaining doctest failure are numerical noise except for one, which I am not sure of. Could be a form of numerical noise. doctest failure left: sage -t --long src/sage/symbolic/expression.pyx # 1 doctest failed sage -t --long src/sage/groups/perm_gps/cubegroup.py # 3 doctests failed sage -t --long src/sage/libs/lcalc/lcalc_Lfunction.pyx # 1 doctest failed sage -t --long src/sage/rings/real_double.pyx # 7 doctests failed sage -t --long src/sage/tests/french_book/float_doctest.py # 1 doctest failed
with the peculiar being: ********************************************************************** File "src/sage/rings/real_double.pyx", line 2296, in sage.rings.real_double.RealDoubleElement.arccos Failed example: i.arccos() == q Expected: True Got: False ********************************************************************** If anyone has an idea. Francois -- 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.
