Hello, How about shipping lbzip2 instead of bzip2? Or at least not building bzip2 if there is already one system-wide, so that it can be replaced with a threaded version? (Were there issues with system bzip2?)
It is most noticeable on the huge polytope database, but will be also useful for other big optional packages and perhaps even for standard ones. Some timing from a 32-core machine with SSD: novoselt@sagenb:~$ time sage -i polytopes_db_4d-1.0.spkg # Pre-downloaded file ... real 0m23.619s ... real 17m51.526s First time is installation, second the overall time that included unpacking. Given that "make distclean && make ptestlong" takes ~40 minutes, it is annoyingly long to spend half of that just to install some precomputed files ;-) Testing (l)bzip2: novoselt@sagenb:~$ time lbzip2 -t polytopes_db_4d-1.0.spkg real 0m40.553s user 20m45.664s sys 0m38.852s novoselt@sagenb:~$ time bzip2 -t polytopes_db_4d-1.0.spkg real 16m55.583s user 16m51.720s sys 0m4.292s Timing on ATLAS spkg (~9Mb): novoselt@sagenb:~/test$ time bzip2 -t atlas-3.10.1.20140210.tar.bz2 real 0m2.635s user 0m2.624s sys 0m0.008s novoselt@sagenb:~/test$ time lbzip2 -t atlas-3.10.1.20140210.tar.bz2 real 0m0.380s user 0m4.652s sys 0m1.472s And a test on a more typical 8-core desktop: novoselt@sage:~$ time lbzip2 -t polytopes_db_4d-1.0.spkg real 1m34.704s user 11m45.152s sys 0m31.300s novoselt@sage:~$ time bzip2 -t polytopes_db_4d-1.0.spkg real 11m28.946s user 11m24.052s sys 0m4.180s Andrey -- 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 http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
