I got the right answer (counting different cores of the same processor as different processors) in all the computers i have checked (with different number of processors and different flavours of linux). All of them are x86 or amd64 architecture. Still have to check how does it work with atom processors with one core but multi thread technology.
Miguel On 23 sep, 07:13, John H Palmieri <jhpalmier...@gmail.com> wrote: > The goal of trac #6283 is 'Make it so NUM_THREADS is set intelligently > instead of idiotically in makefile so doing "make ptest" or "make > ptestlong" doesn't kill some computer'. Right now, NUM_THREADS (set in > SAGE_ROOT/makefile) is used for parallel testing if you do "make > ptest" or "make ptestlong"; the makefile says that in the future, it > could be used for parallel building. One idea is to set NUM_THREADS > equal to the number of cpus, or processors, or cores, or something > like that. More precisely, the idea is to use the output from this: > > sage: import multiprocessing > sage: multiprocessing.cpu_count() > > This seems to give reasonable numbers for my iMac, an ubuntu box I > sometimes use, and sage.math. Does it give bad numbers for your > computer? > > If we can get reasonable numbers using some method like this, then I > propose to change the command "sage -tp N <files>": if N==0, then > change N to be the number of cpus. Then we change NUM_THREADS to 0 at > the top of SAGE_ROOT/makefile, and away we go... > > John --~--~---------~--~----~------------~-------~--~----~ 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 -~----------~----~----~----~------~----~------~--~---
