Martin Rubey wrote:
>
> Martin Rubey <[EMAIL PROTECTED]> writes:
>
> > Dear all,
> >
> > I just committed a rather heavy patch. Please please please test as
> > heavily as
> > you can, especially the interpreter.
>
> And, if possible, on various OS and various lisps.
I just finished the standard tests using sbcl, gcl and Closure CL
(former openmcl) on 2.2 GHz Intel Core 2 Duo running 64-bit Debian. I see
no regressions.
Timing for test suite are:
baseline With optimized SPADCALL
sbcl
real 7m1.724s 5m45.769s
user 6m21.120s 5m24.680s
sys 0m20.709s 0m20.341s
Closure CL
real 8m17.316s 8m17.363s
user 7m55.570s 7m52.558s
sys 0m8.717s 0m8.841s
gcl
real 8m27.313s 8m9.125s
user 7m51.217s 7m47.877s
sys 0m21.593s 0m22.505s
Notes:
1) I give three timings because real time is measured quite
accurately, while user and system time is estimated using sampling and
may contain both random and systematic error. OTOH sum user and real
time is intended to better measure CPU load.
2) Baseline real time for sbcl looks like an outlier -- real time
is significanly longer than sum of user and system times.
3) I did not try changing optimize settings.
4) Baseline contains another patch that I am working on (I re-used
the build tree sitting on the disk for tis test), but I belive that
this patch have negligible impact on speed.
I think conclusion from measurements is clear: the SPADCALL patch
gives significant speedup for sbcl (of order 15-25%) while impact
on gcl and Closure CL is much smaller (probably within measurement
noise).
--
Waldek Hebisch
[EMAIL PROTECTED]
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"FriCAS - computer algebra system" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at
http://groups.google.com/group/fricas-devel?hl=en
-~----------~----~----~----~------~----~------~--~---
