Okay... I was not initializing fricas's gmp support correctly. (It's a bug in gentoo ebuild, and I've reported it upstream.)
So our in-house gmp_wrap takes 242s as well. So it's all good after all. (Side note: shall we put the result of gmp initialization into fricas startup banner?) - Qian On 9/17/24 8:28 PM, Qian Yun wrote:
I'm talking about the official sb-gmp package: https://github.com/sbcl/sbcl/tree/master/contrib/sb-gmp You can enable it in fricas by: )lisp (require :sb-gmp) )lisp (SB-GMP:INSTALL-GMP-FUNS) The test case I am profiling is: )lisp (sb-sprof:start-profiling) integrate(D((1+x^(2/3))^(1/3),x),x) )lisp (progn (sb-sprof:stop-profiling) (sb-sprof:report)) By default, it takes 1326s, with sb-gmp, it takes 242s. It seems that most time is spent in bignum gcd. I'll take a deeper look to see if there are optimization opportunities related with polynomials. - Qian
-- You received this message because you are subscribed to the Google Groups "FriCAS - computer algebra system" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/fricas-devel/2bc9b5ab-31f3-482e-8bea-b4b6f36ab63f%40gmail.com.
