GC refactor ahead
Vasily Chekalkin
bacek at bacek.com
Thu Feb 11 11:44:31 UTC 2010
Patrick R. Michaud wrote:
...
> $ cat fib.nqp
Apparently we need better test now:
bacek at icering:~/src/parrot$ time ./parrot-nqp fib.nqp
fib(28) = 317811
real 0m13.586s
user 0m13.413s
sys 0m0.060s
bacek at icering:~/src/parrot$ time ./parrot-nqp --target=pir fib.nqp
>fib-nqp.pir
real 0m0.273s
user 0m0.256s
sys 0m0.020s
bacek at icering:~/src/parrot$ time ./parrot fib-nqp.pir
fib(28) = 317811
real 0m12.501s
user 0m12.033s
sys 0m0.048s
Less than a second difference after latest PCC improvements. Can anyone
provide test which will favour Generational GC? I'm going to bring Boehm
GC again into Parrot.
--
Bacek
More information about the parrot-dev
mailing list