Re[6]: [Gc] Performance of bdwgc7.2 had degraded compared to 6.8
Ivan Maidanski
ivmai at mail.ru
Mon Nov 29 22:37:22 PST 2010
Hi!
Since I don't know where is the problem, I've prepared a test patch against gc7.1.
So, please compare the speed (against v71/v72a2) of the following: g71+my_patch, gc71+my_patch+(-DSTATIC=).
I assume that: you dont use GC_DEBUG (or any GC_debug_ funcs), GC_ASSERTION, GC_THREADS, "GC_PRINT_STATS" (or any other GC statistic printing routines). I also assume that your target host is Lunux/x86 (or x64). Is this correct?
PS. Due to the mailing list restrictions, I'll send the patch (directly to the recipients) in a separate letter.
Regards.
Wed, 24 Nov 2010 16:52:11 +0100 (CET) Sven Hartrumpf <hartrumpf at gmx.net>:
> I combined all measurements (in seconds) done by Manuel in one table -
> for easier comparison.
>
> Stay away from 7.2 if you can, I would conclude from this table and
> my own benchmarks :-)
>
>
> 7.2a4 7.2a2 7.1 7.0 7.0a7 6.8
> bague 0.76 0.77 0.77 0.76 0.77 0.77
> beval 1.33 1.41 1.29 1.41 1.29 1.31
> boyer 2.23 2.23 2.13 2.14 2.13 2.15
> cgc 0.47 0.48 0.48 0.47 0.48 0.46
> conform 1.91 1.91 1.74 1.72 1.73 1.79
> earley 2.49 2.50 2.08 2.13 2.09 2.23
> fib 0.01 0.01 0.01 0.01 0.01 0.01
> fft 2.51 2.52 2.52 2.50 2.52 2.49
> leval 1.12 1.13 1.05 1.01 1.02 1.09
> maze 1.67 1.40 1.36 1.35 1.26 1.39
> mbrot 7.03 7.05 7.04 7.03 7.05 7.05
> nucleic 1.18 1.20 1.20 1.16 1.16 1.34
> peval 1.46 1.47 1.20 1.19 1.20 1.18
> puzzle 1.96 1.92 1.97 1.96 1.92 1.93
> queens 2.29 2.29 1.55 1.56 1.55 1.44
> qsort 1.65 1.64 1.63 1.62 1.63 1.63
> rgc 1.28 1.28 1.23 1.23 1.24 1.28
> sieve 1.58 1.60 1.44 1.42 1.41 1.51
> traverse 5.14 5.15 3.55 3.60 3.56 3.58
> almabench 1.45 1.45 1.45 1.45 1.45 1.46
> SUM 39.52 39.41 35.69 35.72 35.47 36.09
> _______________________________________________
> Gc mailing list
> Gc at linux.hpl.hp.com
> http://www.hpl.hp.com/hosted/linux/mail-archives/gc/
More information about the Gc
mailing list