Re: Floating point in pango
- From: Jorn Baayen <jorn openedhand com>
- To: Carl Worth <cworth cworth org>
- Cc: Behdad Esfahbod <behdad behdad org>, performance-list gnome org
- Subject: Re: Floating point in pango
- Date: Thu, 13 Jul 2006 18:26:10 +0300
Hi,
On Wed, 2006-07-12 at 06:13 -0700, Carl Worth wrote:
> On Tue, 11 Jul 2006 12:04:38 +0300, Jorn Baayen wrote:
> > >
> > > What do you exactly mean by 30% slowdown? How was that measured?
> >
> > Timing the drawing of text using GTK 2.6 vs GTK 2.8, by connecting two
> > signal handlers to the 'expose-event' of a GtkLabel, one before and one
> > after the default handler, and timing the difference. The difference is
> > about 30% higher using GTK 2.8. I attach the code used.
> > (All this on ARM)
>
> So has anyone done any profiling of this kind of thing yet?
Yeap. Sorry it took me a while to get back.
I have uploaded oprofile profiles of the test program I posted earlier
to:
http://www.o-hand.com/~jorn/pango-benchmarks/
There are profiles against both GTK+ 2.6.8 as well as 2.8.18. For both
pango 1.10.0 and glib 2.10.3 were used.
Some things that jumped to my eye:
o Using GTK+ 2.8.18 the testcase runs about 30% faster.
o pango_renderer_draw_layout_line() jumped from 2.3% to 8.7% in the
pango profile between 2.6.8 and 2.8.18.
o Why 7% of the time spent in GTK+ spent in
_gtk_style_peek_property_value() using 2.8 ?
Thanks,
Jorn
>
> That's what's needed next before we start guessing about things like
> API changes for cairo, etc.
>
> -Carl
--
OpenedHand Ltd.
http://o-hand.com/
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]