[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Re: [Gnash-dev] gnash on mini2440 board

From: strk
Subject: Re: Re: [Gnash-dev] gnash on mini2440 board
Date: Sun, 7 Feb 2010 17:51:13 +0100

On Sun, Feb 07, 2010 at 09:11:45AM +0000, address@hidden wrote:
> On Feb 6, 2010 3:46pm, strk <address@hidden> wrote:

> >Any chance to get more detailed profile ?
> Sorry for not sending it in the first time, here it is now.

> samples % symbol name
> 65 8.1864 void agg::render_scanline_aa_solid<agg::scanline_u8,  
> agg::renderer_base<agg::pixfmt_alpha_blend_rgb_packed<agg::blender_rgb565_pre,
> agg::row_accessor<unsigned char>>>, agg::rgba8>(agg::scanline_u8 const&,  
> agg::renderer_base<agg::pixfmt_alpha_blend_rgb_packed<agg::blender_rgb565_pre,
> agg::row_accessor<unsigned char>>>&, agg::rgba8 const&)
> 48 6.0453 agg::rasterizer_cells_aa<agg::cell_style_aa>::sort_cells()
> 38 4.7859  

I was talking about the libstdc++ part.
It is expected for AGG to be the slowest part of Gnash itself.
Please also consider running trunk for more up-to-date profiling
informations. Not sure about your version but in trunk IIRC
we do already use integer values in the renderer.

Oh, take all my words with many dubts these days as I'm having
hard times with hardware (one of my laptop's VGA card just broke).


  ()   Free GIS & Flash consultant/developer

reply via email to

[Prev in Thread] Current Thread [Next in Thread]