[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [DotGNU]Port of a Java quantum fractal generator.
From: |
Adam Treat |
Subject: |
Re: [DotGNU]Port of a Java quantum fractal generator. |
Date: |
Wed, 20 Nov 2002 23:45:51 -0500 |
User-agent: |
KMail/1.4.7 |
On Wednesday 20 November 2002 04:24 pm, Rhys Weatherley wrote:
> Adam Treat wrote:
> > A. Portable.Net's CSCC -- 'cscc -o qf.exe -l Qt -O3 FractalViewer.cs
> > QPaintBuffer.cs QuantumFractals.cs IQuantumFractal.cs'
> >
> > 1. 1017675/51.455 = 19777.961
> > 2. 1023028/52.562 = 19463.262
> > 3. 1039391/51.836 = 20051.528
> >
> > ~19764.25 datapoints/second
> > ~1.2 seconds to compile
>
> I'm such a dingbat. Whenever it got the length of an array, it
> was calling the "Array.get_Length" method, instead of using the
> shortcut "ldlen" opcode. The overhead of the method call is
> probably what is swamping the performance of the cscc-compiled
> version.
>
> I've added that special case to the compiler now, which should
> improve performance quite a bit. It had a big improvement on
> PNetMark's "Sieve" benchmark after I recompiled it. Retry it
> and let us know how if it helps.
Ahh, yes ... that helps quite a bit. From a quick look it appears cscc is now
producing in the same ballpark as mcs and csc. Thanks Rhys ;)
Adam