Re: [Tinycc-devel] Generated code optimization?

[Zoltán Lörincz]

Hello Anton, 
thank you very much for your help. 

I looked into the home page of VMGen. 
Is it possible to generate only the interpreter without rewriting the
code generator ? Can you tell me how big would be (around) the
generated interpreter if i have a register based machine with 300-400
opcodes?
My problem is that the interpreter has to be as small as possible. I
think if VMGen generates the interpreter based on advanced techniques
it means that it will generate bigger code.
The interpreter will run on an embedded processor (MK68). This
processor is slow and the system has small memory which has to be
managed efficiently.
This is the reason why i am asking all that stuff.

Best regards, 
Zoltan.

On Mon, 21 Feb 2005 14:05:36 +0100 (CET), Anton Ertl
<address@hidden> wrote:
> [This time with cc to tinycc-devel]
> 
> =?ISO-8859-1?Q?Zolt=E1n_L=F6rincz?= wrote:
> >
> > Hello all,
> >
> > last year in November i asked on this newsgroup about how could one
> > generate code for a Virtual Machine. Based on the advices i got i
> > implemented a byte code generator.
> > I used tree general integer registers A,B,C load-store based, right
> > now the code generator doesn't support float/double types.
> > Everything is working fine, but the performance i obtained is really
> > disappointing. The execution of the byte code is 30-45 slower then the
> > original code generated whith tcc.
> 
> [Optimization at the VM level]
> 
> Another (orthogonal) way to speed up things is to optimize the VM,
> using techniques such as threaded code, real register allocation,
> static and dynamic superinstructions.  These things require some
> knowledge, so either you acquire and implement that knowledge
> yourself, or you use Vmgen, which was designed to inject that
> knowledge into the interpreters it generates.
> 
> I offered to help with the use of Vmgen for use in a VM for tcc last
> year, and my offer still stands.
> 
> - anton
> 
>