Fwd: Re: [Shootout-list] Science-related benchmarks (speedoptimisation)

[Dave]

> -----Original Message-----
> From: shootout-list-admin@lists.alioth.debian.org
> [mailto:shootout-list-admin@lists.alioth.debian.org]On Behalf Of skaller
> Sent: Wednesday, April 27, 2005 3:39 AM
> To: Bengt Kleberg
> Cc: shootout
> Subject: Re: Fwd: Re: [Shootout-list] Science-related benchmarks
> (speedoptimisation)
>
>
> On Wed, 2005-04-27 at 16:11, Bengt Kleberg wrote:
>
> > the shootout has a policy of allowing different compile, and start-up,
> > switches for different tests.
> > this does not figure prominently in the faq, and should.
>
> Alternative, the rule should be changed.
> I favour that.
>
> gcc  and gcc -O3 are completely distinct language
> translators from a Standards conformance testing
> viewpoint. You're comparing apples and oranges
> if you allow gcc with different memory models, etc,
> for each test.
>

I completely disagree.. It is common and widespread to suit compiler
switches to the job at hand, or even to the stage of development (debug vs.
'gold' code for example) of the software. That's what the switches are
intended for. It's the job of the compiler implementers to make sure it
compiles the software to function in the same way given the same input with
different switches.

You can disagree w/ the plethora of switches available, the reasoning behind
having so many switches, what they do and even the need for them, but to
limit any compiler (C/++ or not) to what switches can be used to produce the
best result for any program, especially where top performance is important,
really would go against common 'real-world' usage. And would therefore make
the Shootout much less informative to many people.

>
> If you just want to measure 'C', with the best tool,
> then only one of Intel C and gcc compilers should be
> allowed for any particular test. After all, the main
> difference is in the optimising behaviour of the
> back ends, which varies far more by switches on one
> tool than between the tools .. OS ABI's ensure that
> this must be the case on many systems (such as
> C++ ABI for Solaris).
>

A good deal has been made in the past wrt how 'interesting' the tests and
comparisons are. A big part of this for me is seeing different
implementations for the same language run the same code. The comparisons for
different implementations of the same language are very informative, at
least to me, and probably to many other people out there where more than one
implementation is available for their language(s) of interest.

- Dave

> --
> John Skaller, mailto:skaller@users.sf.net
> voice: 061-2-9660-0850,
> snail: PO BOX 401 Glebe NSW 2037 Australia
> Checkout the Felix programming language http://felix.sf.net
>
>
>
>
> _______________________________________________
> Shootout-list mailing list
> Shootout-list@lists.alioth.debian.org
> http://lists.alioth.debian.org/mailman/listinfo/shootout-list
>
>