Re: [Gm2] Re: data alignment is now possible

[Fischlin Andreas]

I concur with Scott on this.

Hardware dependencies are almost by definition not portable. Of course no one 
should not strive for utmost portability and hardware independence. But if you 
need to access the hardware, you might have the more very good reasons to do 
so. Therefore why not allowing to do that? Modula-2 was designed with that in 
mind, since its predecessor Modula(-1) was designed to write operating systems 
with. The goal was to write an OS in a high level language as far away from the 
hardware as possible, yet not preventing anyone to get close to the hardware if 
there should be a strong need to do so. The latter  being expected to always 
being carefully considered and accepted as a last resort only. 

Regards,
Andreas


ETH Zurich
Prof. Dr. Andreas Fischlin
Systems Ecology - Institute of Integrative Biology
CHN E 21.1
Universitaetstrasse 16
8092 Zurich
SWITZERLAND

address@hidden
www.sysecol.ethz.ch

+41 44 633-6090 phone
+41 44 633-1136 fax
+41 79 221-4657 mobile

            Make it as simple as possible, but distrust it!
________________________________________________________________________



On 23/Jun/2010, at 00:56 , SiTex Graphics wrote:

> On Mon, Jun 21, 2010 at 5:36 AM, Iztok Kobal <address@hidden> wrote:
>> As for the others - I would strongly discourage anybody to use packed
>> structures for implementing e.g. communication protocols or HW accesses
> 
> For that narrow set of uses I'd agree.
> 
>> If I were an advisor for GNUM2, I would never support the structure fields
>> packing and clearly state that it is not supported and not to count on it.
> 
> Packed structures can reduce memory use significantly and increase
> cache coherence, which is very important for good performance (at
> least on modern x86 architectures).  I don't see why making that
> option available for architectures where it is advantageous should
> affect those where it is not since the default settings produce
> unpacked structures.
> 
> Regards,
> Scott
> 
> _______________________________________________
> gm2 mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/gm2