Re: [jamboost] Intel toolset for V2

From: Jürgen Hunold (hunold_at_[hidden])
Date: 2004-06-04 02:52:34

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi Volodya !

On Friday 04 June 2004 07:39, Vladimir Prus wrote:
> Michael Stevens wrote:
> > > ... in fact I though there's either vc6, vc7 or vc7.1 mode. I can
> > > add another option, say "native", "full", or even "edg". Just
> > > pick the name ;-)
> >
> > To me it seems very odd to default to putting the Intel compiler in
> > a bug compatible mode. We should really be aiming at ISO standard
> > code. I assume this is just to emulate the current v1 default
> > setting for INTEL_BASE_MSVC_TOOLSET as 'msvc' when regression
> > testing.
>
> Yes, I've tried to be as close to v1 as possible.

Oh, I didn't look at V1 when investigating this. But now I know the 
rationale behind the current setup. 

> > I have an alternative suggestion I use for my own customised
> > toolset files. So that this kind of flag can be permanently set for
> > testing my suggestion is to include them via a single parameter
> > passed to the toolset 'init' function. That way they can be
> > specified in the user-config if necessary for regression testing.
>
> Hey, so one would write
>
> using intel : 8.0vc6 : ...... : -Qvc6 ;
>
> ? That's really interesting. One problem, though, is that
> intel-win.jam has some conditional logic, depending on vc version to
> emulate. But I guess we can just look inside the "extra flags"
> parameter, right?
>
> I think is very neat idea. Other opinions?

I like the current init rule. It should only default to nothing (or at 
least "native" oder maybe better "edg"). And then set some compiler 
options ;-)

I just searched my archives for my intel 7.1 settings for V1 and found:

# STLport does not yet support this.
These were for 7.1 + vc6.0 and STLport-4.5.3 due to the broken 
Vc6.0-Dinkumware-STL. 
Please don't ask me any details, Thomas Witt figured these out.
And we then change 
Warning 1125 is analogous to -Wshadow on gcc and we must remove it due 
to some QT-issues. #70: incomplete type is not allowed is normally a 
warning which Thomas thought should be an error.	

I think that at least some of these should be set when running in 
full-featured EDG-Mode. At least -Qoption,cpp,--no_microsoft_bugs ;-))

Yours, 

Jürgen

- -- 
* Dipl.-Math. Jürgen Hunold ! Institut für Verkehrswesen, Eisenbahnbau
* voice: ++49 511 762-2529 ! und -betrieb, Universität Hannover  
* fax : ++49 511 762-3001 ! Appelstrasse 9a, D-30167 Hannover
* hunold_at_[hidden] ! www.ive.uni-hannover.de
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFAwCpCljbJ/LLrxrYRAkdQAKCeXwpoh1iRPzmLRWvueN2XjqUkuQCbByF6
UaTvZyc2mcK60blkX8MiYkY=
=w6vQ
-----END PGP SIGNATURE-----