setoolkit-developer

[SEToolkit-developer] Work on 3.5.0 continued

[Dagobert M. <da...@ba...>]

Hi,

as you may have notices I continued to work on the 3.5.0 release.
I have already pulled in all include files from Orca which you
proposed.

Alex: Good to see you working on the docs.
Jon: You said something about throrough script testing. Is there
anything you can commit so we can include that in 3.5.0?

Best regargds

   -- Dago

--=20
Dagobert Michelsen (Leiter IT)          Baltic Online Computer GmbH
Firmensitz:   Alter Markt 1-2, 24103 Kiel, Telefon: +49 431 54003-0
Gesch=E4ftsf=FChrer:        Erik Cickovskis, Amtsgericht Kiel, HRB 3756
"Of course computer servers don't need thrust, since they generally
don't go anywhere."  -- Comment in TR on new HP server turbine fans

Re: [SEToolkit-developer] Work on 3.5.0 continued

[Alex K. <ale...@gm...>]

On 20/08/07, Dagobert Michelsen <da...@ba...> wrote:
> Hi,
>
> as you may have notices I continued to work on the 3.5.0 release.
> I have already pulled in all include files from Orca which you
> proposed.
>
> Alex: Good to see you working on the docs.

I spotted you'd been doing something and thought I really must get
around to getting the odds & ends I had comitted!

> Jon: You said something about throrough script testing. Is there
> anything you can commit so we can include that in 3.5.0?
>

Thoughts on 64 bit automagic welcome... from memory we were planning
on some simple wrappers which would do the right thing?

-- 
Alex Kiernan

Re: [SEToolkit-developer] Work on 3.5.0 continued

[Dagobert M. <da...@ba...>]

Hi,

Von "Alex Kiernan" <ale...@gm...> (Tue, 21 Aug 2007 06:00:17 +0100):
> On 20/08/07, Dagobert Michelsen <da...@ba...> wrote:
> > as you may have notices I continued to work on the 3.5.0 release.
> > I have already pulled in all include files from Orca which you
> > proposed.
> >
> > Alex: Good to see you working on the docs.
> 
> I spotted you'd been doing something and thought I really must get
> around to getting the odds & ends I had comitted!

doc/se.1m contains the path to the default include path. This
should be automatically adjusted to $seincludedir which is
set in Makefile.am. Should this be set in configure.ac so
se.1m can be modified in AC_CONFIG_FILES or is there a
better alternative?

> > Jon: You said something about throrough script testing. Is there
> > anything you can commit so we can include that in 3.5.0?
> 
> Thoughts on 64 bit automagic welcome... from memory we were planning
> on some simple wrappers which would do the right thing?

How about the configure-directive
  --with-memory-model=(auto | 32 | 64)   (default=auto)
which adds the 64 bit flags when memory-model=auto and
isainfo -b = 64 or memory-model=64. Compiler flags for
Sun Studio 10, Sun Studio 12 and gcc should be tried then.

Best regards

  -- Dagobert

--
Dagobert Michelsen (Leiter IT)          Baltic Online Computer GmbH
Firmensitz:   Alter Markt 1-2, 24103 Kiel, Telefon: +49 431 54003-0
Geschäftsführer:        Erik Cickovskis, Amtsgericht Kiel, HRB 3756
"Of course computer servers don't need thrust, since they generally
don't go anywhere."  -- Comment in TR on new HP server turbine fans

Re: [SEToolkit-developer] Work on 3.5.0 continued

[Alex K. <ale...@gm...>]

On 21/08/07, Dagobert Michelsen <da...@ba...> wrote:
> Hi,
>
> Von "Alex Kiernan" <ale...@gm...> (Tue, 21 Aug 2007 06:00:17 +0100):
> > On 20/08/07, Dagobert Michelsen <da...@ba...> wrote:
> > > as you may have notices I continued to work on the 3.5.0 release.
> > > I have already pulled in all include files from Orca which you
> > > proposed.
> > >
> > > Alex: Good to see you working on the docs.
> >
> > I spotted you'd been doing something and thought I really must get
> > around to getting the odds & ends I had comitted!
>
> doc/se.1m contains the path to the default include path. This
> should be automatically adjusted to $seincludedir which is
> set in Makefile.am. Should this be set in configure.ac so
> se.1m can be modified in AC_CONFIG_FILES or is there a
> better alternative?
>

Ah, there's a bunch of others (startup scripts etc.) which could do
with similar treatment. I'll stick a bug in for it.

> > > Jon: You said something about throrough script testing. Is there
> > > anything you can commit so we can include that in 3.5.0?
> >
> > Thoughts on 64 bit automagic welcome... from memory we were planning
> > on some simple wrappers which would do the right thing?
>
> How about the configure-directive
>  --with-memory-model=(auto | 32 | 64)   (default=auto)
> which adds the 64 bit flags when memory-model=auto and
> isainfo -b = 64 or memory-model=64. Compiler flags for
> Sun Studio 10, Sun Studio 12 and gcc should be tried then.
>

My big problem is where to put them and how they interact with the
other options.

The naive place to put them is CFLAGS, but that's wrong
(http://www.gnu.org/software/autoconf/manual/autoconf.html#Preset-Output-Variables),
which is where the whole problem gets messy - if I got passed some
flags do I start trying to pick out the pieces, or what. Hence why I'd
be entirely happy to add:

<blink><bold><italic>THE BINARY YOU ARE BUILDING WON'T RUN ON THE
PLATFORM YOU'RE COMPILING IT ON</italic></bold></blink>

type warning to the autoconf and provide folk with a set of wrappers
which pass the right options in (having done the equivalent of a `env
-').

Not that I'm bitter about having fought this kind of battle with
autoconf'd programs a zillion times you understand... :)

-- 
Alex Kiernan