Benjamin,
The current Xcode 3.1.1 DP1 can't compile the ppl cvs.
The response from the ppl developer was...
-----------------------------------------------------------
Jack Howarth wrote:
> The ppl cvs only builds with the newer Apple gcc-4.2 compiler
> in Xcode 3.1. Under the default gcc 4.0.1 compiler in Xcode 3.1
> the build fails with the following compile error...
Dear Jack,
It is a bug of GCC 4.0.1. The minimum version of GCC to
successfully compile and use the CVS HEAD version of the
PPL is 4.0.3 (this was not true until a couple of minutes
ago: I have just committed a fix). Notice that in the past
we identified a miscompilation of the PPL with 4.0.2... so
I really would not go below 4.0.3.
All the best,
Roberto
------------------------------------------------------------
I also recall having one of my bug reports against code
generation of the Apple gcc 4.0.1 compiler closed with a
comment that the compiler was depreciated or such. I agree
that we can just build ppl > 0.9 with gcc-4.2 but my point
was that this will cause gcc44 to be available only on 10.5.
The only alternative would be to leave gcc44 building with
ppl 0.9, however, the FSF maintainers have been quite firm
that gcc 4.4.0 should build against a known version of ppl
to insure consistent code generation. The polyhedral results
are certain to be consistent across all arches for a given
release of ppl but this isn't certain to be true between
releases. Also, using ppl 0.9 (if gcc moves on to ppl > 0.9)
would open the gcc44 package up to all sorts of unexpected
issues (since that combination who likely become untested).
We could submit a radar bug report to Apple and hope they
fix it for Xcode 3.1.1 but I won't hold my breath. Also we
would have to configure in ppl > 0.9 to override the new
gcc >= 4.0.3 requirement.
Jack
On Mon, Aug 11, 2008 at 12:35:04PM -0400, Benjamin Reed wrote:
>
> I don't see why specific package's need for a newer GCC would mean we
> have to force all of fink to it. Isn't that the point of having fink
> versions of gcc so that we don't have to rely on a specific Apple GCC
> for problem packages?
>
> Additionally, are you sure it will really need 4.0.3, since Apple's
> 4.0.1 is really 4.0.1 + a buttload of patches from later versions of gcc?
>
> Having the option of using gcc 4.2 as "gcc" in Fink by default would be
> good, of course, but I don't think jumping to forcing everything is a
> good solution just to solve a few specific package issues.
>
> - --
> Benjamin Reed a.k.a. Ranger Rick
> Fink, KDE, and Mac OS X development
>
> Blog: http://www.raccoonfink.com/
> Music: http://music.raccoonfink.com/
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.9 (Darwin)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
>
> iD8DBQFIoGo4Uu+jZtP2Zf4RAp+SAJkBXAN0t162EOA39eG0nv0bnlsw9gCfdf9o
> OgZwsAqSDqLI74d1xY+Pl6w=
> =+STq
> -----END PGP SIGNATURE-----
-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Fink-devel mailing list
[email protected]
http://news.gmane.org/gmane.os.apple.fink.devel
