Re: a5c7b007eaa675cc904f5285291d2f8a597efc84

Re: a5c7b007eaa675cc904f5285291d2f8a597efc84

[Koen Kooi]

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

Hi,

I just noticed a5c7b007eaa675cc904f5285291d2f8a597efc84 [1], which is a
bit bogus.

I've been building gcc 4.2.x for arm on my ubuntu gutsy machine for a
while now, and I can compile things like octave (which needs a fortran
compiler) and hope to be able to build R and grass soon (which also need
a fortran compiler). After this revision was committed bitbake halted to
a stop and informed me I needed to install gfortran. Hmmm.
I know for sure gfortran isn't need and I suspect the real issue is in
gmp-native or mpfr-native not being picked up correctly.
As a test I compiled gcc natively on my efika (without gfortran, but
with mpfr-dev and gmp-dev installed) and that works and now has a
working fortran compiler. I even compiled octave natively and it runs
quite well.

To sum it up:

~ * gfortran is NOT needed to build a gfortran cross-compiler, something
else is wrong.

regards,

Koen

[1]
http://lists.linuxtogo.org/pipermail/openembedded-commits/2008-January/011406.html

- --
koen@dominion.kabel.utwente.nl will go go away in december 2007, please
use k.kooi@student.utwente.nl instead.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (Darwin)

iD8DBQFHhfc2MkyGM64RGpERAgDYAJ0a0kN10TFp3SlCv08riQk2NXZRHACgsBOU
hYdqF4j+KTr5ZkDOP1dZI/A=
=KswC
-----END PGP SIGNATURE-----

Re: a5c7b007eaa675cc904f5285291d2f8a597efc84

[Holger Freyther]

On Thursday 10 January 2008 11:45:10 Koen Kooi wrote:
> ~ * gfortran is NOT needed to build a gfortran cross-compiler, something
> else is wrong.

Hi,

feel free to disapprove it, there are at least two open bug reports regarding 
fortran that we will have to look into.

z.

Re: a5c7b007eaa675cc904f5285291d2f8a597efc84

[Koen Kooi]

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

Holger Freyther schreef:
| On Thursday 10 January 2008 11:45:10 Koen Kooi wrote:
|> ~ * gfortran is NOT needed to build a gfortran cross-compiler, something
|> else is wrong.
|
| Hi,
|
| feel free to disapprove it, there are at least two open bug reports
regarding
| fortran that we will have to look into.

I don't have a (better) solution on how to pull in the deps needed or
fix the -native ones in OE, so it can stay as it it, I just wanted to
raise some awareness of the issue and why fortran was enabled.

regards,

Koen

- --
koen@dominion.kabel.utwente.nl will go go away in december 2007, please
use k.kooi@student.utwente.nl instead.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (Darwin)

iD8DBQFHhgWhMkyGM64RGpERAmgBAKC1Io3+VFpqG24HTkGELawUFaJeiwCgqkzi
2x2WvpLfnAY5o8cbMs4iZzU=
=65Fg
-----END PGP SIGNATURE-----

Re: a5c7b007eaa675cc904f5285291d2f8a597efc84

[Philip Balister]

[-- Attachment #1: Type: text/plain, Size: 1388 bytes --]

I have gfortran installed already (must be the engineer in me) and I 
also had to add the ugly hack in bug 3363.

Koen is right, there is something funny going on.

Philip

Koen Kooi wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Holger Freyther schreef:
> | On Thursday 10 January 2008 11:45:10 Koen Kooi wrote:
> |> ~ * gfortran is NOT needed to build a gfortran cross-compiler, something
> |> else is wrong.
> |
> | Hi,
> |
> | feel free to disapprove it, there are at least two open bug reports
> regarding
> | fortran that we will have to look into.
> 
> I don't have a (better) solution on how to pull in the deps needed or
> fix the -native ones in OE, so it can stay as it it, I just wanted to
> raise some awareness of the issue and why fortran was enabled.
> 
> regards,
> 
> Koen
> 
> - --
> koen@dominion.kabel.utwente.nl will go go away in december 2007, please
> use k.kooi@student.utwente.nl instead.
> 
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.5 (Darwin)
> 
> iD8DBQFHhgWhMkyGM64RGpERAmgBAKC1Io3+VFpqG24HTkGELawUFaJeiwCgqkzi
> 2x2WvpLfnAY5o8cbMs4iZzU=
> =65Fg
> -----END PGP SIGNATURE-----
> 
> _______________________________________________
> Openembedded-devel mailing list
> Openembedded-devel@lists.openembedded.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
> 

[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/x-pkcs7-signature, Size: 3303 bytes --]

Re: a5c7b007eaa675cc904f5285291d2f8a597efc84

[Leon Woestenberg]

Hello,

On Jan 10, 2008 12:51 PM, Philip Balister <philip@balister.org> wrote:
> I have gfortran installed already (must be the engineer in me) and I
> also had to add the ugly hack in bug 3363.
>
> Koen is right, there is something funny going on.
>
Yup.

Why don't we revert the changes then?

Temporarely raising the host requirements is not a nice work around.

Regards,
-- 
Leon

Re: a5c7b007eaa675cc904f5285291d2f8a597efc84

[Michael 'Mickey' Lauer]

Leon Woestenberg wrote:
> Hello,

> On Jan 10, 2008 12:51 PM, Philip Balister <philip@balister.org> wrote:
>> I have gfortran installed already (must be the engineer in me) and I
>> also had to add the ugly hack in bug 3363.
>>
>> Koen is right, there is something funny going on.
>>
> Yup.

> Why don't we revert the changes then?

I have reverted the change to sanity.bbclass.

> Temporarely raising the host requirements is not a nice work around.

Agreed. Folks, please try to fix it differently.

Regards,

:M:

-- 
Dr. Michael 'Mickey' Lauer | IT-Freelancer | http://www.vanille-media.de

Re: a5c7b007eaa675cc904f5285291d2f8a597efc84

[Leon Woestenberg]

Hello,

On Jan 11, 2008 8:53 PM, Michael 'Mickey' Lauer <mickey@vanille-media.de> wrote:
> Leon Woestenberg wrote:
> > Hello,
>
> > On Jan 10, 2008 12:51 PM, Philip Balister <philip@balister.org> wrote:
> >> I have gfortran installed already (must be the engineer in me) and I
> >> also had to add the ugly hack in bug 3363.
> >>
> >> Koen is right, there is something funny going on.
> >>
> > Yup.
>
> > Why don't we revert the changes then?
>
> I have reverted the change to sanity.bbclass.
>
> > Temporarely raising the host requirements is not a nice work around.
>
> Agreed. Folks, please try to fix it differently.
>
Mickey just reverted the hard requirement. Thanks.

It's not that I do not value the effort for Fortran stuff being added,
but we know that if we don't do stuff properly from the beginning, it
will bite us in the end.

Especially so with the toolchains. They are hard enough to do properly already.

Regards,
-- 
Leon

Re: a5c7b007eaa675cc904f5285291d2f8a597efc84

[Paul Sokolovsky]

Hello Leon,

Friday, January 11, 2008, 10:05:16 PM, you wrote:

[]

> Mickey just reverted the hard requirement. Thanks.

> It's not that I do not value the effort for Fortran stuff being added,
> but we know that if we don't do stuff properly from the beginning, it
> will bite us in the end.

  No, it will be biting us all the way thru! ;-D

> Especially so with the toolchains. They are hard enough to do properly already.

> Regards,



-- 
Best regards,
 Paul                            mailto:pmiscml@gmail.com