[Buildroot] Failing defconfigs due to GCC 6

[Buildroot] Failing defconfigs due to GCC 6

[Arnout Vandecappelle]

 Hi all,

 A number of defconfigs fail to build (usually the kernel, in one case U-Boot)
because of GCC 6. Some of these could be fixed by simply bumping the upstream
kernel; however, we really want defconfigs to be tested on an actual board,
otherwise there's not much point.

 I've put the people who have been involved in these defconfigs in To. Could you
look at these defconfigs and try to fix the build with GCC 6? Anything that
remains unfixed by the end of the month will probably be removed.

beaglebone_qt5_defconfig:
https://gitlab.com/buildroot.org/buildroot/-/jobs/27873600

firefly_rk3288_defconfig:
https://gitlab.com/buildroot.org/buildroot/-/jobs/27873613

firefly_rk3288_demo_defconfig:
https://gitlab.com/buildroot.org/buildroot/-/jobs/27873615

ci40_defconfig:
https://gitlab.com/buildroot.org/buildroot/-/jobs/27873604

armadeus_apf9328_defconfig:
https://gitlab.com/buildroot.org/buildroot/-/jobs/27873568

galileo_defconfig:
https://gitlab.com/buildroot.org/buildroot/-/jobs/27873632

freescale_p1010rdb_pa_defconfig:
https://gitlab.com/buildroot.org/buildroot/-/jobs/27873630


 Regards,
 Arnout
-- 
Arnout Vandecappelle                          arnout at mind be
Senior Embedded Software Architect            +32-16-286500
Essensium/Mind                                http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium           BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint:  7493 020B C7E3 8618 8DEC 222C 82EB F404 F9AC 0DDF

[Buildroot] Failing defconfigs due to GCC 6

[Arnout Vandecappelle]

On 10-08-17 23:01, Arnout Vandecappelle wrote:
>  Hi all,
> 
>  A number of defconfigs fail to build (usually the kernel, in one case U-Boot)
> because of GCC 6. Some of these could be fixed by simply bumping the upstream
> kernel; however, we really want defconfigs to be tested on an actual board,
> otherwise there's not much point.
> 
>  I've put the people who have been involved in these defconfigs in To. Could you
> look at these defconfigs and try to fix the build with GCC 6? Anything that
> remains unfixed by the end of the month will probably be removed.

 Looks like yesterday I missed a few:

riotboard_defconfig:
https://gitlab.com/arnout/buildroot/-/jobs/28979360

telit_evk_pro3_defconfig:
https://gitlab.com/arnout/buildroot/-/jobs/28979373

toradex_apalis_imx6_defconfig:
https://gitlab.com/arnout/buildroot/-/jobs/28979374

ts4900_defconfig:
https://gitlab.com/arnout/buildroot/-/jobs/28979376


 All of these seem to have a DTS in 4.11, so perhaps we can switch to upstream?

 Regards,
 Arnout


-- 
Arnout Vandecappelle                          arnout at mind be
Senior Embedded Software Architect            +32-16-286500
Essensium/Mind                                http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium           BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint:  7493 020B C7E3 8618 8DEC 222C 82EB F404 F9AC 0DDF

[Buildroot] Failing defconfigs due to GCC 6

[Ezequiel Garcia]

On 10 August 2017 at 18:01, Arnout Vandecappelle <arnout@mind.be> wrote:
>
> firefly_rk3288_defconfig:
> https://gitlab.com/buildroot.org/buildroot/-/jobs/27873613
>
> firefly_rk3288_demo_defconfig:
> https://gitlab.com/buildroot.org/buildroot/-/jobs/27873615
>

We no longer have firefly boards. We pushed this upstream
because were supposedly gonna keep the hardware from a
project.

In the end, we had to return the boards.

I think we should remove these if nobody can test.
-- 
Ezequiel Garc?a, VanguardiaSur
www.vanguardiasur.com.ar

[Buildroot] Failing defconfigs due to GCC 6

[Mathieu Audat]

Hi Arnout,

----- Original Message -----
> From: "Arnout Vandecappelle" <arnout@mind.be>
> To: "Buildroot Mailing List" <buildroot@buildroot.org>, "Mathieu Audat" <mathieu.audat@savoirfairelinux.com>, "Fabio
> Estevam" <fabio.estevam@nxp.com>, "Nikolay Dimitrov" <picmaster@mail.bg>, "Fabio Porcedda" <fabio.porcedda@gmail.com>,
> "Sergio Prado" <sergio.prado@e-labworks.com>
> Sent: Friday, 11 August, 2017 09:20:31
> Subject: Re: Failing defconfigs due to GCC 6

> On 10-08-17 23:01, Arnout Vandecappelle wrote:
>>  Hi all,
>> 
>>  A number of defconfigs fail to build (usually the kernel, in one case U-Boot)
>> because of GCC 6. Some of these could be fixed by simply bumping the upstream
>> kernel; however, we really want defconfigs to be tested on an actual board,
>> otherwise there's not much point.
>> 
>>  I've put the people who have been involved in these defconfigs in To. Could you
>> look at these defconfigs and try to fix the build with GCC 6? Anything that
>> remains unfixed by the end of the month will probably be removed.
> 
> Looks like yesterday I missed a few:
> 
> riotboard_defconfig:
> https://gitlab.com/arnout/buildroot/-/jobs/28979360
> 
> telit_evk_pro3_defconfig:
> https://gitlab.com/arnout/buildroot/-/jobs/28979373
> 
> toradex_apalis_imx6_defconfig:
> https://gitlab.com/arnout/buildroot/-/jobs/28979374
> 
> ts4900_defconfig:
> https://gitlab.com/arnout/buildroot/-/jobs/28979376
> 

We still have this board, as you suggested a simple fix would be to bump the kernel.
I'll have a look this week and test it.

> 
> All of these seem to have a DTS in 4.11, so perhaps we can switch to upstream?
>

Indeed, that would be nice :)

> Regards,
> Arnout
> 
> 
> --
> Arnout Vandecappelle                          arnout at mind be
> Senior Embedded Software Architect            +32-16-286500
> Essensium/Mind                                http://www.mind.be
> G.Geenslaan 9, 3001 Leuven, Belgium           BE 872 984 063 RPR Leuven
> LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
> GPG fingerprint:  7493 020B C7E3 8618 8DEC 222C 82EB F404 F9AC 0DDF

-- 
Mathieu Audat
Consultant en logiciel libre | Montr?al, Qc
Savoir-Faire Linux
Bureau: (+1) 514-276-5468 # 356

[Buildroot] Failing defconfigs due to GCC 6

[Kinsella, Ray]

Hi Arnout,

The Galileo and Galileo Gen 2 are supported by the upstream kernel, so that would be simplest approach.
Some minor non-upstream fixes will however be lost along the way.

The biggest obstacle is that I wouldn't be able to do anything on this until at least mid-September (due to travel).
Would be a pity to see Galileo support depreciated in BR.

Ray K

-----Original Message-----
From: buildroot [mailto:buildroot-bounces at busybox.net] On Behalf Of Arnout Vandecappelle
Sent: Thursday, August 10, 2017 2:01 PM
To: Buildroot Mailing List <buildroot@buildroot.org>; Lothar Felten <lothar.felten@gmail.com>; Michael Trimarchi <michael@amarulasolutions.com>; Ezequiel Garc?a <ezequiel@vanguardiasur.com.ar>; Ariel D'Alessandro <ariel@vanguardiasur.com.ar>; Julien Boibessot <julien.boibessot@armadeus.com>; Padraig James Connolly <padraig.connolly@intel.com>; Kinsella, Ray <ray.kinsella@intel.com>; Gustavo Zacarias <gustavo@zacarias.com.ar>
Subject: [Buildroot] Failing defconfigs due to GCC 6

 Hi all,

 A number of defconfigs fail to build (usually the kernel, in one case U-Boot) because of GCC 6. Some of these could be fixed by simply bumping the upstream kernel; however, we really want defconfigs to be tested on an actual board, otherwise there's not much point.

 I've put the people who have been involved in these defconfigs in To. Could you look at these defconfigs and try to fix the build with GCC 6? Anything that remains unfixed by the end of the month will probably be removed.

beaglebone_qt5_defconfig:
https://gitlab.com/buildroot.org/buildroot/-/jobs/27873600

firefly_rk3288_defconfig:
https://gitlab.com/buildroot.org/buildroot/-/jobs/27873613

firefly_rk3288_demo_defconfig:
https://gitlab.com/buildroot.org/buildroot/-/jobs/27873615

ci40_defconfig:
https://gitlab.com/buildroot.org/buildroot/-/jobs/27873604

armadeus_apf9328_defconfig:
https://gitlab.com/buildroot.org/buildroot/-/jobs/27873568

galileo_defconfig:
https://gitlab.com/buildroot.org/buildroot/-/jobs/27873632

freescale_p1010rdb_pa_defconfig:
https://gitlab.com/buildroot.org/buildroot/-/jobs/27873630


 Regards,
 Arnout
-- 
Arnout Vandecappelle                          arnout at mind be
Senior Embedded Software Architect            +32-16-286500
Essensium/Mind                                http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium           BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint:  7493 020B C7E3 8618 8DEC 222C 82EB F404 F9AC 0DDF _______________________________________________
buildroot mailing list
buildroot at busybox.net
http://lists.busybox.net/mailman/listinfo/buildroot

[Buildroot] Failing defconfigs due to GCC 6

[Julien Boibessot]

Hello,

Arnout,

On 10/08/2017 23:01, Arnout Vandecappelle wrote:
>  Hi all,
>
>  A number of defconfigs fail to build (usually the kernel, in one case U-Boot)
> because of GCC 6. Some of these could be fixed by simply bumping the upstream
> kernel; however, we really want defconfigs to be tested on an actual board,
> otherwise there's not much point.
>
>  I've put the people who have been involved in these defconfigs in To. Could you
> look at these defconfigs and try to fix the build with GCC 6? Anything that
> remains unfixed by the end of the month will probably be removed.
>
> ....
>
> armadeus_apf9328_defconfig:
> https://gitlab.com/buildroot.org/buildroot/-/jobs/27873568
>
> ...

As i.MX1 is no more (well) supported in Linux vanilla kernel and APF9328
board is no more recommended for new projects, I think I will send a
patch to remove this defconfig from BR.

OK for you ?

Regards,

[Buildroot] Failing defconfigs due to GCC 6

[Arnout Vandecappelle]

On 16-08-17 12:22, Julien Boibessot wrote:
> Hello,
> 
> Arnout,
> 
> On 10/08/2017 23:01, Arnout Vandecappelle wrote:
>>  Hi all,
>>
>>  A number of defconfigs fail to build (usually the kernel, in one case U-Boot)
>> because of GCC 6. Some of these could be fixed by simply bumping the upstream
>> kernel; however, we really want defconfigs to be tested on an actual board,
>> otherwise there's not much point.
>>
>>  I've put the people who have been involved in these defconfigs in To. Could you
>> look at these defconfigs and try to fix the build with GCC 6? Anything that
>> remains unfixed by the end of the month will probably be removed.
>>
>> ....
>>
>> armadeus_apf9328_defconfig:
>> https://gitlab.com/buildroot.org/buildroot/-/jobs/27873568
>>
>> ...
> 
> As i.MX1 is no more (well) supported in Linux vanilla kernel and APF9328
> board is no more recommended for new projects, I think I will send a
> patch to remove this defconfig from BR.
> 
> OK for you ?

 Yes, thanks.

 Regards,
 Arnout

-- 
Arnout Vandecappelle                          arnout at mind be
Senior Embedded Software Architect            +32-16-286500
Essensium/Mind                                http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium           BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint:  7493 020B C7E3 8618 8DEC 222C 82EB F404 F9AC 0DDF

[Buildroot] Failing defconfigs due to GCC 6

[Arnout Vandecappelle]

On 14-08-17 08:52, Fabio Porcedda wrote:
> If possible i would like to backport the patch to support GCC 6 in the
> mean time.

 That's fine, or alternatively (if you're pretty sure you can find a fundamental
fix eventually) you can just fix GCC to version 5 for now. Up to you.

 Regards,
 Arnout
-- 
Arnout Vandecappelle                          arnout at mind be
Senior Embedded Software Architect            +32-16-286500
Essensium/Mind                                http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium           BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint:  7493 020B C7E3 8618 8DEC 222C 82EB F404 F9AC 0DDF

[Buildroot] Failing defconfigs due to GCC 6

[Arnout Vandecappelle]

On 11-08-17 17:19, Ezequiel Garcia wrote:
> I think we should remove these if nobody can test.

 Could you submit a patch that does that?

 Thanks,
 Arnout

-- 
Arnout Vandecappelle                          arnout at mind be
Senior Embedded Software Architect            +32-16-286500
Essensium/Mind                                http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium           BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint:  7493 020B C7E3 8618 8DEC 222C 82EB F404 F9AC 0DDF

[Buildroot] Failing defconfigs due to GCC 6

[Arnout Vandecappelle]

On 14-08-17 22:38, Kinsella, Ray wrote:
> Hi Arnout,
> 
> The Galileo and Galileo Gen 2 are supported by the upstream kernel, so that would be simplest approach.
> Some minor non-upstream fixes will however be lost along the way.
> 
> The biggest obstacle is that I wouldn't be able to do anything on this until at least mid-September (due to travel).
> Would be a pity to see Galileo support depreciated in BR.

 If you're sure that you can update this defconfig later, I'll for now make a
patch that just fixes GCC to version 5.

 Regards,
 Arnout

-- 
Arnout Vandecappelle                          arnout at mind be
Senior Embedded Software Architect            +32-16-286500
Essensium/Mind                                http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium           BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint:  7493 020B C7E3 8618 8DEC 222C 82EB F404 F9AC 0DDF