Re: Tools for RTEMS 4.12

[Chris Johns]

On 6/11/2015 7:31 am, Sebastian Huber wrote:
> On 06/11/15 15:12, Chris Johns wrote:
>> On 5/11/2015 6:56 am, Gedare Bloom wrote:
>>> >I see no problem with using the newer GCC assuming our development
>>> >cycle is probably still at least 1yr+. Until we get good automation
>>> >this is the case. Also, we are planning to apply to Google Code-In,
>>> >and bumping the tool versions could be a set of tasks. If you would
>>> >like to do one or a few as a sample, then we can have high school
>>> >students do the rest during the GCI program period.
>> I am planing to talk to Joel about the RSB and it's configuration files.
>> The RSB has arrived at an interesting place where I am not sure we want
>> all versions in an RSB. If we strip back the configurations to just the
>> valid ones we will then need to have a specific versions for specific
>> releases. This would simplify the number of files.
> 
> What I noticed during creation of the 4.12 tools that a lot of includes
> are involved which make it difficult to determine what is actually the
> case. I would use more copy and paste, otherwise if you change e.g. the
> basic GCC build file, then you would have to build a lot of stuff to
> test everything. I think its better to use specific files for a specific
> RTEMS version.

The original idea was to up the last number of the base file if the
changes broke a previous version. I do not know how well this works so I
am wondering about just removing them and then have a release branch
that tracks an RTEMS release branch.

> 
>>
>> It would be nice to resolve this before adding 4.12 support.
> 
> The 4.11 branch was created more than three months ago, we should really
> provide a 4.12 tool chain soon and independent of RSB internal issues.
> 

Yes.

Chris
___
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel

Re: Tools for RTEMS 4.12

[Gedare Bloom]

On Fri, Nov 6, 2015 at 9:12 AM, Chris Johns  wrote:
> On 5/11/2015 6:56 am, Gedare Bloom wrote:
>> I see no problem with using the newer GCC assuming our development
>> cycle is probably still at least 1yr+. Until we get good automation
>> this is the case. Also, we are planning to apply to Google Code-In,
>> and bumping the tool versions could be a set of tasks. If you would
>> like to do one or a few as a sample, then we can have high school
>> students do the rest during the GCI program period.
>
> I am planing to talk to Joel about the RSB and it's configuration files.
> The RSB has arrived at an interesting place where I am not sure we want
> all versions in an RSB. If we strip back the configurations to just the
> valid ones we will then need to have a specific versions for specific
> releases. This would simplify the number of files.
>
Sounds good.

> It would be nice to resolve this before adding 4.12 support.
>
And I am still thinking if we do GCI it will make a nice set of tasks.

> Chris
> ___
> devel mailing list
> devel@rtems.org
> http://lists.rtems.org/mailman/listinfo/devel
___
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel

Re: Tools for RTEMS 4.12

[Chris Johns]

On 5/11/2015 6:56 am, Gedare Bloom wrote:
> I see no problem with using the newer GCC assuming our development
> cycle is probably still at least 1yr+. Until we get good automation
> this is the case. Also, we are planning to apply to Google Code-In,
> and bumping the tool versions could be a set of tasks. If you would
> like to do one or a few as a sample, then we can have high school
> students do the rest during the GCI program period.

I am planing to talk to Joel about the RSB and it's configuration files.
The RSB has arrived at an interesting place where I am not sure we want
all versions in an RSB. If we strip back the configurations to just the
valid ones we will then need to have a specific versions for specific
releases. This would simplify the number of files.

It would be nice to resolve this before adding 4.12 support.

Chris
___
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel

Re: Tools for RTEMS 4.12

[Gabriel Ibarra]

Hi,
The issue of recursive calls of callocs was fixed in RTEMS mainline adding
-fno-builtin in the calloc.c compilation, but we made a gcc patch in order
to use -fno-builtin-calloc instead of -fno-builtin, this is more specific
and disables only calloc builtin optimization.

Currently there is an open thread in the gcc mailing list discussing the
best way to solve this issue:
https://gcc.gnu.org/ml/gcc/2015-09/msg00300.html

On Thu, Nov 5, 2015 at 12:09 PM, Daniel Gutson <
daniel.gut...@tallertechnologies.com> wrote:

> On Thu, Nov 5, 2015 at 11:52 AM, Sebastian Huber
> <sebastian.hu...@embedded-brains.de> wrote:
> >
> >
> > On 05/11/15 15:50, Daniel Gutson wrote:
> >>
> >> On Thu, Nov 5, 2015 at 11:41 AM, Sebastian Huber
> >> <sebastian.hu...@embedded-brains.de>  wrote:
> >>>
> >>> >Hello,
> >>> >
> >>> >I would like to add the tools for RTEMS 4.12 to the RSB. The question
> is
> >>> >which GCC version should we use? Since our release process is so slow
> I
> >>> > tend
> >>> >to use GCC 6 since it includes support for OpenMP and C++11 threads
> out
> >>> > of
> >>> >the box. I use it currently for development and it works quite good at
> >>> > least
> >>> >on ARM and PowerPC.
> >>
> >> Out of curiosity: OpenMP on ARM? Could you detail the core name?
> >
> >
> > What do you mean with core name? The Altera Cyclone V and Xilinx Zynq use
> > Cortex-A9 MPCore if this is what you mean.
>
> Yes, thanks.
>
> >
> >>
> >> Thanks,
> >>
> >> Daniel.
> >>
> >> ps: we're successfully using gcc 5.2 with custom patches that make
> >> RTEMS compile. We are currently working with the gcc community to
> >> finish one of them. We can provide them here meanwhile.
> >>
> >
> > What is the problem with 5.2? I didn't experience problems with this GCC
> > branch.
>
> We use C++14. The problems I can remember where the recursive call to
> calloc and the global register.
> Guys? Marcos, Andrés? Please comment.
>
> >
> >
> > --
> > Sebastian Huber, embedded brains GmbH
> >
> > Address : Dornierstr. 4, D-82178 Puchheim, Germany
> > Phone   : +49 89 189 47 41-16
> > Fax : +49 89 189 47 41-09
> > E-Mail  : sebastian.hu...@embedded-brains.de
> > PGP : Public key available on request.
> >
> > Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG.
> >
>
>
>
> --
>
> Daniel F. Gutson
> Chief Engineering Officer, SPD
>
> San Lorenzo 47, 3rd Floor, Office 5
> Córdoba, Argentina
>
> Phone:   +54 351 4217888 / +54 351 4218211
> Skype:dgutson
> LinkedIn: http://ar.linkedin.com/in/danielgutson
> ___
> devel mailing list
> devel@rtems.org
> http://lists.rtems.org/mailman/listinfo/devel
>



-- 



Gabriel Alejandro Ibarra

Software Engineer

San Lorenzo 47, 3rd Floor, Office 5

Córdoba, Argentina

Phone: +54 351 4217888
___
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel

Re: Tools for RTEMS 4.12

[Sebastian Huber]

On 05/11/15 15:50, Daniel Gutson wrote:

On Thu, Nov 5, 2015 at 11:41 AM, Sebastian Huber
<sebastian.hu...@embedded-brains.de>  wrote:

>Hello,
>
>I would like to add the tools for RTEMS 4.12 to the RSB. The question is
>which GCC version should we use? Since our release process is so slow I tend
>to use GCC 6 since it includes support for OpenMP and C++11 threads out of
>the box. I use it currently for development and it works quite good at least
>on ARM and PowerPC.

Out of curiosity: OpenMP on ARM? Could you detail the core name?


What do you mean with core name? The Altera Cyclone V and Xilinx Zynq 
use Cortex-A9 MPCore if this is what you mean.




Thanks,

Daniel.

ps: we're successfully using gcc 5.2 with custom patches that make
RTEMS compile. We are currently working with the gcc community to
finish one of them. We can provide them here meanwhile.



What is the problem with 5.2? I didn't experience problems with this GCC 
branch.


--
Sebastian Huber, embedded brains GmbH

Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone   : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail  : sebastian.hu...@embedded-brains.de
PGP : Public key available on request.

Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG.

___
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel

Re: Tools for RTEMS 4.12

[Gedare Bloom]

Hi Sebastian,

I see no problem with using the newer GCC assuming our development
cycle is probably still at least 1yr+. Until we get good automation
this is the case. Also, we are planning to apply to Google Code-In,
and bumping the tool versions could be a set of tasks. If you would
like to do one or a few as a sample, then we can have high school
students do the rest during the GCI program period.

On Thu, Nov 5, 2015 at 9:41 AM, Sebastian Huber
<sebastian.hu...@embedded-brains.de> wrote:
> Hello,
>
> I would like to add the tools for RTEMS 4.12 to the RSB. The question is
> which GCC version should we use? Since our release process is so slow I tend
> to use GCC 6 since it includes support for OpenMP and C++11 threads out of
> the box. I use it currently for development and it works quite good at least
> on ARM and PowerPC.
>
> --
> Sebastian Huber, embedded brains GmbH
>
> Address : Dornierstr. 4, D-82178 Puchheim, Germany
> Phone   : +49 89 189 47 41-16
> Fax : +49 89 189 47 41-09
> E-Mail  : sebastian.hu...@embedded-brains.de
> PGP : Public key available on request.
>
> Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG.
>
> ___
> devel mailing list
> devel@rtems.org
> http://lists.rtems.org/mailman/listinfo/devel
___
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel

Re: Tools for RTEMS 4.12

[Daniel Gutson]

On Thu, Nov 5, 2015 at 11:52 AM, Sebastian Huber
<sebastian.hu...@embedded-brains.de> wrote:
>
>
> On 05/11/15 15:50, Daniel Gutson wrote:
>>
>> On Thu, Nov 5, 2015 at 11:41 AM, Sebastian Huber
>> <sebastian.hu...@embedded-brains.de>  wrote:
>>>
>>> >Hello,
>>> >
>>> >I would like to add the tools for RTEMS 4.12 to the RSB. The question is
>>> >which GCC version should we use? Since our release process is so slow I
>>> > tend
>>> >to use GCC 6 since it includes support for OpenMP and C++11 threads out
>>> > of
>>> >the box. I use it currently for development and it works quite good at
>>> > least
>>> >on ARM and PowerPC.
>>
>> Out of curiosity: OpenMP on ARM? Could you detail the core name?
>
>
> What do you mean with core name? The Altera Cyclone V and Xilinx Zynq use
> Cortex-A9 MPCore if this is what you mean.

Yes, thanks.

>
>>
>> Thanks,
>>
>> Daniel.
>>
>> ps: we're successfully using gcc 5.2 with custom patches that make
>> RTEMS compile. We are currently working with the gcc community to
>> finish one of them. We can provide them here meanwhile.
>>
>
> What is the problem with 5.2? I didn't experience problems with this GCC
> branch.

We use C++14. The problems I can remember where the recursive call to
calloc and the global register.
Guys? Marcos, Andrés? Please comment.

>
>
> --
> Sebastian Huber, embedded brains GmbH
>
> Address : Dornierstr. 4, D-82178 Puchheim, Germany
> Phone   : +49 89 189 47 41-16
> Fax : +49 89 189 47 41-09
> E-Mail  : sebastian.hu...@embedded-brains.de
> PGP : Public key available on request.
>
> Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG.
>



-- 

Daniel F. Gutson
Chief Engineering Officer, SPD

San Lorenzo 47, 3rd Floor, Office 5
Córdoba, Argentina

Phone:   +54 351 4217888 / +54 351 4218211
Skype:dgutson
LinkedIn: http://ar.linkedin.com/in/danielgutson
___
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel

Re: Tools for RTEMS 4.12

[Daniel Gutson]

On Thu, Nov 5, 2015 at 11:41 AM, Sebastian Huber
<sebastian.hu...@embedded-brains.de> wrote:
> Hello,
>
> I would like to add the tools for RTEMS 4.12 to the RSB. The question is
> which GCC version should we use? Since our release process is so slow I tend
> to use GCC 6 since it includes support for OpenMP and C++11 threads out of
> the box. I use it currently for development and it works quite good at least
> on ARM and PowerPC.

Out of curiosity: OpenMP on ARM? Could you detail the core name?

Thanks,

   Daniel.

ps: we're successfully using gcc 5.2 with custom patches that make
RTEMS compile. We are currently working with the gcc community to
finish one of them. We can provide them here meanwhile.

>
> --
> Sebastian Huber, embedded brains GmbH
>
> Address : Dornierstr. 4, D-82178 Puchheim, Germany
> Phone   : +49 89 189 47 41-16
> Fax : +49 89 189 47 41-09
> E-Mail  : sebastian.hu...@embedded-brains.de
> PGP : Public key available on request.
>
> Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG.
>
> ___
> devel mailing list
> devel@rtems.org
> http://lists.rtems.org/mailman/listinfo/devel



-- 

Daniel F. Gutson
Chief Engineering Officer, SPD

San Lorenzo 47, 3rd Floor, Office 5
Córdoba, Argentina

Phone:   +54 351 4217888 / +54 351 4218211
Skype:dgutson
LinkedIn: http://ar.linkedin.com/in/danielgutson
___
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel