Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Ada on STM32F7 - ARM7 discovery board with touch screen

432 views
Skip to first unread message

H Keller

unread,
Dec 12, 2015, 6:19:57 AM12/12/15
to
We at the Karlsruhe Institute of Technology are dealing with porting the Ada Runtime of GNAT on an ARM7 system realizing a educational project for students.

Unfortunately, AdaCore offers only a sparse support for an ARM4.

The target is a nice STM32F7 with a STM32F746 MCU, some interfaces like analog i/o, ethernet and a nice colour touch screen with a resolution of 480x272.

We intend to awake interest of students for Ada using this discovery kit and would like to welcome any kind of help.

If someone is interested to participate inthat project, please send an emal to hubert...@kit.edu or oliver.s...@kit.edu.

Thanks
Hubert Keller

Nicholas Collin Paul de Gloucester

unread,
Dec 16, 2015, 11:25:59 AM12/16/15
to
On December 12th, 2015, Hubert B. Keller posted on Usenet:
|----------------------------------------------------------------------|
|"We at the Karlsruhe Institute of Technology are dealing with porting |
|the Ada Runtime of GNAT on an ARM7 system realizing a educational |
|project for students. |
| |
|Unfortunately, AdaCore offers only a sparse support for an ARM4. |
| |
|The target is a nice STM32F7 with a STM32F746 MCU, some interfaces |
|like analog i/o, ethernet and a nice colour touch screen with a |
|resolution of 480x272. |
| |
|We intend to awake interest of students for Ada using this discovery |
|kit and would like to welcome any kind of help. |
| |
|If someone is interested to participate inthat project, please send an|
|emal to [. . .] or [. . .]. |
| |
|Thanks |
|Hubert Keller" |
|----------------------------------------------------------------------|


Dear Dr. Keller:

As you know I already built many ARM crosscompilers for a variety of
ARM versions.

Yours sincerely,
Nicholas Collin Paul de Gloucester

P.S. You (including Oliver Schneider) could suffer from a lot of spam
by typing an email address in a newsgroup post.

Luke A. Guest

unread,
Dec 22, 2015, 3:09:13 AM12/22/15
to
H Keller <h...@ci-tec.de> wrote:
> We at the Karlsruhe Institute of Technology are dealing with porting the
> Ada Runtime of GNAT on an ARM7 system realizing a educational project for students.
>
> Unfortunately, AdaCore offers only a sparse support for an ARM4.

ARM4 or ARM7?

Look at Adacore's github, there is an svd2ada program there that will help.

Luke

Nicholas Collin Paul de Gloucester

unread,
Dec 23, 2015, 10:19:22 AM12/23/15
to
On 22nd December 2015, Luke A. Guest sent:
|--------------------------------------------------------------------------|
|"H Keller <[. . .]> wrote: |
|> We at the Karlsruhe Institute of Technology are dealing with porting the|
|> Ada Runtime of GNAT on an ARM7 system realizing a educational |
|project for students. |
|> |
|> Unfortunately, AdaCore offers only a sparse support for an ARM4. |
| |
|ARM4 or ARM7? |
| |
|[. . .]" |
|--------------------------------------------------------------------------|

Maybe ARM4 was a typo. I had received a few emails from him indicating
that he had not decided what type of ARM, but he had indicated a few
ARM-Cortex-M4 platforms which are ARM7 platforms instead of ARM4
platforms.

With best regards,

H Keller

unread,
Dec 23, 2015, 6:30:44 PM12/23/15
to
Sorry, I mentioned that we will working on an ARM7 and that AdaCore offers only support for an ARM4. What may be misunderstanding here?
Best regards
Hubert Keller

Paul Rubin

unread,
Dec 23, 2015, 7:00:31 PM12/23/15
to
H Keller <h...@ci-tec.de> writes:
> Sorry, I mentioned that we will working on an ARM7 and that AdaCore
> offers only support for an ARM4. What may be misunderstanding here?

I wondered if ARM4 actually meant Cortex M4, which uses the ARMv7
architecture. The actual ARMv4 architecture hasn't been used since the
bronze age. There doesn't seem to have even been a core called "ARM4"
like there was ARM7, ARM9, etc.

Simon Wright

unread,
Dec 24, 2015, 4:56:55 AM12/24/15
to
AdaCore's GNAT GPL 2015 supports STM32F4-DISCO, STM32F429-DISCO,
STM32F7-EVAL boards from STMicroelectronics. The first two are
Cortex-M4, the last is Cortex-M7. It looks to me as though all support
the ARMv7-M architecture (this is unclear ...)

MatthiasR

unread,
Dec 24, 2015, 7:35:38 AM12/24/15
to
H Keller wrote:

> Sorry, I mentioned that we will working on an ARM7 and that AdaCore offers
> only support for an ARM4. What may be misunderstanding here? Best regards
> Hubert Keller

Much confusion because of ARM's really confusing naming schemes...

1. There are ARM architectures:
ARMv1
ARMv2
...
ARMv7
ARMv8

2. There are ARM designs/families:
ARM1
...
ARM7
ARM9
ARM11
ARM Cortex-A (A8, A9, A5, A15, A7, A12, A17)
ARM Cortex-M (M0, M3, M4, M7)
ARM Cortex-R

ARM7 belongs to the ARMv3 architecture and was introduced more than 20 years
ago.
Your STM32F7 is not an ARM7, it's an ARM Cortex-M7!
Cortex-M3, -M4 and -M7 are all belonging to the ARMv7 architecture.

https://de.wikipedia.org/wiki/ARM-Architektur
https://en.wikipedia.org/wiki/ARM_architecture

The Cortex-M7 is binary compatible to the -M4. It may be that the compiler
from Adacore doesn't include the latest and greatest optimizations for
Cortex-M7, but apart from that it should work out of the box. Support for
the STM32F7 EVAL board (not the DISCO board) is included in the runtime
delivered with the compiler. Adaption to the DISCO board should be
manageable.

The already mentioned 'svd2ada' project on Github is still work in progress.
Its aim is the auto-generation of register declarations, based on the
manufacturer-supplied CMSIS-SVD files. SVD files should exist for every
Cortex-M microcontroller (though the quality of these files is varying...)

Regards,

Matthias




lamb...@adacore.com

unread,
Jan 12, 2016, 2:45:31 PM1/12/16
to
Le samedi 12 décembre 2015 12:19:57 UTC+1, H Keller a écrit :
> We at the Karlsruhe Institute of Technology are dealing with porting the Ada Runtime of GNAT on an ARM7 system realizing a educational project for students.
>
> Unfortunately, AdaCore offers only a sparse support for an ARM4.

I suggest you read http://blog.adacore.com/porting-the-ada-runtime-to-a-new-arm-board about how to properly amend the default run-time to adapt it to your need in such case.

The GNAT GPL for ARM supports the ARMv7 architecture, which is (as mentioned already) the architecture used by the Cortex-M4 and the Cortex-M7.

Jerry Petrey

unread,
Jan 15, 2016, 2:21:30 PM1/15/16
to
On 1/12/2016 12:45 PM, lamb...@adacore.com wrote:
>
> I suggest you read http://blog.adacore.com/porting-the-ada-runtime-to-a-new-arm-board about how to properly amend the default run-time to adapt it to your need in such case.
>
> The GNAT GPL for ARM supports the ARMv7 architecture, which is (as mentioned already) the architecture used by the Cortex-M4 and the Cortex-M7.
>

Thanks for that great article Jerome! It is very helpful. I managed to
port the runtime to quite a few boards last year including the F469 and
the F7 based ones but it was quite a job. Your article will make that a
lot easier in the future.

Jerry

fabien....@gmail.com

unread,
Jan 15, 2016, 3:23:09 PM1/15/16
to
On Friday, January 15, 2016 at 8:21:30 PM UTC+1, Jerry wrote:
We are also starting repository for drivers: https://github.com/AdaCore/bareboard

0 new messages