BSP lm3s6965

0 views
Skip to first unread message

e...@e-bbes.com

unread,
Feb 20, 2012, 5:55:36 AM2/20/12
to rtems...@rtems.org
Hi all,
anybody tried or is using it on real hardware?
Didn't work out of the box, so I'm trying to figure out where to start.
Sources from git as of yesterday.
_______________________________________________
rtems-users mailing list
rtems...@rtems.org
http://www.rtems.org/mailman/listinfo/rtems-users

Sebastian Huber

unread,
Feb 20, 2012, 6:04:17 AM2/20/12
to rtems...@rtems.org
On 02/20/2012 11:55 AM, e...@e-bbes.com wrote:
> Hi all,
> anybody tried or is using it on real hardware?

No, I used this BSP only on the Qemu simulator.

> Didn't work out of the box, so I'm trying to figure out where to start.
> Sources from git as of yesterday.

You have to add the complete low level initialization.

--
Sebastian Huber, embedded brains GmbH

Address : Obere Lagerstr. 30, D-82178 Puchheim, Germany
Phone : +49 89 18 90 80 79-6
Fax : +49 89 18 90 80 79-9
E-Mail : sebasti...@embedded-brains.de
PGP : Public key available on request.

Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG.

Dmitry Shapovalov

unread,
Feb 20, 2012, 6:08:03 AM2/20/12
to rtems...@rtems.org
i ported it to lm3s8962. you need at least change linkcmds to real memory sizes..

2012/2/20 <e...@e-bbes.com>



--
--
Dmitry Shapoval.

e...@e-bbes.com

unread,
Feb 20, 2012, 6:20:53 AM2/20/12
to rtems...@rtems.org
Zitat von Sebastian Huber <sebasti...@embedded-brains.de>:

> On 02/20/2012 11:55 AM, e...@e-bbes.com wrote:
>> Hi all,
>> anybody tried or is using it on real hardware?
>
> No, I used this BSP only on the Qemu simulator.
>
>> Didn't work out of the box, so I'm trying to figure out where to start.
>> Sources from git as of yesterday.
>
> You have to add the complete low level initialization.

I'm using the original EK-LM3S6965 board from TI, and browsing through
the sources, it looked complete. Could you give me some pointers what
I missed?
(or is missing?)

Thanks in advance

Sebastian Huber

unread,
Feb 20, 2012, 6:22:42 AM2/20/12
to rtems...@rtems.org
On 02/20/2012 12:08 PM, Dmitry Shapovalov wrote:
> i ported it to lm3s8962. you need at least change linkcmds to real memory sizes..

You can send patches for this to <rtems...@rtems.org>. Preferably in the
"git format-patch" format.

--
Sebastian Huber, embedded brains GmbH

Address : Obere Lagerstr. 30, D-82178 Puchheim, Germany
Phone : +49 89 18 90 80 79-6
Fax : +49 89 18 90 80 79-9
E-Mail : sebasti...@embedded-brains.de
PGP : Public key available on request.

Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG.

Sebastian Huber

unread,
Feb 20, 2012, 6:24:29 AM2/20/12
to rtems...@rtems.org
On 02/20/2012 12:20 PM, e...@e-bbes.com wrote:
> Zitat von Sebastian Huber <sebasti...@embedded-brains.de>:
>
>> On 02/20/2012 11:55 AM, e...@e-bbes.com wrote:
>>> Hi all,
>>> anybody tried or is using it on real hardware?
>>
>> No, I used this BSP only on the Qemu simulator.
>>
>>> Didn't work out of the box, so I'm trying to figure out where to start.
>>> Sources from git as of yesterday.
>>
>> You have to add the complete low level initialization.
>
> I'm using the original EK-LM3S6965 board from TI, and browsing through the
> sources, it looked complete. Could you give me some pointers what I missed?
> (or is missing?)
>
> Thanks in advance

Sorry, I never worked with the real hardware. I hacked it together to get it
running on Qemu and did nothing more. You should review the complete BSP sources.

--
Sebastian Huber, embedded brains GmbH

Address : Obere Lagerstr. 30, D-82178 Puchheim, Germany
Phone : +49 89 18 90 80 79-6
Fax : +49 89 18 90 80 79-9
E-Mail : sebasti...@embedded-brains.de
PGP : Public key available on request.

Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG.

Dmitry Shapovalov

unread,
Feb 20, 2012, 6:42:44 AM2/20/12
to rtems...@rtems.org
usart initialisation is missing, memory size is wrong (real board has less memory), and may be system clock initialisation is wrong.
i can post my sources later.
P.S. i switched from RTEMS to freertos, so port is incomplete...

2012/2/20 <e...@e-bbes.com>
Zitat von Sebastian Huber <sebastian.huber@embedded-brains.de>:



--
--
Dmitry Shapoval

e...@e-bbes.com

unread,
Feb 20, 2012, 6:57:46 AM2/20/12
to rtems...@rtems.org
Zitat von Dmitry Shapovalov <dmi...@0fe.ru>:

> usart initialisation is missing, memory size is wrong (real board has less
> memory), and may be system clock initialisation is wrong.
> i can post my sources later.

Please do ...
Thanks in advance ;-)

> P.S. i switched from RTEMS to freertos, so port is incomplete...

Short version of "why" ?

Dmitry Shapovalov

unread,
Feb 20, 2012, 7:08:45 AM2/20/12
to rtems...@rtems.org
we don't plan to add more memory to our device, and 192k of RAM(stm32f4) looks like too small for RTEMS(with ip stack).

2012/2/20 <e...@e-bbes.com>



--
--
Dmitry Shapoval

Sebastian Huber

unread,
Feb 20, 2012, 7:13:47 AM2/20/12
to rtems...@rtems.org
On 02/20/2012 01:08 PM, Dmitry Shapovalov wrote:
> we don't plan to add more memory to our device, and 192k of RAM(stm32f4) looks
> like too small for RTEMS(with ip stack).

Which IP stack do you use on FreeRTOS? You can also use the lwIP stack on
RTEMS. For the basic RTEMS system 192k are more than enough (at around 32k
things get tricky).

--
Sebastian Huber, embedded brains GmbH

Address : Obere Lagerstr. 30, D-82178 Puchheim, Germany
Phone : +49 89 18 90 80 79-6
Fax : +49 89 18 90 80 79-9
E-Mail : sebasti...@embedded-brains.de
PGP : Public key available on request.

Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG.

e...@e-bbes.com

unread,
Feb 20, 2012, 7:18:03 AM2/20/12
to rtems...@rtems.org
Zitat von Dmitry Shapovalov <dmi...@0fe.ru>:

> we don't plan to add more memory to our device, and 192k of RAM(stm32f4)


> looks like too small for RTEMS(with ip stack).

Didn't you mix up something? The lm3s6965 has 256KByte Flash and 64
KBytes RAM ...
The biggest Luminary is (IIRC) the lm3sb9b96 with 512KByte Flash and
96KByte RAM.

e...@e-bbes.com

unread,
Feb 20, 2012, 7:51:16 AM2/20/12
to rtems...@rtems.org
Zitat von e...@e-bbes.com:

> Zitat von Dmitry Shapovalov <dmi...@0fe.ru>:
>
>> we don't plan to add more memory to our device, and 192k of RAM(stm32f4)
>> looks like too small for RTEMS(with ip stack).
>
> Didn't you mix up something? The lm3s6965 has 256KByte Flash and 64
> KBytes RAM ...
> The biggest Luminary is (IIRC) the lm3sb9b96 with 512KByte Flash and
> 96KByte RAM.

Sorry, didn't notice that you started talking about STMicros ;-)

Dmitry Shapovalov

unread,
Feb 20, 2012, 10:33:50 PM2/20/12
to rtems...@rtems.org
we are using lwip on freertos. we are not sure about the required features of our device, so we chose a system with less overhead. for our other project, we continue to use the RTEMS and quite happy.

2012/2/20 Sebastian Huber <sebasti...@embedded-brains.de>
On 02/20/2012 01:08 PM, Dmitry Shapovalov wrote:
we don't plan to add more memory to our device, and 192k of RAM(stm32f4) looks
like too small for RTEMS(with ip stack).

Which IP stack do you use on FreeRTOS?  You can also use the lwIP stack on RTEMS.  For the basic RTEMS system 192k are more than enough (at around 32k things get tricky).


--
Sebastian Huber, embedded brains GmbH

Address : Obere Lagerstr. 30, D-82178 Puchheim, Germany
Phone   : +49 89 18 90 80 79-6
Fax     : +49 89 18 90 80 79-9

PGP     : Public key available on request.

Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG.
_______________________________________________
rtems-users mailing list
rtems...@rtems.org
http://www.rtems.org/mailman/listinfo/rtems-users



--
--
Dmitry Shapoval

Dmitry Shapovalov

unread,
Feb 20, 2012, 10:40:46 PM2/20/12
to rtems...@rtems.org
source code as I promised.
it is a port for lm3s8962, but I do not think they are very different
.

2012/2/20 <e...@e-bbes.com>



--
--
Dmitry Shapoval

Dmitry Shapovalov

unread,
Feb 20, 2012, 10:46:49 PM2/20/12
to rtems...@rtems.org
forgot the link :) http://dl.dropbox.com/u/1285236/rtems_git_dev.tar.bz2

2012/2/21 Dmitry Shapovalov <dmi...@0fe.ru>



--
--
Dmitry Shapoval
Reply all
Reply to author
Forward
0 new messages