Talking with Philipp Sommer last year, he told me that the opal mote (cortex m3u) was working with the tinyos distribution, despite it has not been tested for long periods. Have a look to its github: https://github.com/csiro-wsn/tinyos-csiro-opal , maybe it could help.

Ugo

On 04/23/2014 06:50 PM, András Bíró wrote:
That's intresting, we tried to use the SAM3U code on a prototype board, but almost nothing worked. No uart, one of the spi bus was broken, no sd card. Maybe our SAM3U has different peripherials than yours? I don't have much information on this, a colleage of mine worked on it, I just heard him complaining.

Andris


On Wed, Apr 23, 2014 at 4:26 PM, Thomas Schmid <[email protected] <mailto:[email protected]>> wrote:

    The ARM toolchain doesn't need anything extra for TinyOS. I would
    suggest not to use the Atmel specific one, but the one from
    launchpad (https://launchpad.net/gcc-arm-embedded). I moved away
    from the CodeBench one after they got acquired by Mentor.

    It really is all about writing drivers for those chips. The port
    for the SAM3U and SAM3S are pretty good, though they lack a little
    bit in low-power features (e.g. turning peripherals properly on
    and off).

    Cheers,

    Thomas

    - Thomas



    On Wed, Apr 23, 2014 at 4:26 AM, András Bíró
    <[email protected] <mailto:[email protected]>> wrote:




        On Wed, Apr 23, 2014 at 1:08 AM, Thomas Schmid
        <[email protected] <mailto:[email protected]>> wrote:

            The rf233 driver has been modified and tested, not just
            renamed. But I am not 100% sure if we did all the mods
            mentioned in that app note.

            On Apr 22, 2014 4:48 PM, "Martin Cerveny"
            <[email protected] <mailto:[email protected]>> wrote:

                Hello.

                Thanks for answers.

                On Tue, 22 Apr 2014, András Bíró wrote:

                          Is there some progress or
                    stable/final/tested code for Atmel newer RF chips
                    on ZigBit (ATZB) modules ?
                    We used the ATZB900 and ATZB24 modules without a
                    problem, with a new platform of course. We dropped
                    the atzb24 in
                    favor of the rfa1, and we dropped the atzb900,
                    because it was very hard to get it in larger
                    quantities, it's
                    expensive, hard to solder on, and we want to use
                    the rfa1's better timer stack.


                I am using
                http://www.an-solutions.de/products/900_mhz.html
                (mega1281+rf212) and preparing migration to rf212b.
                And original zigbit/meshbeen (mega1281+rf230).

                          RF:
                          - AT86RF233 - compatible with RF230 ?
                             - Atmel module -
                    http://www.atmel.com/tools/ATZB-RF-233-1-C.aspx
                             Some references:
                             - http://wiesel.ece.utah.edu/projects/10/
                             -
                    git://wiesel.ece.utah.edu/tinyos/tinyos-prod.git
                    <http://wiesel.ece.utah.edu/tinyos/tinyos-prod.git> (branch
                    wiesel)

                    With small modifications:
                    
http://www.atmel.com/ja/jp/Images/Atmel-42198-Migration-from-AT86RF230-to-AT86RF233_AP-Note_AT02601.pdf
                    The problematic parts: PREP_DEEP_SLEEP,
                    TX_AUTO_CRC_ON, probably more, especially with HWACK.


                Is the WIESEL/WREN project driver modified acording to
                this pdf or only "renamed" files & components (Can
                http://wiesel.ece.utah.edu/ answer this question?) ?

                    I don't know much about xmegas, but they seemed
                    very different from atmegas, so it will probably
                    hard work to make
                    the basic drivers. It's probably doesn't worth it:
                    The ARM based MCUs are much more intresting, and
                    it seems the
                    industry moves towards ARM.


                Yes, Xmega is different but it may be possible to
                implement TinyOS on it (with atmel bundles). Yes, ARM
                (Cortex M0+/M3) is probably right way.

                This leads me to new questions:

                What is the timeframe of updating
                https://github.com/tinyos/tinyos-main/tree/master/packaging
                ?

                AVR - to newer toolchain to support new devices
                (TinyOS: AVR binutils 2.22 + gcc 4.6.2 + libc 1.8.0)
                (Atmel: AVR binutils 2.23.2 + gcc 4.8.1 + libc 1.8)
                (
                
http://distribute.atmel.no/tools/opensource/Atmel-AVR-GNU-Toolchain/3.4.3/
                )
                I do some tests on my Solaris/OpenSolaris platform
                with some issues
                ( https://github.com/tinyos/tinyos-main/issues/139 ).

                I saw discussion @dev, any progress/timeline ?


        I'm waiting for Miklos' code review to push the needed
        modifications. The toolchain itself might came out after
        2.1.3, I'm not sure. But after the modifications, you can use
        the binaries from atmel - the official package would be the
        repackaging of that anyway. I'm really impatiant too, at least
        the progmem fixes should be in 2.1.3.


                ARM - add & check support for ARM Cortex (not only Atmel)
                (for example
                
http://distribute.atmel.no/tools/opensource/Atmel-ARM-GNU-Toolchain/4.7.4/
                )

                Does ARM toolchain need special patching for
                TinyOS/nescc ?


        Thomas Schmid might have better answer for this, he's the
        developer for the cortex parts of tinyos. He used CodeBench
        ARM Toolchain. It works, but the ARM drivers are in a very
        early stage, and nobody seems to working on them. We tried to
        make an atmel cortex-m3/cortex-m4 based mote, but developing
        the drivers to a mature state would take a lot of time, so we
        shelved the plan.

        Best,
        Andras Biro
        http://ucmote.com


                Thanks, M.C>






_______________________________________________
Tinyos-help mailing list
[email protected]
https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help


--
Ugo Maria Colesanti
Dipartimento di Ingegneria Informatica, Automatica e Gestionale "Antonio 
Ruberti"
Sapienza Universita' di Roma
Via Ariosto 25, II floor, room A221
00185, Rome
http://wiserver.dis.uniroma1.it/cms/index.php/contacts/13-postdoc/4-ugo-colesanti
Phone:  +39 06 77274056
Fax:    +39 06 77274002

_______________________________________________
Tinyos-help mailing list
[email protected]
https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help

Reply via email to