aethaniel / experimentalcore-sam Goto Github PK

RingBuffer/FIFO should be a cross-architecture service API.
There are different implementations between SAM and SAMD cores which shall me unified and improved.
The one from SAM core exports its variables members instead of methods in order to manage the process.

I know wiki says it is not tested, but integration seems not working, no board is displaying
any tips ?

Can we have a couple of example AS7 projects, maybe blinky and a serial echoing app or something

digital, analog, etc.. headers are still in port_sam at the moment.
This API being the Arduino core framework API, it shall be in arch_common, thus bringing some hardware abstraction.

USB bootloader for G55

Hi,

First of all great job with the Arduino core for the Atmel boards 👍

I got a multiple definition error due to the include in the core_main.cpp being

#include "core_watchdog.c" 

It should be

#include "core_watchdog.h" 

Would recommend shortening the included headers in the core_main.cpp to

#include "Arduino.h"
#if defined(USBCON)
//#include ""
#endif

Best,
Kevin

Reported by @logos-electromechanical :

I got a weird error that I don't understand things well enough to find the source of. I assume it's something in the way the platform.txt recipes are trying to grab the variant archive, but I can't figure out what precisely is failing.
Here's the error message:

arm-none-eabi-gcc: error:
C:......\Arduino\hardware\arachnio\sam\variants\logos_arachnio_due/logos_arachnio_due:
No such file or directory
Error compiling.

Bonjour 😄
I need to get support for SAM4E8E MCU for arduino ide to support a FW for a 3D printer which use this MCU to have an opensource FW

I have started to read docs/ datasheets and realized that I need to redo the variants parts, and the system\CMSIS\Device\ATMEL files.
I am not really familiar with such low level things, but my printer is waiting, so I have to do it.

I just found your repo and looks going to same direction but for other ID, I was wondering if you could give me some direction/advices ? what files need to change and verify ? is datasheet enough to fill all parts ?
I would be more than happy to push the result to your repo once done

Makefiles lack the build of dependencies for each source file.
Adding this process would reduce the human error and reduce the maintenance in Makefiles.

I try to compile "blink" for atmel_samg55_xplained_pro. To do that I set VARIANT_NAME on my environement :

$ export VARIANT_NAME=atmel_samg55_xplained_pro
$ echo $VARIANT_NAME
$ atmel_samg55_xplained_pro

after that on ExperimentalCore-sam/module/libraries/tests/examples/blink I do

$ make

But make keep the default value for VARIANT_NAME...

If I try

$ make VARIENT_NAME=atmel_samg55_xplained_pro

Make keep again the default value...

But it seems travis can do it, Travis (job 28.6) do :

$ make --no-builtin-rules VARIANT_NAME=atmel_samg55_xplained_pro -C /home/travis/build/aethaniel/ExperimentalCore-sam/module/libraries/tests/examples/blink

And it's working.

I don't think having IPAddress, UDP, Server, Client prefixed by 'Core' is a good idea.
They should be reverted to their original names.

btw, I don't know what these files are doing here as they shall belong to a dedicated network library

Hi when use latest gcc-arm-none-eabi-5_3-2016q1-20160330-win32.exe from https://launchpad.net/gcc-arm-embedded
the compilation failed :

error: conflicting types for 'utoa'   
 extern char* utoa( unsigned long value, char *string, int radix );
...
stdlib.h:201:8: note: previous declaration of 'utoa' was here
 char * _EXFUN(utoa,(unsigned, char *, int));

if I use previous version gcc-arm-none-eabi-4_9-2015q3-20150921-win32.exe
I got different warnings (same as travis-ci output) but compilation is Ok

so need to freeze gcc-arm-embedded version or modify makefile to support latest version of gcc-arm-embedded

@LeGeek, @loiclefort, @cmaglie, @logos-electromechanical, @nicolas-rabault
FYI, I have created this wiki page (local) to follow progress:
https://github.com/aethaniel/ExperimentalCore-sam/wiki

tasks will be updated according to atomic commits

(tested on Arduino DUE)

digitalRead does:

  if ( (Ports[g_aPinMap[ulPin].iPort].pGPIO->PIO_PDSR & g_aPinMap[ulPin].ulPin) != 0 )
  {
    return HIGH ;
  }

However if the pin is programmed as output, PIO_PDSR seems not to get the pin state. I discover the ATMEL peripherals but maybe PIO_ODSR should be read instead when the pin is programmed as output.

@cmaglie @LeGeek @loiclefort
I created a branch in this repo named TemplateCore where I have put all 3 current architectures
This is only a draft to identify unique files per arch/port/variants, so only a beginning and certainly not intended to be published.

It would be great if you could take a look and give feedback.

T.

The existing SAM USB device stack has been proven to be slow, specially versus Teensy 3.x.
This function needs time to be spent.