I pulled the USART code from Olimex and ported it to the Yagarto Eclipse environment. In the process, I found an issue with the Olimex code in that it setup the baud rate generator clock with the wrong values.

I used the Jim Lynch "blinkie" example framework as a starting point for the Makefile and other Eclipse environment settings such as the OCD "program" target. After I got the USART code working I tested both ports and they were fine.

I have exported the resulting SAM7P256 USART project as an Eclipse/Yagarto export that you can download from the Sawdust download section.

I started doing a similar port of the Olimex MMC/SD code but have run into some problems with it. I'm working that issue on the SparkFun forums - when I can get that working, I'll write it up, but in the meantime, you can get the ported code in the downloads section of sawdust. It compiles but will not work (for me.)

Update 9/13/09:

Since we could not find an enclosure for the Olimex SAM7 board, we designed one, the Chameleon 1:

I recently bought the Olimex SAM7P256 development board from SparkFun and started learning that architecture and tool set. I will probably eventually use the Rowley dev environment, but started by following Jim Lynch's extremely well written tutorial on setting up the Yagarto Eclipse environment doing cross-development in C to ARM targets.



I used the Olimex USB JTag debugger that is compatible with OCD aware debuggers. With this setup, it's possible to have an Open Source cross-development environment which lets you single-step running code on the SAM7P256 board. It worked like a charm.

With that many moving parts, I expected something to not work right, but I went from 0 to JTAG single-step debugging on the board in about 1/2 hr (if you don't include download times of the pieces.)