On Sun, Jun 21, 2009 at 4:53 AM, David Brownell<davi...@pacbell.net> wrote: > As I said, I won't object to such fixes. But at the same time, it's > worth realizing that it's been quite a few years since much general > purpose code has worked on 16-bit CPUs. Even Intel is working on > moving away from BIOS boot code that must running on 80286 class CPUs. > So I would not put such things on *my* priority list. > > I believe that OpenOCD could be useful for *TARGET* chips that don't > have 32-bit words ... microcontrollers like AVR8, MSP430, and maybe > even some subset of the PIC chaos.
The PIC chaos typically mean PIC12/16 and maybe PIC18. There is no hope to run OpenOCD on those chips as they use ICSP debugging protocol, not JTAG. ICSP debugging protocol is Microchip proprietary 16bit PIC24/dsPIC30/dsPIC33 has a nice architecture. Some of the chips (PIC24/dsPIC33) may have working JTAG but so far the tools all use ICSP debugging. Many of them have errata which render JTAG debugging not usable. 32bit PIC32 is based on MIPS M4K core and does have a working JTAG debugging subsystem. It has another ICSP debugging subsystem. So OpenOCD can be made to work using the JTAG subsystem. -- Xiaofan http://mcuee.blogspot.com _______________________________________________ Openocd-development mailing list Openocd-development@lists.berlios.de https://lists.berlios.de/mailman/listinfo/openocd-development