At 19:05 26-07-05 +0100, you wrote:
>We have just developed a board based on the msp430 f1491pm chip and gout it
>running today. 
>We made a late decision to use the new chip after a TI presentation as it
has an
>on-chip DAC.
>
>To my horror, I discovered that gdbproxy doesn't recognise the chip. I had
assumed
>that it would recognise it as one of the 149x family but that is not the
case.
>
>Is there current/imminent support for this device? A patch that makes it a
generic
>149x would be ok. If not then we must look for another toolchain which
would be
>really sad as we want to stick with gcc is possible.

I don't have a direct solution to this problem. But what keeps me wondering
is why so many people use in-circuit emulators / debuggers for debugging
embedded devices.

The whole 'problem' comes down to verifying software and being able to
locate bugs. 

I have been using the following approach for nearly a decade on different
platforms: I use the serial port to run a command line interface (about 80
lines of code) and implement putchar to make printf work.
The command line interface allows me to execute seperate parts of the
software, dump & alter parts of the memory, input stimuli, etc, etc -the
sky is the limit-. For example, I've made firmware which allows to switch
on debugging in certain parts of the software to keep track on what the
software is doing. This allows customers to provide detailed information
from the field for further analyses.

If things get really complicated, I write and test an algorithm on a PC
first using GCC which offers easier ways to verify software.

Nico Coesel


Reply via email to