Looking for Advice on Selecting Trace Hardware and Software for RTEMS and Cortex-M

[Isaac Gutekunst]

Hi Sebastian,
On 10/23/2015 10:17 AM, Sebastian Huber wrote:
> Hello Isaac,
>
> I didn't use the Lauterbach trace on this platform, but it is quite useful in general.
> Lauterbach offers a low cost trace solution for Cortex-M:
>
> http://www.lauterbach.com/frames.html?microtrace.html
>
That's exactly the one that we are considering. Have you used their tools on other platforms? 
 From the documentation, they all seem quite good.

Isaac
> They provide RTEMS awareness. Due to the instability of internal kernel APIs, this usually
> requires some support from Lauterbach for your current RTEMS version. This is normally done in
> a couple of work days.
>
> On 23/10/15 15:57, Isaac Gutekunst wrote:
>> Hi All,
>>
>> We are considering purchasing trace hardware from Lauterbach as it is the only solution we
>> could find that has an independent trace and and debug tool i.e. isn't tied into a particular
>> compiler tool chain.
>>
>> I have a suspicion that (in theory), any debugger supporting the ARMv7m instruction set and
>> the ARM EABI (especially calling conventions), should be able to debug and trace any Cortex-M
>> processor, regardless of the compiler tool chain used to build the file. Is their any truth
>> to this suspicion?
>>
>>
>> Does anyone have experience with this or other debug and trace tools for Cortex-M that work
>> with RTEMS (or at least GCC and ELF files) and a Linux host?
>>
>> To clarify what I mean by trace, I am interested support for the ARM CoreSight debug
>> architecture including ETM, ITM, ETB, etc that can do instruction level trace and display
>> data from the ITM, potentially doing interesting things like plotting data.
>>
>> Regards,
>>
>