AVR target support

[gsmcmullin]

This is not actively been worked on. This issue is just a placeholder for reference to relevant documentation and/or other references.

[gsmcmullin]

AVR OCD registers are not publicly documented, but have been partially reverse engineered.

See unofficial documentation here:
https://people.ece.cornell.edu/land/courses/ece4760/FinalProjects/s2009/jgs33_rrw32/Final%20Paper/Documentation.html

[aster94]

I am commenting just to express the desire that the BMP one day (hope soon) will support the AVR ecosystem

Also, maybe since the opening of the issue in 2016 there are new possibilities/opportinities to see this achieved

[valerionew]

I also want to express all my intrest towards this feature.
For me, the BMP would become the definitive debugger.

If i'm not mistaken, the old JTAGICE mkII specification is released by Atmel itself (http://www.microchip.com//wwwAppNotes/AppNotes.aspx?appnote=en591238)

And also there's this: http://www.ruemohr.org/docs/debugwire.html

[Testato]

the dW support is very important because it is used on classic 328P (Arduino Uno) so many people may adopt the BMP instead of buy the Atmel-ICE

This is an open implementation:
https://github.com/dcwbrown/dwire-debug

[UweBonnes]

BMP is an open source effort. Why not try to add the support yourself? My main effort is also only programming STM32, so my BMP contributions for STM32 were also only a side effect.

[valerionew]

BMP is an open source effort. Why not try to add the support yourself? My main effort is also only programming STM32, so my BMP contributions for STM32 were also only a side effect.

I'd try to, but that's WAY beyond my skills. I'm an hardware designer, i do little to no software...

[UweBonnes]

https://github.com/UweBonnes/blackmagic/tree/avr as a start can program AT90CANxxx. JTAG ID for other devices and more things need to be added.

[dragonmux]

This issue should be closed (or at least partially addressed) by #992 when we complete that work and get it to land. A complete documentation of the JTAG-PDI system has been undertaken. The limiting factor is parts and a little guesswork as to how exactly certain parts of the debug machinery work which has limited the speed of progress