boards/atmega256rfr2-xpro: initial support

[aabadie]

Contribution description

This PR is adding an initial support for the ATmega256RFR2 Xplained pro board.

Interesting things with this board:

• it allows to test the atmega256rfr2 cpu that has no board support for the moment
• it provides an integrated edbg programmer, compatible with avrdude: the board can be easily flashed and even reset (debug not tested yet).
• it provides a radio (untested yet)

I ran the compile_and_test_for_board.py script and it reports several problems. I'll have a look at that later.

The PR is marked as a draft until the board is fully tested/configured.

Testing procedure

• A green Murdock
• Allowed automatic tests are passing

Issues/PRs references

Based on #12608 to avoid having to update BLACKLIST_BOARD variables that will be removed anyway once merged.

[aabadie]

I went a little further here:

• debugging the board with gdb is now configured. This is possible thanks to the edbg interface and is supported by avarice
• the radio seems to work out-of-the-box
• the failing automatic tests are only tests using the shell. This is due to a long delay between the reset moment and the moment when the shell is ready. I think this will be fixed by gnrc_ipv6: only discard invalid source when assigned to interface [backport 2019.10] #12458

This PR is no longer a draft.

[benpicco]

Looks pretty straightforward, good to hear debugging works!

[aabadie]

Thanks for reviewing @benpicco. If you don't mind, before merging, I'd like to provide more test results with this board.

[aabadie]

I tested a bit the radio with examples/gnrc_networking and it is working well: I was able to ping a RPI with 802.15.4 enabled :)

[aabadie]

Failing automatic tests:

Failures during test:
- [tests/bench_runtime_coreapis](tests/bench_runtime_coreapis/test.failed)
- [tests/bitarithm_timings](tests/bitarithm_timings/test.failed)
- [tests/event_wait_timeout](tests/event_wait_timeout/test.failed)
- [tests/evtimer_msg](tests/evtimer_msg/test.failed)
- [tests/gnrc_ipv6_ext](tests/gnrc_ipv6_ext/test.failed)
- [tests/gnrc_ipv6_ext_frag](tests/gnrc_ipv6_ext_frag/test.failed)
- [tests/gnrc_ipv6_fwd_w_sub](tests/gnrc_ipv6_fwd_w_sub/test.failed)
- [tests/gnrc_ndp](tests/gnrc_ndp/test.failed)
- [tests/gnrc_netif](tests/gnrc_netif/test.failed)
- [tests/gnrc_rpl_srh](tests/gnrc_rpl_srh/test.failed)
- [tests/gnrc_sock_dns](tests/gnrc_sock_dns/test.failed)
- [tests/gnrc_tcp](tests/gnrc_tcp/test.failed)
- [tests/isr_yield_higher](tests/isr_yield_higher/test.failed)
- [tests/libfixmath](tests/libfixmath/test.failed)
- [tests/periph_eeprom](tests/periph_eeprom/test.failed)
- [tests/periph_gpio](tests/periph_gpio/test.failed)
- [tests/pipe](tests/pipe/test.failed)
- [tests/pkg_c25519](tests/pkg_c25519/test.failed)
- [tests/pkg_jsmn](tests/pkg_jsmn/test.failed)
- [tests/pkg_lora-serialization](tests/pkg_lora-serialization/test.failed)
- [tests/pkg_micro-ecc](tests/pkg_micro-ecc/test.failed)
- [tests/ps_schedstatistics](tests/ps_schedstatistics/test.failed)
- [tests/pthread_barrier](tests/pthread_barrier/test.failed)
- [tests/stdin](tests/stdin/test.failed)
- [tests/trickle](tests/trickle/test.failed)
- [tests/xtimer_usleep_short](tests/xtimer_usleep_short/test.failed)

I'm looking into them.

[aabadie]

Too bad, you merged this on a little bit too early: the failing tests revealed a misconfiguration of the clock on that board. Indeed, it provides an external 16MHz oscillator that can be enabled by properly enabling the low fuse byte. I was just testing with this and it's fixing a lot of timeout issues in the tests.

Will provide this in a follow-up PR.

[aabadie]

BTW what is @riot-ci exactly ? who is behind this ?

[maribu]

Oh! I had this page open on my phone and never intended to merge it. I guess it unlocked in my pocket. Sorry!

[maribu]

Btw: The debugger included in this board cannot by chance be connected to other external ATmega boards?

[miri64]

BTW what is @riot-ci exactly ? who is behind this ?

It's the account Murdock uses to access GitHub, so our CI team (@kaspar030, @smlng, ...)

[miri64]

I think I have access as well, somehow.

[miri64]

Why are you asking?

[aabadie]

Why are you asking?

This message appeared in the PR yesterday:
riot-ci pushed a commit to riot-ci/RIOT that referenced this pull request 13 hours ago

Not sure what this is about.

[miri64]

Given these two PRs RIOT-OS/RobotFW-tests#20, RIOT-OS/RobotFW-tests#21, I guess @smlng was logged in as the wrong user again ;-).

[miri64]

I'm wondering however if and why @riot-ci needs push rights to master.

[smlng]

wasn't me, at least I wasn't using the riot-ci user lately (and also shouldn't anyway, other than by mistake). However, that commit or push was not done by me.

[aabadie]

The debugger included in this board cannot by chance be connected to other external ATmega boards?

Something can be soldered on the back of the board but that doesn't seem very easy. Of course there's nothing in their user guide about that.

[maribu]

OK, I see. Thanks for the info!