drivers: eth: native_posix: Enable gPTP support

[jukkar]

Allow gPTP code to be run as a linux process and communicate
with gPTP daemon running in linux host.

Signed-off-by: Jukka Rissanen [email protected]

[codecov-io]

Codecov Report

Merging #8620 into master will decrease coverage by 2.18%.
The diff coverage is 0%.

@@            Coverage Diff             @@
##           master    #8620      +/-   ##
==========================================
- Coverage    54.4%   52.22%   -2.19%     
==========================================
  Files         444      195     -249     
  Lines       42233    24673   -17560     
  Branches     7996     5124    -2872     
==========================================
- Hits        22978    12886   -10092     
+ Misses      16048     9721    -6327     
+ Partials     3207     2066    -1141

Impacted Files	Coverage Δ
drivers/ethernet/eth_native_posix_adapt.c	24.39% <ø> (ø)
drivers/ethernet/eth_native_posix.c	22.1% <0%> (ø)
include/entropy.h	33.33% <0%> (-41.67%)	⬇️
subsys/net/ip/net_stats.h	65.83% <0%> (-34.17%)	⬇️
kernel/stack.c	67.3% <0%> (-25.2%)	⬇️
kernel/init.c	57.35% <0%> (-23.35%)	⬇️
kernel/msg_q.c	77.77% <0%> (-15.47%)	⬇️
include/misc/slist.h	86.66% <0%> (-13.34%)	⬇️
drivers/net/loopback.c	72% <0%> (-12%)	⬇️
kernel/mempool.c	82.05% <0%> (-10.98%)	⬇️
... and 437 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update ad15056...8c26e40. Read the comment docs.

[aescolar]

It would not be too difficult to provide you a version of the host device time which could be offset. If you want tell, and I can add you a couple of functions in the timer driver. Same about the 2 functions below

[jukkar]

Setting the host device time is probably not needed here, at least for now. The reason being that the native_posix target is typically used when testing things and it might not be ok to change the host time in this case. Of course we could make this optional if someone really wishes to have this kind of feature. I am ok if you send a PR that adds this kind of support thou.

[aescolar]

Ok, but note that I did not mean that we would change the real host time, but instead that we would provide you a fake time which would be offset compared to the real host one.

[jukkar]

I misunderstood you then. That kind of fake time would be much welcomed feature so if you have time to add such a feature, I would appreciate that.

[aescolar]

Roger. I added a note in #6044 about it. Hopefully I will have time in the following days to add it.

[aescolar]

Just a couple of comments:

[aescolar]

looking at this, I would actually think it would be better to not access the host time directly, but to have the timer driver provide you the host time, to have just 1 central place for it.

[aescolar]

Just in case it is not obvious: My comments, if you would agree with them, could be addressed after this PR would be merged.

[pfalcon]

@aescolar, I assume that you'd seen the big issues, you gave -1 here. Otherwise, approving, to go on with other gPTP patches.

[jukkar]

Small refactoring because of #8615. This PR also requires #8681 in order to avoid compile warning.