tests: enable regression, upgrade and completion test suites for fedora

[sergiocazzolato]

Enable 3 test suites for fedora.

[Conan-Kudo]

Why are completion files being installed into %{_libexecdir}/snapd?

[sergiocazzolato]

Thanks for reviewing
It is where other scripts are looking for to make snap-completion.
Any suggestion about how to improve it?

[zyga]

@Conan-Kudo Note that those are special files that make snapd implement completion internally. The things you are thinking about about are in the usual location.

[Conan-Kudo]

It's fine, just making sure it's supposed to be there.

[Conan-Kudo]

extention -> extension

[Conan-Kudo]

extention -> extension

[codecov-io]

Codecov Report

Merging #3622 into master will decrease coverage by <.01%.
The diff coverage is n/a.

@@            Coverage Diff            @@
##           master   #3622      +/-   ##
=========================================
- Coverage   75.21%   75.2%   -0.01%     
=========================================
  Files         384     387       +3     
  Lines       33538   33452      -86     
=========================================
- Hits        25226   25159      -67     
+ Misses       6491    6481      -10     
+ Partials     1821    1812       -9

Impacted Files	Coverage Δ
interfaces/builtin/joystick.go	45% <0%> (-25%)	⬇️
interfaces/builtin/framebuffer.go	27.27% <0%> (-21.57%)	⬇️
interfaces/builtin/ubuntu_download_manager.go	31.25% <0%> (-18.75%)	⬇️
interfaces/builtin/time_control.go	57.14% <0%> (-16.55%)	⬇️
interfaces/builtin/physical_memory_observe.go	57.14% <0%> (-16.55%)	⬇️
interfaces/builtin/physical_memory_control.go	57.14% <0%> (-16.55%)	⬇️
interfaces/builtin/io_ports_control.go	61.29% <0%> (-14.32%)	⬇️
interfaces/builtin/hardware_random_control.go	61.53% <0%> (-10.69%)	⬇️
interfaces/builtin/gpio.go	50.98% <0%> (-10.04%)	⬇️
interfaces/builtin/hardware_random_observe.go	61.53% <0%> (-9.9%)	⬇️
... and 57 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 bf28721...ff9c0cb. Read the comment docs.

[zyga]

LGTM with the comment on classic confinement checks. I think this could be done via snapd's API query.

[zyga]

I think it'd be better to ask snapd about it, so that we have to maintain less tables. Also this is likely to change a little (in what it really mans to support it) so this may (hopefully) go away.

[morphis]

We already ask snapd if strict confinement is supported or not. Asking for classic confinement would be just another option.

[morphis]

Why do we bump the number of workers to two? @niemeyer requested this to be a single one for the time being. Anything has changed since then?

[zyga]

Not sure but I think it's because we now have more tests than before.

[zyga]

I think it'd be better to ask snapd about it, so that we have to maintain less tables. Also this is likely to change a little (in what it really mans to support it) so this may (hopefully) go away.

[morphis]

We already ask snapd if strict confinement is supported or not. Asking for classic confinement would be just another option.

[Conan-Kudo]

I'm not a fan of commits that just say "WIP". Can we get some descriptive commit messages for the two "WIP" commits?