Add tests based on nvidia-container-toolkit (New)

[antlassagne]

Description

Implement new nvidia-specific tests.
This first test of this category runs nvidia-smi within a container, to verify that the nvidia-container-toolkit is correctly setup and that the GPU passthrough works as expected.

A passed test: https://certification.canonical.com/hardware/202502-36337/submission/424289/

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 50.50%. Comparing base (5c4aab0) to head (e953fe4).
⚠️ Report is 110 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #1874   +/-   ##
=======================================
  Coverage   50.50%   50.50%           
=======================================
  Files         383      383           
  Lines       41097    41097           
  Branches     6905     6905           
=======================================
  Hits        20757    20757           
  Misses      19594    19594           
  Partials      746      746           

Flag	Coverage Δ
provider-certification-client	57.14% <ø> (ø)
provider-gpgpu	92.61% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[pedro-avalos]

Looks like a good test, but I'd like to learn more about the dependencies and possible setup that this test assumes.

[antlassagne]

From the conversation we had with people involved, we need to discuss a bit more in which provider this (and other nvidia-hardware-specific tests) will go. QA guys want it in the Desktop Preload suite, but that will need some manifest / requires / dependencies discussions.

[antlassagne]

@pedro-avalos From the conversation we had earlier, I assume this PR is ready to be merged. The dependencies are clearly stated (until provisioned directly from here), and GPGPU provider was decided to be the provider we'll use.

I added a small test that verifies if the resource job says the same thing than the manifest. The method is a bit weird due to a bug in checkbox (I discussed this with @Hook25). Do you need anything else from me?

[pedro-avalos]

Looks good, let's just make sure we spell out NVIDIA with all caps

[pedro-avalos]

Thanks for all the changes. I'm excited to have more GPGPU tests coming to checkbox

[antlassagne]

I've made one more change about the detect job, following @Hook25 's guidance.