AMD SNP attestation test enhancement

[bssrikanth]

This patch introduces improvements to the SEV-SNP attestation workflow.
The changes include a new script for installing the snpguest tool from source,
updates to the attestation workflow to align with upstream changes, and
enhancements to the testcase and configuration files.
Specifically, the series:

1. Adds snpguest_install.sh to install the snpguest tool from source.
2. Updates the existing SEV-SNP attestation workflow to align with upstream snpguest tool changes.
3. Enhances the SEV-SNP testcase by supporting snpguest source installation, improving CPU model detection, updating SNP policy values, adding a debug policy variant, improving error handling.

Signed-off-by: Srikanth Aithal [email protected]

[bssrikanth]

@zixi-chen @JinLiul request your review comments.

[bssrikanth]

Hello I will be happy to address any review comments with respect to this PR. Thank you in advance :)

[zixi-chen]

Hello @bssrikanth, I just came back from holidays this week.
Since you want to install snpguest with the upstream repo, this conflicts with our tests with the snpguest package in a RHEL compose. Can you add a new cfg e.g., snp_attestation.cfg, instead of making changes in the original cfg? In this way, different test parameters won't cause any conflict. Please also don't change the main test script name, you could add type = snp_basic_config in your own cfg, and I am happy with your proposed change with snp_basic_config.py.

[bssrikanth]

@zixi-chen Thank you for the review comments.

1. I have renamed snp_attestation.py back to snp_basic_config.py and created a separate configuration file as you suggested.
2. The snpguest_sourcebuild parameter will be evaluated before deciding whether to build snpguest from source. It will only build from source when explicitly set; otherwise, it will follow the default path to avoid disrupting your existing workflow. Please let me know if this is acceptable.

[JinLiul]

The current dict_cpu implementation is incompatible with Turin systems and causes failures. The updated version in this patch resolves this issue and the test case pass in Turin.

[bssrikanth]

@zixi-chen please let me know any comments on the revised version. Happy to address them :)