UnixPB: Add Fedora 40 Support

[steelhead31]

Fixes #3599

Correct Unix playbooks to allow execution on Fedora 40.

Checklist

• commit message has one of the standard prefixes
• playbook changes run through VPC or QPC (if you have access)

Fedora 40: VPC : https://ci.adoptium.net/job/VagrantPlaybookCheck/1974/
Ubuntu 22: VPC : https://ci.adoptium.net/job/VagrantPlaybookCheck/OS=Ubuntu2204,label=vagrant/1975/
Debian 10: VPC : https://ci.adoptium.net/job/VagrantPlaybookCheck/OS=Debian10,label=vagrant/1975/

[sxa]

I'm ok with this, but noting:

• are there any implications on older Fedora versions of completely removing libnss3.so here - I think that was added for some of the security tests.
• Noting that in the nasm playbook we have a mix of syntax: ansible_distribution == "Ubuntu" and ansible_distribution_major_version < "22") vs (ansible_distribution == "Ubuntu" and ansible_distribution_major_version | int >= 24). I definitely have a recollection - possibly ansible version specific - that one is better than the other so maybe we should standardise. I have a feeling it was the | int version that was preferable but don't hold me to that :-)

[steelhead31]

• are there any implications on older Fedora versions of completely removing libnss3.so here - I think that was added for some of the security tests.

For the currently supported versions of Fedora ( 39 & 40 ) the libnss3 libraries are installed via nss-devel & nss-util packages, so removing it has no impact on the currently supported versions.

The vagrant playbook checks run the SSL tests as part of the TEST_JDK flag, so this would highlight any issues for the Fedora 40 run.
 

• Noting that in the nasm playbook we have a mix of syntax: ansible_distribution == "Ubuntu" and ansible_distribution_major_version < "22") vs (ansible_distribution == "Ubuntu" and ansible_distribution_major_version | int >= 24

I've standardised this to be consistent for all the major version checks. I believe the int comparison is a more reliable comparison for numeric values, particularly as using <= with string operations is a bit inconsistent. the string operations are fine for a simple ==.

[sxa]

@steelhead31 I guess this supercedes #3663?

[steelhead31]

@steelhead31 I guess this supercedes #3663?

Yup, I hadnt seen the earlier one.