make sure that TCP_CORK is supported

[ahorek]

Description

according to @dentarg feedback #2345, we shouldn't use TCP_CORK on JRuby until it's properly supported.

It could also break on very old Linux kernels (< 2.6). Let's just make sure that all Socket constants are usable.

Your checklist for this pull request

• I have reviewed the guidelines for contributing to this repository.
• I have added an entry to History.md if this PR fixes a bug or adds a feature. If it doesn't need an entry to HISTORY.md, I have added [changelog skip] or [ci skip] to the pull request title.
• I have added appropriate tests if this PR fixes a bug or adds a feature.
• My pull request is 100 lines added/removed or less so that it can be easily reviewed.
• If this PR doesn't need tests (docs change), I added [ci skip] to the title of the PR.
• If this closes any issues, I have added "Closes #issue" to the PR description or my commit messages.
• I have updated the documentation accordingly.
• All new and existing tests passed, including Rubocop.

[nateberkopec]

Perfect! Thanks for the quick fix.

[headius]

@ahorek @nateberkopec It occurs to me that TCP_CORK is independent of TCP_INFO but they are checked together here. Is the TCP_INFO logic in puma necessary for TCP_CORK in any way?

I ask because JRuby currently still does not bind TCP_INFO mostly because the data it carries is a more complex struct than simply a boolean or integer value. If we bind the constant, then we need to be able to support the struct.

If the TCP_INFO logic were handled separately, TCP_CORK could continue to work on JRuby 9.3 (I have merged jruby/jruby#6366) even if we don't have TCP_INFO fully supported...

[headius]

FWIW @ahorek if you have some cycles I'd love to support TCP_INFO. Along with mapping the structure, there's another set of constants we'd need to have available in JRuby to navigate it.