added more tests for socketpair syscall and fixed an issue with prior socketpair test

[qianxichen233]

Description

Fixes # (issue)

This PR added more tests for the socketpair syscall, and fixed an issue for prior socketpair test that could stuck the entire test in some cases, especially, when a test inside a non-main thread failed, the thread will panic and the main thread will never be able to return from recv_syscall as it is waiting message from an already dead thread. Fix involves creating dedicated thread for each peer and using non-blocking thread join method. Besides the fix of the old test, below is the new tests cased being added:

1. socketpair_syscall called with SOCK_NONBLOCK flag set
2. socketpair_syscall called with SOCK_CLOEXEC flag set
3. socketpair_syscall called with bad inputs to test for error code

Type of change

• New feature (non-breaking change which adds functionality)

How Has This Been Tested?

This PR is a test itself, and currently all the tests were passed.

Checklist:

• My code follows the style guidelines of this project
• I have commented my code, particularly in hard-to-understand areas
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• Any dependent changes have been added to a pull request and/or merged in other modules (native-client, lind-glibc, lind-project)

[Yaxuan-w]

Great job! Tests are comprehensive. LGTM with a minor question

[JustinCappos]

Looks good overall! A few minor comments / request for changes...

[yashaswi2000]

Good work on the tests! , Needs more documentation on what is intended and how we are testing. Missing comments on the original Socketpair syscall in net_calls.rs, and potential error checking improvements there.

[qianxichen233]

@davidge20 Could you review my socketpair PR?

[davidge20]

I learned a lot from reading through your tests! Great work overall!

Just a note, I would appreciate more comments especially at the top of blocks of code that perform something. This would help me out a lot.

[yashaswi2000]

LGTM

[yashaswi2000]

add a panic section after this, mentioning when can this function panic

[qianxichen233]

I did not find any place that might cause panic currently. But I just wonder is the panic something that should occur inside the syscalls? I just feel like whenever something goes wrong, we should always return a syscall error instead of letting it panic directly. A panic inside the syscall would look like a bug to me.

[yashaswi2000]

hmm that's something to think about. ideally I agree with your line of thought, but the purpose of panic is kind of to say something went wrong when it shouldn't. And it might not be related to the logic of the Syscall. @rennergade your thoughts on this?

[rennergade]

Some minor changes requested from Yash and I. Once those are addressed I think this is good to go!

[rennergade]

Great! Approved.