aya: add feature probing for program and map type

[tyrone-wu]

I tested the new integration tests against a ton of ubuntu mainline kernels. Basically on every version where the program/map type was introduced I was able to verify that they were passing as expected.

---

This change is 

[netlify]

✅ Deploy Preview for aya-rs-docs ready!

Built without sensitive environment variables

Name	Link
🔨 Latest commit	f4df1a9
🔍 Latest deploy log	https://app.netlify.com/sites/aya-rs-docs/deploys/67c5f253f8e7c100086d9896
😎 Deploy Preview	https://deploy-preview-1063--aya-rs-docs.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify site configuration.

[mergify]

Hey @alessandrod, this pull request changes the Aya Public API and requires your review.

[mergify]

@tyrone-wu, this pull request is now in conflict and requires a rebase.

[tamird]

@tyrone-wu can you rebase this please?

[mergify]

@tyrone-wu, this pull request is now in conflict and requires a rebase.

[tamird]

@tyrone-wu I rebased this for you.

[tamird]

Reviewed 1 of 51 files at r1, 38 of 47 files at r2, 7 of 12 files at r4, 9 of 11 files at r7, 3 of 6 files at r8, 1 of 1 files at r9, 2 of 8 files at r10, 1 of 4 files at r12, 6 of 9 files at r13, all commit messages.
Reviewable status: 48 of 51 files reviewed, 5 unresolved discussions (waiting on @alessandrod)

---

aya/Cargo.toml line 25 at r13 (raw file):

log = { workspace = true }
object = { workspace = true, features = ["elf", "read_core", "std", "write"] }
once_cell = { workspace = true, features = ["race"] }

back this out? see comment in info.rs

---

aya/src/programs/info.rs line 11 at r13 (raw file):

use aya_obj::generated::{bpf_prog_info, bpf_prog_type};
use once_cell::race::OnceBool;

why can't we use std::cell::OnceCell here?

---

aya/src/programs/info.rs line 125 at r13 (raw file):

        } else {
            Ok(None)
        }

this transformation doesn't look right to me. What am I missing?

Code quote:

    pub fn map_ids(&self) -> Result<Option<Vec<u32>>, ProgramError> {
        if self.0.nr_map_ids > 0 {
            let mut map_ids = vec![0u32; self.0.nr_map_ids as usize];
            bpf_prog_get_info_by_fd(self.fd()?.as_fd(), &mut map_ids)?;
            return Ok(Some(map_ids));
        }

        static CACHE: OnceBool = OnceBool::new();
        if CACHE.get_or_init(|| matches!(is_prog_info_map_ids_supported(), Ok(true))) {
            Ok(Some(vec![]))
        } else {
            Ok(None)
        }

---

aya/src/programs/info.rs line 152 at r13 (raw file):

    pub fn gpl_compatible(&self) -> Option<bool> {
        if self.0.gpl_compatible() != 0 {
            return Some(true);

the previous code checked both conditions, why is it ok to short circuit now?

---

test/integration-test/src/tests/feature_probe.rs line 199 at r7 (raw file):

}

// Back-to-back calls can be flaky and return `EPERM`.

That's weird. Citation?