Add missing definitions for QNX8 for several BPF-related structures and constants

[rafaeling]

Description

These structures are required for working with BPF filters and network interfaces on the QNX8 platform. Their absence currently prevents building or using BPF-related functionality with libc on QNX8 targets.

Sources

Updated src/unix/nto/mod.rs accordingly

Checklist

• [?] Relevant tests in libc-test/semver have been updated
  -> No tests for QNX?
• [*] No placeholder or unstable values like *LAST or *MAX are
  included (see #3131)
• [?] Tested locally (cd libc-test && cargo test --target mytarget); especially relevant for platforms that may not be checked in CI.
  -> even main branch does not work.
  Error: could not execute process /home/user/libc/target/x86_64-pc-nto-qnx800/debug/deps/libc-67e46525c5118905 (never executed)

[tgross35]

This looks fine to me, cc target maintainers @flba-eb @gh-tr @jonathanpallant @japaric to verify things build

[tgross35]

Would you actually mind moving this to the src/new module? We're slowly starting to migrate things there since it's easier to match up with what a user can #include. I believe this would be src/new/nto/net/if.h and src/new/nto/net/bpf.h, unless nto has a different structure. You can either move the existing BPF items over or not, your call.

Then this update to src/new/nto/mod.rs:

pub(crate) mod net {
    pub(crate) mod route;
}

And at

libc/src/new/mod.rs

Lines 171 to 186 in 564e0ef

	// Per-OS headers we export
	cfg_if! {
	if #[cfg(target_os = "android")] {
	pub use sys::socket::*;
	} else if #[cfg(target_os = "linux")] {
	pub use linux::can::bcm::*;
	pub use linux::can::j1939::*;
	pub use linux::can::raw::*;
	pub use linux::can::*;
	pub use linux::keyctl::*;
	#[cfg(target_env = "gnu")]
	pub use net::route::*;
	} else if #[cfg(target_vendor = "apple")] {
	pub use signal::*;
	}
	}

, add a nto arm and re-export everything from the two modules.

After that and the __c_anonymous_ifc_ifcu removal I'll merge this given we haven't heard back from the maintainers but this is straightforward.

[tgross35]

Just noticed something: is this union used at all?

[rafaeling]

Actually not, thanks!! I think I used as reference the linux_like definition, will remove it,

[tgross35]

• [?] Tested locally (cd libc-test && cargo test --target mytarget); especially relevant for platforms that may not be checked in CI.
  -> even main branch does not work.
  Error: could not execute process /home/user/libc/target/x86_64-pc-nto-qnx800/debug/deps/libc-67e46525c5118905 (never executed)

Can you usually cargo run --target ... for nto? The relevant test here woul be cargo test -p libc-test --test ctest --target ... if you want to narrow it down and try executing the binary directly.

[rustbot]

This PR was rebased onto a different main commit. Here's a range-diff highlighting what actually changed.

Rebasing is a normal part of keeping PRs up to date, so no action is needed—this note is just to help reviewers.

[rafaeling]

Would you actually mind moving this to the src/new module? We're slowly starting to migrate things there since it's easier to match up with what a user can #include. I believe this would be src/new/nto/net/if.h and src/new/nto/net/bpf.h, unless nto has a different structure. You can either move the existing BPF items over or not, your call.

Then this update to src/new/nto/mod.rs:

pub(crate) mod net {
    pub(crate) mod route;
}

And at

libc/src/new/mod.rs

Lines 171 to 186 in 564e0ef

	// Per-OS headers we export
	cfg_if! {
	if #[cfg(target_os = "android")] {
	pub use sys::socket::*;
	} else if #[cfg(target_os = "linux")] {
	pub use linux::can::bcm::*;
	pub use linux::can::j1939::*;
	pub use linux::can::raw::*;
	pub use linux::can::*;
	pub use linux::keyctl::*;
	#[cfg(target_env = "gnu")]
	pub use net::route::*;
	} else if #[cfg(target_vendor = "apple")] {
	pub use signal::*;
	}
	}

, add a nto arm and re-export everything from the two modules.

After that and the __c_anonymous_ifc_ifcu removal I'll merge this given we haven't heard back from the maintainers but this is straightforward.

@tgross35 Moved structs to src/new, removed __c_anonymous_ifc_ifcu, and ran ./ci/style.sh.
However, the runners are failing, not sure why.

Sorry for the delay in fixing this! I was away for a bit.
I would be willing to support here gradually move things into src/new going forward.