Added all definitions from linux/ptp-clock.h

[davidv1992]

This still has some trouble passing CI due to two unnamed anonymous unions in ptp_perout_request. Any guidance on how to deal with those in a better way?

[rustbot]

Thanks for the pull request, and welcome! The Rust team is excited to review your changes, and you should hear from @JohnTitor (or someone else) soon.

Please see the contribution instructions for more information. Namely, in order to ensure the minimum review times lag, PR authors and assigned reviewers should ensure that the review label (S-waiting-on-review and S-waiting-on-author) stays updated, invoking these commands when appropriate:

• @rustbot author: the review is finished, PR author should check the comments and take action accordingly
• @rustbot review: the author is ready for a review, this PR will be queued again in the reviewer's queue

[folkertdev]

based on

    #[cfg(libc_union)]
    pub union __c_anonymous_ifr_ifru {
        pub ifru_addr: ::sockaddr,
        pub ifru_dstaddr: ::sockaddr,

these need that #[cfg(libc_union)] pragma. Then at the use site they need a "default" (highest size and alignment of the variants, I guess)

    pub struct ifreq {
        /// interface name, e.g. "en0"
        pub ifr_name: [::c_char; ::IFNAMSIZ],
        #[cfg(libc_union)]
        pub ifr_ifru: __c_anonymous_ifr_ifru,
        #[cfg(not(libc_union))]
        pub ifr_ifru: ::sockaddr,
    }

[bors]

☔ The latest upstream changes (presumably #3376) made this pull request unmergeable. Please resolve the merge conflicts.

[JohnTitor]

Could you resolve the merge conflict and CI failure? Thanks!

[davidv1992]

@JohnTitor I would love to, but I don't know how to fix the CI. Essentially, it is failing because the ptp_perout_request struct contains unnamed anonymous enums (so both the enum and the field that is of the type of the enum don't have a name), and I don't know how to tell the CI that properly.

[JohnTitor]

You can check around:

libc/libc-test/build.rs

Lines 3480 to 3520 in d0ce87b

	cfg.skip_type(move |ty| {
	match ty {
	// FIXME: `sighandler_t` type is incorrect, see:
	// https://github.com/rust-lang/libc/issues/1359
	"sighandler_t" => true,
	// These cannot be tested when "resolv.h" is included and are tested
	// in the `linux_elf.rs` file.
	"Elf64_Phdr" | "Elf32_Phdr" => true,
	// This type is private on Linux. It is implemented as a C `enum`
	// (`c_uint`) and this clashes with the type of the `rlimit` APIs
	// which expect a `c_int` even though both are ABI compatible.
	"__rlimit_resource_t" => true,
	// on Linux, this is a volatile int
	"pthread_spinlock_t" => true,
	// For internal use only, to define architecture specific ioctl constants with a libc
	// specific type.
	"Ioctl" => true,
	// FIXME: requires >= 5.4.1 kernel headers
	"pgn_t" if musl => true,
	"priority_t" if musl => true,
	"name_t" if musl => true,
	t => {
	if musl {
	// LFS64 types have been removed in musl 1.2.4+
	t.ends_with("64") || t.ends_with("64_t")
	} else {
	false
	}
	}
	}
	});
	cfg.skip_struct(move |ty| {
	if ty.starts_with("__c_anonymous_") {
	return true;
	}

[Brijeshkrishna]

libc/libc-test/build.rs

Lines 4180 to 4184 in 3195caf

	(struct_ == "ifconf" && field == "ifc_ifcu") ||
	// glibc uses a single array `uregs` instead of individual fields.
	(struct_ == "user_regs" && arm)
	});

add this

(struct_ == "ptp_perout_request" && field == "anonymous_1")

[davidv1992]

Replaced by #3559