add various ptp_* structs

[folkertdev]

Description

Just the structs (i.e. skipping some constants) from #3865. Hopefully this will be easier to get through CI, and then we can look at the constants.

Sources

Checklist

• Relevant tests in libc-test/semver have been updated
• 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 (sadly that just does not compile)

r? @tgross35

[folkertdev]

the __kernel_fsid_t symbol is not in the semver file, so I didn't add __kernel_clockid_t either. But maybe they should both be added?

[tgross35]

Would you be able to add both? Even though they're dunder types, it's nice to at least make it clear when they disappear.

[folkertdev]

sure, I added them

[tgross35]

  --- stderr
  error: expected `*` or `+`
     --> ../src/macros.rs:248:87
      |
  248 |                 pub $({$constness:ident})* fn $i:ident($($arg:ident: $argty:ty),* $(,)?) -> $ret:ty
      |  

I have no clue why this is showing up in CI here but I am so sick of ctest. Feel free to add a commit that changes the $(,)? to $(,)* with aFIXME(ctest) if it's needed to unblock this.

[folkertdev]

I just removed the ptp_clock_caps struct for now, and that passes CI. So, I think this is good now?

[tgross35]

Hey look, passing CI!

LGTM, could you squash?

[folkertdev]

wow, is this stuff cursed or something?

0.757 Reading package lists...
0.766 E: The repository 'http://security.ubuntu.com/ubuntu mantic-security Release' does not have a Release file.
0.766 E: The repository 'http://archive.ubuntu.com/ubuntu mantic Release' does not have a Release file.
0.766 E: The repository 'http://archive.ubuntu.com/ubuntu mantic-updates Release' does not have a Release file.
0.766 E: The repository 'http://archive.ubuntu.com/ubuntu mantic-backports Release' does not have a Release file.

maybe a rerun just fixes that? anyway that's weird

EDIT: nope, still fails. wth?

[tgross35]

wow, is this stuff cursed or something?

I'm coming to the same conclusion. CI problems have been more prevalent in the past couple of weeks but usually it is s390x that fails.

Don't worry about it, I can retry individual jobs until they pass.

[tgross35]

Not just you, something docker-related must be wonky #4116. I'll retry later today.

[tgross35]

I guess ubuntu cuts off repos after they go EOL, fixed in #4120.

[tgross35]

I accidentally clicked "update with merge" rather than "update with rebase" and can't find any way to undo that. Could you rebase?

It might give me the option again once another PR updates main, will see...

[folkertdev]

sure, rebased and force pushed

[tgross35]

Thanks! Fingers crossed...

[folkertdev]

just out of curiosity: do we know why the spark CI action is that slow? (26min, the others complete <= 3 min). I suspect it's because the test-runner-linux script is used instead of a qemu-sparc invocation (that e.g. powerpc uses). It's not documented in the dockerfile why that approach is chosen.

I'd investigate further but just can't get anything to even build, e.g.

> sh ci/run.sh sparc64-unknown-linux-gnu
+ mirrors_url=https://ci-mirrors.rust-lang.org/libc
+ target=sparc64-unknown-linux-gnu
+ [ -n  ]
+ [ sparc64-unknown-linux-gnu = s390x-unknown-linux-gnu ]
+ cargo test --no-default-features --manifest-path libc-test/Cargo.toml --target sparc64-unknown-linux-gnu
   Compiling libc v1.0.0-alpha.1 (/home/folkertdev/rust/libc)
   Compiling ctest2 v0.4.8
error[E0061]: this function takes 3 arguments but 2 arguments were supplied
    --> /home/folkertdev/.cargo/registry/src/index.crates.io-6f17d22bba15001f/ctest2-0.4.8/src/lib.rs:2525:25
     |
2525 |                 Rc::new(macro_rules::compile(self.parse_sess, item)),
     |                         ^^^^^^^^^^^^^^^^^^^^                  ---- argument #2 of type `&RefCell<Features>` is missing
     |
note: function defined here
    --> /home/folkertdev/.cargo/registry/src/index.crates.io-6f17d22bba15001f/garando_syntax-0.1.0/src/ext/tt/macro_rules.rs:172:8
     |
172  | pub fn compile(sess: &ParseSess, features: &RefCell<Features>, def: &ast::Item) -> SyntaxExtension {
     |        ^^^^^^^
help: provide the argument
     |
2525 |                 Rc::new(macro_rules::compile(self.parse_sess, /* &RefCell<Features> */, item)),
     |                                             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

For more information about this error, try `rustc --explain E0061`.
error: could not compile `ctest2` (lib) due to 1 previous error
warning: build failed, waiting for other jobs to finish...

[tgross35]

I'm not sure exactly what it is but I don't enjoy it very much either, I had an issue open here #4061. I looked through the history before and iirc we use qemu-system rather than qemu-user because there were some crashes or inaccuracies, though I doubt they are still around.

qemu-system is likely part of the slowdown but it doesn't explain everything since both sparc and s390x use it, and s390x isn't all that slow. They do things slightly differently though, I haven't dug into it too much.