rustdoc: Tracking issue for `pointer` and `reference` in intra-doc links

[camelid]

The feature gate for this experimental feature is intra_doc_pointers.

Note

Tracking issues are used to record the overall progress of implementation.
They are also used as hubs connecting to other relevant issues, e.g., bugs or open design questions.

A tracking issue is however not meant for large scale discussion, questions, or bug reports.
Instead, open a dedicated issue for the specific matter or post in the #t-rustdoc Zulip channel.

Discussion comments will get marked as off-topic or deleted.
Repeated discussions on the tracking issue may lead to the tracking issue getting locked.

Steps

• Implementation PR
• Stabilization PR

Implementation history

• FIXME: Find the PR that actually impls these
• Feature-gate pointer and reference in intra-doc links #81015

Historic notes

We may want to put other targets behind the feature gate too, but we at least
want to get pointer behind one before #80181 hits beta.

See the discussion on Zulip for more.

[jhpratt]

Getting an error on nightly from this that originates in stdlib.

Not sure if this needs special-casing, a feature gate added to core, or just the removal of the link. But something certainly isn't correct.

error[E0658]: linking to associated items of raw pointers is experimental
  --> /home/jhpratt/.rustup/toolchains/nightly-x86_64-unknown-linux-gnu/lib/rustlib/src/rust/library/core/src/ptr/metadata.rs:51:23
   |
51 | /// [`to_raw_parts`]: *const::to_raw_parts
   |                       ^^^^^^^^^^^^^^^^^^^^
   |
   = note: see issue #80896 <https://github.com/rust-lang/rust/issues/80896> for more information
   = help: add `#![feature(intra_doc_pointers)]` to the crate attributes to enable
   = note: rustdoc does not allow disambiguating between `*const` and `*mut`, and pointers are unstable until it does

This is the line that the error traces to.

[jhpratt]

cc @jyn514 see above. It looks like you're the one that gated this.

[jhpratt]

Looks like this was noticed before I commented. The issue for it is #82284.

[Nemo157]

Given that #62834 allowed the applied type Vec<T> to link to the type constructor Vec, it feels like *const T/&T and similar syntax should also be supported to link to the pages for the type constructors pointer/reference.

[camelid]

Perhaps, but I imagine that linking to *const T is not that common, and if it's not that common then it doesn't seem that bad to have people use [*const T](pointer).