Rework ptr-to-ref conversion suggestion for method calls

[kadiwa4]

If we have a value z of type *const u8 and try to call z.to_string(), the upstream compiler will show you a note suggesting to call <*const u8>::as_ref first.

This PR extends that:

• The note will only be shown when the method would exist on the corresponding reference type
• It can now suggest any of <*const u8>::as_ref, <*mut u8>::as_ref and <*mut u8>::as_mut, depending on what the method needs.

I didn't introduce a help message because that's not a good idea with unsafe functions (and you'd also need to unwrap the Option<&_> somehow).
People should check the safety requirements.

For the simplest case

fn main() {
    let x = 8u8;
    let z: *const u8 = &x;
    // issue #21596
    println!("{}", z.to_string()); //~ ERROR E0599
}

the output changes like this:

 error[E0599]: `*const u8` doesn't implement `std::fmt::Display`
   --> $DIR/suggest-convert-ptr-to-ref.rs:5:22
    |
 LL |     println!("{}", z.to_string());
    |                      ^^^^^^^^^ `*const u8` cannot be formatted with the default formatter
    |
-   = note: try using `<*const T>::as_ref()` to get a reference to the type behind the pointer: https://doc.rust-lang.org/std/primitive.pointer.html#method.as_ref
-   = note: using `<*const T>::as_ref()` on a pointer which is unaligned or points to invalid or uninitialized memory is undefined behavior
+note: the method `to_string` exists on the type `&u8`
+  --> $SRC_DIR/alloc/src/string.rs:LL:COL
+   = note: you might want to use the unsafe method `<*const T>::as_ref` to get an optional reference to the value behind the pointer
+   = note: read the documentation for `<*const T>::as_ref` and ensure you satisfy its safety preconditions before calling it to avoid undefined behavior: https://doc.rust-lang.org/std/primitive.pointer.html#method.as_ref
    = note: the following trait bounds were not satisfied:
            `*const u8: std::fmt::Display`
            which is required by `*const u8: ToString`

I removed the separate note about the safety requirements because it was incomplete and the linked doc page already has the information you need.

Fixes #83695, but that's more of a side effect. The upstream compiler already suggests the right method name here.

[rustbot]

r? @nnethercote

rustbot has assigned @nnethercote.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

[kadiwa4]

I don't know if this is the right lifetime to use.

[kadiwa4]

r? compiler

[bors]

☔ The latest upstream changes (presumably #123036) made this pull request unmergeable. Please resolve the merge conflicts.

[bors]

☔ The latest upstream changes (presumably #123429) made this pull request unmergeable. Please resolve the merge conflicts.

[estebank]

I removed the separate note about the safety requirements because it was incomplete and the linked doc page already has the information you need.

TBH, I would change it to be

+note: the method `to_string` exists on the type `&u8`
+  --> $SRC_DIR/alloc/src/string.rs:LL:COL
+   = note: you might want to use the unsafe method `<*const T>::as_ref` to get an optional reference to the value behind the pointer
+   = warning: using `<*const T>::as_ref()` on a pointer which is unaligned or points to invalid or uninitialized memory is undefined behavior: https://doc.rust-lang.org/std/primitive.pointer.html#method.as_ref

or

+   = warning: read the documentation for <*const T>::as_ref() and ensure you satisfy its safety preconditions before calling it to avoid undefined behavior: https://doc.rust-lang.org/std/primitive.pointer.html#method.as_ref

In general, we can rely on links to have information for things that are nice to know but nor absolutely required. Most people barely read the errors, even fewer read the linked docs.

If we reliably had <blink> on the terminal, I'd be tempted to use it for this 😛

[bors]

☔ The latest upstream changes (presumably #123339) made this pull request unmergeable. Please resolve the merge conflicts.

[bors]

☔ The latest upstream changes (presumably #123576) made this pull request unmergeable. Please resolve the merge conflicts.

[bors]

☔ The latest upstream changes (presumably #123569) made this pull request unmergeable. Please resolve the merge conflicts.

[bors]

☔ The latest upstream changes (presumably #123708) made this pull request unmergeable. Please resolve the merge conflicts.

[estebank]

@bors r+

[bors]

📌 Commit 3a2a3ae has been approved by estebank

It is now in the queue for this repository.

[bors]

⌛ Testing commit 3a2a3ae with merge e1fa18f...

[bors]

💔 Test failed - checks-actions

[rust-log-analyzer]

The job dist-aarch64-apple failed! Check out the build log: (web) (plain)

Click to see the possible cause of the failure (guessed by this bot)

[estebank]

@bors retry

[bors]

⌛ Testing commit 3a2a3ae with merge f13f37f...

[bors]

☀️ Test successful - checks-actions
Approved by: estebank
Pushing f13f37f to master...

[rust-timer]

Finished benchmarking commit (f13f37f): comparison URL.

Overall result: no relevant changes - no action needed

@rustbot label: -perf-regression

Instruction count

This benchmark run did not return any relevant results for this metric.

Max RSS (memory usage)

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	1.3%	[1.0%, 1.5%]	2
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	1.3%	[1.0%, 1.5%]	2

Cycles

This benchmark run did not return any relevant results for this metric.

Binary size

This benchmark run did not return any relevant results for this metric.

Bootstrap: 676.971s -> 676.139s (-0.12%)
Artifact size: 316.01 MiB -> 316.07 MiB (0.02%)