Add a packed/tagged pointer abstraction and utilize it for ParamEnv
The intent here is mostly just to add the abstraction; I suspect that there are definitely more use cases for it, and we can explore those over time now that there's a (mostly) safe abstraction that can be used in rustc.
Rollup of 7 pull requests
Successful merges:
- #75069 (move const param structural match checks to wfcheck)
- #75587 (mir building: fix some comments)
- #75593 (Adjust installation place for compiler docs)
- #75648 (Make OnceCell<T> transparent to dropck)
- #75649 (Fix intra-doc links for inherent impls that are both lang items and not the default impl)
- #75674 (Move to intra doc links for std::io)
- #75696 (Remove `#[cfg(miri)]` from OnceCell tests)
Failed merges:
r? @ghost
Remove `#[cfg(miri)]` from OnceCell tests
They were carried over from once_cell crate, but they are not entirely
correct (as miri now supports more things), and we don't run miri
tests for std, so let's just remove them.
Maybe one day we'll run miri in std, but then we can just re-install
these attributes.
Move to intra doc links for std::io
Helps with #75080.
@rustbot modify labels: T-doc, A-intra-doc-links, T-rustdoc
r? @jyn514
I had no problems with those files so I added some small links here and there.
Fix intra-doc links for inherent impls that are both lang items and not the default impl
I found in https://github.com/rust-lang/rust/pull/75464#issuecomment-675125984 that `str::to_uppercase()` doesn't resolve while `str::trim()` does. The only real difference is that `to_uppercase` is defined in `alloc`, while trim is defined in `core`. It turns out that rustdoc was ignoring `lang_items.str_alloc_impl()` - it saw them in `collect_trait_impls`, but not for intra-doc links.
This uses the same `impls` for all parts of rustdoc, so that there can be no more inconsistency. It does have the slight downside that the matches are no longer exhaustive but it will be very clear if a new lang item is missed because it will panic when you try to document it (and if you don't document it, does rustdoc really need to know about it?).
~~This needs a test case (probably just `str::to_uppercase`).~~ Added.
This is best reviewed commit-by-commit.
r? @GuillaumeGomez
Adjust installation place for compiler docs
This avoids conflicts when installing with rustup; rustup does not currently
support overlapping installations.
r? @matthiaskrgr
move const param structural match checks to wfcheck
fixes#75047fixes#74950
We currently check for structural match violations inside of `type_of`.
As we need to check the array length when checking if `[NonEq; arr_len]` is structural match, we potentially require the variance of an expression. Computing the variance requires `type_of` for all types though, resulting in a cycle error.
r? @varkor @eddyb
Doc alias checks: ensure only items appearing in search index can use it
Following the discussion in #73721, I added checks to ensure that only items appearing in the search are allowed to have doc alias.
r? @ollie27
Improve codegen for `align_offset`
In this PR the `align_offset` implementation is changed/improved to produce better code in certain scenarios such as when pointer type is has a stride of 1 or when building for low optimisation levels.
While these changes do not achieve the "ideal" codegen referenced in #75579, it gets significantly closer to it. I’m not actually sure if the codegen can actually be much better with this function returning the offset, rather than the aligned pointer.
See the descriptions for separate commits for further information.
They were carried over from once_cell crate, but they are not entirely
correct (as miri now supports more things), and we don't run miri
tests for std, so let's just remove them.
Maybe one day we'll run miri in std, but then we can just re-install
these attributes.
Rollup of 9 pull requests
Successful merges:
- #75038 (See also X-Link mem::{swap, take, replace})
- #75049 (docs(marker/copy): provide example for `&T` being `Copy`)
- #75499 (Fix documentation error)
- #75554 (Fix clashing_extern_declarations stack overflow for recursive types.)
- #75646 (Move to intra doc links for keyword documentation)
- #75652 (Resolve true and false as booleans)
- #75658 (Don't emit "is not a logical operator" error outside of associative expressions)
- #75665 (Add doc examples coverage)
- #75685 (Switch to intra-doc links in /src/sys/unix/ext/*.rs)
Failed merges:
r? @ghost
Switch to intra-doc links in /src/sys/unix/ext/*.rs
Partial fix for #75080
@rustbot modify labels: T-doc, A-intra-doc-links, T-rustdoc
r? @jyn514
These two links are not resolving to either `crate::fs::File...` or `fs::File...`
```
# unix/ext/fs.rs
27: /// [`File::read`]: ../../../../std/fs/struct.File.html#method.read
130: /// [`File::write`]: ../../../../std/fs/struct.File.html#method.write
```
Don't emit "is not a logical operator" error outside of associative expressions
Avoid showing this error where it doesn't make sense by not assuming
"and" and "or" were intended to mean "&&" and "||" until after we decide
to continue parsing input as an associative expression.
Note that the decision of whether or not to continue parsing input as an
associative expression doesn't actually depend on this assumption.
Fixes#75599
---
First time contributor! Let me know if there are any conventions or policies I should be following that I missed here. Thanks :)
Fix clashing_extern_declarations stack overflow for recursive types.
Fixes#75512.
Adds a seen set to `structurally_same_type` to avoid recursing indefinitely for types which contain values of the same type through a pointer or reference.
docs(marker/copy): provide example for `&T` being `Copy`
### Edited 2020-08-16 (most recent)
In the current documentation about the `Copy` marker trait, there is a section
with examples of structs that can implement `Copy`. Currently there is no example for
showing that shared references (`&T`) are also `Copy`.
It is worth to have a dedicated example for `&T` being `Copy`, because shared
references are an integral part of the language and it being `Copy` is not as
intuitive as other types that share this behaviour like `i32` or `bool`.
The example picks up on the previous non-`Copy` struct and shows that
structs can be `Copy`, even when they hold a shared reference to a non-`Copy` type.
-----------------------------------------
### Edited 2020-08-02, 3:28 p.m.
I've just realized that it says "in addition to the **implementors listed below**", which makes this PR kind of "wrong", because `&T` is indeed in the "implementors listed below".
Maybe we can instead show an example with `&T` in the [When can my type be Copy](https://doc.rust-lang.org/std/marker/trait.Copy.html#when-can-my-type-be-copy) section.
What I really want to achieve is that it becomes more obvious that `&T` is also `Copy`, because, I think, it is very valuable to know and it wasn't obvious for me, until I read something about it in a forum post.
What do you think? I would create another PR for that.
**Please feel free to close this PR.**
-----------------------------------
### Original post
In the current documentation about the `Copy` marker trait, there is a section
about "additional implementors", which list additional implementors of the `Copy` trait.
The fact that shared references are also `Copy` is mixed with another point,
which makes it hard to recognize and make it seem not as important.
This clarifies the fact that shared references are also `Copy`, by mentioning it as a
separate item in the list of "additional implementors".
See also X-Link mem::{swap, take, replace}
Since it's easy to end up at one of these functions when you really wanted the other one, cross link them with descriptions of why you'd want to use them.
Don't panic in Vec::shrink_to_fit
We can help the compiler see that `Vec::shrink_to_fit` will never reach the panic case in `RawVec::shrink_to_fit`, just by guarding the call only for cases where the capacity is strictly greater. A capacity less than the length is only possible through an unsafe call to `set_len`, which would break the `Vec` invariants, so `shrink_to_fit` can just ignore that.
This removes the panicking code from the examples in both #71861 and #75636.
Cargo update (almost) all the things!
This runs `cargo update` on our dependency tree, bumping numerous crates. See details in the first commit of this PR.
Several updates were held back intentionally; version_check in particular landed a change in behavior in 0.9.2 (arguably a bug fix, but still will be split into a separate PR).
Promote missing_fragment_specifier to hard error
It has been deny_by_default since 2017 (and warned for some time
before that), so it seems reasonable to promote it.
The specific technical motivation to do this now is to remove a field
from `ParseSess` -- it is a global state, and global state makes
extracting libraries annoying.
Closes#40107
cmake-rs@8141f0e changed the logic for handling asm compiler flags.
This change was pulled in with the cmake 0.1.42 -> 0.1.44 update.
This introduced two new flags to the LLVM build, breaking it:
"-DCMAKE_ASM_FLAGS= -ffunction-sections -fdata-sections -fPIC -m64"
"-DCMAKE_ASM_COMPILER=/usr/bin/cc"
This patch should resolve the breakage by handling it in bootstrap.
Avoid showing this error where it doesn't make sense by not assuming
"and" and "or" were intended to mean "&&" and "||" until after we decide
to continue parsing input as an associative expression.
Note that the decision of whether or not to continue parsing input as an
associative expression doesn't actually depend on this assumption.
Fixes#75599
Don't use `dep_graph.with_ignore` when encoding fn param names
The call to `with_ignore` was added years ago, and should no longer be
necessary with the modern incremental query infrastructure.
I also removed a related FIXME comment for issue #38501, which was
closed months ago.
