diff options
author | 2022-10-03 16:39:37 -0700 | |
---|---|---|
committer | 2022-10-03 16:39:37 -0700 | |
commit | 8aebac82933ff1a7c8eede18cab11e1115e2062b (patch) | |
tree | 8305fe6437fe1db1a5dedf218d53a6bd21120793 | |
parent | Merge tag 'thermal-6.1-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm (diff) | |
parent | MAINTAINERS: Rust (diff) | |
download | wireguard-linux-8aebac82933ff1a7c8eede18cab11e1115e2062b.tar.xz wireguard-linux-8aebac82933ff1a7c8eede18cab11e1115e2062b.zip |
Merge tag 'rust-v6.1-rc1' of https://github.com/Rust-for-Linux/linux
Pull Rust introductory support from Kees Cook:
"The tree has a recent base, but has fundamentally been in linux-next
for a year and a half[1]. It's been updated based on feedback from the
Kernel Maintainer's Summit, and to gain recent Reviewed-by: tags.
Miguel is the primary maintainer, with me helping where needed/wanted.
Our plan is for the tree to switch to the standard non-rebasing
practice once this initial infrastructure series lands.
The contents are the absolute minimum to get Rust code building in the
kernel, with many more interfaces[2] (and drivers - NVMe[3], 9p[4], M1
GPU[5]) on the way.
The initial support of Rust-for-Linux comes in roughly 4 areas:
- Kernel internals (kallsyms expansion for Rust symbols, %pA format)
- Kbuild infrastructure (Rust build rules and support scripts)
- Rust crates and bindings for initial minimum viable build
- Rust kernel documentation and samples
Rust support has been in linux-next for a year and a half now, and the
short log doesn't do justice to the number of people who have
contributed both to the Linux kernel side but also to the upstream
Rust side to support the kernel's needs. Thanks to these 173 people,
and many more, who have been involved in all kinds of ways:
Miguel Ojeda, Wedson Almeida Filho, Alex Gaynor, Boqun Feng, Gary Guo,
Björn Roy Baron, Andreas Hindborg, Adam Bratschi-Kaye, Benno Lossin,
Maciej Falkowski, Finn Behrens, Sven Van Asbroeck, Asahi Lina, FUJITA
Tomonori, John Baublitz, Wei Liu, Geoffrey Thomas, Philip Herron,
Arthur Cohen, David Faust, Antoni Boucher, Philip Li, Yujie Liu,
Jonathan Corbet, Greg Kroah-Hartman, Paul E. McKenney, Josh Triplett,
Kent Overstreet, David Gow, Alice Ryhl, Robin Randhawa, Kees Cook,
Nick Desaulniers, Matthew Wilcox, Linus Walleij, Joe Perches, Michael
Ellerman, Petr Mladek, Masahiro Yamada, Arnaldo Carvalho de Melo,
Andrii Nakryiko, Konstantin Shelekhin, Rasmus Villemoes, Konstantin
Ryabitsev, Stephen Rothwell, Andy Shevchenko, Sergey Senozhatsky, John
Paul Adrian Glaubitz, David Laight, Nathan Chancellor, Jonathan
Cameron, Daniel Latypov, Shuah Khan, Brendan Higgins, Julia Lawall,
Laurent Pinchart, Geert Uytterhoeven, Akira Yokosawa, Pavel Machek,
David S. Miller, John Hawley, James Bottomley, Arnd Bergmann,
Christian Brauner, Dan Robertson, Nicholas Piggin, Zhouyi Zhou, Elena
Zannoni, Jose E. Marchesi, Leon Romanovsky, Will Deacon, Richard
Weinberger, Randy Dunlap, Paolo Bonzini, Roland Dreier, Mark Brown,
Sasha Levin, Ted Ts'o, Steven Rostedt, Jarkko Sakkinen, Michal
Kubecek, Marco Elver, Al Viro, Keith Busch, Johannes Berg, Jan Kara,
David Sterba, Connor Kuehl, Andy Lutomirski, Andrew Lunn, Alexandre
Belloni, Peter Zijlstra, Russell King, Eric W. Biederman, Willy
Tarreau, Christoph Hellwig, Emilio Cobos Álvarez, Christian Poveda,
Mark Rousskov, John Ericson, TennyZhuang, Xuanwo, Daniel Paoliello,
Manish Goregaokar, comex, Josh Stone, Stephan Sokolow, Philipp Krones,
Guillaume Gomez, Joshua Nelson, Mats Larsen, Marc Poulhiès, Samantha
Miller, Esteban Blanc, Martin Schmidt, Martin Rodriguez Reboredo,
Daniel Xu, Viresh Kumar, Bartosz Golaszewski, Vegard Nossum, Milan
Landaverde, Dariusz Sosnowski, Yuki Okushi, Matthew Bakhtiari, Wu
XiangCheng, Tiago Lam, Boris-Chengbiao Zhou, Sumera Priyadarsini,
Viktor Garske, Niklas Mohrin, Nándor István Krácser, Morgan Bartlett,
Miguel Cano, Léo Lanteri Thauvin, Julian Merkle, Andreas Reindl,
Jiapeng Chong, Fox Chen, Douglas Su, Antonio Terceiro, SeongJae Park,
Sergio González Collado, Ngo Iok Ui (Wu Yu Wei), Joshua Abraham,
Milan, Daniel Kolsoi, ahomescu, Manas, Luis Gerhorst, Li Hongyu,
Philipp Gesang, Russell Currey, Jalil David Salamé Messina, Jon Olson,
Raghvender, Angelos, Kaviraj Kanagaraj, Paul Römer, Sladyn Nunes,
Mauro Baladés, Hsiang-Cheng Yang, Abhik Jain, Hongyu Li, Sean Nash,
Yuheng Su, Peng Hao, Anhad Singh, Roel Kluin, Sara Saa, Geert
Stappers, Garrett LeSage, IFo Hancroft, and Linus Torvalds"
Link: https://lwn.net/Articles/849849/ [1]
Link: https://github.com/Rust-for-Linux/linux/commits/rust [2]
Link: https://github.com/metaspace/rust-linux/commit/d88c3744d6cbdf11767e08bad56cbfb67c4c96d0 [3]
Link: https://github.com/wedsonaf/linux/commit/9367032607f7670de0ba1537cf09ab0f4365a338 [4]
Link: https://github.com/AsahiLinux/linux/commits/gpu/rust-wip [5]
* tag 'rust-v6.1-rc1' of https://github.com/Rust-for-Linux/linux: (27 commits)
MAINTAINERS: Rust
samples: add first Rust examples
x86: enable initial Rust support
docs: add Rust documentation
Kbuild: add Rust support
rust: add `.rustfmt.toml`
scripts: add `is_rust_module.sh`
scripts: add `rust_is_available.sh`
scripts: add `generate_rust_target.rs`
scripts: add `generate_rust_analyzer.py`
scripts: decode_stacktrace: demangle Rust symbols
scripts: checkpatch: enable language-independent checks for Rust
scripts: checkpatch: diagnose uses of `%pA` in the C side as errors
vsprintf: add new `%pA` format specifier
rust: export generated symbols
rust: add `kernel` crate
rust: add `bindings` crate
rust: add `macros` crate
rust: add `compiler_builtins` crate
rust: adapt `alloc` crate to the kernel
...
89 files changed, 12552 insertions, 51 deletions
diff --git a/.gitignore b/.gitignore index 265959544978..5da004814678 100644 --- a/.gitignore +++ b/.gitignore @@ -37,6 +37,8 @@ *.o *.o.* *.patch +*.rmeta +*.rsi *.s *.so *.so.dbg @@ -97,6 +99,7 @@ modules.order !.gitattributes !.gitignore !.mailmap +!.rustfmt.toml # # Generated include files @@ -162,3 +165,6 @@ x509.genkey # Documentation toolchain sphinx_*/ + +# Rust analyzer configuration +/rust-project.json diff --git a/.rustfmt.toml b/.rustfmt.toml new file mode 100644 index 000000000000..3de5cc497465 --- /dev/null +++ b/.rustfmt.toml @@ -0,0 +1,12 @@ +edition = "2021" +newline_style = "Unix" + +# Unstable options that help catching some mistakes in formatting and that we may want to enable +# when they become stable. +# +# They are kept here since they are useful to run from time to time. +#format_code_in_doc_comments = true +#reorder_impl_items = true +#comment_width = 100 +#wrap_comments = true +#normalize_comments = true diff --git a/Documentation/core-api/printk-formats.rst b/Documentation/core-api/printk-formats.rst index 5e89497ba314..dbe1aacc79d0 100644 --- a/Documentation/core-api/printk-formats.rst +++ b/Documentation/core-api/printk-formats.rst @@ -625,6 +625,16 @@ Examples:: %p4cc Y10 little-endian (0x20303159) %p4cc NV12 big-endian (0xb231564e) +Rust +---- + +:: + + %pA + +Only intended to be used from Rust code to format ``core::fmt::Arguments``. +Do *not* use it from C. + Thanks ====== diff --git a/Documentation/doc-guide/kernel-doc.rst b/Documentation/doc-guide/kernel-doc.rst index 9c779bd7a751..1dcbd7332476 100644 --- a/Documentation/doc-guide/kernel-doc.rst +++ b/Documentation/doc-guide/kernel-doc.rst @@ -14,6 +14,9 @@ when it is embedded in source files. reasons. The kernel source contains tens of thousands of kernel-doc comments. Please stick to the style described here. +.. note:: kernel-doc does not cover Rust code: please see + Documentation/rust/general-information.rst instead. + The kernel-doc structure is extracted from the comments, and proper `Sphinx C Domain`_ function and type descriptions with anchors are generated from them. The descriptions are filtered for special kernel-doc diff --git a/Documentation/index.rst b/Documentation/index.rst index 85eab6e990ab..bf6aa681c960 100644 --- a/Documentation/index.rst +++ b/Documentation/index.rst @@ -58,6 +58,7 @@ Various other manuals with useful information for all kernel developers. trace/index fault-injection/index livepatch/index + rust/index User-oriented documentation diff --git a/Documentation/kbuild/kbuild.rst b/Documentation/kbuild/kbuild.rst index ef19b9c13523..08f575e6236c 100644 --- a/Documentation/kbuild/kbuild.rst +++ b/Documentation/kbuild/kbuild.rst @@ -48,6 +48,10 @@ KCFLAGS ------- Additional options to the C compiler (for built-in and modules). +KRUSTFLAGS +---------- +Additional options to the Rust compiler (for built-in and modules). + CFLAGS_KERNEL ------------- Additional options for $(CC) when used to compile @@ -57,6 +61,15 @@ CFLAGS_MODULE ------------- Additional module specific options to use for $(CC). +RUSTFLAGS_KERNEL +---------------- +Additional options for $(RUSTC) when used to compile +code that is compiled as built-in. + +RUSTFLAGS_MODULE +---------------- +Additional module specific options to use for $(RUSTC). + LDFLAGS_MODULE -------------- Additional options used for $(LD) when linking modules. @@ -69,6 +82,10 @@ HOSTCXXFLAGS ------------ Additional flags to be passed to $(HOSTCXX) when building host programs. +HOSTRUSTFLAGS +------------- +Additional flags to be passed to $(HOSTRUSTC) when building host programs. + HOSTLDFLAGS ----------- Additional flags to be passed when linking host programs. diff --git a/Documentation/kbuild/makefiles.rst b/Documentation/kbuild/makefiles.rst index 11a296e52d68..5ea1e72d89c8 100644 --- a/Documentation/kbuild/makefiles.rst +++ b/Documentation/kbuild/makefiles.rst @@ -29,8 +29,9 @@ This document describes the Linux kernel Makefiles. --- 4.1 Simple Host Program --- 4.2 Composite Host Programs --- 4.3 Using C++ for host programs - --- 4.4 Controlling compiler options for host programs - --- 4.5 When host programs are actually built + --- 4.4 Using Rust for host programs + --- 4.5 Controlling compiler options for host programs + --- 4.6 When host programs are actually built === 5 Userspace Program support --- 5.1 Simple Userspace Program @@ -835,7 +836,24 @@ Both possibilities are described in the following. qconf-cxxobjs := qconf.o qconf-objs := check.o -4.4 Controlling compiler options for host programs +4.4 Using Rust for host programs +-------------------------------- + + Kbuild offers support for host programs written in Rust. However, + since a Rust toolchain is not mandatory for kernel compilation, + it may only be used in scenarios where Rust is required to be + available (e.g. when ``CONFIG_RUST`` is enabled). + + Example:: + + hostprogs := target + target-rust := y + + Kbuild will compile ``target`` using ``target.rs`` as the crate root, + located in the same directory as the ``Makefile``. The crate may + consist of several source files (see ``samples/rust/hostprogs``). + +4.5 Controlling compiler options for host programs -------------------------------------------------- When compiling host programs, it is possible to set specific flags. @@ -867,7 +885,7 @@ Both possibilities are described in the following. When linking qconf, it will be passed the extra option "-L$(QTDIR)/lib". -4.5 When host programs are actually built +4.6 When host programs are actually built ----------------------------------------- Kbuild will only build host-programs when they are referenced @@ -1181,6 +1199,17 @@ When kbuild executes, the following steps are followed (roughly): The first example utilises the trick that a config option expands to 'y' when selected. + KBUILD_RUSTFLAGS + $(RUSTC) compiler flags + + Default value - see top level Makefile + Append or modify as required per architecture. + + Often, the KBUILD_RUSTFLAGS variable depends on the configuration. + + Note that target specification file generation (for ``--target``) + is handled in ``scripts/generate_rust_target.rs``. + KBUILD_AFLAGS_KERNEL Assembler options specific for built-in @@ -1208,6 +1237,19 @@ When kbuild executes, the following steps are followed (roughly): are used for $(CC). From commandline CFLAGS_MODULE shall be used (see kbuild.rst). + KBUILD_RUSTFLAGS_KERNEL + $(RUSTC) options specific for built-in + + $(KBUILD_RUSTFLAGS_KERNEL) contains extra Rust compiler flags used to + compile resident kernel code. + + KBUILD_RUSTFLAGS_MODULE + Options for $(RUSTC) when building modules + + $(KBUILD_RUSTFLAGS_MODULE) is used to add arch-specific options that + are used for $(RUSTC). + From commandline RUSTFLAGS_MODULE shall be used (see kbuild.rst). + KBUILD_LDFLAGS_MODULE Options for $(LD) when linking modules diff --git a/Documentation/process/changes.rst b/Documentation/process/changes.rst index 19c286c23786..9a90197989dd 100644 --- a/Documentation/process/changes.rst +++ b/Documentation/process/changes.rst @@ -31,6 +31,8 @@ you probably needn't concern yourself with pcmciautils. ====================== =============== ======================================== GNU C 5.1 gcc --version Clang/LLVM (optional) 11.0.0 clang --version +Rust (optional) 1.62.0 rustc --version +bindgen (optional) 0.56.0 bindgen --version GNU make 3.81 make --version bash 4.2 bash --version binutils 2.23 ld -v @@ -80,6 +82,29 @@ kernels. Older releases aren't guaranteed to work, and we may drop workarounds from the kernel that were used to support older versions. Please see additional docs on :ref:`Building Linux with Clang/LLVM <kbuild_llvm>`. +Rust (optional) +--------------- + +A particular version of the Rust toolchain is required. Newer versions may or +may not work because the kernel depends on some unstable Rust features, for +the moment. + +Each Rust toolchain comes with several "components", some of which are required +(like ``rustc``) and some that are optional. The ``rust-src`` component (which +is optional) needs to be installed to build the kernel. Other components are +useful for developing. + +Please see Documentation/rust/quick-start.rst for instructions on how to +satisfy the build requirements of Rust support. In particular, the ``Makefile`` +target ``rustavailable`` is useful to check why the Rust toolchain may not +be detected. + +bindgen (optional) +------------------ + +``bindgen`` is used to generate the Rust bindings to the C side of the kernel. +It depends on ``libclang``. + Make ---- @@ -348,6 +373,12 @@ Sphinx Please see :ref:`sphinx_install` in :ref:`Documentation/doc-guide/sphinx.rst <sphinxdoc>` for details about Sphinx requirements. +rustdoc +------- + +``rustdoc`` is used to generate the documentation for Rust code. Please see +Documentation/rust/general-information.rst for more information. + Getting updated software ======================== @@ -364,6 +395,16 @@ Clang/LLVM - :ref:`Getting LLVM <getting_llvm>`. +Rust +---- + +- Documentation/rust/quick-start.rst. + +bindgen +------- + +- Documentation/rust/quick-start.rst. + Make ---- diff --git a/Documentation/rust/arch-support.rst b/Documentation/rust/arch-support.rst new file mode 100644 index 000000000000..6982b63775da --- /dev/null +++ b/Documentation/rust/arch-support.rst @@ -0,0 +1,19 @@ +.. SPDX-License-Identifier: GPL-2.0 + +Arch Support +============ + +Currently, the Rust compiler (``rustc``) uses LLVM for code generation, +which limits the supported architectures that can be targeted. In addition, +support for building the kernel with LLVM/Clang varies (please see +Documentation/kbuild/llvm.rst). This support is needed for ``bindgen`` +which uses ``libclang``. + +Below is a general summary of architectures that currently work. Level of +support corresponds to ``S`` values in the ``MAINTAINERS`` file. + +============ ================ ============================================== +Architecture Level of support Constraints +============ ================ ============================================== +``x86`` Maintained ``x86_64`` only. +============ ================ ============================================== diff --git a/Documentation/rust/coding-guidelines.rst b/Documentation/rust/coding-guidelines.rst new file mode 100644 index 000000000000..aa8ed082613e --- /dev/null +++ b/Documentation/rust/coding-guidelines.rst @@ -0,0 +1,216 @@ +.. SPDX-License-Identifier: GPL-2.0 + +Coding Guidelines +================= + +This document describes how to write Rust code in the kernel. + + +Style & formatting +------------------ + +The code should be formatted using ``rustfmt``. In this way, a person +contributing from time to time to the kernel does not need to learn and +remember one more style guide. More importantly, reviewers and maintainers +do not need to spend time pointing out style issues anymore, and thus +less patch roundtrips may be needed to land a change. + +.. note:: Conventions on comments and documentation are not checked by + ``rustfmt``. Thus those are still needed to be taken care of. + +The default settings of ``rustfmt`` are used. This means the idiomatic Rust +style is followed. For instance, 4 spaces are used for indentation rather +than tabs. + +It is convenient to instruct editors/IDEs to format while typing, +when saving or at commit time. However, if for some reason reformatting +the entire kernel Rust sources is needed at some point, the following can be +run:: + + make LLVM=1 rustfmt + +It is also possible to check if everything is formatted (printing a diff +otherwise), for instance for a CI, with:: + + make LLVM=1 rustfmtcheck + +Like ``clang-format`` for the rest of the kernel, ``rustfmt`` works on +individual files, and does not require a kernel configuration. Sometimes it may +even work with broken code. + + +Comments +-------- + +"Normal" comments (i.e. ``//``, rather than code documentation which starts +with ``///`` or ``//!``) are written in Markdown the same way as documentation +comments are, even though they will not be rendered. This improves consistency, +simplifies the rules and allows to move content between the two kinds of +comments more easily. For instance: + +.. code-block:: rust + + // `object` is ready to be handled now. + f(object); + +Furthermore, just like documentation, comments are capitalized at the beginning +of a sentence and ended with a period (even if it is a single sentence). This +includes ``// SAFETY:``, ``// TODO:`` and other "tagged" comments, e.g.: + +.. code-block:: rust + + // FIXME: The error should be handled properly. + +Comments should not be used for documentation purposes: comments are intended +for implementation details, not users. This distinction is useful even if the +reader of the source file is both an implementor and a user of an API. In fact, +sometimes it is useful to use both comments and documentation at the same time. +For instance, for a ``TODO`` list or to comment on the documentation itself. +For the latter case, comments can be inserted in the middle; that is, closer to +the line of documentation to be commented. For any other case, comments are +written after the documentation, e.g.: + +.. code-block:: rust + + /// Returns a new [`Foo`]. + /// + /// # Examples + /// + // TODO: Find a better example. + /// ``` + /// let foo = f(42); + /// ``` + // FIXME: Use fallible approach. + pub fn f(x: i32) -> Foo { + // ... + } + +One special kind of comments are the ``// SAFETY:`` comments. These must appear +before every ``unsafe`` block, and they explain why the code inside the block is +correct/sound, i.e. why it cannot trigger undefined behavior in any case, e.g.: + +.. code-block:: rust + + // SAFETY: `p` is valid by the safety requirements. + unsafe { *p = 0; } + +``// SAFETY:`` comments are not to be confused with the ``# Safety`` sections +in code documentation. ``# Safety`` sections specify the contract that callers +(for functions) or implementors (for traits) need to abide by. ``// SAFETY:`` +comments show why a call (for functions) or implementation (for traits) actually +respects the preconditions stated in a ``# Safety`` section or the language +reference. + + +Code documentation +------------------ + +Rust kernel code is not documented like C kernel code (i.e. via kernel-doc). +Instead, the usual system for documenting Rust code is used: the ``rustdoc`` +tool, which uses Markdown (a lightweight markup language). + +To learn Markdown, there are many guides available out there. For instance, +the one at: + + https://commonmark.org/help/ + +This is how a well-documented Rust function may look like: + +.. code-block:: rust + + /// Returns the contained [`Some`] value, consuming the `self` value, + /// without checking that the value is not [`None`]. + /// + /// # Safety + /// + /// Calling this method on [`None`] is *[undefined behavior]*. + /// + /// [undefined behavior]: https://doc.rust-lang.org/reference/behavior-considered-undefined.html + /// + /// # Examples + /// + /// ``` + /// let x = Some("air"); + /// assert_eq!(unsafe { x.unwrap_unchecked() }, "air"); + /// ``` + pub unsafe fn unwrap_unchecked(self) -> T { + match self { + Some(val) => val, + + // SAFETY: The safety contract must be upheld by the caller. + None => unsafe { hint::unreachable_unchecked() }, + } + } + +This example showcases a few ``rustdoc`` features and some conventions followed +in the kernel: + + - The first paragraph must be a single sentence briefly describing what + the documented item does. Further explanations must go in extra paragraphs. + + - Unsafe functions must document their safety preconditions under + a ``# Safety`` section. + + - While not shown here, if a function may panic, the conditions under which + that happens must be described under a ``# Panics`` section. + + Please note that panicking should be very rare and used only with a good + reason. In almost all cases, a fallible approach should be used, typically + returning a ``Result``. + + - If providing examples of usage would help readers, they must be written in + a section called ``# Examples``. + + - Rust items (functions, types, constants...) must be linked appropriately + (``rustdoc`` will create a link automatically). + + - Any ``unsafe`` block must be preceded by a ``// SAFETY:`` comment + describing why the code inside is sound. + + While sometimes the reason might look trivial and therefore unneeded, + writing these comments is not just a good way of documenting what has been + taken into account, but most importantly, it provides a way to know that + there are no *extra* implicit constraints. + +To learn more about how to write documentation for Rust and extra features, +please take a look at the ``rustdoc`` book at: + + https://doc.rust-lang.org/rustdoc/how-to-write-documentation.html + + +Naming +------ + +Rust kernel code follows the usual Rust naming conventions: + + https://rust-lang.github.io/api-guidelines/naming.html + +When existing C concepts (e.g. macros, functions, objects...) are wrapped into +a Rust abstraction, a name as close as reasonably possible to the C side should +be used in order to avoid confusion and to improve readability when switching +back and forth between the C and Rust sides. For instance, macros such as +``pr_info`` from C are named the same in the Rust side. + +Having said that, casing should be adjusted to follow the Rust naming +conventions, and namespacing introduced by modules and types should not be +repeated in the item names. For instance, when wrapping constants like: + +.. code-block:: c + + #define GPIO_LINE_DIRECTION_IN 0 + #define GPIO_LINE_DIRECTION_OUT 1 + +The equivalent in Rust may look like (ignoring documentation): + +.. code-block:: rust + + pub mod gpio { + pub enum LineDirection { + In = bindings::GPIO_LINE_DIRECTION_IN as _, + Out = bindings::GPIO_LINE_DIRECTION_OUT as _, + } + } + +That is, the equivalent of ``GPIO_LINE_DIRECTION_IN`` would be referred to as +``gpio::LineDirection::In``. In particular, it should not be named +``gpio::gpio_line_direction::GPIO_LINE_DIRECTION_IN``. diff --git a/Documentation/rust/general-information.rst b/Documentation/rust/general-information.rst new file mode 100644 index 000000000000..49029ee82e55 --- /dev/null +++ b/Documentation/rust/general-information.rst @@ -0,0 +1,79 @@ +.. SPDX-License-Identifier: GPL-2.0 + +General Information +=================== + +This document contains useful information to know when working with +the Rust support in the kernel. + + +Code documentation +------------------ + +Rust kernel code is documented using ``rustdoc``, its built-in documentation +generator. + +The generated HTML docs include integrated search, linked items (e.g. types, +functions, constants), source code, etc. They may be read at (TODO: link when +in mainline and generated alongside the rest of the documentation): + + http://kernel.org/ + +The docs can also be easily generated and read locally. This is quite fast +(same order as compiling the code itself) and no special tools or environment +are needed. This has the added advantage that they will be tailored to +the particular kernel configuration used. To generate them, use the ``rustdoc`` +target with the same invocation used for compilation, e.g.:: + + make LLVM=1 rustdoc + +To read the docs locally in your web browser, run e.g.:: + + xdg-open rust/doc/kernel/index.html + +To learn about how to write the documentation, please see coding-guidelines.rst. + + +Extra lints +----------- + +While ``rustc`` is a very helpful compiler, some extra lints and analyses are +available via ``clippy``, a Rust linter. To enable it, pass ``CLIPPY=1`` to +the same invocation used for compilation, e.g.:: + + make LLVM=1 CLIPPY=1 + +Please note that Clippy may change code generation, thus it should not be +enabled while building a production kernel. + + +Abstractions vs. bindings +------------------------- + +Abstractions are Rust code wrapping kernel functionality from the C side. + +In order to use functions and types from the C side, bindings are created. +Bindings are the declarations for Rust of those functions and types from +the C side. + +For instance, one may write a ``Mutex`` abstraction in Rust which wraps +a ``struct mutex`` from the C side and calls its functions through the bindings. + +Abstractions are not available for all the kernel internal APIs and concepts, +but it is intended that coverage is expanded as time goes on. "Leaf" modules +(e.g. drivers) should not use the C bindings directly. Instead, subsystems +should provide as-safe-as-possible abstractions as needed. + + +Conditional compilation +----------------------- + +Rust code has access to conditional compilation based on the kernel +configuration: + +.. code-block:: rust + + #[cfg(CONFIG_X)] // Enabled (`y` or `m`) + #[cfg(CONFIG_X="y")] // Enabled as a built-in (`y`) + #[cfg(CONFIG_X="m")] // Enabled as a module (`m`) + #[cfg(not(CONFIG_X))] // Disabled diff --git a/Documentation/rust/index.rst b/Documentation/rust/index.rst new file mode 100644 index 000000000000..4ae8c66b94fa --- /dev/null +++ b/Documentation/rust/index.rst @@ -0,0 +1,22 @@ +.. SPDX-License-Identifier: GPL-2.0 + +Rust +==== + +Documentation related to Rust within the kernel. To start using Rust +in the kernel, please read the quick-start.rst guide. + +.. toctree:: + :maxdepth: 1 + + quick-start + general-information + coding-guidelines + arch-support + +.. only:: subproject and html + + Indices + ======= + + * :ref:`genindex` diff --git a/Documentation/rust/quick-start.rst b/Documentation/rust/quick-start.rst new file mode 100644 index 000000000000..13b7744b1e27 --- /dev/null +++ b/Documentation/rust/quick-start.rst @@ -0,0 +1,232 @@ +.. SPDX-License-Identifier: GPL-2.0 + +Quick Start +=========== + +This document describes how to get started with kernel development in Rust. + + +Requirements: Building +---------------------- + +This section explains how to fetch the tools needed for building. + +Some of these requirements might be available from Linux distributions +under names like ``rustc``, ``rust-src``, ``rust-bindgen``, etc. However, +at the time of writing, they are likely not to be recent enough unless +the distribution tracks the latest releases. + +To easily check whether the requirements are met, the following target +can be used:: + + make LLVM=1 rustavailable + +This triggers the same logic used by Kconfig to determine whether +``RUST_IS_AVAILABLE`` should be enabled; but it also explains why not +if that is the case. + + +rustc +***** + +A particular version of the Rust compiler is required. Newer versions may or +may not work because, for the moment, the kernel depends on some unstable +Rust features. + +If ``rustup`` is being used, enter the checked out source code directory +and run:: + + rustup override set $(scripts/min-tool-version.sh rustc) + +Otherwise, fetch a standalone installer or install ``rustup`` from: + + https://www.rust-lang.org + + +Rust standard library source +**************************** + +The Rust standard library source is required because the build system will +cross-compile ``core`` and ``alloc``. + +If ``rustup`` is being used, run:: + + rustup component add rust-src + +The components are installed per toolchain, thus upgrading the Rust compiler +version later on requires re-adding the component. + +Otherwise, if a standalone installer is used, the Rust repository may be cloned +into the installation folder of the toolchain:: + + git clone --recurse-submodules \ + --branch $(scripts/min-tool-version.sh rustc) \ + https://github.com/rust-lang/rust \ + $(rustc --print sysroot)/lib/rustlib/src/rust + +In this case, upgrading the Rust compiler version later on requires manually +updating this clone. + + +libclang +******** + +``libclang`` (part of LLVM) is used by ``bindgen`` to understand the C code +in the kernel, which means LLVM needs to be installed; like when the kernel +is compiled with ``CC=clang`` or ``LLVM=1``. + +Linux distributions are likely to have a suitable one available, so it is +best to check that first. + +There are also some binaries for several systems and architectures uploaded at: + + https://releases.llvm.org/download.html + +Otherwise, building LLVM takes quite a while, but it is not a complex process: + + https://llvm.org/docs/GettingStarted.html#getting-the-source-code-and-building-llvm + +Please see Documentation/kbuild/llvm.rst for more information and further ways +to fetch pre-built releases and distribution packages. + + +bindgen +******* + +The bindings to the C side of the kernel are generated at build time using +the ``bindgen`` tool. A particular version is required. + +Install it via (note that this will download and build the tool from source):: + + cargo install --locked --version $(scripts/min-tool-version.sh bindgen) bindgen + + +Requirements: Developing +------------------------ + +This section explains how to fetch the tools needed for developing. That is, +they are not needed when just building the kernel. + + +rustfmt +******* + +The ``rustfmt`` tool is used to automatically format all the Rust kernel code, +including the generated C bindings (for details, please see +coding-guidelines.rst). + +If ``rustup`` is being used, its ``default`` profile already installs the tool, +thus nothing needs to be done. If another profile is being used, the component +can be installed manually:: + + rustup component add rustfmt + +The standalone installers also come with ``rustfmt``. + + +clippy +****** + +``clippy`` is a Rust linter. Running it provides extra warnings for Rust code. +It can be run by passing ``CLIPPY=1`` to ``make`` (for details, please see +general-information.rst). + +If ``rustup`` is being used, its ``default`` profile already installs the tool, +thus nothing needs to be done. If another profile is being used, the component +can be installed manually:: + + rustup component add clippy + +The standalone installers also come with ``clippy``. + + +cargo +***** + +``cargo`` is the Rust native build system. It is currently required to run +the tests since it is used to build a custom standard library that contains +the facilities provided by the custom ``alloc`` in the kernel. The tests can +be run using the ``rusttest`` Make target. + +If ``rustup`` is being used, all the profiles already install the tool, +thus nothing needs to be done. + +The standalone installers also come with ``cargo``. + + +rustdoc +******* + +``rustdoc`` is the documentation tool for Rust. It generates pretty HTML +documentation for Rust code (for details, please see +general-information.rst). + +``rustdoc`` is also used to test the examples provided in documented Rust code +(called doctests or documentation tests). The ``rusttest`` Make target uses +this feature. + +If ``rustup`` is being used, all the profiles already install the tool, +thus nothing needs to be done. + +The standalone installers also come with ``rustdoc``. + + +rust-analyzer +************* + +The `rust-analyzer <https://rust-analyzer.github.io/>`_ language server can +be used with many editors to enable syntax highlighting, completion, go to +definition, and other features. + +``rust-analyzer`` needs a configuration file, ``rust-project.json``, which +can be generated by the ``rust-analyzer`` Make target. + + +Configuration +------------- + +``Rust support`` (``CONFIG_RUST``) needs to be enabled in the ``General setup`` +menu. The option is only shown if a suitable Rust toolchain is found (see +above), as long as the other requirements are met. In turn, this will make +visible the rest of options that depend on Rust. + +Afterwards, go to:: + + Kernel hacking + -> Sample kernel code + -> Rust samples + +And enable some sample modules either as built-in or as loadable. + + +Building +-------- + +Building a kernel with a complete LLVM toolchain is the best supported setup +at the moment. That is:: + + make LLVM=1 + +For architectures that do not support a full LLVM toolchain, use:: + + make CC=clang + +Using GCC also works for some configurations, but it is very experimental at +the moment. + + +Hacking +------- + +To dive deeper, take a look at the source code of the samples +at ``samples/rust/``, the Rust support code under ``rust/`` and +the ``Rust hacking`` menu under ``Kernel hacking``. + +If GDB/Binutils is used and Rust symbols are not getting demangled, the reason +is the toolchain does not support Rust's new v0 mangling scheme yet. +There are a few ways out: + + - Install a newer release (GDB >= 10.2, Binutils >= 2.36). + + - Some versions of GDB (e.g. vanilla GDB 10.1) are able to use + the pre-demangled names embedded in the debug info (``CONFIG_DEBUG_INFO``). diff --git a/MAINTAINERS b/MAINTAINERS index dcf675bbfbe9..c82daad19c3d 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -17755,6 +17755,24 @@ F: include/rv/ F: kernel/trace/rv/ F: tools/verification/ +RUST +M: Miguel Ojeda <ojeda@kernel.org> +M: Alex Gaynor <alex.gaynor@gmail.com> +M: Wedson Almeida Filho <wedsonaf@gmail.com> +R: Boqun Feng <boqun.feng@gmail.com> +R: Gary Guo <gary@garyguo.net> +R: Björn Roy Baron <bjorn3_gh@protonmail.com> +L: rust-for-linux@vger.kernel.org +S: Supported +W: https://github.com/Rust-for-Linux/linux +B: https://github.com/Rust-for-Linux/linux/issues +T: git https://github.com/Rust-for-Linux/linux.git rust-next +F: Documentation/rust/ +F: rust/ +F: samples/rust/ +F: scripts/*rust* +K: \b(?i:rust)\b + RXRPC SOCKETS (AF_RXRPC) M: David Howells <dhowells@redhat.com> M: Marc Dionne <marc.dionne@auristor.com> @@ -120,6 +120,15 @@ endif export KBUILD_CHECKSRC +# Enable "clippy" (a linter) as part of the Rust compilation. +# +# Use 'make CLIPPY=1' to enable it. +ifeq ("$(origin CLIPPY)", "command line") + KBUILD_CLIPPY := $(CLIPPY) +endif + +export KBUILD_CLIPPY + # Use make M=dir or set the environment variable KBUILD_EXTMOD to specify the # directory of external module to build. Setting M= takes precedence. ifeq ("$(origin M)", "command line") @@ -270,14 +279,14 @@ no-dot-config-targets := $(clean-targets) \ cscope gtags TAGS tags help% %docs check% coccicheck \ $(version_h) headers headers_% archheaders archscripts \ %asm-generic kernelversion %src-pkg dt_binding_check \ - outputmakefile + outputmakefile rustavailable rustfmt rustfmtcheck # Installation targets should not require compiler. Unfortunately, vdso_install # is an exception where build artifacts may be updated. This must be fixed. no-compiler-targets := $(no-dot-config-targets) install dtbs_install \ headers_install modules_install kernelrelease image_name no-sync-config-targets := $(no-dot-config-targets) %install kernelrelease \ image_name -single-targets := %.a %.i %.ko %.lds %.ll %.lst %.mod %.o %.s %.symtypes %/ +single-targets := %.a %.i %.rsi %.ko %.lds %.ll %.lst %.mod %.o %.s %.symtypes %/ config-build := mixed-build := @@ -439,6 +448,7 @@ else HOSTCC = gcc HOSTCXX = g++ endif +HOSTRUSTC = rustc HOSTPKG_CONFIG = pkg-config KBUILD_USERHOSTCFLAGS := -Wall -Wmissing-prototypes -Wstrict-prototypes \ @@ -447,8 +457,26 @@ KBUILD_USERHOSTCFLAGS := -Wall -Wmissing-prototypes -Wstrict-prototypes \ KBUILD_USERCFLAGS := $(KBUILD_USERHOSTCFLAGS) $(USERCFLAGS) KBUILD_USERLDFLAGS := $(USERLDFLAGS) +# These flags apply to all Rust code in the tree, including the kernel and +# host programs. +export rust_common_flags := --edition=2021 \ + -Zbinary_dep_depinfo=y \ + -Dunsafe_op_in_unsafe_fn -Drust_2018_idioms \ + -Dunreachable_pub -Dnon_ascii_idents \ + -Wmissing_docs \ + -Drustdoc::missing_crate_level_docs \ + -Dclippy::correctness -Dclippy::style \ + -Dclippy::suspicious -Dclippy::complexity \ + -Dclippy::perf \ + -Dclippy::let_unit_value -Dclippy::mut_mut \ + -Dclippy::needless_bitwise_bool \ + -Dclippy::needless_continue \ + -Wclippy::dbg_macro + KBUILD_HOSTCFLAGS := $(KBUILD_USERHOSTCFLAGS) $(HOST_LFS_CFLAGS) $(HOSTCFLAGS) KBUILD_HOSTCXXFLAGS := -Wall -O2 $(HOST_LFS_CFLAGS) $(HOSTCXXFLAGS) +KBUILD_HOSTRUSTFLAGS := $(rust_common_flags) -O -Cstrip=debuginfo \ + -Zallow-features= $(HOSTRUSTFLAGS) KBUILD_HOSTLDFLAGS := $(HOST_LFS_LDFLAGS) $(HOSTLDFLAGS) KBUILD_HOSTLDLIBS := $(HOST_LFS_LIBS) $(HOSTLDLIBS) @@ -473,6 +501,12 @@ OBJDUMP = $(CROSS_COMPILE)objdump READELF = $(CROSS_COMPILE)readelf STRIP = $(CROSS_COMPILE)strip endif +RUSTC = rustc +RUSTDOC = rustdoc +RUSTFMT = rustfmt +CLIPPY_DRIVER = clippy-driver +BINDGEN = bindgen +CARGO = cargo PAHOLE = pahole RESOLVE_BTFIDS = $(objtree)/tools/bpf/resolve_btfids/resolve_btfids LEX = flex @@ -498,9 +532,11 @@ CHECKFLAGS := -D__linux__ -Dlinux -D__STDC__ -Dunix -D__unix__ \ -Wbitwise -Wno-return-void -Wno-unknown-attribute $(CF) NOSTDINC_FLAGS := CFLAGS_MODULE = +RUSTFLAGS_MODULE = AFLAGS_MODULE = LDFLAGS_MODULE = CFLAGS_KERNEL = +RUSTFLAGS_KERNEL = AFLAGS_KERNEL = LDFLAGS_vmlinux = @@ -529,15 +565,43 @@ KBUILD_CFLAGS := -Wall -Wundef -Werror=strict-prototypes -Wno-trigraphs \ -Werror=return-type -Wno-format-security \ -std=gnu11 KBUILD_CPPFLAGS := -D__KERNEL__ +KBUILD_RUSTFLAGS := $(rust_common_flags) \ + --target=$(objtree)/rust/target.json \ + -Cpanic=abort -Cembed-bitcode=n -Clto=n \ + -Cforce-unwind-tables=n -Ccodegen-units=1 \ + -Csymbol-mangling-version=v0 \ + -Crelocation-model=static \ + -Zfunction-sections=n \ + -Dclippy::float_arithmetic + KBUILD_AFLAGS_KERNEL := KBUILD_CFLAGS_KERNEL := +KBUILD_RUSTFLAGS_KERNEL := KBUILD_AFLAGS_MODULE := -DMODULE KBUILD_CFLAGS_MODULE := -DMODULE +KBUILD_RUSTFLAGS_MODULE := --cfg MODULE KBUILD_LDFLAGS_MODULE := KBUILD_LDFLAGS := CLANG_FLAGS := +ifeq ($(KBUILD_CLIPPY),1) + RUSTC_OR_CLIPPY_QUIET := CLIPPY + RUSTC_OR_CLIPPY = $(CLIPPY_DRIVER) +else + RUSTC_OR_CLIPPY_QUIET := RUSTC + RUSTC_OR_CLIPPY = $(RUSTC) +endif + +ifdef RUST_LIB_SRC + export RUST_LIB_SRC +endif + +# Allows the usage of unstable features in stable compilers. +export RUSTC_BOOTSTRAP := 1 + export ARCH SRCARCH CONFIG_SHELL BASH HOSTCC KBUILD_HOSTCFLAGS CROSS_COMPILE LD CC HOSTPKG_CONFIG +export RUSTC RUSTDOC RUSTFMT RUSTC_OR_CLIPPY_QUIET RUSTC_OR_CLIPPY BINDGEN CARGO +export HOSTRUSTC KBUILD_HOSTRUSTFLAGS export CPP AR NM STRIP OBJCOPY OBJDUMP READELF PAHOLE RESOLVE_BTFIDS LEX YACC AWK INSTALLKERNEL export PERL PYTHON3 CHECK CHECKFLAGS MAKE UTS_MACHINE HOSTCXX export KGZIP KBZIP2 KLZOP LZMA LZ4 XZ ZSTD @@ -546,9 +610,10 @@ export KBUILD_USERCFLAGS KBUILD_USERLDFLAGS export KBUILD_CPPFLAGS NOSTDINC_FLAGS LINUXINCLUDE OBJCOPYFLAGS KBUILD_LDFLAGS export KBUILD_CFLAGS CFLAGS_KERNEL CFLAGS_MODULE +export KBUILD_RUSTFLAGS RUSTFLAGS_KERNEL RUSTFLAGS_MODULE export KBUILD_AFLAGS AFLAGS_KERNEL AFLAGS_MODULE -export KBUILD_AFLAGS_MODULE KBUILD_CFLAGS_MODULE KBUILD_LDFLAGS_MODULE -export KBUILD_AFLAGS_KERNEL KBUILD_CFLAGS_KERNEL +export KBUILD_AFLAGS_MODULE KBUILD_CFLAGS_MODULE KBUILD_RUSTFLAGS_MODULE KBUILD_LDFLAGS_MODULE +export KBUILD_AFLAGS_KERNEL KBUILD_CFLAGS_KERNEL KBUILD_RUSTFLAGS_KERNEL export PAHOLE_FLAGS # Files to ignore in find ... statements @@ -729,7 +794,7 @@ $(KCONFIG_CONFIG): # # Do not use $(call cmd,...) here. That would suppress prompts from syncconfig, # so you cannot notice that Kconfig is waiting for the user input. -%/config/auto.conf %/config/auto.conf.cmd %/generated/autoconf.h: $(KCONFIG_CONFIG) +%/config/auto.conf %/config/auto.conf.cmd %/generated/autoconf.h %/generated/rustc_cfg: $(KCONFIG_CONFIG) $(Q)$(kecho) " SYNC $@" $(Q)$(MAKE) -f $(srctree)/Makefile syncconfig else # !may-sync-config @@ -758,10 +823,17 @@ KBUILD_CFLAGS += $(call cc-disable-warning, address-of-packed-member) ifdef CONFIG_CC_OPTIMIZE_FOR_PERFORMANCE KBUILD_CFLAGS += -O2 +KBUILD_RUSTFLAGS += -Copt-level=2 else ifdef CONFIG_CC_OPTIMIZE_FOR_SIZE KBUILD_CFLAGS += -Os +KBUILD_RUSTFLAGS += -Copt-level=s endif +# Always set `debug-assertions` and `overflow-checks` because their default +# depends on `opt-level` and `debug-assertions`, respectively. +KBUILD_RUSTFLAGS += -Cdebug-assertions=$(if $(CONFIG_RUST_DEBUG_ASSERTIONS),y,n) +KBUILD_RUSTFLAGS += -Coverflow-checks=$(if $(CONFIG_RUST_OVERFLOW_CHECKS),y,n) + # Tell gcc to never replace conditional load with a non-conditional one ifdef CONFIG_CC_IS_GCC # gcc-10 renamed --param=allow-store-data-races=0 to @@ -792,6 +864,9 @@ KBUILD_CFLAGS-$(CONFIG_WERROR) += -Werror KBUILD_CFLAGS-$(CONFIG_CC_NO_ARRAY_BOUNDS) += -Wno-array-bounds KBUILD_CFLAGS += $(KBUILD_CFLAGS-y) $(CONFIG_CC_IMPLICIT_FALLTHROUGH) +KBUILD_RUSTFLAGS-$(CONFIG_WERROR) += -Dwarnings +KBUILD_RUSTFLAGS += $(KBUILD_RUSTFLAGS-y) + ifdef CONFIG_CC_IS_CLANG KBUILD_CPPFLAGS += -Qunused-arguments # The kernel builds with '-std=gnu11' so use of GNU extensions is acceptable. @@ -812,12 +887,15 @@ KBUILD_CFLAGS += $(call cc-disable-warning, dangling-pointer) ifdef CONFIG_FRAME_POINTER KBUILD_CFLAGS += -fno-omit-frame-pointer -fno-optimize-sibling-calls +KBUILD_RUSTFLAGS += -Cforce-frame-pointers=y else # Some targets (ARM with Thumb2, for example), can't be built with frame # pointers. For those, we don't have FUNCTION_TRACER automatically # select FRAME_POINTER. However, FUNCTION_TRACER adds -pg, and this is # incompatible with -fomit-frame-pointer with current GCC, so we don't use # -fomit-frame-pointer with FUNCTION_TRACER. +# In the Rust target specification, "frame-pointer" is set explicitly +# to "may-omit". ifndef CONFIG_FUNCTION_TRACER KBUILD_CFLAGS += -fomit-frame-pointer endif @@ -882,8 +960,10 @@ ifdef CONFIG_DEBUG_SECTION_MISMATCH KBUILD_CFLAGS += -fno-inline-functions-called-once endif +# `rustc`'s `-Zfunction-sections` applies to data too (as of 1.59.0). ifdef CONFIG_LD_DEAD_CODE_DATA_ELIMINATION KBUILD_CFLAGS_KERNEL += -ffunction-sections -fdata-sections +KBUILD_RUSTFLAGS_KERNEL += -Zfunction-sections=y LDFLAGS_vmlinux += --gc-sections endif @@ -1026,10 +1106,11 @@ include $(addprefix $(srctree)/, $(include-y)) # Do not add $(call cc-option,...) below this line. When you build the kernel # from the clean source tree, the GCC plugins do not exist at this point. -# Add user supplied CPPFLAGS, AFLAGS and CFLAGS as the last assignments +# Add user supplied CPPFLAGS, AFLAGS, CFLAGS and RUSTFLAGS as the last assignments KBUILD_CPPFLAGS += $(KCPPFLAGS) KBUILD_AFLAGS += $(KAFLAGS) KBUILD_CFLAGS += $(KCFLAGS) +KBUILD_RUSTFLAGS += $(KRUSTFLAGS) KBUILD_LDFLAGS_MODULE += --build-id=sha1 LDFLAGS_vmlinux += --build-id=sha1 @@ -1104,6 +1185,7 @@ ifeq ($(KBUILD_EXTMOD),) core-y += kernel/ certs/ mm/ fs/ ipc/ security/ crypto/ core-$(CONFIG_BLOCK) += block/ core-$(CONFIG_IO_URING) += io_uring/ +core-$(CONFIG_RUST) += rust/ vmlinux-dirs := $(patsubst %/,%,$(filter %/, \ $(core-y) $(core-m) $(drivers-y) $(drivers-m) \ @@ -1206,6 +1288,10 @@ prepare0: archprepare # All the preparing.. prepare: prepare0 +ifdef CONFIG_RUST + $(Q)$(CONFIG_SHELL) $(srctree)/scripts/rust_is_available.sh -v + $(Q)$(MAKE) $(build)=rust +endif PHONY += remove-stale-files remove-stale-files: @@ -1499,7 +1585,7 @@ endif # CONFIG_MODULES # Directories & files removed with 'make clean' CLEAN_FILES += include/ksym vmlinux.symvers modules-only.symvers \ modules.builtin modules.builtin.modinfo modules.nsdeps \ - compile_commands.json .thinlto-cache + compile_commands.json .thinlto-cache rust/test rust/doc # Directories & files removed with 'make mrproper' MRPROPER_FILES += include/config include/generated \ @@ -1510,7 +1596,8 @@ MRPROPER_FILES += include/config include/generated \ certs/signing_key.pem \ certs/x509.genkey \ vmlinux-gdb.py \ - *.spec + *.spec \ + rust/target.json rust/libmacros.so # clean - Delete most, but leave enough to build external modules # @@ -1535,6 +1622,9 @@ $(mrproper-dirs): mrproper: clean $(mrproper-dirs) $(call cmd,rmfiles) + @find . $(RCS_FIND_IGNORE) \ + \( -name '*.rmeta' \) \ + -type f -print | xargs rm -f # distclean # @@ -1622,6 +1712,24 @@ help: @echo ' kselftest-merge - Merge all the config dependencies of' @echo ' kselftest to existing .config.' @echo '' + @echo 'Rust targets:' + @echo ' rustavailable - Checks whether the Rust toolchain is' + @echo ' available and, if not, explains why.' + @echo ' rustfmt - Reformat all the Rust code in the kernel' + @echo ' rustfmtcheck - Checks if all the Rust code in the kernel' + @echo ' is formatted, printing a diff otherwise.' + @echo ' rustdoc - Generate Rust documentation' + @echo ' (requires kernel .config)' + @echo ' rusttest - Runs the Rust tests' + @echo ' (requires kernel .config; downloads external repos)' + @echo ' rust-analyzer - Generate rust-project.json rust-analyzer support file' + @echo ' (requires kernel .config)' + @echo ' dir/file.[os] - Build specified target only' + @echo ' dir/file.rsi - Build macro expanded source, similar to C preprocessing.' + @echo ' Run with RUSTFMT=n to skip reformatting if needed.' + @echo ' The output is not intended to be compilable.' + @echo ' dir/file.ll - Build the LLVM assembly file' + @echo '' @$(if $(dtstree), \ echo 'Devicetree:'; \ echo '* dtbs - Build device tree blobs for enabled boards'; \ @@ -1694,6 +1802,52 @@ PHONY += $(DOC_TARGETS) $(DOC_TARGETS): $(Q)$(MAKE) $(build)=Documentation $@ + +# Rust targets +# --------------------------------------------------------------------------- + +# "Is Rust available?" target +PHONY += rustavailable +rustavailable: + $(Q)$(CONFIG_SHELL) $(srctree)/scripts/rust_is_available.sh -v && echo "Rust is available!" + +# Documentation target +# +# Using the singular to avoid running afoul of `no-dot-config-targets`. +PHONY += rustdoc +rustdoc: prepare + $(Q)$(MAKE) $(build)=rust $@ + +# Testing target +PHONY += rusttest +rusttest: prepare + $(Q)$(MAKE) $(build)=rust $@ + +# Formatting targets +PHONY += rustfmt rustfmtcheck + +# We skip `rust/alloc` since we want to minimize the diff w.r.t. upstream. +# +# We match using absolute paths since `find` does not resolve them +# when matching, which is a problem when e.g. `srctree` is `..`. +# We `grep` afterwards in order to remove the directory entry itself. +rustfmt: + $(Q)find $(abs_srctree) -type f -name '*.rs' \ + -o -path $(abs_srctree)/rust/alloc -prune \ + -o -path $(abs_objtree)/rust/test -prune \ + | grep -Fv $(abs_srctree)/rust/alloc \ + | grep -Fv $(abs_objtree)/rust/test \ + | grep -Fv generated \ + | xargs $(RUSTFMT) $(rustfmt_flags) + +rustfmtcheck: rustfmt_flags = --check +rustfmtcheck: rustfmt + +# IDE support targets +PHONY += rust-analyzer +rust-analyzer: + $(Q)$(MAKE) $(build)=rust $@ + # Misc # --------------------------------------------------------------------------- @@ -1861,7 +2015,7 @@ $(clean-dirs): clean: $(clean-dirs) $(call cmd,rmfiles) @find $(or $(KBUILD_EXTMOD), .) $(RCS_FIND_IGNORE) \ - \( -name '*.[aios]' -o -name '*.ko' -o -name '.*.cmd' \ + \( -name '*.[aios]' -o -name '*.rsi' -o -name '*.ko' -o -name '.*.cmd' \ -o -name '*.ko.*' \ -o -name '*.dtb' -o -name '*.dtbo' -o -name '*.dtb.S' -o -name '*.dt.yaml' \ -o -name '*.dwo' -o -name '*.lst' \ diff --git a/arch/Kconfig b/arch/Kconfig index 8b311e400ec1..d9b4ae0fc805 100644 --- a/arch/Kconfig +++ b/arch/Kconfig @@ -355,6 +355,12 @@ config HAVE_RSEQ This symbol should be selected by an architecture if it supports an implementation of restartable sequences. +config HAVE_RUST + bool + help + This symbol should be selected by an architecture if it + supports Rust. + config HAVE_FUNCTION_ARG_ACCESS_API bool help diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index f9920f1341c8..3ca198742b10 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -257,6 +257,7 @@ config X86 select HAVE_STATIC_CALL_INLINE if HAVE_OBJTOOL select HAVE_PREEMPT_DYNAMIC_CALL select HAVE_RSEQ + select HAVE_RUST if X86_64 select HAVE_SYSCALL_TRACEPOINTS select HAVE_UACCESS_VALIDATION if HAVE_OBJTOOL select HAVE_UNSTABLE_SCHED_CLOCK diff --git a/arch/x86/Makefile b/arch/x86/Makefile index bafbd905e6e7..2d7e640674c6 100644 --- a/arch/x86/Makefile +++ b/arch/x86/Makefile @@ -68,6 +68,7 @@ export BITS # https://gcc.gnu.org/bugzilla/show_bug.cgi?id=53383 # KBUILD_CFLAGS += -mno-sse -mno-mmx -mno-sse2 -mno-3dnow -mno-avx +KBUILD_RUSTFLAGS += -Ctarget-feature=-sse,-sse2,-sse3,-ssse3,-sse4.1,-sse4.2,-avx,-avx2 ifeq ($(CONFIG_X86_KERNEL_IBT),y) # @@ -155,8 +156,17 @@ else cflags-$(CONFIG_GENERIC_CPU) += -mtune=generic KBUILD_CFLAGS += $(cflags-y) + rustflags-$(CONFIG_MK8) += -Ctarget-cpu=k8 + rustflags-$(CONFIG_MPSC) += -Ctarget-cpu=nocona + rustflags-$(CONFIG_MCORE2) += -Ctarget-cpu=core2 + rustflags-$(CONFIG_MATOM) += -Ctarget-cpu=atom + rustflags-$(CONFIG_GENERIC_CPU) += -Ztune-cpu=generic + KBUILD_RUSTFLAGS += $(rustflags-y) + KBUILD_CFLAGS += -mno-red-zone KBUILD_CFLAGS += -mcmodel=kernel + KBUILD_RUSTFLAGS += -Cno-redzone=y + KBUILD_RUSTFLAGS += -Ccode-model=kernel endif # diff --git a/include/linux/compiler_types.h b/include/linux/compiler_types.h index 4f2a819fd60a..50b3f6b9502e 100644 --- a/include/linux/compiler_types.h +++ b/include/linux/compiler_types.h @@ -4,8 +4,12 @@ #ifndef __ASSEMBLY__ +/* + * Skipped when running bindgen due to a libclang issue; + * see https://github.com/rust-lang/rust-bindgen/issues/2244. + */ #if defined(CONFIG_DEBUG_INFO_BTF) && defined(CONFIG_PAHOLE_HAS_BTF_TAG) && \ - __has_attribute(btf_type_tag) + __has_attribute(btf_type_tag) && !defined(__BINDGEN__) # define BTF_TYPE_TAG(value) __attribute__((btf_type_tag(#value))) #else # define BTF_TYPE_TAG(value) /* nothing */ diff --git a/include/linux/kallsyms.h b/include/linux/kallsyms.h index ad39636e0c3f..649faac31ddb 100644 --- a/include/linux/kallsyms.h +++ b/include/linux/kallsyms.h @@ -15,7 +15,7 @@ #include <asm/sections.h> -#define KSYM_NAME_LEN 128 +#define KSYM_NAME_LEN 512 #define KSYM_SYMBOL_LEN (sizeof("%s+%#lx/%#lx [%s %s]") + \ (KSYM_NAME_LEN - 1) + \ 2*(BITS_PER_LONG*3/10) + (MODULE_NAME_LEN - 1) + \ diff --git a/init/Kconfig b/init/Kconfig index 532362fcfe31..a078cb026523 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -60,6 +60,17 @@ config LLD_VERSION default $(ld-version) if LD_IS_LLD default 0 +config RUST_IS_AVAILABLE + def_bool $(success,$(srctree)/scripts/rust_is_available.sh) + help + This shows whether a suitable Rust toolchain is available (found). + + Please see Documentation/rust/quick-start.rst for instructions on how + to satify the build requirements of Rust support. + + In particular, the Makefile target 'rustavailable' is useful to check + why the Rust toolchain is not being detected. + config CC_CAN_LINK bool default $(success,$(srctree)/scripts/cc-can-link.sh $(CC) $(CLANG_FLAGS) $(USERCFLAGS) $(USERLDFLAGS) $(m64-flag)) if 64BIT @@ -147,7 +158,8 @@ config WERROR default COMPILE_TEST help A kernel build should not cause any compiler warnings, and this - enables the '-Werror' flag to enforce that rule by default. + enables the '-Werror' (for C) and '-Dwarnings' (for Rust) flags + to enforce that rule by default. However, if you have a new (or very old) compiler with odd and unusual warnings, or you have some architecture with problems, @@ -1899,6 +1911,38 @@ config PROFILING Say Y here to enable the extended profiling support mechanisms used by profilers. +config RUST + bool "Rust support" + depends on HAVE_RUST + depends on RUST_IS_AVAILABLE + depends on !MODVERSIONS + depends on !GCC_PLUGINS + depends on !RANDSTRUCT + depends on !DEBUG_INFO_BTF + select CONSTRUCTORS + help + Enables Rust support in the kernel. + + This allows other Rust-related options, like drivers written in Rust, + to be selected. + + It is also required to be able to load external kernel modules + written in Rust. + + See Documentation/rust/ for more information. + + If unsure, say N. + +config RUSTC_VERSION_TEXT + string + depends on RUST + default $(shell,command -v $(RUSTC) >/dev/null 2>&1 && $(RUSTC) --version || echo n) + +config BINDGEN_VERSION_TEXT + string + depends on RUST + default $(shell,command -v $(BINDGEN) >/dev/null 2>&1 && $(BINDGEN) --version || echo n) + # # Place an empty function call at each tracepoint site. Can be # dynamically changed for a probe function. diff --git a/kernel/configs/rust.config b/kernel/configs/rust.config new file mode 100644 index 000000000000..38a7c5362c9c --- /dev/null +++ b/kernel/configs/rust.config @@ -0,0 +1 @@ +CONFIG_RUST=y diff --git a/kernel/kallsyms.c b/kernel/kallsyms.c index 3e7e2c2ad2f7..fc5e26348d25 100644 --- a/kernel/kallsyms.c +++ b/kernel/kallsyms.c @@ -50,12 +50,20 @@ static unsigned int kallsyms_expand_symbol(unsigned int off, data = &kallsyms_names[off]; len = *data; data++; + off++; + + /* If MSB is 1, it is a "big" symbol, so needs an additional byte. */ + if ((len & 0x80) != 0) { + len = (len & 0x7F) | (*data << 7); + data++; + off++; + } /* * Update the offset to return the offset for the next symbol on * the compressed stream. */ - off += len + 1; + off += len; /* * For every byte on the compressed symbol data, copy the table @@ -108,7 +116,7 @@ static char kallsyms_get_symbol_type(unsigned int off) static unsigned int get_symbol_offset(unsigned long pos) { const u8 *name; - int i; + int i, len; /* * Use the closest marker we have. We have markers every 256 positions, @@ -122,8 +130,18 @@ static unsigned int get_symbol_offset(unsigned long pos) * so we just need to add the len to the current pointer for every * symbol we wish to skip. */ - for (i = 0; i < (pos & 0xFF); i++) - name = name + (*name) + 1; + for (i = 0; i < (pos & 0xFF); i++) { + len = *name; + + /* + * If MSB is 1, it is a "big" symbol, so we need to look into + * the next byte (and skip it, too). + */ + if ((len & 0x80) != 0) + len = ((len & 0x7F) | (name[1] << 7)) + 1; + + name = name + len + 1; + } return name - kallsyms_names; } diff --git a/kernel/livepatch/core.c b/kernel/livepatch/core.c index bc475e62279d..ec06ce59d728 100644 --- a/kernel/livepatch/core.c +++ b/kernel/livepatch/core.c @@ -213,7 +213,7 @@ static int klp_resolve_symbols(Elf_Shdr *sechdrs, const char *strtab, * we use the smallest/strictest upper bound possible (56, based on * the current definition of MODULE_NAME_LEN) to prevent overflows. */ - BUILD_BUG_ON(MODULE_NAME_LEN < 56 || KSYM_NAME_LEN != 128); + BUILD_BUG_ON(MODULE_NAME_LEN < 56 || KSYM_NAME_LEN != 512); relas = (Elf_Rela *) relasec->sh_addr; /* For each rela in this klp relocation section */ @@ -227,7 +227,7 @@ static int klp_resolve_symbols(Elf_Shdr *sechdrs, const char *strtab, /* Format: .klp.sym.sym_objname.sym_name,sympos */ cnt = sscanf(strtab + sym->st_name, - ".klp.sym.%55[^.].%127[^,],%lu", + ".klp.sym.%55[^.].%511[^,],%lu", sym_objname, sym_name, &sympos); if (cnt != 3) { pr_err("symbol %s has an incorrectly formatted name\n", diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index d3e5f36bb01e..e62271da937f 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -2710,6 +2710,40 @@ config HYPERV_TESTING endmenu # "Kernel Testing and Coverage" +menu "Rust hacking" + +config RUST_DEBUG_ASSERTIONS + bool "Debug assertions" + depends on RUST + help + Enables rustc's `-Cdebug-assertions` codegen option. + + This flag lets you turn `cfg(debug_assertions)` conditional + compilation on or off. This can be used to enable extra debugging + code in development but not in production. For example, it controls + the behavior of the standard library's `debug_assert!` macro. + + Note that this will apply to all Rust code, including `core`. + + If unsure, say N. + +config RUST_OVERFLOW_CHECKS + bool "Overflow checks" + default y + depends on RUST + help + Enables rustc's `-Coverflow-checks` codegen option. + + This flag allows you to control the behavior of runtime integer + overflow. When overflow-checks are enabled, a Rust panic will occur + on overflow. + + Note that this will apply to all Rust code, including `core`. + + If unsure, say Y. + +endmenu # "Rust" + source "Documentation/Kconfig" endmenu # Kernel hacking diff --git a/lib/vsprintf.c b/lib/vsprintf.c index 3c1853a9d1c0..c414a8d9f1ea 100644 --- a/lib/vsprintf.c +++ b/lib/vsprintf.c @@ -2246,6 +2246,9 @@ int __init no_hash_pointers_enable(char *str) } early_param("no_hash_pointers", no_hash_pointers_enable); +/* Used for Rust formatting ('%pA'). */ +char *rust_fmt_argument(char *buf, char *end, void *ptr); + /* * Show a '%p' thing. A kernel extension is that the '%p' is followed * by an extra set of alphanumeric characters that are extended format @@ -2372,6 +2375,10 @@ early_param("no_hash_pointers", no_hash_pointers_enable); * * Note: The default behaviour (unadorned %p) is to hash the address, * rendering it useful as a unique identifier. + * + * There is also a '%pA' format specifier, but it is only intended to be used + * from Rust code to format core::fmt::Arguments. Do *not* use it from C. + * See rust/kernel/print.rs for details. */ static noinline_for_stack char *pointer(const char *fmt, char *buf, char *end, void *ptr, @@ -2444,6 +2451,12 @@ char *pointer(const char *fmt, char *buf, char *end, void *ptr, return device_node_string(buf, end, ptr, spec, fmt + 1); case 'f': return fwnode_string(buf, end, ptr, spec, fmt + 1); + case 'A': + if (!IS_ENABLED(CONFIG_RUST)) { + WARN_ONCE(1, "Please remove %%pA from non-Rust code\n"); + return error_string(buf, end, "(%pA?)", spec); + } + return rust_fmt_argument(buf, end, ptr); case 'x': return pointer_string(buf, end, ptr, spec); case 'e': diff --git a/rust/.gitignore b/rust/.gitignore new file mode 100644 index 000000000000..9bd1af8e05a1 --- /dev/null +++ b/rust/.gitignore @@ -0,0 +1,8 @@ +# SPDX-License-Identifier: GPL-2.0 + +target.json +bindings_generated.rs +bindings_helpers_generated.rs +exports_*_generated.h +doc/ +test/ diff --git a/rust/Makefile b/rust/Makefile new file mode 100644 index 000000000000..7700d3853404 --- /dev/null +++ b/rust/Makefile @@ -0,0 +1,381 @@ +# SPDX-License-Identifier: GPL-2.0 + +always-$(CONFIG_RUST) += target.json +no-clean-files += target.json + +obj-$(CONFIG_RUST) += core.o compiler_builtins.o +always-$(CONFIG_RUST) += exports_core_generated.h + +# Missing prototypes are expected in the helpers since these are exported +# for Rust only, thus there is no header nor prototypes. +obj-$(CONFIG_RUST) += helpers.o +CFLAGS_REMOVE_helpers.o = -Wmissing-prototypes -Wmissing-declarations + +always-$(CONFIG_RUST) += libmacros.so +no-clean-files += libmacros.so + +always-$(CONFIG_RUST) += bindings/bindings_generated.rs bindings/bindings_helpers_generated.rs +obj-$(CONFIG_RUST) += alloc.o bindings.o kernel.o +always-$(CONFIG_RUST) += exports_alloc_generated.h exports_bindings_generated.h \ + exports_kernel_generated.h + +obj-$(CONFIG_RUST) += exports.o + +# Avoids running `$(RUSTC)` for the sysroot when it may not be available. +ifdef CONFIG_RUST + +# `$(rust_flags)` is passed in case the user added `--sysroot`. +rustc_sysroot := $(shell $(RUSTC) $(rust_flags) --print sysroot) +rustc_host_target := $(shell $(RUSTC) --version --verbose | grep -F 'host: ' | cut -d' ' -f2) +RUST_LIB_SRC ?= $(rustc_sysroot)/lib/rustlib/src/rust/library + +ifeq ($(quiet),silent_) +cargo_quiet=-q +rust_test_quiet=-q +rustdoc_test_quiet=--test-args -q +else ifeq ($(quiet),quiet_) +rust_test_quiet=-q +rustdoc_test_quiet=--test-args -q +else +cargo_quiet=--verbose +endif + +core-cfgs = \ + --cfg no_fp_fmt_parse + +alloc-cfgs = \ + --cfg no_fmt \ + --cfg no_global_oom_handling \ + --cfg no_macros \ + --cfg no_rc \ + --cfg no_str \ + --cfg no_string \ + --cfg no_sync \ + --cfg no_thin + +quiet_cmd_rustdoc = RUSTDOC $(if $(rustdoc_host),H, ) $< + cmd_rustdoc = \ + OBJTREE=$(abspath $(objtree)) \ + $(RUSTDOC) $(if $(rustdoc_host),$(rust_common_flags),$(rust_flags)) \ + $(rustc_target_flags) -L$(objtree)/$(obj) \ + --output $(objtree)/$(obj)/doc \ + --crate-name $(subst rustdoc-,,$@) \ + @$(objtree)/include/generated/rustc_cfg $< + +# The `html_logo_url` and `html_favicon_url` forms of the `doc` attribute +# can be used to specify a custom logo. However: +# - The given value is used as-is, thus it cannot be relative or a local file +# (unlike the non-custom case) since the generated docs have subfolders. +# - It requires adding it to every crate. +# - It requires changing `core` which comes from the sysroot. +# +# Using `-Zcrate-attr` would solve the last two points, but not the first. +# The https://github.com/rust-lang/rfcs/pull/3226 RFC suggests two new +# command-like flags to solve the issue. Meanwhile, we use the non-custom case +# and then retouch the generated files. +rustdoc: rustdoc-core rustdoc-macros rustdoc-compiler_builtins \ + rustdoc-alloc rustdoc-kernel + $(Q)cp $(srctree)/Documentation/images/logo.svg $(objtree)/$(obj)/doc + $(Q)cp $(srctree)/Documentation/images/COPYING-logo $(objtree)/$(obj)/doc + $(Q)find $(objtree)/$(obj)/doc -name '*.html' -type f -print0 | xargs -0 sed -Ei \ + -e 's:rust-logo\.svg:logo.svg:g' \ + -e 's:rust-logo\.png:logo.svg:g' \ + -e 's:favicon\.svg:logo.svg:g' \ + -e 's:<link rel="alternate icon" type="image/png" href="[./]*favicon-(16x16|32x32)\.png">::g' + $(Q)echo '.logo-container > img { object-fit: contain; }' \ + >> $(objtree)/$(obj)/doc/rustdoc.css + +rustdoc-macros: private rustdoc_host = yes +rustdoc-macros: private rustc_target_flags = --crate-type proc-macro \ + --extern proc_macro +rustdoc-macros: $(src)/macros/lib.rs FORCE + $(call if_changed,rustdoc) + +rustdoc-core: private rustc_target_flags = $(core-cfgs) +rustdoc-core: $(RUST_LIB_SRC)/core/src/lib.rs FORCE + $(call if_changed,rustdoc) + +rustdoc-compiler_builtins: $(src)/compiler_builtins.rs rustdoc-core FORCE + $(call if_changed,rustdoc) + +# We need to allow `rustdoc::broken_intra_doc_links` because some +# `no_global_oom_handling` functions refer to non-`no_global_oom_handling` +# functions. Ideally `rustdoc` would have a way to distinguish broken links +# due to things that are "configured out" vs. entirely non-existing ones. +rustdoc-alloc: private rustc_target_flags = $(alloc-cfgs) \ + -Arustdoc::broken_intra_doc_links +rustdoc-alloc: $(src)/alloc/lib.rs rustdoc-core rustdoc-compiler_builtins FORCE + $(call if_changed,rustdoc) + +rustdoc-kernel: private rustc_target_flags = --extern alloc \ + --extern macros=$(objtree)/$(obj)/libmacros.so \ + --extern bindings +rustdoc-kernel: $(src)/kernel/lib.rs rustdoc-core rustdoc-macros \ + rustdoc-compiler_builtins rustdoc-alloc $(obj)/libmacros.so \ + $(obj)/bindings.o FORCE + $(call if_changed,rustdoc) + +quiet_cmd_rustc_test_library = RUSTC TL $< + cmd_rustc_test_library = \ + OBJTREE=$(abspath $(objtree)) \ + $(RUSTC) $(rust_common_flags) \ + @$(objtree)/include/generated/rustc_cfg $(rustc_target_flags) \ + --crate-type $(if $(rustc_test_library_proc),proc-macro,rlib) \ + --out-dir $(objtree)/$(obj)/test --cfg testlib \ + --sysroot $(objtree)/$(obj)/test/sysroot \ + -L$(objtree)/$(obj)/test \ + --crate-name $(subst rusttest-,,$(subst rusttestlib-,,$@)) $< + +rusttestlib-macros: private rustc_target_flags = --extern proc_macro +rusttestlib-macros: private rustc_test_library_proc = yes +rusttestlib-macros: $(src)/macros/lib.rs rusttest-prepare FORCE + $(call if_changed,rustc_test_library) + +rusttestlib-bindings: $(src)/bindings/lib.rs rusttest-prepare FORCE + $(call if_changed,rustc_test_library) + +quiet_cmd_rustdoc_test = RUSTDOC T $< + cmd_rustdoc_test = \ + OBJTREE=$(abspath $(objtree)) \ + $(RUSTDOC) --test $(rust_common_flags) \ + @$(objtree)/include/generated/rustc_cfg \ + $(rustc_target_flags) $(rustdoc_test_target_flags) \ + --sysroot $(objtree)/$(obj)/test/sysroot $(rustdoc_test_quiet) \ + -L$(objtree)/$(obj)/test --output $(objtree)/$(obj)/doc \ + --crate-name $(subst rusttest-,,$@) $< + +# We cannot use `-Zpanic-abort-tests` because some tests are dynamic, +# so for the moment we skip `-Cpanic=abort`. +quiet_cmd_rustc_test = RUSTC T $< + cmd_rustc_test = \ + OBJTREE=$(abspath $(objtree)) \ + $(RUSTC) --test $(rust_common_flags) \ + @$(objtree)/include/generated/rustc_cfg \ + $(rustc_target_flags) --out-dir $(objtree)/$(obj)/test \ + --sysroot $(objtree)/$(obj)/test/sysroot \ + -L$(objtree)/$(obj)/test \ + --crate-name $(subst rusttest-,,$@) $<; \ + $(objtree)/$(obj)/test/$(subst rusttest-,,$@) $(rust_test_quiet) \ + $(rustc_test_run_flags) + +rusttest: rusttest-macros rusttest-kernel + +# This prepares a custom sysroot with our custom `alloc` instead of +# the standard one. +# +# This requires several hacks: +# - Unlike `core` and `alloc`, `std` depends on more than a dozen crates, +# including third-party crates that need to be downloaded, plus custom +# `build.rs` steps. Thus hardcoding things here is not maintainable. +# - `cargo` knows how to build the standard library, but it is an unstable +# feature so far (`-Zbuild-std`). +# - `cargo` only considers the use case of building the standard library +# to use it in a given package. Thus we need to create a dummy package +# and pick the generated libraries from there. +# - Since we only keep a subset of upstream `alloc` in-tree, we need +# to recreate it on the fly by putting our sources on top. +# - The usual ways of modifying the dependency graph in `cargo` do not seem +# to apply for the `-Zbuild-std` steps, thus we have to mislead it +# by modifying the sources in the sysroot. +# - To avoid messing with the user's Rust installation, we create a clone +# of the sysroot. However, `cargo` ignores `RUSTFLAGS` in the `-Zbuild-std` +# steps, thus we use a wrapper binary passed via `RUSTC` to pass the flag. +# +# In the future, we hope to avoid the whole ordeal by either: +# - Making the `test` crate not depend on `std` (either improving upstream +# or having our own custom crate). +# - Making the tests run in kernel space (requires the previous point). +# - Making `std` and friends be more like a "normal" crate, so that +# `-Zbuild-std` and related hacks are not needed. +quiet_cmd_rustsysroot = RUSTSYSROOT + cmd_rustsysroot = \ + rm -rf $(objtree)/$(obj)/test; \ + mkdir -p $(objtree)/$(obj)/test; \ + cp -a $(rustc_sysroot) $(objtree)/$(obj)/test/sysroot; \ + cp -r $(srctree)/$(src)/alloc/* \ + $(objtree)/$(obj)/test/sysroot/lib/rustlib/src/rust/library/alloc/src; \ + echo '\#!/bin/sh' > $(objtree)/$(obj)/test/rustc_sysroot; \ + echo "$(RUSTC) --sysroot=$(abspath $(objtree)/$(obj)/test/sysroot) \"\$$@\"" \ + >> $(objtree)/$(obj)/test/rustc_sysroot; \ + chmod u+x $(objtree)/$(obj)/test/rustc_sysroot; \ + $(CARGO) -q new $(objtree)/$(obj)/test/dummy; \ + RUSTC=$(objtree)/$(obj)/test/rustc_sysroot $(CARGO) $(cargo_quiet) \ + test -Zbuild-std --target $(rustc_host_target) \ + --manifest-path $(objtree)/$(obj)/test/dummy/Cargo.toml; \ + rm $(objtree)/$(obj)/test/sysroot/lib/rustlib/$(rustc_host_target)/lib/*; \ + cp $(objtree)/$(obj)/test/dummy/target/$(rustc_host_target)/debug/deps/* \ + $(objtree)/$(obj)/test/sysroot/lib/rustlib/$(rustc_host_target)/lib + +rusttest-prepare: FORCE + $(call if_changed,rustsysroot) + +rusttest-macros: private rustc_target_flags = --extern proc_macro +rusttest-macros: private rustdoc_test_target_flags = --crate-type proc-macro +rusttest-macros: $(src)/macros/lib.rs rusttest-prepare FORCE + $(call if_changed,rustc_test) + $(call if_changed,rustdoc_test) + +rusttest-kernel: private rustc_target_flags = --extern alloc \ + --extern macros --extern bindings +rusttest-kernel: $(src)/kernel/lib.rs rusttest-prepare \ + rusttestlib-macros rusttestlib-bindings FORCE + $(call if_changed,rustc_test) + $(call if_changed,rustc_test_library) + +filechk_rust_target = $(objtree)/scripts/generate_rust_target < $< + +$(obj)/target.json: $(objtree)/include/config/auto.conf FORCE + $(call filechk,rust_target) + +ifdef CONFIG_CC_IS_CLANG +bindgen_c_flags = $(c_flags) +else +# bindgen relies on libclang to parse C. Ideally, bindgen would support a GCC +# plugin backend and/or the Clang driver would be perfectly compatible with GCC. +# +# For the moment, here we are tweaking the flags on the fly. This is a hack, +# and some kernel configurations may not work (e.g. `GCC_PLUGIN_RANDSTRUCT` +# if we end up using one of those structs). +bindgen_skip_c_flags := -mno-fp-ret-in-387 -mpreferred-stack-boundary=% \ + -mskip-rax-setup -mgeneral-regs-only -msign-return-address=% \ + -mindirect-branch=thunk-extern -mindirect-branch-register \ + -mfunction-return=thunk-extern -mrecord-mcount -mabi=lp64 \ + -mindirect-branch-cs-prefix -mstack-protector-guard% -mtraceback=no \ + -mno-pointers-to-nested-functions -mno-string \ + -mno-strict-align -mstrict-align \ + -fconserve-stack -falign-jumps=% -falign-loops=% \ + -femit-struct-debug-baseonly -fno-ipa-cp-clone -fno-ipa-sra \ + -fno-partial-inlining -fplugin-arg-arm_ssp_per_task_plugin-% \ + -fno-reorder-blocks -fno-allow-store-data-races -fasan-shadow-offset=% \ + -fzero-call-used-regs=% -fno-stack-clash-protection \ + -fno-inline-functions-called-once \ + --param=% --param asan-% + +# Derived from `scripts/Makefile.clang`. +BINDGEN_TARGET_x86 := x86_64-linux-gnu +BINDGEN_TARGET := $(BINDGEN_TARGET_$(SRCARCH)) + +# All warnings are inhibited since GCC builds are very experimental, +# many GCC warnings are not supported by Clang, they may only appear in +# some configurations, with new GCC versions, etc. +bindgen_extra_c_flags = -w --target=$(BINDGEN_TARGET) + +bindgen_c_flags = $(filter-out $(bindgen_skip_c_flags), $(c_flags)) \ + $(bindgen_extra_c_flags) +endif + +ifdef CONFIG_LTO +bindgen_c_flags_lto = $(filter-out $(CC_FLAGS_LTO), $(bindgen_c_flags)) +else +bindgen_c_flags_lto = $(bindgen_c_flags) +endif + +bindgen_c_flags_final = $(bindgen_c_flags_lto) -D__BINDGEN__ + +quiet_cmd_bindgen = BINDGEN $@ + cmd_bindgen = \ + $(BINDGEN) $< $(bindgen_target_flags) \ + --use-core --with-derive-default --ctypes-prefix core::ffi --no-layout-tests \ + --no-debug '.*' \ + --size_t-is-usize -o $@ -- $(bindgen_c_flags_final) -DMODULE \ + $(bindgen_target_cflags) $(bindgen_target_extra) + +$(obj)/bindings/bindings_generated.rs: private bindgen_target_flags = \ + $(shell grep -v '^\#\|^$$' $(srctree)/$(src)/bindgen_parameters) +$(obj)/bindings/bindings_generated.rs: $(src)/bindings/bindings_helper.h \ + $(src)/bindgen_parameters FORCE + $(call if_changed_dep,bindgen) + +# See `CFLAGS_REMOVE_helpers.o` above. In addition, Clang on C does not warn +# with `-Wmissing-declarations` (unlike GCC), so it is not strictly needed here +# given it is `libclang`; but for consistency, future Clang changes and/or +# a potential future GCC backend for `bindgen`, we disable it too. +$(obj)/bindings/bindings_helpers_generated.rs: private bindgen_target_flags = \ + --blacklist-type '.*' --whitelist-var '' \ + --whitelist-function 'rust_helper_.*' +$(obj)/bindings/bindings_helpers_generated.rs: private bindgen_target_cflags = \ + -I$(objtree)/$(obj) -Wno-missing-prototypes -Wno-missing-declarations +$(obj)/bindings/bindings_helpers_generated.rs: private bindgen_target_extra = ; \ + sed -Ei 's/pub fn rust_helper_([a-zA-Z0-9_]*)/#[link_name="rust_helper_\1"]\n pub fn \1/g' $@ +$(obj)/bindings/bindings_helpers_generated.rs: $(src)/helpers.c FORCE + $(call if_changed_dep,bindgen) + +quiet_cmd_exports = EXPORTS $@ + cmd_exports = \ + $(NM) -p --defined-only $< \ + | grep -E ' (T|R|D) ' | cut -d ' ' -f 3 \ + | xargs -Isymbol \ + echo 'EXPORT_SYMBOL_RUST_GPL(symbol);' > $@ + +$(obj)/exports_core_generated.h: $(obj)/core.o FORCE + $(call if_changed,exports) + +$(obj)/exports_alloc_generated.h: $(obj)/alloc.o FORCE + $(call if_changed,exports) + +$(obj)/exports_bindings_generated.h: $(obj)/bindings.o FORCE + $(call if_changed,exports) + +$(obj)/exports_kernel_generated.h: $(obj)/kernel.o FORCE + $(call if_changed,exports) + +quiet_cmd_rustc_procmacro = $(RUSTC_OR_CLIPPY_QUIET) P $@ + cmd_rustc_procmacro = \ + $(RUSTC_OR_CLIPPY) $(rust_common_flags) \ + --emit=dep-info,link --extern proc_macro \ + --crate-type proc-macro --out-dir $(objtree)/$(obj) \ + --crate-name $(patsubst lib%.so,%,$(notdir $@)) $<; \ + mv $(objtree)/$(obj)/$(patsubst lib%.so,%,$(notdir $@)).d $(depfile); \ + sed -i '/^\#/d' $(depfile) + +# Procedural macros can only be used with the `rustc` that compiled it. +# Therefore, to get `libmacros.so` automatically recompiled when the compiler +# version changes, we add `core.o` as a dependency (even if it is not needed). +$(obj)/libmacros.so: $(src)/macros/lib.rs $(obj)/core.o FORCE + $(call if_changed_dep,rustc_procmacro) + +quiet_cmd_rustc_library = $(if $(skip_clippy),RUSTC,$(RUSTC_OR_CLIPPY_QUIET)) L $@ + cmd_rustc_library = \ + OBJTREE=$(abspath $(objtree)) \ + $(if $(skip_clippy),$(RUSTC),$(RUSTC_OR_CLIPPY)) \ + $(filter-out $(skip_flags),$(rust_flags) $(rustc_target_flags)) \ + --emit=dep-info,obj,metadata --crate-type rlib \ + --out-dir $(objtree)/$(obj) -L$(objtree)/$(obj) \ + --crate-name $(patsubst %.o,%,$(notdir $@)) $<; \ + mv $(objtree)/$(obj)/$(patsubst %.o,%,$(notdir $@)).d $(depfile); \ + sed -i '/^\#/d' $(depfile) \ + $(if $(rustc_objcopy),;$(OBJCOPY) $(rustc_objcopy) $@) + +rust-analyzer: + $(Q)$(srctree)/scripts/generate_rust_analyzer.py $(srctree) $(objtree) \ + $(RUST_LIB_SRC) > $(objtree)/rust-project.json + +$(obj)/core.o: private skip_clippy = 1 +$(obj)/core.o: private skip_flags = -Dunreachable_pub +$(obj)/core.o: private rustc_target_flags = $(core-cfgs) +$(obj)/core.o: $(RUST_LIB_SRC)/core/src/lib.rs $(obj)/target.json FORCE + $(call if_changed_dep,rustc_library) + +$(obj)/compiler_builtins.o: private rustc_objcopy = -w -W '__*' +$(obj)/compiler_builtins.o: $(src)/compiler_builtins.rs $(obj)/core.o FORCE + $(call if_changed_dep,rustc_library) + +$(obj)/alloc.o: private skip_clippy = 1 +$(obj)/alloc.o: private skip_flags = -Dunreachable_pub +$(obj)/alloc.o: private rustc_target_flags = $(alloc-cfgs) +$(obj)/alloc.o: $(src)/alloc/lib.rs $(obj)/compiler_builtins.o FORCE + $(call if_changed_dep,rustc_library) + +$(obj)/bindings.o: $(src)/bindings/lib.rs \ + $(obj)/compiler_builtins.o \ + $(obj)/bindings/bindings_generated.rs \ + $(obj)/bindings/bindings_helpers_generated.rs FORCE + $(call if_changed_dep,rustc_library) + +$(obj)/kernel.o: private rustc_target_flags = --extern alloc \ + --extern macros --extern bindings +$(obj)/kernel.o: $(src)/kernel/lib.rs $(obj)/alloc.o \ + $(obj)/libmacros.so $(obj)/bindings.o FORCE + $(call if_changed_dep,rustc_library) + +endif # CONFIG_RUST diff --git a/rust/alloc/README.md b/rust/alloc/README.md new file mode 100644 index 000000000000..c89c753720b5 --- /dev/null +++ b/rust/alloc/README.md @@ -0,0 +1,33 @@ +# `alloc` + +These source files come from the Rust standard library, hosted in +the <https://github.com/rust-lang/rust> repository, licensed under +"Apache-2.0 OR MIT" and adapted for kernel use. For copyright details, +see <https://github.com/rust-lang/rust/blob/master/COPYRIGHT>. + +Please note that these files should be kept as close as possible to +upstream. In general, only additions should be performed (e.g. new +methods). Eventually, changes should make it into upstream so that, +at some point, this fork can be dropped from the kernel tree. + + +## Rationale + +On one hand, kernel folks wanted to keep `alloc` in-tree to have more +freedom in both workflow and actual features if actually needed +(e.g. receiver types if we ended up using them), which is reasonable. + +On the other hand, Rust folks wanted to keep `alloc` as close as +upstream as possible and avoid as much divergence as possible, which +is also reasonable. + +We agreed on a middle-ground: we would keep a subset of `alloc` +in-tree that would be as small and as close as possible to upstream. +Then, upstream can start adding the functions that we add to `alloc` +etc., until we reach a point where the kernel already knows exactly +what it needs in `alloc` and all the new methods are merged into +upstream, so that we can drop `alloc` from the kernel tree and go back +to using the upstream one. + +By doing this, the kernel can go a bit faster now, and Rust can +slowly incorporate and discuss the changes as needed. diff --git a/rust/alloc/alloc.rs b/rust/alloc/alloc.rs new file mode 100644 index 000000000000..ca224a541770 --- /dev/null +++ b/rust/alloc/alloc.rs @@ -0,0 +1,440 @@ +// SPDX-License-Identifier: Apache-2.0 OR MIT + +//! Memory allocation APIs + +#![stable(feature = "alloc_module", since = "1.28.0")] + +#[cfg(not(test))] +use core::intrinsics; +use core::intrinsics::{min_align_of_val, size_of_val}; + +use core::ptr::Unique; +#[cfg(not(test))] +use core::ptr::{self, NonNull}; + +#[stable(feature = "alloc_module", since = "1.28.0")] +#[doc(inline)] +pub use core::alloc::*; + +use core::marker::Destruct; + +#[cfg(test)] +mod tests; + +extern "Rust" { + // These are the magic symbols to call the global allocator. rustc generates + // them to call `__rg_alloc` etc. if there is a `#[global_allocator]` attribute + // (the code expanding that attribute macro generates those functions), or to call + // the default implementations in libstd (`__rdl_alloc` etc. in `library/std/src/alloc.rs`) + // otherwise. + // The rustc fork of LLVM also special-cases these function names to be able to optimize them + // like `malloc`, `realloc`, and `free`, respectively. + #[rustc_allocator] + #[rustc_allocator_nounwind] + fn __rust_alloc(size: usize, align: usize) -> *mut u8; + #[rustc_allocator_nounwind] + fn __rust_dealloc(ptr: *mut u8, size: usize, align: usize); + #[rustc_allocator_nounwind] + fn __rust_realloc(ptr: *mut u8, old_size: usize, align: usize, new_size: usize) -> *mut u8; + #[rustc_allocator_nounwind] + fn __rust_alloc_zeroed(size: usize, align: usize) -> *mut u8; +} + +/// The global memory allocator. +/// +/// This type implements the [`Allocator`] trait by forwarding calls +/// to the allocator registered with the `#[global_allocator]` attribute +/// if there is one, or the `std` crate’s default. +/// +/// Note: while this type is unstable, the functionality it provides can be +/// accessed through the [free functions in `alloc`](self#functions). +#[unstable(feature = "allocator_api", issue = "32838")] +#[derive(Copy, Clone, Default, Debug)] +#[cfg(not(test))] +pub struct Global; + +#[cfg(test)] +pub use std::alloc::Global; + +/// Allocate memory with the global allocator. +/// +/// This function forwards calls to the [`GlobalAlloc::alloc`] method +/// of the allocator registered with the `#[global_allocator]` attribute +/// if there is one, or the `std` crate’s default. +/// +/// This function is expected to be deprecated in favor of the `alloc` method +/// of the [`Global`] type when it and the [`Allocator`] trait become stable. +/// +/// # Safety +/// +/// See [`GlobalAlloc::alloc`]. +/// +/// # Examples +/// +/// ``` +/// use std::alloc::{alloc, dealloc, Layout}; +/// +/// unsafe { +/// let layout = Layout::new::<u16>(); +/// let ptr = alloc(layout); +/// +/// *(ptr as *mut u16) = 42; +/// assert_eq!(*(ptr as *mut u16), 42); +/// +/// dealloc(ptr, layout); +/// } +/// ``` +#[stable(feature = "global_alloc", since = "1.28.0")] +#[must_use = "losing the pointer will leak memory"] +#[inline] +pub unsafe fn alloc(layout: Layout) -> *mut u8 { + unsafe { __rust_alloc(layout.size(), layout.align()) } +} + +/// Deallocate memory with the global allocator. +/// +/// This function forwards calls to the [`GlobalAlloc::dealloc`] method +/// of the allocator registered with the `#[global_allocator]` attribute +/// if there is one, or the `std` crate’s default. +/// +/// This function is expected to be deprecated in favor of the `dealloc` method +/// of the [`Global`] type when it and the [`Allocator`] trait become stable. +/// +/// # Safety +/// +/// See [`GlobalAlloc::dealloc`]. +#[stable(feature = "global_alloc", since = "1.28.0")] +#[inline] +pub unsafe fn dealloc(ptr: *mut u8, layout: Layout) { + unsafe { __rust_dealloc(ptr, layout.size(), layout.align()) } +} + +/// Reallocate memory with the global allocator. +/// +/// This function forwards calls to the [`GlobalAlloc::realloc`] method +/// of the allocator registered with the `#[global_allocator]` attribute +/// if there is one, or the `std` crate’s default. +/// +/// This function is expected to be deprecated in favor of the `realloc` method +/// of the [`Global`] type when it and the [`Allocator`] trait become stable. +/// +/// # Safety +/// +/// See [`GlobalAlloc::realloc`]. +#[stable(feature = "global_alloc", since = "1.28.0")] +#[must_use = "losing the pointer will leak memory"] +#[inline] +pub unsafe fn realloc(ptr: *mut u8, layout: Layout, new_size: usize) -> *mut u8 { + unsafe { __rust_realloc(ptr, layout.size(), layout.align(), new_size) } +} + +/// Allocate zero-initialized memory with the global allocator. +/// +/// This function forwards calls to the [`GlobalAlloc::alloc_zeroed`] method +/// of the allocator registered with the `#[global_allocator]` attribute +/// if there is one, or the `std` crate’s default. +/// +/// This function is expected to be deprecated in favor of the `alloc_zeroed` method +/// of the [`Global`] type when it and the [`Allocator`] trait become stable. +/// +/// # Safety +/// +/// See [`GlobalAlloc::alloc_zeroed`]. +/// +/// # Examples +/// +/// ``` +/// use std::alloc::{alloc_zeroed, dealloc, Layout}; +/// +/// unsafe { +/// let layout = Layout::new::<u16>(); +/// let ptr = alloc_zeroed(layout); +/// +/// assert_eq!(*(ptr as *mut u16), 0); +/// +/// dealloc(ptr, layout); +/// } +/// ``` +#[stable(feature = "global_alloc", since = "1.28.0")] +#[must_use = "losing the pointer will leak memory"] +#[inline] +pub unsafe fn alloc_zeroed(layout: Layout) -> *mut u8 { + unsafe { __rust_alloc_zeroed(layout.size(), layout.align()) } +} + +#[cfg(not(test))] +impl Global { + #[inline] + fn alloc_impl(&self, layout: Layout, zeroed: bool) -> Result<NonNull<[u8]>, AllocError> { + match layout.size() { + 0 => Ok(NonNull::slice_from_raw_parts(layout.dangling(), 0)), + // SAFETY: `layout` is non-zero in size, + size => unsafe { + let raw_ptr = if zeroed { alloc_zeroed(layout) } else { alloc(layout) }; + let ptr = NonNull::new(raw_ptr).ok_or(AllocError)?; + Ok(NonNull::slice_from_raw_parts(ptr, size)) + }, + } + } + + // SAFETY: Same as `Allocator::grow` + #[inline] + unsafe fn grow_impl( + &self, + ptr: NonNull<u8>, + old_layout: Layout, + new_layout: Layout, + zeroed: bool, + ) -> Result<NonNull<[u8]>, AllocError> { + debug_assert!( + new_layout.size() >= old_layout.size(), + "`new_layout.size()` must be greater than or equal to `old_layout.size()`" + ); + + match old_layout.size() { + 0 => self.alloc_impl(new_layout, zeroed), + + // SAFETY: `new_size` is non-zero as `old_size` is greater than or equal to `new_size` + // as required by safety conditions. Other conditions must be upheld by the caller + old_size if old_layout.align() == new_layout.align() => unsafe { + let new_size = new_layout.size(); + + // `realloc` probably checks for `new_size >= old_layout.size()` or something similar. + intrinsics::assume(new_size >= old_layout.size()); + + let raw_ptr = realloc(ptr.as_ptr(), old_layout, new_size); + let ptr = NonNull::new(raw_ptr).ok_or(AllocError)?; + if zeroed { + raw_ptr.add(old_size).write_bytes(0, new_size - old_size); + } + Ok(NonNull::slice_from_raw_parts(ptr, new_size)) + }, + + // SAFETY: because `new_layout.size()` must be greater than or equal to `old_size`, + // both the old and new memory allocation are valid for reads and writes for `old_size` + // bytes. Also, because the old allocation wasn't yet deallocated, it cannot overlap + // `new_ptr`. Thus, the call to `copy_nonoverlapping` is safe. The safety contract + // for `dealloc` must be upheld by the caller. + old_size => unsafe { + let new_ptr = self.alloc_impl(new_layout, zeroed)?; + ptr::copy_nonoverlapping(ptr.as_ptr(), new_ptr.as_mut_ptr(), old_size); + self.deallocate(ptr, old_layout); + Ok(new_ptr) + }, + } + } +} + +#[unstable(feature = "allocator_api", issue = "32838")] +#[cfg(not(test))] +unsafe impl Allocator for Global { + #[inline] + fn allocate(&self, layout: Layout) -> Result<NonNull<[u8]>, AllocError> { + self.alloc_impl(layout, false) + } + + #[inline] + fn allocate_zeroed(&self, layout: Layout) -> Result<NonNull<[u8]>, AllocError> { + self.alloc_impl(layout, true) + } + + #[inline] + unsafe fn deallocate(&self, ptr: NonNull<u8>, layout: Layout) { + if layout.size() != 0 { + // SAFETY: `layout` is non-zero in size, + // other conditions must be upheld by the caller + unsafe { dealloc(ptr.as_ptr(), layout) } + } + } + + #[inline] + unsafe fn grow( + &self, + ptr: NonNull<u8>, + old_layout: Layout, + new_layout: Layout, + ) -> Result<NonNull<[u8]>, AllocError> { + // SAFETY: all conditions must be upheld by the caller + unsafe { self.grow_impl(ptr, old_layout, new_layout, false) } + } + + #[inline] + unsafe fn grow_zeroed( + &self, + ptr: NonNull<u8>, + old_layout: Layout, + new_layout: Layout, + ) -> Result<NonNull<[u8]>, AllocError> { + // SAFETY: all conditions must be upheld by the caller + unsafe { self.grow_impl(ptr, old_layout, new_layout, true) } + } + + #[inline] + unsafe fn shrink( + &self, + ptr: NonNull<u8>, + old_layout: Layout, + new_layout: Layout, + ) -> Result<NonNull<[u8]>, AllocError> { + debug_assert!( + new_layout.size() <= old_layout.size(), + "`new_layout.size()` must be smaller than or equal to `old_layout.size()`" + ); + + match new_layout.size() { + // SAFETY: conditions must be upheld by the caller + 0 => unsafe { + self.deallocate(ptr, old_layout); + Ok(NonNull::slice_from_raw_parts(new_layout.dangling(), 0)) + }, + + // SAFETY: `new_size` is non-zero. Other conditions must be upheld by the caller + new_size if old_layout.align() == new_layout.align() => unsafe { + // `realloc` probably checks for `new_size <= old_layout.size()` or something similar. + intrinsics::assume(new_size <= old_layout.size()); + + let raw_ptr = realloc(ptr.as_ptr(), old_layout, new_size); + let ptr = NonNull::new(raw_ptr).ok_or(AllocError)?; + Ok(NonNull::slice_from_raw_parts(ptr, new_size)) + }, + + // SAFETY: because `new_size` must be smaller than or equal to `old_layout.size()`, + // both the old and new memory allocation are valid for reads and writes for `new_size` + // bytes. Also, because the old allocation wasn't yet deallocated, it cannot overlap + // `new_ptr`. Thus, the call to `copy_nonoverlapping` is safe. The safety contract + // for `dealloc` must be upheld by the caller. + new_size => unsafe { + let new_ptr = self.allocate(new_layout)?; + ptr::copy_nonoverlapping(ptr.as_ptr(), new_ptr.as_mut_ptr(), new_size); + self.deallocate(ptr, old_layout); + Ok(new_ptr) + }, + } + } +} + +/// The allocator for unique pointers. +#[cfg(all(not(no_global_oom_handling), not(test)))] +#[lang = "exchange_malloc"] +#[inline] +unsafe fn exchange_malloc(size: usize, align: usize) -> *mut u8 { + let layout = unsafe { Layout::from_size_align_unchecked(size, align) }; + match Global.allocate(layout) { + Ok(ptr) => ptr.as_mut_ptr(), + Err(_) => handle_alloc_error(layout), + } +} + +#[cfg_attr(not(test), lang = "box_free")] +#[inline] +#[rustc_const_unstable(feature = "const_box", issue = "92521")] +// This signature has to be the same as `Box`, otherwise an ICE will happen. +// When an additional parameter to `Box` is added (like `A: Allocator`), this has to be added here as +// well. +// For example if `Box` is changed to `struct Box<T: ?Sized, A: Allocator>(Unique<T>, A)`, +// this function has to be changed to `fn box_free<T: ?Sized, A: Allocator>(Unique<T>, A)` as well. +pub(crate) const unsafe fn box_free<T: ?Sized, A: ~const Allocator + ~const Destruct>( + ptr: Unique<T>, + alloc: A, +) { + unsafe { + let size = size_of_val(ptr.as_ref()); + let align = min_align_of_val(ptr.as_ref()); + let layout = Layout::from_size_align_unchecked(size, align); + alloc.deallocate(From::from(ptr.cast()), layout) + } +} + +// # Allocation error handler + +#[cfg(not(no_global_oom_handling))] +extern "Rust" { + // This is the magic symbol to call the global alloc error handler. rustc generates + // it to call `__rg_oom` if there is a `#[alloc_error_handler]`, or to call the + // default implementations below (`__rdl_oom`) otherwise. + fn __rust_alloc_error_handler(size: usize, align: usize) -> !; +} + +/// Abort on memory allocation error or failure. +/// +/// Callers of memory allocation APIs wishing to abort computation +/// in response to an allocation error are encouraged to call this function, +/// rather than directly invoking `panic!` or similar. +/// +/// The default behavior of this function is to print a message to standard error +/// and abort the process. +/// It can be replaced with [`set_alloc_error_hook`] and [`take_alloc_error_hook`]. +/// +/// [`set_alloc_error_hook`]: ../../std/alloc/fn.set_alloc_error_hook.html +/// [`take_alloc_error_hook`]: ../../std/alloc/fn.take_alloc_error_hook.html +#[stable(feature = "global_alloc", since = "1.28.0")] +#[rustc_const_unstable(feature = "const_alloc_error", issue = "92523")] +#[cfg(all(not(no_global_oom_handling), not(test)))] +#[cold] +pub const fn handle_alloc_error(layout: Layout) -> ! { + const fn ct_error(_: Layout) -> ! { + panic!("allocation failed"); + } + + fn rt_error(layout: Layout) -> ! { + unsafe { + __rust_alloc_error_handler(layout.size(), layout.align()); + } + } + + unsafe { core::intrinsics::const_eval_select((layout,), ct_error, rt_error) } +} + +// For alloc test `std::alloc::handle_alloc_error` can be used directly. +#[cfg(all(not(no_global_oom_handling), test))] +pub use std::alloc::handle_alloc_error; + +#[cfg(all(not(no_global_oom_handling), not(test)))] +#[doc(hidden)] +#[allow(unused_attributes)] +#[unstable(feature = "alloc_internals", issue = "none")] +pub mod __alloc_error_handler { + use crate::alloc::Layout; + + // called via generated `__rust_alloc_error_handler` + + // if there is no `#[alloc_error_handler]` + #[rustc_std_internal_symbol] + pub unsafe extern "C-unwind" fn __rdl_oom(size: usize, _align: usize) -> ! { + panic!("memory allocation of {size} bytes failed") + } + + // if there is an `#[alloc_error_handler]` + #[rustc_std_internal_symbol] + pub unsafe extern "C-unwind" fn __rg_oom(size: usize, align: usize) -> ! { + let layout = unsafe { Layout::from_size_align_unchecked(size, align) }; + extern "Rust" { + #[lang = "oom"] + fn oom_impl(layout: Layout) -> !; + } + unsafe { oom_impl(layout) } + } +} + +/// Specialize clones into pre-allocated, uninitialized memory. +/// Used by `Box::clone` and `Rc`/`Arc::make_mut`. +pub(crate) trait WriteCloneIntoRaw: Sized { + unsafe fn write_clone_into_raw(&self, target: *mut Self); +} + +impl<T: Clone> WriteCloneIntoRaw for T { + #[inline] + default unsafe fn write_clone_into_raw(&self, target: *mut Self) { + // Having allocated *first* may allow the optimizer to create + // the cloned value in-place, skipping the local and move. + unsafe { target.write(self.clone()) }; + } +} + +impl<T: Copy> WriteCloneIntoRaw for T { + #[inline] + unsafe fn write_clone_into_raw(&self, target: *mut Self) { + // We can always copy in-place, without ever involving a local value. + unsafe { target.copy_from_nonoverlapping(self, 1) }; + } +} diff --git a/rust/alloc/borrow.rs b/rust/alloc/borrow.rs new file mode 100644 index 000000000000..dde4957200d4 --- /dev/null +++ b/rust/alloc/borrow.rs @@ -0,0 +1,498 @@ +// SPDX-License-Identifier: Apache-2.0 OR MIT + +//! A module for working with borrowed data. + +#![stable(feature = "rust1", since = "1.0.0")] + +use core::cmp::Ordering; +use core::hash::{Hash, Hasher}; +use core::ops::Deref; +#[cfg(not(no_global_oom_handling))] +use core::ops::{Add, AddAssign}; + +#[stable(feature = "rust1", since = "1.0.0")] +pub use core::borrow::{Borrow, BorrowMut}; + +use core::fmt; +#[cfg(not(no_global_oom_handling))] +use crate::string::String; + +use Cow::*; + +#[stable(feature = "rust1", since = "1.0.0")] +impl<'a, B: ?Sized> Borrow<B> for Cow<'a, B> +where + B: ToOwned, + <B as ToOwned>::Owned: 'a, +{ + fn borrow(&self) -> &B { + &**self + } +} + +/// A generalization of `Clone` to borrowed data. +/// +/// Some types make it possible to go from borrowed to owned, usually by +/// implementing the `Clone` trait. But `Clone` works only for going from `&T` +/// to `T`. The `ToOwned` trait generalizes `Clone` to construct owned data +/// from any borrow of a given type. +#[cfg_attr(not(test), rustc_diagnostic_item = "ToOwned")] +#[stable(feature = "rust1", since = "1.0.0")] +pub trait ToOwned { + /// The resulting type after obtaining ownership. + #[stable(feature = "rust1", since = "1.0.0")] + type Owned: Borrow<Self>; + + /// Creates owned data from borrowed data, usually by cloning. + /// + /// # Examples + /// + /// Basic usage: + /// + /// ``` + /// let s: &str = "a"; + /// let ss: String = s.to_owned(); + /// + /// let v: &[i32] = &[1, 2]; + /// let vv: Vec<i32> = v.to_owned(); + /// ``` + #[stable(feature = "rust1", since = "1.0.0")] + #[must_use = "cloning is often expensive and is not expected to have side effects"] + fn to_owned(&self) -> Self::Owned; + + /// Uses borrowed data to replace owned data, usually by cloning. + /// + /// This is borrow-generalized version of `Clone::clone_from`. + /// + /// # Examples + /// + /// Basic usage: + /// + /// ``` + /// # #![feature(toowned_clone_into)] + /// let mut s: String = String::new(); + /// "hello".clone_into(&mut s); + /// + /// let mut v: Vec<i32> = Vec::new(); + /// [1, 2][..].clone_into(&mut v); + /// ``` + #[unstable(feature = "toowned_clone_into", reason = "recently added", issue = "41263")] + fn clone_into(&self, target: &mut Self::Owned) { + *target = self.to_owned(); + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<T> ToOwned for T +where + T: Clone, +{ + type Owned = T; + fn to_owned(&self) -> T { + self.clone() + } + + fn clone_into(&self, target: &mut T) { + target.clone_from(self); + } +} + +/// A clone-on-write smart pointer. +/// +/// The type `Cow` is a smart pointer providing clone-on-write functionality: it +/// can enclose and provide immutable access to borrowed data, and clone the +/// data lazily when mutation or ownership is required. The type is designed to +/// work with general borrowed data via the `Borrow` trait. +/// +/// `Cow` implements `Deref`, which means that you can call +/// non-mutating methods directly on the data it encloses. If mutation +/// is desired, `to_mut` will obtain a mutable reference to an owned +/// value, cloning if necessary. +/// +/// If you need reference-counting pointers, note that +/// [`Rc::make_mut`][crate::rc::Rc::make_mut] and +/// [`Arc::make_mut`][crate::sync::Arc::make_mut] can provide clone-on-write +/// functionality as well. +/// +/// # Examples +/// +/// ``` +/// use std::borrow::Cow; +/// +/// fn abs_all(input: &mut Cow<[i32]>) { +/// for i in 0..input.len() { +/// let v = input[i]; +/// if v < 0 { +/// // Clones into a vector if not already owned. +/// input.to_mut()[i] = -v; +/// } +/// } +/// } +/// +/// // No clone occurs because `input` doesn't need to be mutated. +/// let slice = [0, 1, 2]; +/// let mut input = Cow::from(&slice[..]); +/// abs_all(&mut input); +/// +/// // Clone occurs because `input` needs to be mutated. +/// let slice = [-1, 0, 1]; +/// let mut input = Cow::from(&slice[..]); +/// abs_all(&mut input); +/// +/// // No clone occurs because `input` is already owned. +/// let mut input = Cow::from(vec![-1, 0, 1]); +/// abs_all(&mut input); +/// ``` +/// +/// Another example showing how to keep `Cow` in a struct: +/// +/// ``` +/// use std::borrow::Cow; +/// +/// struct Items<'a, X: 'a> where [X]: ToOwned<Owned = Vec<X>> { +/// values: Cow<'a, [X]>, +/// } +/// +/// impl<'a, X: Clone + 'a> Items<'a, X> where [X]: ToOwned<Owned = Vec<X>> { +/// fn new(v: Cow<'a, [X]>) -> Self { +/// Items { values: v } +/// } +/// } +/// +/// // Creates a container from borrowed values of a slice +/// let readonly = [1, 2]; +/// let borrowed = Items::new((&readonly[..]).into()); +/// match borrowed { +/// Items { values: Cow::Borrowed(b) } => println!("borrowed {b:?}"), +/// _ => panic!("expect borrowed value"), +/// } +/// +/// let mut clone_on_write = borrowed; +/// // Mutates the data from slice into owned vec and pushes a new value on top +/// clone_on_write.values.to_mut().push(3); +/// println!("clone_on_write = {:?}", clone_on_write.values); +/// +/// // The data was mutated. Let's check it out. +/// match clone_on_write { +/// Items { values: Cow::Owned(_) } => println!("clone_on_write contains owned data"), +/// _ => panic!("expect owned data"), +/// } +/// ``` +#[stable(feature = "rust1", since = "1.0.0")] +#[cfg_attr(not(test), rustc_diagnostic_item = "Cow")] +pub enum Cow<'a, B: ?Sized + 'a> +where + B: ToOwned, +{ + /// Borrowed data. + #[stable(feature = "rust1", since = "1.0.0")] + Borrowed(#[stable(feature = "rust1", since = "1.0.0")] &'a B), + + /// Owned data. + #[stable(feature = "rust1", since = "1.0.0")] + Owned(#[stable(feature = "rust1", since = "1.0.0")] <B as ToOwned>::Owned), +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<B: ?Sized + ToOwned> Clone for Cow<'_, B> { + fn clone(&self) -> Self { + match *self { + Borrowed(b) => Borrowed(b), + Owned(ref o) => { + let b: &B = o.borrow(); + Owned(b.to_owned()) + } + } + } + + fn clone_from(&mut self, source: &Self) { + match (self, source) { + (&mut Owned(ref mut dest), &Owned(ref o)) => o.borrow().clone_into(dest), + (t, s) => *t = s.clone(), + } + } +} + +impl<B: ?Sized + ToOwned> Cow<'_, B> { + /// Returns true if the data is borrowed, i.e. if `to_mut` would require additional work. + /// + /// # Examples + /// + /// ``` + /// #![feature(cow_is_borrowed)] + /// use std::borrow::Cow; + /// + /// let cow = Cow::Borrowed("moo"); + /// assert!(cow.is_borrowed()); + /// + /// let bull: Cow<'_, str> = Cow::Owned("...moo?".to_string()); + /// assert!(!bull.is_borrowed()); + /// ``` + #[unstable(feature = "cow_is_borrowed", issue = "65143")] + #[rustc_const_unstable(feature = "const_cow_is_borrowed", issue = "65143")] + pub const fn is_borrowed(&self) -> bool { + match *self { + Borrowed(_) => true, + Owned(_) => false, + } + } + + /// Returns true if the data is owned, i.e. if `to_mut` would be a no-op. + /// + /// # Examples + /// + /// ``` + /// #![feature(cow_is_borrowed)] + /// use std::borrow::Cow; + /// + /// let cow: Cow<'_, str> = Cow::Owned("moo".to_string()); + /// assert!(cow.is_owned()); + /// + /// let bull = Cow::Borrowed("...moo?"); + /// assert!(!bull.is_owned()); + /// ``` + #[unstable(feature = "cow_is_borrowed", issue = "65143")] + #[rustc_const_unstable(feature = "const_cow_is_borrowed", issue = "65143")] + pub const fn is_owned(&self) -> bool { + !self.is_borrowed() + } + + /// Acquires a mutable reference to the owned form of the data. + /// + /// Clones the data if it is not already owned. + /// + /// # Examples + /// + /// ``` + /// use std::borrow::Cow; + /// + /// let mut cow = Cow::Borrowed("foo"); + /// cow.to_mut().make_ascii_uppercase(); + /// + /// assert_eq!( + /// cow, + /// Cow::Owned(String::from("FOO")) as Cow<str> + /// ); + /// ``` + #[stable(feature = "rust1", since = "1.0.0")] + pub fn to_mut(&mut self) -> &mut <B as ToOwned>::Owned { + match *self { + Borrowed(borrowed) => { + *self = Owned(borrowed.to_owned()); + match *self { + Borrowed(..) => unreachable!(), + Owned(ref mut owned) => owned, + } + } + Owned(ref mut owned) => owned, + } + } + + /// Extracts the owned data. + /// + /// Clones the data if it is not already owned. + /// + /// # Examples + /// + /// Calling `into_owned` on a `Cow::Borrowed` returns a clone of the borrowed data: + /// + /// ``` + /// use std::borrow::Cow; + /// + /// let s = "Hello world!"; + /// let cow = Cow::Borrowed(s); + /// + /// assert_eq!( + /// cow.into_owned(), + /// String::from(s) + /// ); + /// ``` + /// + /// Calling `into_owned` on a `Cow::Owned` returns the owned data. The data is moved out of the + /// `Cow` without being cloned. + /// + /// ``` + /// use std::borrow::Cow; + /// + /// let s = "Hello world!"; + /// let cow: Cow<str> = Cow::Owned(String::from(s)); + /// + /// assert_eq!( + /// cow.into_owned(), + /// String::from(s) + /// ); + /// ``` + #[stable(feature = "rust1", since = "1.0.0")] + pub fn into_owned(self) -> <B as ToOwned>::Owned { + match self { + Borrowed(borrowed) => borrowed.to_owned(), + Owned(owned) => owned, + } + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +#[rustc_const_unstable(feature = "const_deref", issue = "88955")] +impl<B: ?Sized + ToOwned> const Deref for Cow<'_, B> +where + B::Owned: ~const Borrow<B>, +{ + type Target = B; + + fn deref(&self) -> &B { + match *self { + Borrowed(borrowed) => borrowed, + Owned(ref owned) => owned.borrow(), + } + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<B: ?Sized> Eq for Cow<'_, B> where B: Eq + ToOwned {} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<B: ?Sized> Ord for Cow<'_, B> +where + B: Ord + ToOwned, +{ + #[inline] + fn cmp(&self, other: &Self) -> Ordering { + Ord::cmp(&**self, &**other) + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<'a, 'b, B: ?Sized, C: ?Sized> PartialEq<Cow<'b, C>> for Cow<'a, B> +where + B: PartialEq<C> + ToOwned, + C: ToOwned, +{ + #[inline] + fn eq(&self, other: &Cow<'b, C>) -> bool { + PartialEq::eq(&**self, &**other) + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<'a, B: ?Sized> PartialOrd for Cow<'a, B> +where + B: PartialOrd + ToOwned, +{ + #[inline] + fn partial_cmp(&self, other: &Cow<'a, B>) -> Option<Ordering> { + PartialOrd::partial_cmp(&**self, &**other) + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<B: ?Sized> fmt::Debug for Cow<'_, B> +where + B: fmt::Debug + ToOwned<Owned: fmt::Debug>, +{ + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + match *self { + Borrowed(ref b) => fmt::Debug::fmt(b, f), + Owned(ref o) => fmt::Debug::fmt(o, f), + } + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<B: ?Sized> fmt::Display for Cow<'_, B> +where + B: fmt::Display + ToOwned<Owned: fmt::Display>, +{ + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + match *self { + Borrowed(ref b) => fmt::Display::fmt(b, f), + Owned(ref o) => fmt::Display::fmt(o, f), + } + } +} + +#[stable(feature = "default", since = "1.11.0")] +impl<B: ?Sized> Default for Cow<'_, B> +where + B: ToOwned<Owned: Default>, +{ + /// Creates an owned Cow<'a, B> with the default value for the contained owned value. + fn default() -> Self { + Owned(<B as ToOwned>::Owned::default()) + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<B: ?Sized> Hash for Cow<'_, B> +where + B: Hash + ToOwned, +{ + #[inline] + fn hash<H: Hasher>(&self, state: &mut H) { + Hash::hash(&**self, state) + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<T: ?Sized + ToOwned> AsRef<T> for Cow<'_, T> { + fn as_ref(&self) -> &T { + self + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "cow_add", since = "1.14.0")] +impl<'a> Add<&'a str> for Cow<'a, str> { + type Output = Cow<'a, str>; + + #[inline] + fn add(mut self, rhs: &'a str) -> Self::Output { + self += rhs; + self + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "cow_add", since = "1.14.0")] +impl<'a> Add<Cow<'a, str>> for Cow<'a, str> { + type Output = Cow<'a, str>; + + #[inline] + fn add(mut self, rhs: Cow<'a, str>) -> Self::Output { + self += rhs; + self + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "cow_add", since = "1.14.0")] +impl<'a> AddAssign<&'a str> for Cow<'a, str> { + fn add_assign(&mut self, rhs: &'a str) { + if self.is_empty() { + *self = Cow::Borrowed(rhs) + } else if !rhs.is_empty() { + if let Cow::Borrowed(lhs) = *self { + let mut s = String::with_capacity(lhs.len() + rhs.len()); + s.push_str(lhs); + *self = Cow::Owned(s); + } + self.to_mut().push_str(rhs); + } + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "cow_add", since = "1.14.0")] +impl<'a> AddAssign<Cow<'a, str>> for Cow<'a, str> { + fn add_assign(&mut self, rhs: Cow<'a, str>) { + if self.is_empty() { + *self = rhs + } else if !rhs.is_empty() { + if let Cow::Borrowed(lhs) = *self { + let mut s = String::with_capacity(lhs.len() + rhs.len()); + s.push_str(lhs); + *self = Cow::Owned(s); + } + self.to_mut().push_str(&rhs); + } + } +} diff --git a/rust/alloc/boxed.rs b/rust/alloc/boxed.rs new file mode 100644 index 000000000000..dcfe87b14f3a --- /dev/null +++ b/rust/alloc/boxed.rs @@ -0,0 +1,2028 @@ +// SPDX-License-Identifier: Apache-2.0 OR MIT + +//! A pointer type for heap allocation. +//! +//! [`Box<T>`], casually referred to as a 'box', provides the simplest form of +//! heap allocation in Rust. Boxes provide ownership for this allocation, and +//! drop their contents when they go out of scope. Boxes also ensure that they +//! never allocate more than `isize::MAX` bytes. +//! +//! # Examples +//! +//! Move a value from the stack to the heap by creating a [`Box`]: +//! +//! ``` +//! let val: u8 = 5; +//! let boxed: Box<u8> = Box::new(val); +//! ``` +//! +//! Move a value from a [`Box`] back to the stack by [dereferencing]: +//! +//! ``` +//! let boxed: Box<u8> = Box::new(5); +//! let val: u8 = *boxed; +//! ``` +//! +//! Creating a recursive data structure: +//! +//! ``` +//! #[derive(Debug)] +//! enum List<T> { +//! Cons(T, Box<List<T>>), +//! Nil, +//! } +//! +//! let list: List<i32> = List::Cons(1, Box::new(List::Cons(2, Box::new(List::Nil)))); +//! println!("{list:?}"); +//! ``` +//! +//! This will print `Cons(1, Cons(2, Nil))`. +//! +//! Recursive structures must be boxed, because if the definition of `Cons` +//! looked like this: +//! +//! ```compile_fail,E0072 +//! # enum List<T> { +//! Cons(T, List<T>), +//! # } +//! ``` +//! +//! It wouldn't work. This is because the size of a `List` depends on how many +//! elements are in the list, and so we don't know how much memory to allocate +//! for a `Cons`. By introducing a [`Box<T>`], which has a defined size, we know how +//! big `Cons` needs to be. +//! +//! # Memory layout +//! +//! For non-zero-sized values, a [`Box`] will use the [`Global`] allocator for +//! its allocation. It is valid to convert both ways between a [`Box`] and a +//! raw pointer allocated with the [`Global`] allocator, given that the +//! [`Layout`] used with the allocator is correct for the type. More precisely, +//! a `value: *mut T` that has been allocated with the [`Global`] allocator +//! with `Layout::for_value(&*value)` may be converted into a box using +//! [`Box::<T>::from_raw(value)`]. Conversely, the memory backing a `value: *mut +//! T` obtained from [`Box::<T>::into_raw`] may be deallocated using the +//! [`Global`] allocator with [`Layout::for_value(&*value)`]. +//! +//! For zero-sized values, the `Box` pointer still has to be [valid] for reads +//! and writes and sufficiently aligned. In particular, casting any aligned +//! non-zero integer literal to a raw pointer produces a valid pointer, but a +//! pointer pointing into previously allocated memory that since got freed is +//! not valid. The recommended way to build a Box to a ZST if `Box::new` cannot +//! be used is to use [`ptr::NonNull::dangling`]. +//! +//! So long as `T: Sized`, a `Box<T>` is guaranteed to be represented +//! as a single pointer and is also ABI-compatible with C pointers +//! (i.e. the C type `T*`). This means that if you have extern "C" +//! Rust functions that will be called from C, you can define those +//! Rust functions using `Box<T>` types, and use `T*` as corresponding +//! type on the C side. As an example, consider this C header which +//! declares functions that create and destroy some kind of `Foo` +//! value: +//! +//! ```c +//! /* C header */ +//! +//! /* Returns ownership to the caller */ +//! struct Foo* foo_new(void); +//! +//! /* Takes ownership from the caller; no-op when invoked with null */ +//! void foo_delete(struct Foo*); +//! ``` +//! +//! These two functions might be implemented in Rust as follows. Here, the +//! `struct Foo*` type from C is translated to `Box<Foo>`, which captures +//! the ownership constraints. Note also that the nullable argument to +//! `foo_delete` is represented in Rust as `Option<Box<Foo>>`, since `Box<Foo>` +//! cannot be null. +//! +//! ``` +//! #[repr(C)] +//! pub struct Foo; +//! +//! #[no_mangle] +//! pub extern "C" fn foo_new() -> Box<Foo> { +//! Box::new(Foo) +//! } +//! +//! #[no_mangle] +//! pub extern "C" fn foo_delete(_: Option<Box<Foo>>) {} +//! ``` +//! +//! Even though `Box<T>` has the same representation and C ABI as a C pointer, +//! this does not mean that you can convert an arbitrary `T*` into a `Box<T>` +//! and expect things to work. `Box<T>` values will always be fully aligned, +//! non-null pointers. Moreover, the destructor for `Box<T>` will attempt to +//! free the value with the global allocator. In general, the best practice +//! is to only use `Box<T>` for pointers that originated from the global +//! allocator. +//! +//! **Important.** At least at present, you should avoid using +//! `Box<T>` types for functions that are defined in C but invoked +//! from Rust. In those cases, you should directly mirror the C types +//! as closely as possible. Using types like `Box<T>` where the C +//! definition is just using `T*` can lead to undefined behavior, as +//! described in [rust-lang/unsafe-code-guidelines#198][ucg#198]. +//! +//! [ucg#198]: https://github.com/rust-lang/unsafe-code-guidelines/issues/198 +//! [dereferencing]: core::ops::Deref +//! [`Box::<T>::from_raw(value)`]: Box::from_raw +//! [`Global`]: crate::alloc::Global +//! [`Layout`]: crate::alloc::Layout +//! [`Layout::for_value(&*value)`]: crate::alloc::Layout::for_value +//! [valid]: ptr#safety + +#![stable(feature = "rust1", since = "1.0.0")] + +use core::any::Any; +use core::async_iter::AsyncIterator; +use core::borrow; +use core::cmp::Ordering; +use core::convert::{From, TryFrom}; +use core::fmt; +use core::future::Future; +use core::hash::{Hash, Hasher}; +#[cfg(not(no_global_oom_handling))] +use core::iter::FromIterator; +use core::iter::{FusedIterator, Iterator}; +use core::marker::{Destruct, Unpin, Unsize}; +use core::mem; +use core::ops::{ + CoerceUnsized, Deref, DerefMut, DispatchFromDyn, Generator, GeneratorState, Receiver, +}; +use core::pin::Pin; +use core::ptr::{self, Unique}; +use core::task::{Context, Poll}; + +#[cfg(not(no_global_oom_handling))] +use crate::alloc::{handle_alloc_error, WriteCloneIntoRaw}; +use crate::alloc::{AllocError, Allocator, Global, Layout}; +#[cfg(not(no_global_oom_handling))] +use crate::borrow::Cow; +use crate::raw_vec::RawVec; +#[cfg(not(no_global_oom_handling))] +use crate::str::from_boxed_utf8_unchecked; +#[cfg(not(no_global_oom_handling))] +use crate::vec::Vec; + +#[cfg(not(no_thin))] +#[unstable(feature = "thin_box", issue = "92791")] +pub use thin::ThinBox; + +#[cfg(not(no_thin))] +mod thin; + +/// A pointer type for heap allocation. +/// +/// See the [module-level documentation](../../std/boxed/index.html) for more. +#[lang = "owned_box"] +#[fundamental] +#[stable(feature = "rust1", since = "1.0.0")] +// The declaration of the `Box` struct must be kept in sync with the +// `alloc::alloc::box_free` function or ICEs will happen. See the comment +// on `box_free` for more details. +pub struct Box< + T: ?Sized, + #[unstable(feature = "allocator_api", issue = "32838")] A: Allocator = Global, +>(Unique<T>, A); + +impl<T> Box<T> { + /// Allocates memory on the heap and then places `x` into it. + /// + /// This doesn't actually allocate if `T` is zero-sized. + /// + /// # Examples + /// + /// ``` + /// let five = Box::new(5); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[inline(always)] + #[stable(feature = "rust1", since = "1.0.0")] + #[must_use] + pub fn new(x: T) -> Self { + box x + } + + /// Constructs a new box with uninitialized contents. + /// + /// # Examples + /// + /// ``` + /// #![feature(new_uninit)] + /// + /// let mut five = Box::<u32>::new_uninit(); + /// + /// let five = unsafe { + /// // Deferred initialization: + /// five.as_mut_ptr().write(5); + /// + /// five.assume_init() + /// }; + /// + /// assert_eq!(*five, 5) + /// ``` + #[cfg(not(no_global_oom_handling))] + #[unstable(feature = "new_uninit", issue = "63291")] + #[must_use] + #[inline] + pub fn new_uninit() -> Box<mem::MaybeUninit<T>> { + Self::new_uninit_in(Global) + } + + /// Constructs a new `Box` with uninitialized contents, with the memory + /// being filled with `0` bytes. + /// + /// See [`MaybeUninit::zeroed`][zeroed] for examples of correct and incorrect usage + /// of this method. + /// + /// # Examples + /// + /// ``` + /// #![feature(new_uninit)] + /// + /// let zero = Box::<u32>::new_zeroed(); + /// let zero = unsafe { zero.assume_init() }; + /// + /// assert_eq!(*zero, 0) + /// ``` + /// + /// [zeroed]: mem::MaybeUninit::zeroed + #[cfg(not(no_global_oom_handling))] + #[inline] + #[unstable(feature = "new_uninit", issue = "63291")] + #[must_use] + pub fn new_zeroed() -> Box<mem::MaybeUninit<T>> { + Self::new_zeroed_in(Global) + } + + /// Constructs a new `Pin<Box<T>>`. If `T` does not implement `Unpin`, then + /// `x` will be pinned in memory and unable to be moved. + #[cfg(not(no_global_oom_handling))] + #[stable(feature = "pin", since = "1.33.0")] + #[must_use] + #[inline(always)] + pub fn pin(x: T) -> Pin<Box<T>> { + (box x).into() + } + + /// Allocates memory on the heap then places `x` into it, + /// returning an error if the allocation fails + /// + /// This doesn't actually allocate if `T` is zero-sized. + /// + /// # Examples + /// + /// ``` + /// #![feature(allocator_api)] + /// + /// let five = Box::try_new(5)?; + /// # Ok::<(), std::alloc::AllocError>(()) + /// ``` + #[unstable(feature = "allocator_api", issue = "32838")] + #[inline] + pub fn try_new(x: T) -> Result<Self, AllocError> { + Self::try_new_in(x, Global) + } + + /// Constructs a new box with uninitialized contents on the heap, + /// returning an error if the allocation fails + /// + /// # Examples + /// + /// ``` + /// #![feature(allocator_api, new_uninit)] + /// + /// let mut five = Box::<u32>::try_new_uninit()?; + /// + /// let five = unsafe { + /// // Deferred initialization: + /// five.as_mut_ptr().write(5); + /// + /// five.assume_init() + /// }; + /// + /// assert_eq!(*five, 5); + /// # Ok::<(), std::alloc::AllocError>(()) + /// ``` + #[unstable(feature = "allocator_api", issue = "32838")] + // #[unstable(feature = "new_uninit", issue = "63291")] + #[inline] + pub fn try_new_uninit() -> Result<Box<mem::MaybeUninit<T>>, AllocError> { + Box::try_new_uninit_in(Global) + } + + /// Constructs a new `Box` with uninitialized contents, with the memory + /// being filled with `0` bytes on the heap + /// + /// See [`MaybeUninit::zeroed`][zeroed] for examples of correct and incorrect usage + /// of this method. + /// + /// # Examples + /// + /// ``` + /// #![feature(allocator_api, new_uninit)] + /// + /// let zero = Box::<u32>::try_new_zeroed()?; + /// let zero = unsafe { zero.assume_init() }; + /// + /// assert_eq!(*zero, 0); + /// # Ok::<(), std::alloc::AllocError>(()) + /// ``` + /// + /// [zeroed]: mem::MaybeUninit::zeroed + #[unstable(feature = "allocator_api", issue = "32838")] + // #[unstable(feature = "new_uninit", issue = "63291")] + #[inline] + pub fn try_new_zeroed() -> Result<Box<mem::MaybeUninit<T>>, AllocError> { + Box::try_new_zeroed_in(Global) + } +} + +impl<T, A: Allocator> Box<T, A> { + /// Allocates memory in the given allocator then places `x` into it. + /// + /// This doesn't actually allocate if `T` is zero-sized. + /// + /// # Examples + /// + /// ``` + /// #![feature(allocator_api)] + /// + /// use std::alloc::System; + /// + /// let five = Box::new_in(5, System); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[unstable(feature = "allocator_api", issue = "32838")] + #[rustc_const_unstable(feature = "const_box", issue = "92521")] + #[must_use] + #[inline] + pub const fn new_in(x: T, alloc: A) -> Self + where + A: ~const Allocator + ~const Destruct, + { + let mut boxed = Self::new_uninit_in(alloc); + unsafe { + boxed.as_mut_ptr().write(x); + boxed.assume_init() + } + } + + /// Allocates memory in the given allocator then places `x` into it, + /// returning an error if the allocation fails + /// + /// This doesn't actually allocate if `T` is zero-sized. + /// + /// # Examples + /// + /// ``` + /// #![feature(allocator_api)] + /// + /// use std::alloc::System; + /// + /// let five = Box::try_new_in(5, System)?; + /// # Ok::<(), std::alloc::AllocError>(()) + /// ``` + #[unstable(feature = "allocator_api", issue = "32838")] + #[rustc_const_unstable(feature = "const_box", issue = "92521")] + #[inline] + pub const fn try_new_in(x: T, alloc: A) -> Result<Self, AllocError> + where + T: ~const Destruct, + A: ~const Allocator + ~const Destruct, + { + let mut boxed = Self::try_new_uninit_in(alloc)?; + unsafe { + boxed.as_mut_ptr().write(x); + Ok(boxed.assume_init()) + } + } + + /// Constructs a new box with uninitialized contents in the provided allocator. + /// + /// # Examples + /// + /// ``` + /// #![feature(allocator_api, new_uninit)] + /// + /// use std::alloc::System; + /// + /// let mut five = Box::<u32, _>::new_uninit_in(System); + /// + /// let five = unsafe { + /// // Deferred initialization: + /// five.as_mut_ptr().write(5); + /// + /// five.assume_init() + /// }; + /// + /// assert_eq!(*five, 5) + /// ``` + #[unstable(feature = "allocator_api", issue = "32838")] + #[rustc_const_unstable(feature = "const_box", issue = "92521")] + #[cfg(not(no_global_oom_handling))] + #[must_use] + // #[unstable(feature = "new_uninit", issue = "63291")] + pub const fn new_uninit_in(alloc: A) -> Box<mem::MaybeUninit<T>, A> + where + A: ~const Allocator + ~const Destruct, + { + let layout = Layout::new::<mem::MaybeUninit<T>>(); + // NOTE: Prefer match over unwrap_or_else since closure sometimes not inlineable. + // That would make code size bigger. + match Box::try_new_uninit_in(alloc) { + Ok(m) => m, + Err(_) => handle_alloc_error(layout), + } + } + + /// Constructs a new box with uninitialized contents in the provided allocator, + /// returning an error if the allocation fails + /// + /// # Examples + /// + /// ``` + /// #![feature(allocator_api, new_uninit)] + /// + /// use std::alloc::System; + /// + /// let mut five = Box::<u32, _>::try_new_uninit_in(System)?; + /// + /// let five = unsafe { + /// // Deferred initialization: + /// five.as_mut_ptr().write(5); + /// + /// five.assume_init() + /// }; + /// + /// assert_eq!(*five, 5); + /// # Ok::<(), std::alloc::AllocError>(()) + /// ``` + #[unstable(feature = "allocator_api", issue = "32838")] + // #[unstable(feature = "new_uninit", issue = "63291")] + #[rustc_const_unstable(feature = "const_box", issue = "92521")] + pub const fn try_new_uninit_in(alloc: A) -> Result<Box<mem::MaybeUninit<T>, A>, AllocError> + where + A: ~const Allocator + ~const Destruct, + { + let layout = Layout::new::<mem::MaybeUninit<T>>(); + let ptr = alloc.allocate(layout)?.cast(); + unsafe { Ok(Box::from_raw_in(ptr.as_ptr(), alloc)) } + } + + /// Constructs a new `Box` with uninitialized contents, with the memory + /// being filled with `0` bytes in the provided allocator. + /// + /// See [`MaybeUninit::zeroed`][zeroed] for examples of correct and incorrect usage + /// of this method. + /// + /// # Examples + /// + /// ``` + /// #![feature(allocator_api, new_uninit)] + /// + /// use std::alloc::System; + /// + /// let zero = Box::<u32, _>::new_zeroed_in(System); + /// let zero = unsafe { zero.assume_init() }; + /// + /// assert_eq!(*zero, 0) + /// ``` + /// + /// [zeroed]: mem::MaybeUninit::zeroed + #[unstable(feature = "allocator_api", issue = "32838")] + #[rustc_const_unstable(feature = "const_box", issue = "92521")] + #[cfg(not(no_global_oom_handling))] + // #[unstable(feature = "new_uninit", issue = "63291")] + #[must_use] + pub const fn new_zeroed_in(alloc: A) -> Box<mem::MaybeUninit<T>, A> + where + A: ~const Allocator + ~const Destruct, + { + let layout = Layout::new::<mem::MaybeUninit<T>>(); + // NOTE: Prefer match over unwrap_or_else since closure sometimes not inlineable. + // That would make code size bigger. + match Box::try_new_zeroed_in(alloc) { + Ok(m) => m, + Err(_) => handle_alloc_error(layout), + } + } + + /// Constructs a new `Box` with uninitialized contents, with the memory + /// being filled with `0` bytes in the provided allocator, + /// returning an error if the allocation fails, + /// + /// See [`MaybeUninit::zeroed`][zeroed] for examples of correct and incorrect usage + /// of this method. + /// + /// # Examples + /// + /// ``` + /// #![feature(allocator_api, new_uninit)] + /// + /// use std::alloc::System; + /// + /// let zero = Box::<u32, _>::try_new_zeroed_in(System)?; + /// let zero = unsafe { zero.assume_init() }; + /// + /// assert_eq!(*zero, 0); + /// # Ok::<(), std::alloc::AllocError>(()) + /// ``` + /// + /// [zeroed]: mem::MaybeUninit::zeroed + #[unstable(feature = "allocator_api", issue = "32838")] + // #[unstable(feature = "new_uninit", issue = "63291")] + #[rustc_const_unstable(feature = "const_box", issue = "92521")] + pub const fn try_new_zeroed_in(alloc: A) -> Result<Box<mem::MaybeUninit<T>, A>, AllocError> + where + A: ~const Allocator + ~const Destruct, + { + let layout = Layout::new::<mem::MaybeUninit<T>>(); + let ptr = alloc.allocate_zeroed(layout)?.cast(); + unsafe { Ok(Box::from_raw_in(ptr.as_ptr(), alloc)) } + } + + /// Constructs a new `Pin<Box<T, A>>`. If `T` does not implement `Unpin`, then + /// `x` will be pinned in memory and unable to be moved. + #[cfg(not(no_global_oom_handling))] + #[unstable(feature = "allocator_api", issue = "32838")] + #[rustc_const_unstable(feature = "const_box", issue = "92521")] + #[must_use] + #[inline(always)] + pub const fn pin_in(x: T, alloc: A) -> Pin<Self> + where + A: 'static + ~const Allocator + ~const Destruct, + { + Self::into_pin(Self::new_in(x, alloc)) + } + + /// Converts a `Box<T>` into a `Box<[T]>` + /// + /// This conversion does not allocate on the heap and happens in place. + #[unstable(feature = "box_into_boxed_slice", issue = "71582")] + #[rustc_const_unstable(feature = "const_box", issue = "92521")] + pub const fn into_boxed_slice(boxed: Self) -> Box<[T], A> { + let (raw, alloc) = Box::into_raw_with_allocator(boxed); + unsafe { Box::from_raw_in(raw as *mut [T; 1], alloc) } + } + + /// Consumes the `Box`, returning the wrapped value. + /// + /// # Examples + /// + /// ``` + /// #![feature(box_into_inner)] + /// + /// let c = Box::new(5); + /// + /// assert_eq!(Box::into_inner(c), 5); + /// ``` + #[unstable(feature = "box_into_inner", issue = "80437")] + #[rustc_const_unstable(feature = "const_box", issue = "92521")] + #[inline] + pub const fn into_inner(boxed: Self) -> T + where + Self: ~const Destruct, + { + *boxed + } +} + +impl<T> Box<[T]> { + /// Constructs a new boxed slice with uninitialized contents. + /// + /// # Examples + /// + /// ``` + /// #![feature(new_uninit)] + /// + /// let mut values = Box::<[u32]>::new_uninit_slice(3); + /// + /// let values = unsafe { + /// // Deferred initialization: + /// values[0].as_mut_ptr().write(1); + /// values[1].as_mut_ptr().write(2); + /// values[2].as_mut_ptr().write(3); + /// + /// values.assume_init() + /// }; + /// + /// assert_eq!(*values, [1, 2, 3]) + /// ``` + #[cfg(not(no_global_oom_handling))] + #[unstable(feature = "new_uninit", issue = "63291")] + #[must_use] + pub fn new_uninit_slice(len: usize) -> Box<[mem::MaybeUninit<T>]> { + unsafe { RawVec::with_capacity(len).into_box(len) } + } + + /// Constructs a new boxed slice with uninitialized contents, with the memory + /// being filled with `0` bytes. + /// + /// See [`MaybeUninit::zeroed`][zeroed] for examples of correct and incorrect usage + /// of this method. + /// + /// # Examples + /// + /// ``` + /// #![feature(new_uninit)] + /// + /// let values = Box::<[u32]>::new_zeroed_slice(3); + /// let values = unsafe { values.assume_init() }; + /// + /// assert_eq!(*values, [0, 0, 0]) + /// ``` + /// + /// [zeroed]: mem::MaybeUninit::zeroed + #[cfg(not(no_global_oom_handling))] + #[unstable(feature = "new_uninit", issue = "63291")] + #[must_use] + pub fn new_zeroed_slice(len: usize) -> Box<[mem::MaybeUninit<T>]> { + unsafe { RawVec::with_capacity_zeroed(len).into_box(len) } + } + + /// Constructs a new boxed slice with uninitialized contents. Returns an error if + /// the allocation fails + /// + /// # Examples + /// + /// ``` + /// #![feature(allocator_api, new_uninit)] + /// + /// let mut values = Box::<[u32]>::try_new_uninit_slice(3)?; + /// let values = unsafe { + /// // Deferred initialization: + /// values[0].as_mut_ptr().write(1); + /// values[1].as_mut_ptr().write(2); + /// values[2].as_mut_ptr().write(3); + /// values.assume_init() + /// }; + /// + /// assert_eq!(*values, [1, 2, 3]); + /// # Ok::<(), std::alloc::AllocError>(()) + /// ``` + #[unstable(feature = "allocator_api", issue = "32838")] + #[inline] + pub fn try_new_uninit_slice(len: usize) -> Result<Box<[mem::MaybeUninit<T>]>, AllocError> { + unsafe { + let layout = match Layout::array::<mem::MaybeUninit<T>>(len) { + Ok(l) => l, + Err(_) => return Err(AllocError), + }; + let ptr = Global.allocate(layout)?; + Ok(RawVec::from_raw_parts_in(ptr.as_mut_ptr() as *mut _, len, Global).into_box(len)) + } + } + + /// Constructs a new boxed slice with uninitialized contents, with the memory + /// being filled with `0` bytes. Returns an error if the allocation fails + /// + /// See [`MaybeUninit::zeroed`][zeroed] for examples of correct and incorrect usage + /// of this method. + /// + /// # Examples + /// + /// ``` + /// #![feature(allocator_api, new_uninit)] + /// + /// let values = Box::<[u32]>::try_new_zeroed_slice(3)?; + /// let values = unsafe { values.assume_init() }; + /// + /// assert_eq!(*values, [0, 0, 0]); + /// # Ok::<(), std::alloc::AllocError>(()) + /// ``` + /// + /// [zeroed]: mem::MaybeUninit::zeroed + #[unstable(feature = "allocator_api", issue = "32838")] + #[inline] + pub fn try_new_zeroed_slice(len: usize) -> Result<Box<[mem::MaybeUninit<T>]>, AllocError> { + unsafe { + let layout = match Layout::array::<mem::MaybeUninit<T>>(len) { + Ok(l) => l, + Err(_) => return Err(AllocError), + }; + let ptr = Global.allocate_zeroed(layout)?; + Ok(RawVec::from_raw_parts_in(ptr.as_mut_ptr() as *mut _, len, Global).into_box(len)) + } + } +} + +impl<T, A: Allocator> Box<[T], A> { + /// Constructs a new boxed slice with uninitialized contents in the provided allocator. + /// + /// # Examples + /// + /// ``` + /// #![feature(allocator_api, new_uninit)] + /// + /// use std::alloc::System; + /// + /// let mut values = Box::<[u32], _>::new_uninit_slice_in(3, System); + /// + /// let values = unsafe { + /// // Deferred initialization: + /// values[0].as_mut_ptr().write(1); + /// values[1].as_mut_ptr().write(2); + /// values[2].as_mut_ptr().write(3); + /// + /// values.assume_init() + /// }; + /// + /// assert_eq!(*values, [1, 2, 3]) + /// ``` + #[cfg(not(no_global_oom_handling))] + #[unstable(feature = "allocator_api", issue = "32838")] + // #[unstable(feature = "new_uninit", issue = "63291")] + #[must_use] + pub fn new_uninit_slice_in(len: usize, alloc: A) -> Box<[mem::MaybeUninit<T>], A> { + unsafe { RawVec::with_capacity_in(len, alloc).into_box(len) } + } + + /// Constructs a new boxed slice with uninitialized contents in the provided allocator, + /// with the memory being filled with `0` bytes. + /// + /// See [`MaybeUninit::zeroed`][zeroed] for examples of correct and incorrect usage + /// of this method. + /// + /// # Examples + /// + /// ``` + /// #![feature(allocator_api, new_uninit)] + /// + /// use std::alloc::System; + /// + /// let values = Box::<[u32], _>::new_zeroed_slice_in(3, System); + /// let values = unsafe { values.assume_init() }; + /// + /// assert_eq!(*values, [0, 0, 0]) + /// ``` + /// + /// [zeroed]: mem::MaybeUninit::zeroed + #[cfg(not(no_global_oom_handling))] + #[unstable(feature = "allocator_api", issue = "32838")] + // #[unstable(feature = "new_uninit", issue = "63291")] + #[must_use] + pub fn new_zeroed_slice_in(len: usize, alloc: A) -> Box<[mem::MaybeUninit<T>], A> { + unsafe { RawVec::with_capacity_zeroed_in(len, alloc).into_box(len) } + } +} + +impl<T, A: Allocator> Box<mem::MaybeUninit<T>, A> { + /// Converts to `Box<T, A>`. + /// + /// # Safety + /// + /// As with [`MaybeUninit::assume_init`], + /// it is up to the caller to guarantee that the value + /// really is in an initialized state. + /// Calling this when the content is not yet fully initialized + /// causes immediate undefined behavior. + /// + /// [`MaybeUninit::assume_init`]: mem::MaybeUninit::assume_init + /// + /// # Examples + /// + /// ``` + /// #![feature(new_uninit)] + /// + /// let mut five = Box::<u32>::new_uninit(); + /// + /// let five: Box<u32> = unsafe { + /// // Deferred initialization: + /// five.as_mut_ptr().write(5); + /// + /// five.assume_init() + /// }; + /// + /// assert_eq!(*five, 5) + /// ``` + #[unstable(feature = "new_uninit", issue = "63291")] + #[rustc_const_unstable(feature = "const_box", issue = "92521")] + #[inline] + pub const unsafe fn assume_init(self) -> Box<T, A> { + let (raw, alloc) = Box::into_raw_with_allocator(self); + unsafe { Box::from_raw_in(raw as *mut T, alloc) } + } + + /// Writes the value and converts to `Box<T, A>`. + /// + /// This method converts the box similarly to [`Box::assume_init`] but + /// writes `value` into it before conversion thus guaranteeing safety. + /// In some scenarios use of this method may improve performance because + /// the compiler may be able to optimize copying from stack. + /// + /// # Examples + /// + /// ``` + /// #![feature(new_uninit)] + /// + /// let big_box = Box::<[usize; 1024]>::new_uninit(); + /// + /// let mut array = [0; 1024]; + /// for (i, place) in array.iter_mut().enumerate() { + /// *place = i; + /// } + /// + /// // The optimizer may be able to elide this copy, so previous code writes + /// // to heap directly. + /// let big_box = Box::write(big_box, array); + /// + /// for (i, x) in big_box.iter().enumerate() { + /// assert_eq!(*x, i); + /// } + /// ``` + #[unstable(feature = "new_uninit", issue = "63291")] + #[rustc_const_unstable(feature = "const_box", issue = "92521")] + #[inline] + pub const fn write(mut boxed: Self, value: T) -> Box<T, A> { + unsafe { + (*boxed).write(value); + boxed.assume_init() + } + } +} + +impl<T, A: Allocator> Box<[mem::MaybeUninit<T>], A> { + /// Converts to `Box<[T], A>`. + /// + /// # Safety + /// + /// As with [`MaybeUninit::assume_init`], + /// it is up to the caller to guarantee that the values + /// really are in an initialized state. + /// Calling this when the content is not yet fully initialized + /// causes immediate undefined behavior. + /// + /// [`MaybeUninit::assume_init`]: mem::MaybeUninit::assume_init + /// + /// # Examples + /// + /// ``` + /// #![feature(new_uninit)] + /// + /// let mut values = Box::<[u32]>::new_uninit_slice(3); + /// + /// let values = unsafe { + /// // Deferred initialization: + /// values[0].as_mut_ptr().write(1); + /// values[1].as_mut_ptr().write(2); + /// values[2].as_mut_ptr().write(3); + /// + /// values.assume_init() + /// }; + /// + /// assert_eq!(*values, [1, 2, 3]) + /// ``` + #[unstable(feature = "new_uninit", issue = "63291")] + #[inline] + pub unsafe fn assume_init(self) -> Box<[T], A> { + let (raw, alloc) = Box::into_raw_with_allocator(self); + unsafe { Box::from_raw_in(raw as *mut [T], alloc) } + } +} + +impl<T: ?Sized> Box<T> { + /// Constructs a box from a raw pointer. + /// + /// After calling this function, the raw pointer is owned by the + /// resulting `Box`. Specifically, the `Box` destructor will call + /// the destructor of `T` and free the allocated memory. For this + /// to be safe, the memory must have been allocated in accordance + /// with the [memory layout] used by `Box` . + /// + /// # Safety + /// + /// This function is unsafe because improper use may lead to + /// memory problems. For example, a double-free may occur if the + /// function is called twice on the same raw pointer. + /// + /// The safety conditions are described in the [memory layout] section. + /// + /// # Examples + /// + /// Recreate a `Box` which was previously converted to a raw pointer + /// using [`Box::into_raw`]: + /// ``` + /// let x = Box::new(5); + /// let ptr = Box::into_raw(x); + /// let x = unsafe { Box::from_raw(ptr) }; + /// ``` + /// Manually create a `Box` from scratch by using the global allocator: + /// ``` + /// use std::alloc::{alloc, Layout}; + /// + /// unsafe { + /// let ptr = alloc(Layout::new::<i32>()) as *mut i32; + /// // In general .write is required to avoid attempting to destruct + /// // the (uninitialized) previous contents of `ptr`, though for this + /// // simple example `*ptr = 5` would have worked as well. + /// ptr.write(5); + /// let x = Box::from_raw(ptr); + /// } + /// ``` + /// + /// [memory layout]: self#memory-layout + /// [`Layout`]: crate::Layout + #[stable(feature = "box_raw", since = "1.4.0")] + #[inline] + pub unsafe fn from_raw(raw: *mut T) -> Self { + unsafe { Self::from_raw_in(raw, Global) } + } +} + +impl<T: ?Sized, A: Allocator> Box<T, A> { + /// Constructs a box from a raw pointer in the given allocator. + /// + /// After calling this function, the raw pointer is owned by the + /// resulting `Box`. Specifically, the `Box` destructor will call + /// the destructor of `T` and free the allocated memory. For this + /// to be safe, the memory must have been allocated in accordance + /// with the [memory layout] used by `Box` . + /// + /// # Safety + /// + /// This function is unsafe because improper use may lead to + /// memory problems. For example, a double-free may occur if the + /// function is called twice on the same raw pointer. + /// + /// + /// # Examples + /// + /// Recreate a `Box` which was previously converted to a raw pointer + /// using [`Box::into_raw_with_allocator`]: + /// ``` + /// #![feature(allocator_api)] + /// + /// use std::alloc::System; + /// + /// let x = Box::new_in(5, System); + /// let (ptr, alloc) = Box::into_raw_with_allocator(x); + /// let x = unsafe { Box::from_raw_in(ptr, alloc) }; + /// ``` + /// Manually create a `Box` from scratch by using the system allocator: + /// ``` + /// #![feature(allocator_api, slice_ptr_get)] + /// + /// use std::alloc::{Allocator, Layout, System}; + /// + /// unsafe { + /// let ptr = System.allocate(Layout::new::<i32>())?.as_mut_ptr() as *mut i32; + /// // In general .write is required to avoid attempting to destruct + /// // the (uninitialized) previous contents of `ptr`, though for this + /// // simple example `*ptr = 5` would have worked as well. + /// ptr.write(5); + /// let x = Box::from_raw_in(ptr, System); + /// } + /// # Ok::<(), std::alloc::AllocError>(()) + /// ``` + /// + /// [memory layout]: self#memory-layout + /// [`Layout`]: crate::Layout + #[unstable(feature = "allocator_api", issue = "32838")] + #[rustc_const_unstable(feature = "const_box", issue = "92521")] + #[inline] + pub const unsafe fn from_raw_in(raw: *mut T, alloc: A) -> Self { + Box(unsafe { Unique::new_unchecked(raw) }, alloc) + } + + /// Consumes the `Box`, returning a wrapped raw pointer. + /// + /// The pointer will be properly aligned and non-null. + /// + /// After calling this function, the caller is responsible for the + /// memory previously managed by the `Box`. In particular, the + /// caller should properly destroy `T` and release the memory, taking + /// into account the [memory layout] used by `Box`. The easiest way to + /// do this is to convert the raw pointer back into a `Box` with the + /// [`Box::from_raw`] function, allowing the `Box` destructor to perform + /// the cleanup. + /// + /// Note: this is an associated function, which means that you have + /// to call it as `Box::into_raw(b)` instead of `b.into_raw()`. This + /// is so that there is no conflict with a method on the inner type. + /// + /// # Examples + /// Converting the raw pointer back into a `Box` with [`Box::from_raw`] + /// for automatic cleanup: + /// ``` + /// let x = Box::new(String::from("Hello")); + /// let ptr = Box::into_raw(x); + /// let x = unsafe { Box::from_raw(ptr) }; + /// ``` + /// Manual cleanup by explicitly running the destructor and deallocating + /// the memory: + /// ``` + /// use std::alloc::{dealloc, Layout}; + /// use std::ptr; + /// + /// let x = Box::new(String::from("Hello")); + /// let p = Box::into_raw(x); + /// unsafe { + /// ptr::drop_in_place(p); + /// dealloc(p as *mut u8, Layout::new::<String>()); + /// } + /// ``` + /// + /// [memory layout]: self#memory-layout + #[stable(feature = "box_raw", since = "1.4.0")] + #[inline] + pub fn into_raw(b: Self) -> *mut T { + Self::into_raw_with_allocator(b).0 + } + + /// Consumes the `Box`, returning a wrapped raw pointer and the allocator. + /// + /// The pointer will be properly aligned and non-null. + /// + /// After calling this function, the caller is responsible for the + /// memory previously managed by the `Box`. In particular, the + /// caller should properly destroy `T` and release the memory, taking + /// into account the [memory layout] used by `Box`. The easiest way to + /// do this is to convert the raw pointer back into a `Box` with the + /// [`Box::from_raw_in`] function, allowing the `Box` destructor to perform + /// the cleanup. + /// + /// Note: this is an associated function, which means that you have + /// to call it as `Box::into_raw_with_allocator(b)` instead of `b.into_raw_with_allocator()`. This + /// is so that there is no conflict with a method on the inner type. + /// + /// # Examples + /// Converting the raw pointer back into a `Box` with [`Box::from_raw_in`] + /// for automatic cleanup: + /// ``` + /// #![feature(allocator_api)] + /// + /// use std::alloc::System; + /// + /// let x = Box::new_in(String::from("Hello"), System); + /// let (ptr, alloc) = Box::into_raw_with_allocator(x); + /// let x = unsafe { Box::from_raw_in(ptr, alloc) }; + /// ``` + /// Manual cleanup by explicitly running the destructor and deallocating + /// the memory: + /// ``` + /// #![feature(allocator_api)] + /// + /// use std::alloc::{Allocator, Layout, System}; + /// use std::ptr::{self, NonNull}; + /// + /// let x = Box::new_in(String::from("Hello"), System); + /// let (ptr, alloc) = Box::into_raw_with_allocator(x); + /// unsafe { + /// ptr::drop_in_place(ptr); + /// let non_null = NonNull::new_unchecked(ptr); + /// alloc.deallocate(non_null.cast(), Layout::new::<String>()); + /// } + /// ``` + /// + /// [memory layout]: self#memory-layout + #[unstable(feature = "allocator_api", issue = "32838")] + #[rustc_const_unstable(feature = "const_box", issue = "92521")] + #[inline] + pub const fn into_raw_with_allocator(b: Self) -> (*mut T, A) { + let (leaked, alloc) = Box::into_unique(b); + (leaked.as_ptr(), alloc) + } + + #[unstable( + feature = "ptr_internals", + issue = "none", + reason = "use `Box::leak(b).into()` or `Unique::from(Box::leak(b))` instead" + )] + #[rustc_const_unstable(feature = "const_box", issue = "92521")] + #[inline] + #[doc(hidden)] + pub const fn into_unique(b: Self) -> (Unique<T>, A) { + // Box is recognized as a "unique pointer" by Stacked Borrows, but internally it is a + // raw pointer for the type system. Turning it directly into a raw pointer would not be + // recognized as "releasing" the unique pointer to permit aliased raw accesses, + // so all raw pointer methods have to go through `Box::leak`. Turning *that* to a raw pointer + // behaves correctly. + let alloc = unsafe { ptr::read(&b.1) }; + (Unique::from(Box::leak(b)), alloc) + } + + /// Returns a reference to the underlying allocator. + /// + /// Note: this is an associated function, which means that you have + /// to call it as `Box::allocator(&b)` instead of `b.allocator()`. This + /// is so that there is no conflict with a method on the inner type. + #[unstable(feature = "allocator_api", issue = "32838")] + #[rustc_const_unstable(feature = "const_box", issue = "92521")] + #[inline] + pub const fn allocator(b: &Self) -> &A { + &b.1 + } + + /// Consumes and leaks the `Box`, returning a mutable reference, + /// `&'a mut T`. Note that the type `T` must outlive the chosen lifetime + /// `'a`. If the type has only static references, or none at all, then this + /// may be chosen to be `'static`. + /// + /// This function is mainly useful for data that lives for the remainder of + /// the program's life. Dropping the returned reference will cause a memory + /// leak. If this is not acceptable, the reference should first be wrapped + /// with the [`Box::from_raw`] function producing a `Box`. This `Box` can + /// then be dropped which will properly destroy `T` and release the + /// allocated memory. + /// + /// Note: this is an associated function, which means that you have + /// to call it as `Box::leak(b)` instead of `b.leak()`. This + /// is so that there is no conflict with a method on the inner type. + /// + /// # Examples + /// + /// Simple usage: + /// + /// ``` + /// let x = Box::new(41); + /// let static_ref: &'static mut usize = Box::leak(x); + /// *static_ref += 1; + /// assert_eq!(*static_ref, 42); + /// ``` + /// + /// Unsized data: + /// + /// ``` + /// let x = vec![1, 2, 3].into_boxed_slice(); + /// let static_ref = Box::leak(x); + /// static_ref[0] = 4; + /// assert_eq!(*static_ref, [4, 2, 3]); + /// ``` + #[stable(feature = "box_leak", since = "1.26.0")] + #[rustc_const_unstable(feature = "const_box", issue = "92521")] + #[inline] + pub const fn leak<'a>(b: Self) -> &'a mut T + where + A: 'a, + { + unsafe { &mut *mem::ManuallyDrop::new(b).0.as_ptr() } + } + + /// Converts a `Box<T>` into a `Pin<Box<T>>` + /// + /// This conversion does not allocate on the heap and happens in place. + /// + /// This is also available via [`From`]. + #[unstable(feature = "box_into_pin", issue = "62370")] + #[rustc_const_unstable(feature = "const_box", issue = "92521")] + pub const fn into_pin(boxed: Self) -> Pin<Self> + where + A: 'static, + { + // It's not possible to move or replace the insides of a `Pin<Box<T>>` + // when `T: !Unpin`, so it's safe to pin it directly without any + // additional requirements. + unsafe { Pin::new_unchecked(boxed) } + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +unsafe impl<#[may_dangle] T: ?Sized, A: Allocator> Drop for Box<T, A> { + fn drop(&mut self) { + // FIXME: Do nothing, drop is currently performed by compiler. + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "rust1", since = "1.0.0")] +impl<T: Default> Default for Box<T> { + /// Creates a `Box<T>`, with the `Default` value for T. + fn default() -> Self { + box T::default() + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "rust1", since = "1.0.0")] +#[rustc_const_unstable(feature = "const_default_impls", issue = "87864")] +impl<T> const Default for Box<[T]> { + fn default() -> Self { + let ptr: Unique<[T]> = Unique::<[T; 0]>::dangling(); + Box(ptr, Global) + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "default_box_extra", since = "1.17.0")] +#[rustc_const_unstable(feature = "const_default_impls", issue = "87864")] +impl const Default for Box<str> { + fn default() -> Self { + // SAFETY: This is the same as `Unique::cast<U>` but with an unsized `U = str`. + let ptr: Unique<str> = unsafe { + let bytes: Unique<[u8]> = Unique::<[u8; 0]>::dangling(); + Unique::new_unchecked(bytes.as_ptr() as *mut str) + }; + Box(ptr, Global) + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "rust1", since = "1.0.0")] +impl<T: Clone, A: Allocator + Clone> Clone for Box<T, A> { + /// Returns a new box with a `clone()` of this box's contents. + /// + /// # Examples + /// + /// ``` + /// let x = Box::new(5); + /// let y = x.clone(); + /// + /// // The value is the same + /// assert_eq!(x, y); + /// + /// // But they are unique objects + /// assert_ne!(&*x as *const i32, &*y as *const i32); + /// ``` + #[inline] + fn clone(&self) -> Self { + // Pre-allocate memory to allow writing the cloned value directly. + let mut boxed = Self::new_uninit_in(self.1.clone()); + unsafe { + (**self).write_clone_into_raw(boxed.as_mut_ptr()); + boxed.assume_init() + } + } + + /// Copies `source`'s contents into `self` without creating a new allocation. + /// + /// # Examples + /// + /// ``` + /// let x = Box::new(5); + /// let mut y = Box::new(10); + /// let yp: *const i32 = &*y; + /// + /// y.clone_from(&x); + /// + /// // The value is the same + /// assert_eq!(x, y); + /// + /// // And no allocation occurred + /// assert_eq!(yp, &*y); + /// ``` + #[inline] + fn clone_from(&mut self, source: &Self) { + (**self).clone_from(&(**source)); + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "box_slice_clone", since = "1.3.0")] +impl Clone for Box<str> { + fn clone(&self) -> Self { + // this makes a copy of the data + let buf: Box<[u8]> = self.as_bytes().into(); + unsafe { from_boxed_utf8_unchecked(buf) } + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<T: ?Sized + PartialEq, A: Allocator> PartialEq for Box<T, A> { + #[inline] + fn eq(&self, other: &Self) -> bool { + PartialEq::eq(&**self, &**other) + } + #[inline] + fn ne(&self, other: &Self) -> bool { + PartialEq::ne(&**self, &**other) + } +} +#[stable(feature = "rust1", since = "1.0.0")] +impl<T: ?Sized + PartialOrd, A: Allocator> PartialOrd for Box<T, A> { + #[inline] + fn partial_cmp(&self, other: &Self) -> Option<Ordering> { + PartialOrd::partial_cmp(&**self, &**other) + } + #[inline] + fn lt(&self, other: &Self) -> bool { + PartialOrd::lt(&**self, &**other) + } + #[inline] + fn le(&self, other: &Self) -> bool { + PartialOrd::le(&**self, &**other) + } + #[inline] + fn ge(&self, other: &Self) -> bool { + PartialOrd::ge(&**self, &**other) + } + #[inline] + fn gt(&self, other: &Self) -> bool { + PartialOrd::gt(&**self, &**other) + } +} +#[stable(feature = "rust1", since = "1.0.0")] +impl<T: ?Sized + Ord, A: Allocator> Ord for Box<T, A> { + #[inline] + fn cmp(&self, other: &Self) -> Ordering { + Ord::cmp(&**self, &**other) + } +} +#[stable(feature = "rust1", since = "1.0.0")] +impl<T: ?Sized + Eq, A: Allocator> Eq for Box<T, A> {} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<T: ?Sized + Hash, A: Allocator> Hash for Box<T, A> { + fn hash<H: Hasher>(&self, state: &mut H) { + (**self).hash(state); + } +} + +#[stable(feature = "indirect_hasher_impl", since = "1.22.0")] +impl<T: ?Sized + Hasher, A: Allocator> Hasher for Box<T, A> { + fn finish(&self) -> u64 { + (**self).finish() + } + fn write(&mut self, bytes: &[u8]) { + (**self).write(bytes) + } + fn write_u8(&mut self, i: u8) { + (**self).write_u8(i) + } + fn write_u16(&mut self, i: u16) { + (**self).write_u16(i) + } + fn write_u32(&mut self, i: u32) { + (**self).write_u32(i) + } + fn write_u64(&mut self, i: u64) { + (**self).write_u64(i) + } + fn write_u128(&mut self, i: u128) { + (**self).write_u128(i) + } + fn write_usize(&mut self, i: usize) { + (**self).write_usize(i) + } + fn write_i8(&mut self, i: i8) { + (**self).write_i8(i) + } + fn write_i16(&mut self, i: i16) { + (**self).write_i16(i) + } + fn write_i32(&mut self, i: i32) { + (**self).write_i32(i) + } + fn write_i64(&mut self, i: i64) { + (**self).write_i64(i) + } + fn write_i128(&mut self, i: i128) { + (**self).write_i128(i) + } + fn write_isize(&mut self, i: isize) { + (**self).write_isize(i) + } + fn write_length_prefix(&mut self, len: usize) { + (**self).write_length_prefix(len) + } + fn write_str(&mut self, s: &str) { + (**self).write_str(s) + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "from_for_ptrs", since = "1.6.0")] +impl<T> From<T> for Box<T> { + /// Converts a `T` into a `Box<T>` + /// + /// The conversion allocates on the heap and moves `t` + /// from the stack into it. + /// + /// # Examples + /// + /// ```rust + /// let x = 5; + /// let boxed = Box::new(5); + /// + /// assert_eq!(Box::from(x), boxed); + /// ``` + fn from(t: T) -> Self { + Box::new(t) + } +} + +#[stable(feature = "pin", since = "1.33.0")] +#[rustc_const_unstable(feature = "const_box", issue = "92521")] +impl<T: ?Sized, A: Allocator> const From<Box<T, A>> for Pin<Box<T, A>> +where + A: 'static, +{ + /// Converts a `Box<T>` into a `Pin<Box<T>>` + /// + /// This conversion does not allocate on the heap and happens in place. + fn from(boxed: Box<T, A>) -> Self { + Box::into_pin(boxed) + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "box_from_slice", since = "1.17.0")] +impl<T: Copy> From<&[T]> for Box<[T]> { + /// Converts a `&[T]` into a `Box<[T]>` + /// + /// This conversion allocates on the heap + /// and performs a copy of `slice`. + /// + /// # Examples + /// ```rust + /// // create a &[u8] which will be used to create a Box<[u8]> + /// let slice: &[u8] = &[104, 101, 108, 108, 111]; + /// let boxed_slice: Box<[u8]> = Box::from(slice); + /// + /// println!("{boxed_slice:?}"); + /// ``` + fn from(slice: &[T]) -> Box<[T]> { + let len = slice.len(); + let buf = RawVec::with_capacity(len); + unsafe { + ptr::copy_nonoverlapping(slice.as_ptr(), buf.ptr(), len); + buf.into_box(slice.len()).assume_init() + } + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "box_from_cow", since = "1.45.0")] +impl<T: Copy> From<Cow<'_, [T]>> for Box<[T]> { + /// Converts a `Cow<'_, [T]>` into a `Box<[T]>` + /// + /// When `cow` is the `Cow::Borrowed` variant, this + /// conversion allocates on the heap and copies the + /// underlying slice. Otherwise, it will try to reuse the owned + /// `Vec`'s allocation. + #[inline] + fn from(cow: Cow<'_, [T]>) -> Box<[T]> { + match cow { + Cow::Borrowed(slice) => Box::from(slice), + Cow::Owned(slice) => Box::from(slice), + } + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "box_from_slice", since = "1.17.0")] +impl From<&str> for Box<str> { + /// Converts a `&str` into a `Box<str>` + /// + /// This conversion allocates on the heap + /// and performs a copy of `s`. + /// + /// # Examples + /// + /// ```rust + /// let boxed: Box<str> = Box::from("hello"); + /// println!("{boxed}"); + /// ``` + #[inline] + fn from(s: &str) -> Box<str> { + unsafe { from_boxed_utf8_unchecked(Box::from(s.as_bytes())) } + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "box_from_cow", since = "1.45.0")] +impl From<Cow<'_, str>> for Box<str> { + /// Converts a `Cow<'_, str>` into a `Box<str>` + /// + /// When `cow` is the `Cow::Borrowed` variant, this + /// conversion allocates on the heap and copies the + /// underlying `str`. Otherwise, it will try to reuse the owned + /// `String`'s allocation. + /// + /// # Examples + /// + /// ```rust + /// use std::borrow::Cow; + /// + /// let unboxed = Cow::Borrowed("hello"); + /// let boxed: Box<str> = Box::from(unboxed); + /// println!("{boxed}"); + /// ``` + /// + /// ```rust + /// # use std::borrow::Cow; + /// let unboxed = Cow::Owned("hello".to_string()); + /// let boxed: Box<str> = Box::from(unboxed); + /// println!("{boxed}"); + /// ``` + #[inline] + fn from(cow: Cow<'_, str>) -> Box<str> { + match cow { + Cow::Borrowed(s) => Box::from(s), + Cow::Owned(s) => Box::from(s), + } + } +} + +#[stable(feature = "boxed_str_conv", since = "1.19.0")] +impl<A: Allocator> From<Box<str, A>> for Box<[u8], A> { + /// Converts a `Box<str>` into a `Box<[u8]>` + /// + /// This conversion does not allocate on the heap and happens in place. + /// + /// # Examples + /// ```rust + /// // create a Box<str> which will be used to create a Box<[u8]> + /// let boxed: Box<str> = Box::from("hello"); + /// let boxed_str: Box<[u8]> = Box::from(boxed); + /// + /// // create a &[u8] which will be used to create a Box<[u8]> + /// let slice: &[u8] = &[104, 101, 108, 108, 111]; + /// let boxed_slice = Box::from(slice); + /// + /// assert_eq!(boxed_slice, boxed_str); + /// ``` + #[inline] + fn from(s: Box<str, A>) -> Self { + let (raw, alloc) = Box::into_raw_with_allocator(s); + unsafe { Box::from_raw_in(raw as *mut [u8], alloc) } + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "box_from_array", since = "1.45.0")] +impl<T, const N: usize> From<[T; N]> for Box<[T]> { + /// Converts a `[T; N]` into a `Box<[T]>` + /// + /// This conversion moves the array to newly heap-allocated memory. + /// + /// # Examples + /// + /// ```rust + /// let boxed: Box<[u8]> = Box::from([4, 2]); + /// println!("{boxed:?}"); + /// ``` + fn from(array: [T; N]) -> Box<[T]> { + box array + } +} + +#[stable(feature = "boxed_slice_try_from", since = "1.43.0")] +impl<T, const N: usize> TryFrom<Box<[T]>> for Box<[T; N]> { + type Error = Box<[T]>; + + /// Attempts to convert a `Box<[T]>` into a `Box<[T; N]>`. + /// + /// The conversion occurs in-place and does not require a + /// new memory allocation. + /// + /// # Errors + /// + /// Returns the old `Box<[T]>` in the `Err` variant if + /// `boxed_slice.len()` does not equal `N`. + fn try_from(boxed_slice: Box<[T]>) -> Result<Self, Self::Error> { + if boxed_slice.len() == N { + Ok(unsafe { Box::from_raw(Box::into_raw(boxed_slice) as *mut [T; N]) }) + } else { + Err(boxed_slice) + } + } +} + +impl<A: Allocator> Box<dyn Any, A> { + /// Attempt to downcast the box to a concrete type. + /// + /// # Examples + /// + /// ``` + /// use std::any::Any; + /// + /// fn print_if_string(value: Box<dyn Any>) { + /// if let Ok(string) = value.downcast::<String>() { + /// println!("String ({}): {}", string.len(), string); + /// } + /// } + /// + /// let my_string = "Hello World".to_string(); + /// print_if_string(Box::new(my_string)); + /// print_if_string(Box::new(0i8)); + /// ``` + #[inline] + #[stable(feature = "rust1", since = "1.0.0")] + pub fn downcast<T: Any>(self) -> Result<Box<T, A>, Self> { + if self.is::<T>() { unsafe { Ok(self.downcast_unchecked::<T>()) } } else { Err(self) } + } + + /// Downcasts the box to a concrete type. + /// + /// For a safe alternative see [`downcast`]. + /// + /// # Examples + /// + /// ``` + /// #![feature(downcast_unchecked)] + /// + /// use std::any::Any; + /// + /// let x: Box<dyn Any> = Box::new(1_usize); + /// + /// unsafe { + /// assert_eq!(*x.downcast_unchecked::<usize>(), 1); + /// } + /// ``` + /// + /// # Safety + /// + /// The contained value must be of type `T`. Calling this method + /// with the incorrect type is *undefined behavior*. + /// + /// [`downcast`]: Self::downcast + #[inline] + #[unstable(feature = "downcast_unchecked", issue = "90850")] + pub unsafe fn downcast_unchecked<T: Any>(self) -> Box<T, A> { + debug_assert!(self.is::<T>()); + unsafe { + let (raw, alloc): (*mut dyn Any, _) = Box::into_raw_with_allocator(self); + Box::from_raw_in(raw as *mut T, alloc) + } + } +} + +impl<A: Allocator> Box<dyn Any + Send, A> { + /// Attempt to downcast the box to a concrete type. + /// + /// # Examples + /// + /// ``` + /// use std::any::Any; + /// + /// fn print_if_string(value: Box<dyn Any + Send>) { + /// if let Ok(string) = value.downcast::<String>() { + /// println!("String ({}): {}", string.len(), string); + /// } + /// } + /// + /// let my_string = "Hello World".to_string(); + /// print_if_string(Box::new(my_string)); + /// print_if_string(Box::new(0i8)); + /// ``` + #[inline] + #[stable(feature = "rust1", since = "1.0.0")] + pub fn downcast<T: Any>(self) -> Result<Box<T, A>, Self> { + if self.is::<T>() { unsafe { Ok(self.downcast_unchecked::<T>()) } } else { Err(self) } + } + + /// Downcasts the box to a concrete type. + /// + /// For a safe alternative see [`downcast`]. + /// + /// # Examples + /// + /// ``` + /// #![feature(downcast_unchecked)] + /// + /// use std::any::Any; + /// + /// let x: Box<dyn Any + Send> = Box::new(1_usize); + /// + /// unsafe { + /// assert_eq!(*x.downcast_unchecked::<usize>(), 1); + /// } + /// ``` + /// + /// # Safety + /// + /// The contained value must be of type `T`. Calling this method + /// with the incorrect type is *undefined behavior*. + /// + /// [`downcast`]: Self::downcast + #[inline] + #[unstable(feature = "downcast_unchecked", issue = "90850")] + pub unsafe fn downcast_unchecked<T: Any>(self) -> Box<T, A> { + debug_assert!(self.is::<T>()); + unsafe { + let (raw, alloc): (*mut (dyn Any + Send), _) = Box::into_raw_with_allocator(self); + Box::from_raw_in(raw as *mut T, alloc) + } + } +} + +impl<A: Allocator> Box<dyn Any + Send + Sync, A> { + /// Attempt to downcast the box to a concrete type. + /// + /// # Examples + /// + /// ``` + /// use std::any::Any; + /// + /// fn print_if_string(value: Box<dyn Any + Send + Sync>) { + /// if let Ok(string) = value.downcast::<String>() { + /// println!("String ({}): {}", string.len(), string); + /// } + /// } + /// + /// let my_string = "Hello World".to_string(); + /// print_if_string(Box::new(my_string)); + /// print_if_string(Box::new(0i8)); + /// ``` + #[inline] + #[stable(feature = "box_send_sync_any_downcast", since = "1.51.0")] + pub fn downcast<T: Any>(self) -> Result<Box<T, A>, Self> { + if self.is::<T>() { unsafe { Ok(self.downcast_unchecked::<T>()) } } else { Err(self) } + } + + /// Downcasts the box to a concrete type. + /// + /// For a safe alternative see [`downcast`]. + /// + /// # Examples + /// + /// ``` + /// #![feature(downcast_unchecked)] + /// + /// use std::any::Any; + /// + /// let x: Box<dyn Any + Send + Sync> = Box::new(1_usize); + /// + /// unsafe { + /// assert_eq!(*x.downcast_unchecked::<usize>(), 1); + /// } + /// ``` + /// + /// # Safety + /// + /// The contained value must be of type `T`. Calling this method + /// with the incorrect type is *undefined behavior*. + /// + /// [`downcast`]: Self::downcast + #[inline] + #[unstable(feature = "downcast_unchecked", issue = "90850")] + pub unsafe fn downcast_unchecked<T: Any>(self) -> Box<T, A> { + debug_assert!(self.is::<T>()); + unsafe { + let (raw, alloc): (*mut (dyn Any + Send + Sync), _) = + Box::into_raw_with_allocator(self); + Box::from_raw_in(raw as *mut T, alloc) + } + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<T: fmt::Display + ?Sized, A: Allocator> fmt::Display for Box<T, A> { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + fmt::Display::fmt(&**self, f) + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<T: fmt::Debug + ?Sized, A: Allocator> fmt::Debug for Box<T, A> { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + fmt::Debug::fmt(&**self, f) + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<T: ?Sized, A: Allocator> fmt::Pointer for Box<T, A> { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + // It's not possible to extract the inner Uniq directly from the Box, + // instead we cast it to a *const which aliases the Unique + let ptr: *const T = &**self; + fmt::Pointer::fmt(&ptr, f) + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +#[rustc_const_unstable(feature = "const_box", issue = "92521")] +impl<T: ?Sized, A: Allocator> const Deref for Box<T, A> { + type Target = T; + + fn deref(&self) -> &T { + &**self + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +#[rustc_const_unstable(feature = "const_box", issue = "92521")] +impl<T: ?Sized, A: Allocator> const DerefMut for Box<T, A> { + fn deref_mut(&mut self) -> &mut T { + &mut **self + } +} + +#[unstable(feature = "receiver_trait", issue = "none")] +impl<T: ?Sized, A: Allocator> Receiver for Box<T, A> {} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<I: Iterator + ?Sized, A: Allocator> Iterator for Box<I, A> { + type Item = I::Item; + fn next(&mut self) -> Option<I::Item> { + (**self).next() + } + fn size_hint(&self) -> (usize, Option<usize>) { + (**self).size_hint() + } + fn nth(&mut self, n: usize) -> Option<I::Item> { + (**self).nth(n) + } + fn last(self) -> Option<I::Item> { + BoxIter::last(self) + } +} + +trait BoxIter { + type Item; + fn last(self) -> Option<Self::Item>; +} + +impl<I: Iterator + ?Sized, A: Allocator> BoxIter for Box<I, A> { + type Item = I::Item; + default fn last(self) -> Option<I::Item> { + #[inline] + fn some<T>(_: Option<T>, x: T) -> Option<T> { + Some(x) + } + + self.fold(None, some) + } +} + +/// Specialization for sized `I`s that uses `I`s implementation of `last()` +/// instead of the default. +#[stable(feature = "rust1", since = "1.0.0")] +impl<I: Iterator, A: Allocator> BoxIter for Box<I, A> { + fn last(self) -> Option<I::Item> { + (*self).last() + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<I: DoubleEndedIterator + ?Sized, A: Allocator> DoubleEndedIterator for Box<I, A> { + fn next_back(&mut self) -> Option<I::Item> { + (**self).next_back() + } + fn nth_back(&mut self, n: usize) -> Option<I::Item> { + (**self).nth_back(n) + } +} +#[stable(feature = "rust1", since = "1.0.0")] +impl<I: ExactSizeIterator + ?Sized, A: Allocator> ExactSizeIterator for Box<I, A> { + fn len(&self) -> usize { + (**self).len() + } + fn is_empty(&self) -> bool { + (**self).is_empty() + } +} + +#[stable(feature = "fused", since = "1.26.0")] +impl<I: FusedIterator + ?Sized, A: Allocator> FusedIterator for Box<I, A> {} + +#[stable(feature = "boxed_closure_impls", since = "1.35.0")] +impl<Args, F: FnOnce<Args> + ?Sized, A: Allocator> FnOnce<Args> for Box<F, A> { + type Output = <F as FnOnce<Args>>::Output; + + extern "rust-call" fn call_once(self, args: Args) -> Self::Output { + <F as FnOnce<Args>>::call_once(*self, args) + } +} + +#[stable(feature = "boxed_closure_impls", since = "1.35.0")] +impl<Args, F: FnMut<Args> + ?Sized, A: Allocator> FnMut<Args> for Box<F, A> { + extern "rust-call" fn call_mut(&mut self, args: Args) -> Self::Output { + <F as FnMut<Args>>::call_mut(self, args) + } +} + +#[stable(feature = "boxed_closure_impls", since = "1.35.0")] +impl<Args, F: Fn<Args> + ?Sized, A: Allocator> Fn<Args> for Box<F, A> { + extern "rust-call" fn call(&self, args: Args) -> Self::Output { + <F as Fn<Args>>::call(self, args) + } +} + +#[unstable(feature = "coerce_unsized", issue = "27732")] +impl<T: ?Sized + Unsize<U>, U: ?Sized, A: Allocator> CoerceUnsized<Box<U, A>> for Box<T, A> {} + +#[unstable(feature = "dispatch_from_dyn", issue = "none")] +impl<T: ?Sized + Unsize<U>, U: ?Sized> DispatchFromDyn<Box<U>> for Box<T, Global> {} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "boxed_slice_from_iter", since = "1.32.0")] +impl<I> FromIterator<I> for Box<[I]> { + fn from_iter<T: IntoIterator<Item = I>>(iter: T) -> Self { + iter.into_iter().collect::<Vec<_>>().into_boxed_slice() + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "box_slice_clone", since = "1.3.0")] +impl<T: Clone, A: Allocator + Clone> Clone for Box<[T], A> { + fn clone(&self) -> Self { + let alloc = Box::allocator(self).clone(); + self.to_vec_in(alloc).into_boxed_slice() + } + + fn clone_from(&mut self, other: &Self) { + if self.len() == other.len() { + self.clone_from_slice(&other); + } else { + *self = other.clone(); + } + } +} + +#[stable(feature = "box_borrow", since = "1.1.0")] +impl<T: ?Sized, A: Allocator> borrow::Borrow<T> for Box<T, A> { + fn borrow(&self) -> &T { + &**self + } +} + +#[stable(feature = "box_borrow", since = "1.1.0")] +impl<T: ?Sized, A: Allocator> borrow::BorrowMut<T> for Box<T, A> { + fn borrow_mut(&mut self) -> &mut T { + &mut **self + } +} + +#[stable(since = "1.5.0", feature = "smart_ptr_as_ref")] +impl<T: ?Sized, A: Allocator> AsRef<T> for Box<T, A> { + fn as_ref(&self) -> &T { + &**self + } +} + +#[stable(since = "1.5.0", feature = "smart_ptr_as_ref")] +impl<T: ?Sized, A: Allocator> AsMut<T> for Box<T, A> { + fn as_mut(&mut self) -> &mut T { + &mut **self + } +} + +/* Nota bene + * + * We could have chosen not to add this impl, and instead have written a + * function of Pin<Box<T>> to Pin<T>. Such a function would not be sound, + * because Box<T> implements Unpin even when T does not, as a result of + * this impl. + * + * We chose this API instead of the alternative for a few reasons: + * - Logically, it is helpful to understand pinning in regard to the + * memory region being pointed to. For this reason none of the + * standard library pointer types support projecting through a pin + * (Box<T> is the only pointer type in std for which this would be + * safe.) + * - It is in practice very useful to have Box<T> be unconditionally + * Unpin because of trait objects, for which the structural auto + * trait functionality does not apply (e.g., Box<dyn Foo> would + * otherwise not be Unpin). + * + * Another type with the same semantics as Box but only a conditional + * implementation of `Unpin` (where `T: Unpin`) would be valid/safe, and + * could have a method to project a Pin<T> from it. + */ +#[stable(feature = "pin", since = "1.33.0")] +#[rustc_const_unstable(feature = "const_box", issue = "92521")] +impl<T: ?Sized, A: Allocator> const Unpin for Box<T, A> where A: 'static {} + +#[unstable(feature = "generator_trait", issue = "43122")] +impl<G: ?Sized + Generator<R> + Unpin, R, A: Allocator> Generator<R> for Box<G, A> +where + A: 'static, +{ + type Yield = G::Yield; + type Return = G::Return; + + fn resume(mut self: Pin<&mut Self>, arg: R) -> GeneratorState<Self::Yield, Self::Return> { + G::resume(Pin::new(&mut *self), arg) + } +} + +#[unstable(feature = "generator_trait", issue = "43122")] +impl<G: ?Sized + Generator<R>, R, A: Allocator> Generator<R> for Pin<Box<G, A>> +where + A: 'static, +{ + type Yield = G::Yield; + type Return = G::Return; + + fn resume(mut self: Pin<&mut Self>, arg: R) -> GeneratorState<Self::Yield, Self::Return> { + G::resume((*self).as_mut(), arg) + } +} + +#[stable(feature = "futures_api", since = "1.36.0")] +impl<F: ?Sized + Future + Unpin, A: Allocator> Future for Box<F, A> +where + A: 'static, +{ + type Output = F::Output; + + fn poll(mut self: Pin<&mut Self>, cx: &mut Context<'_>) -> Poll<Self::Output> { + F::poll(Pin::new(&mut *self), cx) + } +} + +#[unstable(feature = "async_iterator", issue = "79024")] +impl<S: ?Sized + AsyncIterator + Unpin> AsyncIterator for Box<S> { + type Item = S::Item; + + fn poll_next(mut self: Pin<&mut Self>, cx: &mut Context<'_>) -> Poll<Option<Self::Item>> { + Pin::new(&mut **self).poll_next(cx) + } + + fn size_hint(&self) -> (usize, Option<usize>) { + (**self).size_hint() + } +} diff --git a/rust/alloc/collections/mod.rs b/rust/alloc/collections/mod.rs new file mode 100644 index 000000000000..1eec265b28f8 --- /dev/null +++ b/rust/alloc/collections/mod.rs @@ -0,0 +1,156 @@ +// SPDX-License-Identifier: Apache-2.0 OR MIT + +//! Collection types. + +#![stable(feature = "rust1", since = "1.0.0")] + +#[cfg(not(no_global_oom_handling))] +pub mod binary_heap; +#[cfg(not(no_global_oom_handling))] +mod btree; +#[cfg(not(no_global_oom_handling))] +pub mod linked_list; +#[cfg(not(no_global_oom_handling))] +pub mod vec_deque; + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "rust1", since = "1.0.0")] +pub mod btree_map { + //! An ordered map based on a B-Tree. + #[stable(feature = "rust1", since = "1.0.0")] + pub use super::btree::map::*; +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "rust1", since = "1.0.0")] +pub mod btree_set { + //! An ordered set based on a B-Tree. + #[stable(feature = "rust1", since = "1.0.0")] + pub use super::btree::set::*; +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "rust1", since = "1.0.0")] +#[doc(no_inline)] +pub use binary_heap::BinaryHeap; + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "rust1", since = "1.0.0")] +#[doc(no_inline)] +pub use btree_map::BTreeMap; + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "rust1", since = "1.0.0")] +#[doc(no_inline)] +pub use btree_set::BTreeSet; + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "rust1", since = "1.0.0")] +#[doc(no_inline)] +pub use linked_list::LinkedList; + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "rust1", since = "1.0.0")] +#[doc(no_inline)] +pub use vec_deque::VecDeque; + +use crate::alloc::{Layout, LayoutError}; +use core::fmt::Display; + +/// The error type for `try_reserve` methods. +#[derive(Clone, PartialEq, Eq, Debug)] +#[stable(feature = "try_reserve", since = "1.57.0")] +pub struct TryReserveError { + kind: TryReserveErrorKind, +} + +impl TryReserveError { + /// Details about the allocation that caused the error + #[inline] + #[must_use] + #[unstable( + feature = "try_reserve_kind", + reason = "Uncertain how much info should be exposed", + issue = "48043" + )] + pub fn kind(&self) -> TryReserveErrorKind { + self.kind.clone() + } +} + +/// Details of the allocation that caused a `TryReserveError` +#[derive(Clone, PartialEq, Eq, Debug)] +#[unstable( + feature = "try_reserve_kind", + reason = "Uncertain how much info should be exposed", + issue = "48043" +)] +pub enum TryReserveErrorKind { + /// Error due to the computed capacity exceeding the collection's maximum + /// (usually `isize::MAX` bytes). + CapacityOverflow, + + /// The memory allocator returned an error + AllocError { + /// The layout of allocation request that failed + layout: Layout, + + #[doc(hidden)] + #[unstable( + feature = "container_error_extra", + issue = "none", + reason = "\ + Enable exposing the allocator’s custom error value \ + if an associated type is added in the future: \ + https://github.com/rust-lang/wg-allocators/issues/23" + )] + non_exhaustive: (), + }, +} + +#[unstable( + feature = "try_reserve_kind", + reason = "Uncertain how much info should be exposed", + issue = "48043" +)] +impl From<TryReserveErrorKind> for TryReserveError { + #[inline] + fn from(kind: TryReserveErrorKind) -> Self { + Self { kind } + } +} + +#[unstable(feature = "try_reserve_kind", reason = "new API", issue = "48043")] +impl From<LayoutError> for TryReserveErrorKind { + /// Always evaluates to [`TryReserveErrorKind::CapacityOverflow`]. + #[inline] + fn from(_: LayoutError) -> Self { + TryReserveErrorKind::CapacityOverflow + } +} + +#[stable(feature = "try_reserve", since = "1.57.0")] +impl Display for TryReserveError { + fn fmt( + &self, + fmt: &mut core::fmt::Formatter<'_>, + ) -> core::result::Result<(), core::fmt::Error> { + fmt.write_str("memory allocation failed")?; + let reason = match self.kind { + TryReserveErrorKind::CapacityOverflow => { + " because the computed capacity exceeded the collection's maximum" + } + TryReserveErrorKind::AllocError { .. } => { + " because the memory allocator returned a error" + } + }; + fmt.write_str(reason) + } +} + +/// An intermediate trait for specialization of `Extend`. +#[doc(hidden)] +trait SpecExtend<I: IntoIterator> { + /// Extends `self` with the contents of the given iterator. + fn spec_extend(&mut self, iter: I); +} diff --git a/rust/alloc/lib.rs b/rust/alloc/lib.rs new file mode 100644 index 000000000000..233bcd5e4654 --- /dev/null +++ b/rust/alloc/lib.rs @@ -0,0 +1,244 @@ +// SPDX-License-Identifier: Apache-2.0 OR MIT + +//! # The Rust core allocation and collections library +//! +//! This library provides smart pointers and collections for managing +//! heap-allocated values. +//! +//! This library, like libcore, normally doesn’t need to be used directly +//! since its contents are re-exported in the [`std` crate](../std/index.html). +//! Crates that use the `#![no_std]` attribute however will typically +//! not depend on `std`, so they’d use this crate instead. +//! +//! ## Boxed values +//! +//! The [`Box`] type is a smart pointer type. There can only be one owner of a +//! [`Box`], and the owner can decide to mutate the contents, which live on the +//! heap. +//! +//! This type can be sent among threads efficiently as the size of a `Box` value +//! is the same as that of a pointer. Tree-like data structures are often built +//! with boxes because each node often has only one owner, the parent. +//! +//! ## Reference counted pointers +//! +//! The [`Rc`] type is a non-threadsafe reference-counted pointer type intended +//! for sharing memory within a thread. An [`Rc`] pointer wraps a type, `T`, and +//! only allows access to `&T`, a shared reference. +//! +//! This type is useful when inherited mutability (such as using [`Box`]) is too +//! constraining for an application, and is often paired with the [`Cell`] or +//! [`RefCell`] types in order to allow mutation. +//! +//! ## Atomically reference counted pointers +//! +//! The [`Arc`] type is the threadsafe equivalent of the [`Rc`] type. It +//! provides all the same functionality of [`Rc`], except it requires that the +//! contained type `T` is shareable. Additionally, [`Arc<T>`][`Arc`] is itself +//! sendable while [`Rc<T>`][`Rc`] is not. +//! +//! This type allows for shared access to the contained data, and is often +//! paired with synchronization primitives such as mutexes to allow mutation of +//! shared resources. +//! +//! ## Collections +//! +//! Implementations of the most common general purpose data structures are +//! defined in this library. They are re-exported through the +//! [standard collections library](../std/collections/index.html). +//! +//! ## Heap interfaces +//! +//! The [`alloc`](alloc/index.html) module defines the low-level interface to the +//! default global allocator. It is not compatible with the libc allocator API. +//! +//! [`Arc`]: sync +//! [`Box`]: boxed +//! [`Cell`]: core::cell +//! [`Rc`]: rc +//! [`RefCell`]: core::cell + +// To run liballoc tests without x.py without ending up with two copies of liballoc, Miri needs to be +// able to "empty" this crate. See <https://github.com/rust-lang/miri-test-libstd/issues/4>. +// rustc itself never sets the feature, so this line has no affect there. +#![cfg(any(not(feature = "miri-test-libstd"), test, doctest))] +#![allow(unused_attributes)] +#![stable(feature = "alloc", since = "1.36.0")] +#![doc( + html_playground_url = "https://play.rust-lang.org/", + issue_tracker_base_url = "https://github.com/rust-lang/rust/issues/", + test(no_crate_inject, attr(allow(unused_variables), deny(warnings))) +)] +#![doc(cfg_hide( + not(test), + not(any(test, bootstrap)), + any(not(feature = "miri-test-libstd"), test, doctest), + no_global_oom_handling, + not(no_global_oom_handling), + target_has_atomic = "ptr" +))] +#![no_std] +#![needs_allocator] +// +// Lints: +#![deny(unsafe_op_in_unsafe_fn)] +#![warn(deprecated_in_future)] +#![warn(missing_debug_implementations)] +#![warn(missing_docs)] +#![allow(explicit_outlives_requirements)] +// +// Library features: +#![cfg_attr(not(no_global_oom_handling), feature(alloc_c_string))] +#![feature(alloc_layout_extra)] +#![feature(allocator_api)] +#![feature(array_chunks)] +#![feature(array_methods)] +#![feature(array_windows)] +#![feature(assert_matches)] +#![feature(async_iterator)] +#![feature(coerce_unsized)] +#![cfg_attr(not(no_global_oom_handling), feature(const_alloc_error))] +#![feature(const_box)] +#![cfg_attr(not(no_global_oom_handling), feature(const_btree_new))] +#![feature(const_cow_is_borrowed)] +#![feature(const_convert)] +#![feature(const_size_of_val)] +#![feature(const_align_of_val)] +#![feature(const_ptr_read)] +#![feature(const_maybe_uninit_write)] +#![feature(const_maybe_uninit_as_mut_ptr)] +#![feature(const_refs_to_cell)] +#![feature(core_c_str)] +#![feature(core_intrinsics)] +#![feature(core_ffi_c)] +#![feature(const_eval_select)] +#![feature(const_pin)] +#![feature(cstr_from_bytes_until_nul)] +#![feature(dispatch_from_dyn)] +#![feature(exact_size_is_empty)] +#![feature(extend_one)] +#![feature(fmt_internals)] +#![feature(fn_traits)] +#![feature(hasher_prefixfree_extras)] +#![feature(inplace_iteration)] +#![feature(iter_advance_by)] +#![feature(layout_for_ptr)] +#![feature(maybe_uninit_slice)] +#![cfg_attr(test, feature(new_uninit))] +#![feature(nonnull_slice_from_raw_parts)] +#![feature(pattern)] +#![feature(ptr_internals)] +#![feature(ptr_metadata)] +#![feature(ptr_sub_ptr)] +#![feature(receiver_trait)] +#![feature(set_ptr_value)] +#![feature(slice_group_by)] +#![feature(slice_ptr_get)] +#![feature(slice_ptr_len)] +#![feature(slice_range)] +#![feature(str_internals)] +#![feature(strict_provenance)] +#![feature(trusted_len)] +#![feature(trusted_random_access)] +#![feature(try_trait_v2)] +#![feature(unchecked_math)] +#![feature(unicode_internals)] +#![feature(unsize)] +// +// Language features: +#![feature(allocator_internals)] +#![feature(allow_internal_unstable)] +#![feature(associated_type_bounds)] +#![feature(box_syntax)] +#![feature(cfg_sanitize)] +#![feature(const_deref)] +#![feature(const_mut_refs)] +#![feature(const_ptr_write)] +#![feature(const_precise_live_drops)] +#![feature(const_trait_impl)] +#![feature(const_try)] +#![feature(dropck_eyepatch)] +#![feature(exclusive_range_pattern)] +#![feature(fundamental)] +#![cfg_attr(not(test), feature(generator_trait))] +#![feature(hashmap_internals)] +#![feature(lang_items)] +#![feature(let_else)] +#![feature(min_specialization)] +#![feature(negative_impls)] +#![feature(never_type)] +#![feature(nll)] // Not necessary, but here to test the `nll` feature. +#![feature(rustc_allow_const_fn_unstable)] +#![feature(rustc_attrs)] +#![feature(slice_internals)] +#![feature(staged_api)] +#![cfg_attr(test, feature(test))] +#![feature(unboxed_closures)] +#![feature(unsized_fn_params)] +#![feature(c_unwind)] +// +// Rustdoc features: +#![feature(doc_cfg)] +#![feature(doc_cfg_hide)] +// Technically, this is a bug in rustdoc: rustdoc sees the documentation on `#[lang = slice_alloc]` +// blocks is for `&[T]`, which also has documentation using this feature in `core`, and gets mad +// that the feature-gate isn't enabled. Ideally, it wouldn't check for the feature gate for docs +// from other crates, but since this can only appear for lang items, it doesn't seem worth fixing. +#![feature(intra_doc_pointers)] + +// Allow testing this library +#[cfg(test)] +#[macro_use] +extern crate std; +#[cfg(test)] +extern crate test; + +// Module with internal macros used by other modules (needs to be included before other modules). +#[cfg(not(no_macros))] +#[macro_use] +mod macros; + +mod raw_vec; + +// Heaps provided for low-level allocation strategies + +pub mod alloc; + +// Primitive types using the heaps above + +// Need to conditionally define the mod from `boxed.rs` to avoid +// duplicating the lang-items when building in test cfg; but also need +// to allow code to have `use boxed::Box;` declarations. +#[cfg(not(test))] +pub mod boxed; +#[cfg(test)] +mod boxed { + pub use std::boxed::Box; +} +pub mod borrow; +pub mod collections; +#[cfg(not(no_global_oom_handling))] +pub mod ffi; +#[cfg(not(no_fmt))] +pub mod fmt; +#[cfg(not(no_rc))] +pub mod rc; +pub mod slice; +#[cfg(not(no_str))] +pub mod str; +#[cfg(not(no_string))] +pub mod string; +#[cfg(not(no_sync))] +#[cfg(target_has_atomic = "ptr")] +pub mod sync; +#[cfg(all(not(no_global_oom_handling), target_has_atomic = "ptr"))] +pub mod task; +#[cfg(test)] +mod tests; +pub mod vec; + +#[doc(hidden)] +#[unstable(feature = "liballoc_internals", issue = "none", reason = "implementation detail")] +pub mod __export { + pub use core::format_args; +} diff --git a/rust/alloc/raw_vec.rs b/rust/alloc/raw_vec.rs new file mode 100644 index 000000000000..daf5f2da7168 --- /dev/null +++ b/rust/alloc/raw_vec.rs @@ -0,0 +1,527 @@ +// SPDX-License-Identifier: Apache-2.0 OR MIT + +#![unstable(feature = "raw_vec_internals", reason = "unstable const warnings", issue = "none")] + +use core::alloc::LayoutError; +use core::cmp; +use core::intrinsics; +use core::mem::{self, ManuallyDrop, MaybeUninit}; +use core::ops::Drop; +use core::ptr::{self, NonNull, Unique}; +use core::slice; + +#[cfg(not(no_global_oom_handling))] +use crate::alloc::handle_alloc_error; +use crate::alloc::{Allocator, Global, Layout}; +use crate::boxed::Box; +use crate::collections::TryReserveError; +use crate::collections::TryReserveErrorKind::*; + +#[cfg(test)] +mod tests; + +#[cfg(not(no_global_oom_handling))] +enum AllocInit { + /// The contents of the new memory are uninitialized. + Uninitialized, + /// The new memory is guaranteed to be zeroed. + Zeroed, +} + +/// A low-level utility for more ergonomically allocating, reallocating, and deallocating +/// a buffer of memory on the heap without having to worry about all the corner cases +/// involved. This type is excellent for building your own data structures like Vec and VecDeque. +/// In particular: +/// +/// * Produces `Unique::dangling()` on zero-sized types. +/// * Produces `Unique::dangling()` on zero-length allocations. +/// * Avoids freeing `Unique::dangling()`. +/// * Catches all overflows in capacity computations (promotes them to "capacity overflow" panics). +/// * Guards against 32-bit systems allocating more than isize::MAX bytes. +/// * Guards against overflowing your length. +/// * Calls `handle_alloc_error` for fallible allocations. +/// * Contains a `ptr::Unique` and thus endows the user with all related benefits. +/// * Uses the excess returned from the allocator to use the largest available capacity. +/// +/// This type does not in anyway inspect the memory that it manages. When dropped it *will* +/// free its memory, but it *won't* try to drop its contents. It is up to the user of `RawVec` +/// to handle the actual things *stored* inside of a `RawVec`. +/// +/// Note that the excess of a zero-sized types is always infinite, so `capacity()` always returns +/// `usize::MAX`. This means that you need to be careful when round-tripping this type with a +/// `Box<[T]>`, since `capacity()` won't yield the length. +#[allow(missing_debug_implementations)] +pub(crate) struct RawVec<T, A: Allocator = Global> { + ptr: Unique<T>, + cap: usize, + alloc: A, +} + +impl<T> RawVec<T, Global> { + /// HACK(Centril): This exists because stable `const fn` can only call stable `const fn`, so + /// they cannot call `Self::new()`. + /// + /// If you change `RawVec<T>::new` or dependencies, please take care to not introduce anything + /// that would truly const-call something unstable. + pub const NEW: Self = Self::new(); + + /// Creates the biggest possible `RawVec` (on the system heap) + /// without allocating. If `T` has positive size, then this makes a + /// `RawVec` with capacity `0`. If `T` is zero-sized, then it makes a + /// `RawVec` with capacity `usize::MAX`. Useful for implementing + /// delayed allocation. + #[must_use] + pub const fn new() -> Self { + Self::new_in(Global) + } + + /// Creates a `RawVec` (on the system heap) with exactly the + /// capacity and alignment requirements for a `[T; capacity]`. This is + /// equivalent to calling `RawVec::new` when `capacity` is `0` or `T` is + /// zero-sized. Note that if `T` is zero-sized this means you will + /// *not* get a `RawVec` with the requested capacity. + /// + /// # Panics + /// + /// Panics if the requested capacity exceeds `isize::MAX` bytes. + /// + /// # Aborts + /// + /// Aborts on OOM. + #[cfg(not(any(no_global_oom_handling, test)))] + #[must_use] + #[inline] + pub fn with_capacity(capacity: usize) -> Self { + Self::with_capacity_in(capacity, Global) + } + + /// Like `with_capacity`, but guarantees the buffer is zeroed. + #[cfg(not(any(no_global_oom_handling, test)))] + #[must_use] + #[inline] + pub fn with_capacity_zeroed(capacity: usize) -> Self { + Self::with_capacity_zeroed_in(capacity, Global) + } +} + +impl<T, A: Allocator> RawVec<T, A> { + // Tiny Vecs are dumb. Skip to: + // - 8 if the element size is 1, because any heap allocators is likely + // to round up a request of less than 8 bytes to at least 8 bytes. + // - 4 if elements are moderate-sized (<= 1 KiB). + // - 1 otherwise, to avoid wasting too much space for very short Vecs. + pub(crate) const MIN_NON_ZERO_CAP: usize = if mem::size_of::<T>() == 1 { + 8 + } else if mem::size_of::<T>() <= 1024 { + 4 + } else { + 1 + }; + + /// Like `new`, but parameterized over the choice of allocator for + /// the returned `RawVec`. + pub const fn new_in(alloc: A) -> Self { + // `cap: 0` means "unallocated". zero-sized types are ignored. + Self { ptr: Unique::dangling(), cap: 0, alloc } + } + + /// Like `with_capacity`, but parameterized over the choice of + /// allocator for the returned `RawVec`. + #[cfg(not(no_global_oom_handling))] + #[inline] + pub fn with_capacity_in(capacity: usize, alloc: A) -> Self { + Self::allocate_in(capacity, AllocInit::Uninitialized, alloc) + } + + /// Like `with_capacity_zeroed`, but parameterized over the choice + /// of allocator for the returned `RawVec`. + #[cfg(not(no_global_oom_handling))] + #[inline] + pub fn with_capacity_zeroed_in(capacity: usize, alloc: A) -> Self { + Self::allocate_in(capacity, AllocInit::Zeroed, alloc) + } + + /// Converts the entire buffer into `Box<[MaybeUninit<T>]>` with the specified `len`. + /// + /// Note that this will correctly reconstitute any `cap` changes + /// that may have been performed. (See description of type for details.) + /// + /// # Safety + /// + /// * `len` must be greater than or equal to the most recently requested capacity, and + /// * `len` must be less than or equal to `self.capacity()`. + /// + /// Note, that the requested capacity and `self.capacity()` could differ, as + /// an allocator could overallocate and return a greater memory block than requested. + pub unsafe fn into_box(self, len: usize) -> Box<[MaybeUninit<T>], A> { + // Sanity-check one half of the safety requirement (we cannot check the other half). + debug_assert!( + len <= self.capacity(), + "`len` must be smaller than or equal to `self.capacity()`" + ); + + let me = ManuallyDrop::new(self); + unsafe { + let slice = slice::from_raw_parts_mut(me.ptr() as *mut MaybeUninit<T>, len); + Box::from_raw_in(slice, ptr::read(&me.alloc)) + } + } + + #[cfg(not(no_global_oom_handling))] + fn allocate_in(capacity: usize, init: AllocInit, alloc: A) -> Self { + // Don't allocate here because `Drop` will not deallocate when `capacity` is 0. + if mem::size_of::<T>() == 0 || capacity == 0 { + Self::new_in(alloc) + } else { + // We avoid `unwrap_or_else` here because it bloats the amount of + // LLVM IR generated. + let layout = match Layout::array::<T>(capacity) { + Ok(layout) => layout, + Err(_) => capacity_overflow(), + }; + match alloc_guard(layout.size()) { + Ok(_) => {} + Err(_) => capacity_overflow(), + } + let result = match init { + AllocInit::Uninitialized => alloc.allocate(layout), + AllocInit::Zeroed => alloc.allocate_zeroed(layout), + }; + let ptr = match result { + Ok(ptr) => ptr, + Err(_) => handle_alloc_error(layout), + }; + + // Allocators currently return a `NonNull<[u8]>` whose length + // matches the size requested. If that ever changes, the capacity + // here should change to `ptr.len() / mem::size_of::<T>()`. + Self { + ptr: unsafe { Unique::new_unchecked(ptr.cast().as_ptr()) }, + cap: capacity, + alloc, + } + } + } + + /// Reconstitutes a `RawVec` from a pointer, capacity, and allocator. + /// + /// # Safety + /// + /// The `ptr` must be allocated (via the given allocator `alloc`), and with the given + /// `capacity`. + /// The `capacity` cannot exceed `isize::MAX` for sized types. (only a concern on 32-bit + /// systems). ZST vectors may have a capacity up to `usize::MAX`. + /// If the `ptr` and `capacity` come from a `RawVec` created via `alloc`, then this is + /// guaranteed. + #[inline] + pub unsafe fn from_raw_parts_in(ptr: *mut T, capacity: usize, alloc: A) -> Self { + Self { ptr: unsafe { Unique::new_unchecked(ptr) }, cap: capacity, alloc } + } + + /// Gets a raw pointer to the start of the allocation. Note that this is + /// `Unique::dangling()` if `capacity == 0` or `T` is zero-sized. In the former case, you must + /// be careful. + #[inline] + pub fn ptr(&self) -> *mut T { + self.ptr.as_ptr() + } + + /// Gets the capacity of the allocation. + /// + /// This will always be `usize::MAX` if `T` is zero-sized. + #[inline(always)] + pub fn capacity(&self) -> usize { + if mem::size_of::<T>() == 0 { usize::MAX } else { self.cap } + } + + /// Returns a shared reference to the allocator backing this `RawVec`. + pub fn allocator(&self) -> &A { + &self.alloc + } + + fn current_memory(&self) -> Option<(NonNull<u8>, Layout)> { + if mem::size_of::<T>() == 0 || self.cap == 0 { + None + } else { + // We have an allocated chunk of memory, so we can bypass runtime + // checks to get our current layout. + unsafe { + let layout = Layout::array::<T>(self.cap).unwrap_unchecked(); + Some((self.ptr.cast().into(), layout)) + } + } + } + + /// Ensures that the buffer contains at least enough space to hold `len + + /// additional` elements. If it doesn't already have enough capacity, will + /// reallocate enough space plus comfortable slack space to get amortized + /// *O*(1) behavior. Will limit this behavior if it would needlessly cause + /// itself to panic. + /// + /// If `len` exceeds `self.capacity()`, this may fail to actually allocate + /// the requested space. This is not really unsafe, but the unsafe + /// code *you* write that relies on the behavior of this function may break. + /// + /// This is ideal for implementing a bulk-push operation like `extend`. + /// + /// # Panics + /// + /// Panics if the new capacity exceeds `isize::MAX` bytes. + /// + /// # Aborts + /// + /// Aborts on OOM. + #[cfg(not(no_global_oom_handling))] + #[inline] + pub fn reserve(&mut self, len: usize, additional: usize) { + // Callers expect this function to be very cheap when there is already sufficient capacity. + // Therefore, we move all the resizing and error-handling logic from grow_amortized and + // handle_reserve behind a call, while making sure that this function is likely to be + // inlined as just a comparison and a call if the comparison fails. + #[cold] + fn do_reserve_and_handle<T, A: Allocator>( + slf: &mut RawVec<T, A>, + len: usize, + additional: usize, + ) { + handle_reserve(slf.grow_amortized(len, additional)); + } + + if self.needs_to_grow(len, additional) { + do_reserve_and_handle(self, len, additional); + } + } + + /// A specialized version of `reserve()` used only by the hot and + /// oft-instantiated `Vec::push()`, which does its own capacity check. + #[cfg(not(no_global_oom_handling))] + #[inline(never)] + pub fn reserve_for_push(&mut self, len: usize) { + handle_reserve(self.grow_amortized(len, 1)); + } + + /// The same as `reserve`, but returns on errors instead of panicking or aborting. + pub fn try_reserve(&mut self, len: usize, additional: usize) -> Result<(), TryReserveError> { + if self.needs_to_grow(len, additional) { + self.grow_amortized(len, additional) + } else { + Ok(()) + } + } + + /// The same as `reserve_for_push`, but returns on errors instead of panicking or aborting. + #[inline(never)] + pub fn try_reserve_for_push(&mut self, len: usize) -> Result<(), TryReserveError> { + self.grow_amortized(len, 1) + } + + /// Ensures that the buffer contains at least enough space to hold `len + + /// additional` elements. If it doesn't already, will reallocate the + /// minimum possible amount of memory necessary. Generally this will be + /// exactly the amount of memory necessary, but in principle the allocator + /// is free to give back more than we asked for. + /// + /// If `len` exceeds `self.capacity()`, this may fail to actually allocate + /// the requested space. This is not really unsafe, but the unsafe code + /// *you* write that relies on the behavior of this function may break. + /// + /// # Panics + /// + /// Panics if the new capacity exceeds `isize::MAX` bytes. + /// + /// # Aborts + /// + /// Aborts on OOM. + #[cfg(not(no_global_oom_handling))] + pub fn reserve_exact(&mut self, len: usize, additional: usize) { + handle_reserve(self.try_reserve_exact(len, additional)); + } + + /// The same as `reserve_exact`, but returns on errors instead of panicking or aborting. + pub fn try_reserve_exact( + &mut self, + len: usize, + additional: usize, + ) -> Result<(), TryReserveError> { + if self.needs_to_grow(len, additional) { self.grow_exact(len, additional) } else { Ok(()) } + } + + /// Shrinks the buffer down to the specified capacity. If the given amount + /// is 0, actually completely deallocates. + /// + /// # Panics + /// + /// Panics if the given amount is *larger* than the current capacity. + /// + /// # Aborts + /// + /// Aborts on OOM. + #[cfg(not(no_global_oom_handling))] + pub fn shrink_to_fit(&mut self, cap: usize) { + handle_reserve(self.shrink(cap)); + } +} + +impl<T, A: Allocator> RawVec<T, A> { + /// Returns if the buffer needs to grow to fulfill the needed extra capacity. + /// Mainly used to make inlining reserve-calls possible without inlining `grow`. + fn needs_to_grow(&self, len: usize, additional: usize) -> bool { + additional > self.capacity().wrapping_sub(len) + } + + fn set_ptr_and_cap(&mut self, ptr: NonNull<[u8]>, cap: usize) { + // Allocators currently return a `NonNull<[u8]>` whose length matches + // the size requested. If that ever changes, the capacity here should + // change to `ptr.len() / mem::size_of::<T>()`. + self.ptr = unsafe { Unique::new_unchecked(ptr.cast().as_ptr()) }; + self.cap = cap; + } + + // This method is usually instantiated many times. So we want it to be as + // small as possible, to improve compile times. But we also want as much of + // its contents to be statically computable as possible, to make the + // generated code run faster. Therefore, this method is carefully written + // so that all of the code that depends on `T` is within it, while as much + // of the code that doesn't depend on `T` as possible is in functions that + // are non-generic over `T`. + fn grow_amortized(&mut self, len: usize, additional: usize) -> Result<(), TryReserveError> { + // This is ensured by the calling contexts. + debug_assert!(additional > 0); + + if mem::size_of::<T>() == 0 { + // Since we return a capacity of `usize::MAX` when `elem_size` is + // 0, getting to here necessarily means the `RawVec` is overfull. + return Err(CapacityOverflow.into()); + } + + // Nothing we can really do about these checks, sadly. + let required_cap = len.checked_add(additional).ok_or(CapacityOverflow)?; + + // This guarantees exponential growth. The doubling cannot overflow + // because `cap <= isize::MAX` and the type of `cap` is `usize`. + let cap = cmp::max(self.cap * 2, required_cap); + let cap = cmp::max(Self::MIN_NON_ZERO_CAP, cap); + + let new_layout = Layout::array::<T>(cap); + + // `finish_grow` is non-generic over `T`. + let ptr = finish_grow(new_layout, self.current_memory(), &mut self.alloc)?; + self.set_ptr_and_cap(ptr, cap); + Ok(()) + } + + // The constraints on this method are much the same as those on + // `grow_amortized`, but this method is usually instantiated less often so + // it's less critical. + fn grow_exact(&mut self, len: usize, additional: usize) -> Result<(), TryReserveError> { + if mem::size_of::<T>() == 0 { + // Since we return a capacity of `usize::MAX` when the type size is + // 0, getting to here necessarily means the `RawVec` is overfull. + return Err(CapacityOverflow.into()); + } + + let cap = len.checked_add(additional).ok_or(CapacityOverflow)?; + let new_layout = Layout::array::<T>(cap); + + // `finish_grow` is non-generic over `T`. + let ptr = finish_grow(new_layout, self.current_memory(), &mut self.alloc)?; + self.set_ptr_and_cap(ptr, cap); + Ok(()) + } + + #[allow(dead_code)] + fn shrink(&mut self, cap: usize) -> Result<(), TryReserveError> { + assert!(cap <= self.capacity(), "Tried to shrink to a larger capacity"); + + let (ptr, layout) = if let Some(mem) = self.current_memory() { mem } else { return Ok(()) }; + + let ptr = unsafe { + // `Layout::array` cannot overflow here because it would have + // overflowed earlier when capacity was larger. + let new_layout = Layout::array::<T>(cap).unwrap_unchecked(); + self.alloc + .shrink(ptr, layout, new_layout) + .map_err(|_| AllocError { layout: new_layout, non_exhaustive: () })? + }; + self.set_ptr_and_cap(ptr, cap); + Ok(()) + } +} + +// This function is outside `RawVec` to minimize compile times. See the comment +// above `RawVec::grow_amortized` for details. (The `A` parameter isn't +// significant, because the number of different `A` types seen in practice is +// much smaller than the number of `T` types.) +#[inline(never)] +fn finish_grow<A>( + new_layout: Result<Layout, LayoutError>, + current_memory: Option<(NonNull<u8>, Layout)>, + alloc: &mut A, +) -> Result<NonNull<[u8]>, TryReserveError> +where + A: Allocator, +{ + // Check for the error here to minimize the size of `RawVec::grow_*`. + let new_layout = new_layout.map_err(|_| CapacityOverflow)?; + + alloc_guard(new_layout.size())?; + + let memory = if let Some((ptr, old_layout)) = current_memory { + debug_assert_eq!(old_layout.align(), new_layout.align()); + unsafe { + // The allocator checks for alignment equality + intrinsics::assume(old_layout.align() == new_layout.align()); + alloc.grow(ptr, old_layout, new_layout) + } + } else { + alloc.allocate(new_layout) + }; + + memory.map_err(|_| AllocError { layout: new_layout, non_exhaustive: () }.into()) +} + +unsafe impl<#[may_dangle] T, A: Allocator> Drop for RawVec<T, A> { + /// Frees the memory owned by the `RawVec` *without* trying to drop its contents. + fn drop(&mut self) { + if let Some((ptr, layout)) = self.current_memory() { + unsafe { self.alloc.deallocate(ptr, layout) } + } + } +} + +// Central function for reserve error handling. +#[cfg(not(no_global_oom_handling))] +#[inline] +fn handle_reserve(result: Result<(), TryReserveError>) { + match result.map_err(|e| e.kind()) { + Err(CapacityOverflow) => capacity_overflow(), + Err(AllocError { layout, .. }) => handle_alloc_error(layout), + Ok(()) => { /* yay */ } + } +} + +// We need to guarantee the following: +// * We don't ever allocate `> isize::MAX` byte-size objects. +// * We don't overflow `usize::MAX` and actually allocate too little. +// +// On 64-bit we just need to check for overflow since trying to allocate +// `> isize::MAX` bytes will surely fail. On 32-bit and 16-bit we need to add +// an extra guard for this in case we're running on a platform which can use +// all 4GB in user-space, e.g., PAE or x32. + +#[inline] +fn alloc_guard(alloc_size: usize) -> Result<(), TryReserveError> { + if usize::BITS < 64 && alloc_size > isize::MAX as usize { + Err(CapacityOverflow.into()) + } else { + Ok(()) + } +} + +// One central function responsible for reporting capacity overflows. This'll +// ensure that the code generation related to these panics is minimal as there's +// only one location which panics rather than a bunch throughout the module. +#[cfg(not(no_global_oom_handling))] +fn capacity_overflow() -> ! { + panic!("capacity overflow"); +} diff --git a/rust/alloc/slice.rs b/rust/alloc/slice.rs new file mode 100644 index 000000000000..e444e97fa145 --- /dev/null +++ b/rust/alloc/slice.rs @@ -0,0 +1,1204 @@ +// SPDX-License-Identifier: Apache-2.0 OR MIT + +//! A dynamically-sized view into a contiguous sequence, `[T]`. +//! +//! *[See also the slice primitive type](slice).* +//! +//! Slices are a view into a block of memory represented as a pointer and a +//! length. +//! +//! ``` +//! // slicing a Vec +//! let vec = vec![1, 2, 3]; +//! let int_slice = &vec[..]; +//! // coercing an array to a slice +//! let str_slice: &[&str] = &["one", "two", "three"]; +//! ``` +//! +//! Slices are either mutable or shared. The shared slice type is `&[T]`, +//! while the mutable slice type is `&mut [T]`, where `T` represents the element +//! type. For example, you can mutate the block of memory that a mutable slice +//! points to: +//! +//! ``` +//! let x = &mut [1, 2, 3]; +//! x[1] = 7; +//! assert_eq!(x, &[1, 7, 3]); +//! ``` +//! +//! Here are some of the things this module contains: +//! +//! ## Structs +//! +//! There are several structs that are useful for slices, such as [`Iter`], which +//! represents iteration over a slice. +//! +//! ## Trait Implementations +//! +//! There are several implementations of common traits for slices. Some examples +//! include: +//! +//! * [`Clone`] +//! * [`Eq`], [`Ord`] - for slices whose element type are [`Eq`] or [`Ord`]. +//! * [`Hash`] - for slices whose element type is [`Hash`]. +//! +//! ## Iteration +//! +//! The slices implement `IntoIterator`. The iterator yields references to the +//! slice elements. +//! +//! ``` +//! let numbers = &[0, 1, 2]; +//! for n in numbers { +//! println!("{n} is a number!"); +//! } +//! ``` +//! +//! The mutable slice yields mutable references to the elements: +//! +//! ``` +//! let mut scores = [7, 8, 9]; +//! for score in &mut scores[..] { +//! *score += 1; +//! } +//! ``` +//! +//! This iterator yields mutable references to the slice's elements, so while +//! the element type of the slice is `i32`, the element type of the iterator is +//! `&mut i32`. +//! +//! * [`.iter`] and [`.iter_mut`] are the explicit methods to return the default +//! iterators. +//! * Further methods that return iterators are [`.split`], [`.splitn`], +//! [`.chunks`], [`.windows`] and more. +//! +//! [`Hash`]: core::hash::Hash +//! [`.iter`]: slice::iter +//! [`.iter_mut`]: slice::iter_mut +//! [`.split`]: slice::split +//! [`.splitn`]: slice::splitn +//! [`.chunks`]: slice::chunks +//! [`.windows`]: slice::windows +#![stable(feature = "rust1", since = "1.0.0")] +// Many of the usings in this module are only used in the test configuration. +// It's cleaner to just turn off the unused_imports warning than to fix them. +#![cfg_attr(test, allow(unused_imports, dead_code))] + +use core::borrow::{Borrow, BorrowMut}; +#[cfg(not(no_global_oom_handling))] +use core::cmp::Ordering::{self, Less}; +#[cfg(not(no_global_oom_handling))] +use core::mem; +#[cfg(not(no_global_oom_handling))] +use core::mem::size_of; +#[cfg(not(no_global_oom_handling))] +use core::ptr; + +use crate::alloc::Allocator; +#[cfg(not(no_global_oom_handling))] +use crate::alloc::Global; +#[cfg(not(no_global_oom_handling))] +use crate::borrow::ToOwned; +use crate::boxed::Box; +use crate::vec::Vec; + +#[unstable(feature = "slice_range", issue = "76393")] +pub use core::slice::range; +#[unstable(feature = "array_chunks", issue = "74985")] +pub use core::slice::ArrayChunks; +#[unstable(feature = "array_chunks", issue = "74985")] +pub use core::slice::ArrayChunksMut; +#[unstable(feature = "array_windows", issue = "75027")] +pub use core::slice::ArrayWindows; +#[stable(feature = "inherent_ascii_escape", since = "1.60.0")] +pub use core::slice::EscapeAscii; +#[stable(feature = "slice_get_slice", since = "1.28.0")] +pub use core::slice::SliceIndex; +#[stable(feature = "from_ref", since = "1.28.0")] +pub use core::slice::{from_mut, from_ref}; +#[stable(feature = "rust1", since = "1.0.0")] +pub use core::slice::{from_raw_parts, from_raw_parts_mut}; +#[stable(feature = "rust1", since = "1.0.0")] +pub use core::slice::{Chunks, Windows}; +#[stable(feature = "chunks_exact", since = "1.31.0")] +pub use core::slice::{ChunksExact, ChunksExactMut}; +#[stable(feature = "rust1", since = "1.0.0")] +pub use core::slice::{ChunksMut, Split, SplitMut}; +#[unstable(feature = "slice_group_by", issue = "80552")] +pub use core::slice::{GroupBy, GroupByMut}; +#[stable(feature = "rust1", since = "1.0.0")] +pub use core::slice::{Iter, IterMut}; +#[stable(feature = "rchunks", since = "1.31.0")] +pub use core::slice::{RChunks, RChunksExact, RChunksExactMut, RChunksMut}; +#[stable(feature = "slice_rsplit", since = "1.27.0")] +pub use core::slice::{RSplit, RSplitMut}; +#[stable(feature = "rust1", since = "1.0.0")] +pub use core::slice::{RSplitN, RSplitNMut, SplitN, SplitNMut}; +#[stable(feature = "split_inclusive", since = "1.51.0")] +pub use core::slice::{SplitInclusive, SplitInclusiveMut}; + +//////////////////////////////////////////////////////////////////////////////// +// Basic slice extension methods +//////////////////////////////////////////////////////////////////////////////// + +// HACK(japaric) needed for the implementation of `vec!` macro during testing +// N.B., see the `hack` module in this file for more details. +#[cfg(test)] +pub use hack::into_vec; + +// HACK(japaric) needed for the implementation of `Vec::clone` during testing +// N.B., see the `hack` module in this file for more details. +#[cfg(test)] +pub use hack::to_vec; + +// HACK(japaric): With cfg(test) `impl [T]` is not available, these three +// functions are actually methods that are in `impl [T]` but not in +// `core::slice::SliceExt` - we need to supply these functions for the +// `test_permutations` test +pub(crate) mod hack { + use core::alloc::Allocator; + + use crate::boxed::Box; + use crate::vec::Vec; + + // We shouldn't add inline attribute to this since this is used in + // `vec!` macro mostly and causes perf regression. See #71204 for + // discussion and perf results. + pub fn into_vec<T, A: Allocator>(b: Box<[T], A>) -> Vec<T, A> { + unsafe { + let len = b.len(); + let (b, alloc) = Box::into_raw_with_allocator(b); + Vec::from_raw_parts_in(b as *mut T, len, len, alloc) + } + } + + #[cfg(not(no_global_oom_handling))] + #[inline] + pub fn to_vec<T: ConvertVec, A: Allocator>(s: &[T], alloc: A) -> Vec<T, A> { + T::to_vec(s, alloc) + } + + #[cfg(not(no_global_oom_handling))] + pub trait ConvertVec { + fn to_vec<A: Allocator>(s: &[Self], alloc: A) -> Vec<Self, A> + where + Self: Sized; + } + + #[cfg(not(no_global_oom_handling))] + impl<T: Clone> ConvertVec for T { + #[inline] + default fn to_vec<A: Allocator>(s: &[Self], alloc: A) -> Vec<Self, A> { + struct DropGuard<'a, T, A: Allocator> { + vec: &'a mut Vec<T, A>, + num_init: usize, + } + impl<'a, T, A: Allocator> Drop for DropGuard<'a, T, A> { + #[inline] + fn drop(&mut self) { + // SAFETY: + // items were marked initialized in the loop below + unsafe { + self.vec.set_len(self.num_init); + } + } + } + let mut vec = Vec::with_capacity_in(s.len(), alloc); + let mut guard = DropGuard { vec: &mut vec, num_init: 0 }; + let slots = guard.vec.spare_capacity_mut(); + // .take(slots.len()) is necessary for LLVM to remove bounds checks + // and has better codegen than zip. + for (i, b) in s.iter().enumerate().take(slots.len()) { + guard.num_init = i; + slots[i].write(b.clone()); + } + core::mem::forget(guard); + // SAFETY: + // the vec was allocated and initialized above to at least this length. + unsafe { + vec.set_len(s.len()); + } + vec + } + } + + #[cfg(not(no_global_oom_handling))] + impl<T: Copy> ConvertVec for T { + #[inline] + fn to_vec<A: Allocator>(s: &[Self], alloc: A) -> Vec<Self, A> { + let mut v = Vec::with_capacity_in(s.len(), alloc); + // SAFETY: + // allocated above with the capacity of `s`, and initialize to `s.len()` in + // ptr::copy_to_non_overlapping below. + unsafe { + s.as_ptr().copy_to_nonoverlapping(v.as_mut_ptr(), s.len()); + v.set_len(s.len()); + } + v + } + } +} + +#[cfg(not(test))] +impl<T> [T] { + /// Sorts the slice. + /// + /// This sort is stable (i.e., does not reorder equal elements) and *O*(*n* \* log(*n*)) worst-case. + /// + /// When applicable, unstable sorting is preferred because it is generally faster than stable + /// sorting and it doesn't allocate auxiliary memory. + /// See [`sort_unstable`](slice::sort_unstable). + /// + /// # Current implementation + /// + /// The current algorithm is an adaptive, iterative merge sort inspired by + /// [timsort](https://en.wikipedia.org/wiki/Timsort). + /// It is designed to be very fast in cases where the slice is nearly sorted, or consists of + /// two or more sorted sequences concatenated one after another. + /// + /// Also, it allocates temporary storage half the size of `self`, but for short slices a + /// non-allocating insertion sort is used instead. + /// + /// # Examples + /// + /// ``` + /// let mut v = [-5, 4, 1, -3, 2]; + /// + /// v.sort(); + /// assert!(v == [-5, -3, 1, 2, 4]); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[rustc_allow_incoherent_impl] + #[stable(feature = "rust1", since = "1.0.0")] + #[inline] + pub fn sort(&mut self) + where + T: Ord, + { + merge_sort(self, |a, b| a.lt(b)); + } + + /// Sorts the slice with a comparator function. + /// + /// This sort is stable (i.e., does not reorder equal elements) and *O*(*n* \* log(*n*)) worst-case. + /// + /// The comparator function must define a total ordering for the elements in the slice. If + /// the ordering is not total, the order of the elements is unspecified. An order is a + /// total order if it is (for all `a`, `b` and `c`): + /// + /// * total and antisymmetric: exactly one of `a < b`, `a == b` or `a > b` is true, and + /// * transitive, `a < b` and `b < c` implies `a < c`. The same must hold for both `==` and `>`. + /// + /// For example, while [`f64`] doesn't implement [`Ord`] because `NaN != NaN`, we can use + /// `partial_cmp` as our sort function when we know the slice doesn't contain a `NaN`. + /// + /// ``` + /// let mut floats = [5f64, 4.0, 1.0, 3.0, 2.0]; + /// floats.sort_by(|a, b| a.partial_cmp(b).unwrap()); + /// assert_eq!(floats, [1.0, 2.0, 3.0, 4.0, 5.0]); + /// ``` + /// + /// When applicable, unstable sorting is preferred because it is generally faster than stable + /// sorting and it doesn't allocate auxiliary memory. + /// See [`sort_unstable_by`](slice::sort_unstable_by). + /// + /// # Current implementation + /// + /// The current algorithm is an adaptive, iterative merge sort inspired by + /// [timsort](https://en.wikipedia.org/wiki/Timsort). + /// It is designed to be very fast in cases where the slice is nearly sorted, or consists of + /// two or more sorted sequences concatenated one after another. + /// + /// Also, it allocates temporary storage half the size of `self`, but for short slices a + /// non-allocating insertion sort is used instead. + /// + /// # Examples + /// + /// ``` + /// let mut v = [5, 4, 1, 3, 2]; + /// v.sort_by(|a, b| a.cmp(b)); + /// assert!(v == [1, 2, 3, 4, 5]); + /// + /// // reverse sorting + /// v.sort_by(|a, b| b.cmp(a)); + /// assert!(v == [5, 4, 3, 2, 1]); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[rustc_allow_incoherent_impl] + #[stable(feature = "rust1", since = "1.0.0")] + #[inline] + pub fn sort_by<F>(&mut self, mut compare: F) + where + F: FnMut(&T, &T) -> Ordering, + { + merge_sort(self, |a, b| compare(a, b) == Less); + } + + /// Sorts the slice with a key extraction function. + /// + /// This sort is stable (i.e., does not reorder equal elements) and *O*(*m* \* *n* \* log(*n*)) + /// worst-case, where the key function is *O*(*m*). + /// + /// For expensive key functions (e.g. functions that are not simple property accesses or + /// basic operations), [`sort_by_cached_key`](slice::sort_by_cached_key) is likely to be + /// significantly faster, as it does not recompute element keys. + /// + /// When applicable, unstable sorting is preferred because it is generally faster than stable + /// sorting and it doesn't allocate auxiliary memory. + /// See [`sort_unstable_by_key`](slice::sort_unstable_by_key). + /// + /// # Current implementation + /// + /// The current algorithm is an adaptive, iterative merge sort inspired by + /// [timsort](https://en.wikipedia.org/wiki/Timsort). + /// It is designed to be very fast in cases where the slice is nearly sorted, or consists of + /// two or more sorted sequences concatenated one after another. + /// + /// Also, it allocates temporary storage half the size of `self`, but for short slices a + /// non-allocating insertion sort is used instead. + /// + /// # Examples + /// + /// ``` + /// let mut v = [-5i32, 4, 1, -3, 2]; + /// + /// v.sort_by_key(|k| k.abs()); + /// assert!(v == [1, 2, -3, 4, -5]); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[rustc_allow_incoherent_impl] + #[stable(feature = "slice_sort_by_key", since = "1.7.0")] + #[inline] + pub fn sort_by_key<K, F>(&mut self, mut f: F) + where + F: FnMut(&T) -> K, + K: Ord, + { + merge_sort(self, |a, b| f(a).lt(&f(b))); + } + + /// Sorts the slice with a key extraction function. + /// + /// During sorting, the key function is called at most once per element, by using + /// temporary storage to remember the results of key evaluation. + /// The order of calls to the key function is unspecified and may change in future versions + /// of the standard library. + /// + /// This sort is stable (i.e., does not reorder equal elements) and *O*(*m* \* *n* + *n* \* log(*n*)) + /// worst-case, where the key function is *O*(*m*). + /// + /// For simple key functions (e.g., functions that are property accesses or + /// basic operations), [`sort_by_key`](slice::sort_by_key) is likely to be + /// faster. + /// + /// # Current implementation + /// + /// The current algorithm is based on [pattern-defeating quicksort][pdqsort] by Orson Peters, + /// which combines the fast average case of randomized quicksort with the fast worst case of + /// heapsort, while achieving linear time on slices with certain patterns. It uses some + /// randomization to avoid degenerate cases, but with a fixed seed to always provide + /// deterministic behavior. + /// + /// In the worst case, the algorithm allocates temporary storage in a `Vec<(K, usize)>` the + /// length of the slice. + /// + /// # Examples + /// + /// ``` + /// let mut v = [-5i32, 4, 32, -3, 2]; + /// + /// v.sort_by_cached_key(|k| k.to_string()); + /// assert!(v == [-3, -5, 2, 32, 4]); + /// ``` + /// + /// [pdqsort]: https://github.com/orlp/pdqsort + #[cfg(not(no_global_oom_handling))] + #[rustc_allow_incoherent_impl] + #[stable(feature = "slice_sort_by_cached_key", since = "1.34.0")] + #[inline] + pub fn sort_by_cached_key<K, F>(&mut self, f: F) + where + F: FnMut(&T) -> K, + K: Ord, + { + // Helper macro for indexing our vector by the smallest possible type, to reduce allocation. + macro_rules! sort_by_key { + ($t:ty, $slice:ident, $f:ident) => {{ + let mut indices: Vec<_> = + $slice.iter().map($f).enumerate().map(|(i, k)| (k, i as $t)).collect(); + // The elements of `indices` are unique, as they are indexed, so any sort will be + // stable with respect to the original slice. We use `sort_unstable` here because + // it requires less memory allocation. + indices.sort_unstable(); + for i in 0..$slice.len() { + let mut index = indices[i].1; + while (index as usize) < i { + index = indices[index as usize].1; + } + indices[i].1 = index; + $slice.swap(i, index as usize); + } + }}; + } + + let sz_u8 = mem::size_of::<(K, u8)>(); + let sz_u16 = mem::size_of::<(K, u16)>(); + let sz_u32 = mem::size_of::<(K, u32)>(); + let sz_usize = mem::size_of::<(K, usize)>(); + + let len = self.len(); + if len < 2 { + return; + } + if sz_u8 < sz_u16 && len <= (u8::MAX as usize) { + return sort_by_key!(u8, self, f); + } + if sz_u16 < sz_u32 && len <= (u16::MAX as usize) { + return sort_by_key!(u16, self, f); + } + if sz_u32 < sz_usize && len <= (u32::MAX as usize) { + return sort_by_key!(u32, self, f); + } + sort_by_key!(usize, self, f) + } + + /// Copies `self` into a new `Vec`. + /// + /// # Examples + /// + /// ``` + /// let s = [10, 40, 30]; + /// let x = s.to_vec(); + /// // Here, `s` and `x` can be modified independently. + /// ``` + #[cfg(not(no_global_oom_handling))] + #[rustc_allow_incoherent_impl] + #[rustc_conversion_suggestion] + #[stable(feature = "rust1", since = "1.0.0")] + #[inline] + pub fn to_vec(&self) -> Vec<T> + where + T: Clone, + { + self.to_vec_in(Global) + } + + /// Copies `self` into a new `Vec` with an allocator. + /// + /// # Examples + /// + /// ``` + /// #![feature(allocator_api)] + /// + /// use std::alloc::System; + /// + /// let s = [10, 40, 30]; + /// let x = s.to_vec_in(System); + /// // Here, `s` and `x` can be modified independently. + /// ``` + #[cfg(not(no_global_oom_handling))] + #[rustc_allow_incoherent_impl] + #[inline] + #[unstable(feature = "allocator_api", issue = "32838")] + pub fn to_vec_in<A: Allocator>(&self, alloc: A) -> Vec<T, A> + where + T: Clone, + { + // N.B., see the `hack` module in this file for more details. + hack::to_vec(self, alloc) + } + + /// Converts `self` into a vector without clones or allocation. + /// + /// The resulting vector can be converted back into a box via + /// `Vec<T>`'s `into_boxed_slice` method. + /// + /// # Examples + /// + /// ``` + /// let s: Box<[i32]> = Box::new([10, 40, 30]); + /// let x = s.into_vec(); + /// // `s` cannot be used anymore because it has been converted into `x`. + /// + /// assert_eq!(x, vec![10, 40, 30]); + /// ``` + #[rustc_allow_incoherent_impl] + #[stable(feature = "rust1", since = "1.0.0")] + #[inline] + pub fn into_vec<A: Allocator>(self: Box<Self, A>) -> Vec<T, A> { + // N.B., see the `hack` module in this file for more details. + hack::into_vec(self) + } + + /// Creates a vector by repeating a slice `n` times. + /// + /// # Panics + /// + /// This function will panic if the capacity would overflow. + /// + /// # Examples + /// + /// Basic usage: + /// + /// ``` + /// assert_eq!([1, 2].repeat(3), vec![1, 2, 1, 2, 1, 2]); + /// ``` + /// + /// A panic upon overflow: + /// + /// ```should_panic + /// // this will panic at runtime + /// b"0123456789abcdef".repeat(usize::MAX); + /// ``` + #[rustc_allow_incoherent_impl] + #[cfg(not(no_global_oom_handling))] + #[stable(feature = "repeat_generic_slice", since = "1.40.0")] + pub fn repeat(&self, n: usize) -> Vec<T> + where + T: Copy, + { + if n == 0 { + return Vec::new(); + } + + // If `n` is larger than zero, it can be split as + // `n = 2^expn + rem (2^expn > rem, expn >= 0, rem >= 0)`. + // `2^expn` is the number represented by the leftmost '1' bit of `n`, + // and `rem` is the remaining part of `n`. + + // Using `Vec` to access `set_len()`. + let capacity = self.len().checked_mul(n).expect("capacity overflow"); + let mut buf = Vec::with_capacity(capacity); + + // `2^expn` repetition is done by doubling `buf` `expn`-times. + buf.extend(self); + { + let mut m = n >> 1; + // If `m > 0`, there are remaining bits up to the leftmost '1'. + while m > 0 { + // `buf.extend(buf)`: + unsafe { + ptr::copy_nonoverlapping( + buf.as_ptr(), + (buf.as_mut_ptr() as *mut T).add(buf.len()), + buf.len(), + ); + // `buf` has capacity of `self.len() * n`. + let buf_len = buf.len(); + buf.set_len(buf_len * 2); + } + + m >>= 1; + } + } + + // `rem` (`= n - 2^expn`) repetition is done by copying + // first `rem` repetitions from `buf` itself. + let rem_len = capacity - buf.len(); // `self.len() * rem` + if rem_len > 0 { + // `buf.extend(buf[0 .. rem_len])`: + unsafe { + // This is non-overlapping since `2^expn > rem`. + ptr::copy_nonoverlapping( + buf.as_ptr(), + (buf.as_mut_ptr() as *mut T).add(buf.len()), + rem_len, + ); + // `buf.len() + rem_len` equals to `buf.capacity()` (`= self.len() * n`). + buf.set_len(capacity); + } + } + buf + } + + /// Flattens a slice of `T` into a single value `Self::Output`. + /// + /// # Examples + /// + /// ``` + /// assert_eq!(["hello", "world"].concat(), "helloworld"); + /// assert_eq!([[1, 2], [3, 4]].concat(), [1, 2, 3, 4]); + /// ``` + #[rustc_allow_incoherent_impl] + #[stable(feature = "rust1", since = "1.0.0")] + pub fn concat<Item: ?Sized>(&self) -> <Self as Concat<Item>>::Output + where + Self: Concat<Item>, + { + Concat::concat(self) + } + + /// Flattens a slice of `T` into a single value `Self::Output`, placing a + /// given separator between each. + /// + /// # Examples + /// + /// ``` + /// assert_eq!(["hello", "world"].join(" "), "hello world"); + /// assert_eq!([[1, 2], [3, 4]].join(&0), [1, 2, 0, 3, 4]); + /// assert_eq!([[1, 2], [3, 4]].join(&[0, 0][..]), [1, 2, 0, 0, 3, 4]); + /// ``` + #[rustc_allow_incoherent_impl] + #[stable(feature = "rename_connect_to_join", since = "1.3.0")] + pub fn join<Separator>(&self, sep: Separator) -> <Self as Join<Separator>>::Output + where + Self: Join<Separator>, + { + Join::join(self, sep) + } + + /// Flattens a slice of `T` into a single value `Self::Output`, placing a + /// given separator between each. + /// + /// # Examples + /// + /// ``` + /// # #![allow(deprecated)] + /// assert_eq!(["hello", "world"].connect(" "), "hello world"); + /// assert_eq!([[1, 2], [3, 4]].connect(&0), [1, 2, 0, 3, 4]); + /// ``` + #[rustc_allow_incoherent_impl] + #[stable(feature = "rust1", since = "1.0.0")] + #[deprecated(since = "1.3.0", note = "renamed to join")] + pub fn connect<Separator>(&self, sep: Separator) -> <Self as Join<Separator>>::Output + where + Self: Join<Separator>, + { + Join::join(self, sep) + } +} + +#[cfg(not(test))] +impl [u8] { + /// Returns a vector containing a copy of this slice where each byte + /// is mapped to its ASCII upper case equivalent. + /// + /// ASCII letters 'a' to 'z' are mapped to 'A' to 'Z', + /// but non-ASCII letters are unchanged. + /// + /// To uppercase the value in-place, use [`make_ascii_uppercase`]. + /// + /// [`make_ascii_uppercase`]: slice::make_ascii_uppercase + #[cfg(not(no_global_oom_handling))] + #[rustc_allow_incoherent_impl] + #[must_use = "this returns the uppercase bytes as a new Vec, \ + without modifying the original"] + #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")] + #[inline] + pub fn to_ascii_uppercase(&self) -> Vec<u8> { + let mut me = self.to_vec(); + me.make_ascii_uppercase(); + me + } + + /// Returns a vector containing a copy of this slice where each byte + /// is mapped to its ASCII lower case equivalent. + /// + /// ASCII letters 'A' to 'Z' are mapped to 'a' to 'z', + /// but non-ASCII letters are unchanged. + /// + /// To lowercase the value in-place, use [`make_ascii_lowercase`]. + /// + /// [`make_ascii_lowercase`]: slice::make_ascii_lowercase + #[cfg(not(no_global_oom_handling))] + #[rustc_allow_incoherent_impl] + #[must_use = "this returns the lowercase bytes as a new Vec, \ + without modifying the original"] + #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")] + #[inline] + pub fn to_ascii_lowercase(&self) -> Vec<u8> { + let mut me = self.to_vec(); + me.make_ascii_lowercase(); + me + } +} + +//////////////////////////////////////////////////////////////////////////////// +// Extension traits for slices over specific kinds of data +//////////////////////////////////////////////////////////////////////////////// + +/// Helper trait for [`[T]::concat`](slice::concat). +/// +/// Note: the `Item` type parameter is not used in this trait, +/// but it allows impls to be more generic. +/// Without it, we get this error: +/// +/// ```error +/// error[E0207]: the type parameter `T` is not constrained by the impl trait, self type, or predica +/// --> src/liballoc/slice.rs:608:6 +/// | +/// 608 | impl<T: Clone, V: Borrow<[T]>> Concat for [V] { +/// | ^ unconstrained type parameter +/// ``` +/// +/// This is because there could exist `V` types with multiple `Borrow<[_]>` impls, +/// such that multiple `T` types would apply: +/// +/// ``` +/// # #[allow(dead_code)] +/// pub struct Foo(Vec<u32>, Vec<String>); +/// +/// impl std::borrow::Borrow<[u32]> for Foo { +/// fn borrow(&self) -> &[u32] { &self.0 } +/// } +/// +/// impl std::borrow::Borrow<[String]> for Foo { +/// fn borrow(&self) -> &[String] { &self.1 } +/// } +/// ``` +#[unstable(feature = "slice_concat_trait", issue = "27747")] +pub trait Concat<Item: ?Sized> { + #[unstable(feature = "slice_concat_trait", issue = "27747")] + /// The resulting type after concatenation + type Output; + + /// Implementation of [`[T]::concat`](slice::concat) + #[unstable(feature = "slice_concat_trait", issue = "27747")] + fn concat(slice: &Self) -> Self::Output; +} + +/// Helper trait for [`[T]::join`](slice::join) +#[unstable(feature = "slice_concat_trait", issue = "27747")] +pub trait Join<Separator> { + #[unstable(feature = "slice_concat_trait", issue = "27747")] + /// The resulting type after concatenation + type Output; + + /// Implementation of [`[T]::join`](slice::join) + #[unstable(feature = "slice_concat_trait", issue = "27747")] + fn join(slice: &Self, sep: Separator) -> Self::Output; +} + +#[cfg(not(no_global_oom_handling))] +#[unstable(feature = "slice_concat_ext", issue = "27747")] +impl<T: Clone, V: Borrow<[T]>> Concat<T> for [V] { + type Output = Vec<T>; + + fn concat(slice: &Self) -> Vec<T> { + let size = slice.iter().map(|slice| slice.borrow().len()).sum(); + let mut result = Vec::with_capacity(size); + for v in slice { + result.extend_from_slice(v.borrow()) + } + result + } +} + +#[cfg(not(no_global_oom_handling))] +#[unstable(feature = "slice_concat_ext", issue = "27747")] +impl<T: Clone, V: Borrow<[T]>> Join<&T> for [V] { + type Output = Vec<T>; + + fn join(slice: &Self, sep: &T) -> Vec<T> { + let mut iter = slice.iter(); + let first = match iter.next() { + Some(first) => first, + None => return vec![], + }; + let size = slice.iter().map(|v| v.borrow().len()).sum::<usize>() + slice.len() - 1; + let mut result = Vec::with_capacity(size); + result.extend_from_slice(first.borrow()); + + for v in iter { + result.push(sep.clone()); + result.extend_from_slice(v.borrow()) + } + result + } +} + +#[cfg(not(no_global_oom_handling))] +#[unstable(feature = "slice_concat_ext", issue = "27747")] +impl<T: Clone, V: Borrow<[T]>> Join<&[T]> for [V] { + type Output = Vec<T>; + + fn join(slice: &Self, sep: &[T]) -> Vec<T> { + let mut iter = slice.iter(); + let first = match iter.next() { + Some(first) => first, + None => return vec![], + }; + let size = + slice.iter().map(|v| v.borrow().len()).sum::<usize>() + sep.len() * (slice.len() - 1); + let mut result = Vec::with_capacity(size); + result.extend_from_slice(first.borrow()); + + for v in iter { + result.extend_from_slice(sep); + result.extend_from_slice(v.borrow()) + } + result + } +} + +//////////////////////////////////////////////////////////////////////////////// +// Standard trait implementations for slices +//////////////////////////////////////////////////////////////////////////////// + +#[stable(feature = "rust1", since = "1.0.0")] +impl<T> Borrow<[T]> for Vec<T> { + fn borrow(&self) -> &[T] { + &self[..] + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<T> BorrowMut<[T]> for Vec<T> { + fn borrow_mut(&mut self) -> &mut [T] { + &mut self[..] + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "rust1", since = "1.0.0")] +impl<T: Clone> ToOwned for [T] { + type Owned = Vec<T>; + #[cfg(not(test))] + fn to_owned(&self) -> Vec<T> { + self.to_vec() + } + + #[cfg(test)] + fn to_owned(&self) -> Vec<T> { + hack::to_vec(self, Global) + } + + fn clone_into(&self, target: &mut Vec<T>) { + // drop anything in target that will not be overwritten + target.truncate(self.len()); + + // target.len <= self.len due to the truncate above, so the + // slices here are always in-bounds. + let (init, tail) = self.split_at(target.len()); + + // reuse the contained values' allocations/resources. + target.clone_from_slice(init); + target.extend_from_slice(tail); + } +} + +//////////////////////////////////////////////////////////////////////////////// +// Sorting +//////////////////////////////////////////////////////////////////////////////// + +/// Inserts `v[0]` into pre-sorted sequence `v[1..]` so that whole `v[..]` becomes sorted. +/// +/// This is the integral subroutine of insertion sort. +#[cfg(not(no_global_oom_handling))] +fn insert_head<T, F>(v: &mut [T], is_less: &mut F) +where + F: FnMut(&T, &T) -> bool, +{ + if v.len() >= 2 && is_less(&v[1], &v[0]) { + unsafe { + // There are three ways to implement insertion here: + // + // 1. Swap adjacent elements until the first one gets to its final destination. + // However, this way we copy data around more than is necessary. If elements are big + // structures (costly to copy), this method will be slow. + // + // 2. Iterate until the right place for the first element is found. Then shift the + // elements succeeding it to make room for it and finally place it into the + // remaining hole. This is a good method. + // + // 3. Copy the first element into a temporary variable. Iterate until the right place + // for it is found. As we go along, copy every traversed element into the slot + // preceding it. Finally, copy data from the temporary variable into the remaining + // hole. This method is very good. Benchmarks demonstrated slightly better + // performance than with the 2nd method. + // + // All methods were benchmarked, and the 3rd showed best results. So we chose that one. + let tmp = mem::ManuallyDrop::new(ptr::read(&v[0])); + + // Intermediate state of the insertion process is always tracked by `hole`, which + // serves two purposes: + // 1. Protects integrity of `v` from panics in `is_less`. + // 2. Fills the remaining hole in `v` in the end. + // + // Panic safety: + // + // If `is_less` panics at any point during the process, `hole` will get dropped and + // fill the hole in `v` with `tmp`, thus ensuring that `v` still holds every object it + // initially held exactly once. + let mut hole = InsertionHole { src: &*tmp, dest: &mut v[1] }; + ptr::copy_nonoverlapping(&v[1], &mut v[0], 1); + + for i in 2..v.len() { + if !is_less(&v[i], &*tmp) { + break; + } + ptr::copy_nonoverlapping(&v[i], &mut v[i - 1], 1); + hole.dest = &mut v[i]; + } + // `hole` gets dropped and thus copies `tmp` into the remaining hole in `v`. + } + } + + // When dropped, copies from `src` into `dest`. + struct InsertionHole<T> { + src: *const T, + dest: *mut T, + } + + impl<T> Drop for InsertionHole<T> { + fn drop(&mut self) { + unsafe { + ptr::copy_nonoverlapping(self.src, self.dest, 1); + } + } + } +} + +/// Merges non-decreasing runs `v[..mid]` and `v[mid..]` using `buf` as temporary storage, and +/// stores the result into `v[..]`. +/// +/// # Safety +/// +/// The two slices must be non-empty and `mid` must be in bounds. Buffer `buf` must be long enough +/// to hold a copy of the shorter slice. Also, `T` must not be a zero-sized type. +#[cfg(not(no_global_oom_handling))] +unsafe fn merge<T, F>(v: &mut [T], mid: usize, buf: *mut T, is_less: &mut F) +where + F: FnMut(&T, &T) -> bool, +{ + let len = v.len(); + let v = v.as_mut_ptr(); + let (v_mid, v_end) = unsafe { (v.add(mid), v.add(len)) }; + + // The merge process first copies the shorter run into `buf`. Then it traces the newly copied + // run and the longer run forwards (or backwards), comparing their next unconsumed elements and + // copying the lesser (or greater) one into `v`. + // + // As soon as the shorter run is fully consumed, the process is done. If the longer run gets + // consumed first, then we must copy whatever is left of the shorter run into the remaining + // hole in `v`. + // + // Intermediate state of the process is always tracked by `hole`, which serves two purposes: + // 1. Protects integrity of `v` from panics in `is_less`. + // 2. Fills the remaining hole in `v` if the longer run gets consumed first. + // + // Panic safety: + // + // If `is_less` panics at any point during the process, `hole` will get dropped and fill the + // hole in `v` with the unconsumed range in `buf`, thus ensuring that `v` still holds every + // object it initially held exactly once. + let mut hole; + + if mid <= len - mid { + // The left run is shorter. + unsafe { + ptr::copy_nonoverlapping(v, buf, mid); + hole = MergeHole { start: buf, end: buf.add(mid), dest: v }; + } + + // Initially, these pointers point to the beginnings of their arrays. + let left = &mut hole.start; + let mut right = v_mid; + let out = &mut hole.dest; + + while *left < hole.end && right < v_end { + // Consume the lesser side. + // If equal, prefer the left run to maintain stability. + unsafe { + let to_copy = if is_less(&*right, &**left) { + get_and_increment(&mut right) + } else { + get_and_increment(left) + }; + ptr::copy_nonoverlapping(to_copy, get_and_increment(out), 1); + } + } + } else { + // The right run is shorter. + unsafe { + ptr::copy_nonoverlapping(v_mid, buf, len - mid); + hole = MergeHole { start: buf, end: buf.add(len - mid), dest: v_mid }; + } + + // Initially, these pointers point past the ends of their arrays. + let left = &mut hole.dest; + let right = &mut hole.end; + let mut out = v_end; + + while v < *left && buf < *right { + // Consume the greater side. + // If equal, prefer the right run to maintain stability. + unsafe { + let to_copy = if is_less(&*right.offset(-1), &*left.offset(-1)) { + decrement_and_get(left) + } else { + decrement_and_get(right) + }; + ptr::copy_nonoverlapping(to_copy, decrement_and_get(&mut out), 1); + } + } + } + // Finally, `hole` gets dropped. If the shorter run was not fully consumed, whatever remains of + // it will now be copied into the hole in `v`. + + unsafe fn get_and_increment<T>(ptr: &mut *mut T) -> *mut T { + let old = *ptr; + *ptr = unsafe { ptr.offset(1) }; + old + } + + unsafe fn decrement_and_get<T>(ptr: &mut *mut T) -> *mut T { + *ptr = unsafe { ptr.offset(-1) }; + *ptr + } + + // When dropped, copies the range `start..end` into `dest..`. + struct MergeHole<T> { + start: *mut T, + end: *mut T, + dest: *mut T, + } + + impl<T> Drop for MergeHole<T> { + fn drop(&mut self) { + // `T` is not a zero-sized type, and these are pointers into a slice's elements. + unsafe { + let len = self.end.sub_ptr(self.start); + ptr::copy_nonoverlapping(self.start, self.dest, len); + } + } + } +} + +/// This merge sort borrows some (but not all) ideas from TimSort, which is described in detail +/// [here](https://github.com/python/cpython/blob/main/Objects/listsort.txt). +/// +/// The algorithm identifies strictly descending and non-descending subsequences, which are called +/// natural runs. There is a stack of pending runs yet to be merged. Each newly found run is pushed +/// onto the stack, and then some pairs of adjacent runs are merged until these two invariants are +/// satisfied: +/// +/// 1. for every `i` in `1..runs.len()`: `runs[i - 1].len > runs[i].len` +/// 2. for every `i` in `2..runs.len()`: `runs[i - 2].len > runs[i - 1].len + runs[i].len` +/// +/// The invariants ensure that the total running time is *O*(*n* \* log(*n*)) worst-case. +#[cfg(not(no_global_oom_handling))] +fn merge_sort<T, F>(v: &mut [T], mut is_less: F) +where + F: FnMut(&T, &T) -> bool, +{ + // Slices of up to this length get sorted using insertion sort. + const MAX_INSERTION: usize = 20; + // Very short runs are extended using insertion sort to span at least this many elements. + const MIN_RUN: usize = 10; + + // Sorting has no meaningful behavior on zero-sized types. + if size_of::<T>() == 0 { + return; + } + + let len = v.len(); + + // Short arrays get sorted in-place via insertion sort to avoid allocations. + if len <= MAX_INSERTION { + if len >= 2 { + for i in (0..len - 1).rev() { + insert_head(&mut v[i..], &mut is_less); + } + } + return; + } + + // Allocate a buffer to use as scratch memory. We keep the length 0 so we can keep in it + // shallow copies of the contents of `v` without risking the dtors running on copies if + // `is_less` panics. When merging two sorted runs, this buffer holds a copy of the shorter run, + // which will always have length at most `len / 2`. + let mut buf = Vec::with_capacity(len / 2); + + // In order to identify natural runs in `v`, we traverse it backwards. That might seem like a + // strange decision, but consider the fact that merges more often go in the opposite direction + // (forwards). According to benchmarks, merging forwards is slightly faster than merging + // backwards. To conclude, identifying runs by traversing backwards improves performance. + let mut runs = vec![]; + let mut end = len; + while end > 0 { + // Find the next natural run, and reverse it if it's strictly descending. + let mut start = end - 1; + if start > 0 { + start -= 1; + unsafe { + if is_less(v.get_unchecked(start + 1), v.get_unchecked(start)) { + while start > 0 && is_less(v.get_unchecked(start), v.get_unchecked(start - 1)) { + start -= 1; + } + v[start..end].reverse(); + } else { + while start > 0 && !is_less(v.get_unchecked(start), v.get_unchecked(start - 1)) + { + start -= 1; + } + } + } + } + + // Insert some more elements into the run if it's too short. Insertion sort is faster than + // merge sort on short sequences, so this significantly improves performance. + while start > 0 && end - start < MIN_RUN { + start -= 1; + insert_head(&mut v[start..end], &mut is_less); + } + + // Push this run onto the stack. + runs.push(Run { start, len: end - start }); + end = start; + + // Merge some pairs of adjacent runs to satisfy the invariants. + while let Some(r) = collapse(&runs) { + let left = runs[r + 1]; + let right = runs[r]; + unsafe { + merge( + &mut v[left.start..right.start + right.len], + left.len, + buf.as_mut_ptr(), + &mut is_less, + ); + } + runs[r] = Run { start: left.start, len: left.len + right.len }; + runs.remove(r + 1); + } + } + + // Finally, exactly one run must remain in the stack. + debug_assert!(runs.len() == 1 && runs[0].start == 0 && runs[0].len == len); + + // Examines the stack of runs and identifies the next pair of runs to merge. More specifically, + // if `Some(r)` is returned, that means `runs[r]` and `runs[r + 1]` must be merged next. If the + // algorithm should continue building a new run instead, `None` is returned. + // + // TimSort is infamous for its buggy implementations, as described here: + // http://envisage-project.eu/timsort-specification-and-verification/ + // + // The gist of the story is: we must enforce the invariants on the top four runs on the stack. + // Enforcing them on just top three is not sufficient to ensure that the invariants will still + // hold for *all* runs in the stack. + // + // This function correctly checks invariants for the top four runs. Additionally, if the top + // run starts at index 0, it will always demand a merge operation until the stack is fully + // collapsed, in order to complete the sort. + #[inline] + fn collapse(runs: &[Run]) -> Option<usize> { + let n = runs.len(); + if n >= 2 + && (runs[n - 1].start == 0 + || runs[n - 2].len <= runs[n - 1].len + || (n >= 3 && runs[n - 3].len <= runs[n - 2].len + runs[n - 1].len) + || (n >= 4 && runs[n - 4].len <= runs[n - 3].len + runs[n - 2].len)) + { + if n >= 3 && runs[n - 3].len < runs[n - 1].len { Some(n - 3) } else { Some(n - 2) } + } else { + None + } + } + + #[derive(Clone, Copy)] + struct Run { + start: usize, + len: usize, + } +} diff --git a/rust/alloc/vec/drain.rs b/rust/alloc/vec/drain.rs new file mode 100644 index 000000000000..b6a5f98e4fcd --- /dev/null +++ b/rust/alloc/vec/drain.rs @@ -0,0 +1,186 @@ +// SPDX-License-Identifier: Apache-2.0 OR MIT + +use crate::alloc::{Allocator, Global}; +use core::fmt; +use core::iter::{FusedIterator, TrustedLen}; +use core::mem; +use core::ptr::{self, NonNull}; +use core::slice::{self}; + +use super::Vec; + +/// A draining iterator for `Vec<T>`. +/// +/// This `struct` is created by [`Vec::drain`]. +/// See its documentation for more. +/// +/// # Example +/// +/// ``` +/// let mut v = vec![0, 1, 2]; +/// let iter: std::vec::Drain<_> = v.drain(..); +/// ``` +#[stable(feature = "drain", since = "1.6.0")] +pub struct Drain< + 'a, + T: 'a, + #[unstable(feature = "allocator_api", issue = "32838")] A: Allocator + 'a = Global, +> { + /// Index of tail to preserve + pub(super) tail_start: usize, + /// Length of tail + pub(super) tail_len: usize, + /// Current remaining range to remove + pub(super) iter: slice::Iter<'a, T>, + pub(super) vec: NonNull<Vec<T, A>>, +} + +#[stable(feature = "collection_debug", since = "1.17.0")] +impl<T: fmt::Debug, A: Allocator> fmt::Debug for Drain<'_, T, A> { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + f.debug_tuple("Drain").field(&self.iter.as_slice()).finish() + } +} + +impl<'a, T, A: Allocator> Drain<'a, T, A> { + /// Returns the remaining items of this iterator as a slice. + /// + /// # Examples + /// + /// ``` + /// let mut vec = vec!['a', 'b', 'c']; + /// let mut drain = vec.drain(..); + /// assert_eq!(drain.as_slice(), &['a', 'b', 'c']); + /// let _ = drain.next().unwrap(); + /// assert_eq!(drain.as_slice(), &['b', 'c']); + /// ``` + #[must_use] + #[stable(feature = "vec_drain_as_slice", since = "1.46.0")] + pub fn as_slice(&self) -> &[T] { + self.iter.as_slice() + } + + /// Returns a reference to the underlying allocator. + #[unstable(feature = "allocator_api", issue = "32838")] + #[must_use] + #[inline] + pub fn allocator(&self) -> &A { + unsafe { self.vec.as_ref().allocator() } + } +} + +#[stable(feature = "vec_drain_as_slice", since = "1.46.0")] +impl<'a, T, A: Allocator> AsRef<[T]> for Drain<'a, T, A> { + fn as_ref(&self) -> &[T] { + self.as_slice() + } +} + +#[stable(feature = "drain", since = "1.6.0")] +unsafe impl<T: Sync, A: Sync + Allocator> Sync for Drain<'_, T, A> {} +#[stable(feature = "drain", since = "1.6.0")] +unsafe impl<T: Send, A: Send + Allocator> Send for Drain<'_, T, A> {} + +#[stable(feature = "drain", since = "1.6.0")] +impl<T, A: Allocator> Iterator for Drain<'_, T, A> { + type Item = T; + + #[inline] + fn next(&mut self) -> Option<T> { + self.iter.next().map(|elt| unsafe { ptr::read(elt as *const _) }) + } + + fn size_hint(&self) -> (usize, Option<usize>) { + self.iter.size_hint() + } +} + +#[stable(feature = "drain", since = "1.6.0")] +impl<T, A: Allocator> DoubleEndedIterator for Drain<'_, T, A> { + #[inline] + fn next_back(&mut self) -> Option<T> { + self.iter.next_back().map(|elt| unsafe { ptr::read(elt as *const _) }) + } +} + +#[stable(feature = "drain", since = "1.6.0")] +impl<T, A: Allocator> Drop for Drain<'_, T, A> { + fn drop(&mut self) { + /// Moves back the un-`Drain`ed elements to restore the original `Vec`. + struct DropGuard<'r, 'a, T, A: Allocator>(&'r mut Drain<'a, T, A>); + + impl<'r, 'a, T, A: Allocator> Drop for DropGuard<'r, 'a, T, A> { + fn drop(&mut self) { + if self.0.tail_len > 0 { + unsafe { + let source_vec = self.0.vec.as_mut(); + // memmove back untouched tail, update to new length + let start = source_vec.len(); + let tail = self.0.tail_start; + if tail != start { + let src = source_vec.as_ptr().add(tail); + let dst = source_vec.as_mut_ptr().add(start); + ptr::copy(src, dst, self.0.tail_len); + } + source_vec.set_len(start + self.0.tail_len); + } + } + } + } + + let iter = mem::replace(&mut self.iter, (&mut []).iter()); + let drop_len = iter.len(); + + let mut vec = self.vec; + + if mem::size_of::<T>() == 0 { + // ZSTs have no identity, so we don't need to move them around, we only need to drop the correct amount. + // this can be achieved by manipulating the Vec length instead of moving values out from `iter`. + unsafe { + let vec = vec.as_mut(); + let old_len = vec.len(); + vec.set_len(old_len + drop_len + self.tail_len); + vec.truncate(old_len + self.tail_len); + } + + return; + } + + // ensure elements are moved back into their appropriate places, even when drop_in_place panics + let _guard = DropGuard(self); + + if drop_len == 0 { + return; + } + + // as_slice() must only be called when iter.len() is > 0 because + // vec::Splice modifies vec::Drain fields and may grow the vec which would invalidate + // the iterator's internal pointers. Creating a reference to deallocated memory + // is invalid even when it is zero-length + let drop_ptr = iter.as_slice().as_ptr(); + + unsafe { + // drop_ptr comes from a slice::Iter which only gives us a &[T] but for drop_in_place + // a pointer with mutable provenance is necessary. Therefore we must reconstruct + // it from the original vec but also avoid creating a &mut to the front since that could + // invalidate raw pointers to it which some unsafe code might rely on. + let vec_ptr = vec.as_mut().as_mut_ptr(); + let drop_offset = drop_ptr.sub_ptr(vec_ptr); + let to_drop = ptr::slice_from_raw_parts_mut(vec_ptr.add(drop_offset), drop_len); + ptr::drop_in_place(to_drop); + } + } +} + +#[stable(feature = "drain", since = "1.6.0")] +impl<T, A: Allocator> ExactSizeIterator for Drain<'_, T, A> { + fn is_empty(&self) -> bool { + self.iter.is_empty() + } +} + +#[unstable(feature = "trusted_len", issue = "37572")] +unsafe impl<T, A: Allocator> TrustedLen for Drain<'_, T, A> {} + +#[stable(feature = "fused", since = "1.26.0")] +impl<T, A: Allocator> FusedIterator for Drain<'_, T, A> {} diff --git a/rust/alloc/vec/drain_filter.rs b/rust/alloc/vec/drain_filter.rs new file mode 100644 index 000000000000..b04fce041622 --- /dev/null +++ b/rust/alloc/vec/drain_filter.rs @@ -0,0 +1,145 @@ +// SPDX-License-Identifier: Apache-2.0 OR MIT + +use crate::alloc::{Allocator, Global}; +use core::ptr::{self}; +use core::slice::{self}; + +use super::Vec; + +/// An iterator which uses a closure to determine if an element should be removed. +/// +/// This struct is created by [`Vec::drain_filter`]. +/// See its documentation for more. +/// +/// # Example +/// +/// ``` +/// #![feature(drain_filter)] +/// +/// let mut v = vec![0, 1, 2]; +/// let iter: std::vec::DrainFilter<_, _> = v.drain_filter(|x| *x % 2 == 0); +/// ``` +#[unstable(feature = "drain_filter", reason = "recently added", issue = "43244")] +#[derive(Debug)] +pub struct DrainFilter< + 'a, + T, + F, + #[unstable(feature = "allocator_api", issue = "32838")] A: Allocator = Global, +> where + F: FnMut(&mut T) -> bool, +{ + pub(super) vec: &'a mut Vec<T, A>, + /// The index of the item that will be inspected by the next call to `next`. + pub(super) idx: usize, + /// The number of items that have been drained (removed) thus far. + pub(super) del: usize, + /// The original length of `vec` prior to draining. + pub(super) old_len: usize, + /// The filter test predicate. + pub(super) pred: F, + /// A flag that indicates a panic has occurred in the filter test predicate. + /// This is used as a hint in the drop implementation to prevent consumption + /// of the remainder of the `DrainFilter`. Any unprocessed items will be + /// backshifted in the `vec`, but no further items will be dropped or + /// tested by the filter predicate. + pub(super) panic_flag: bool, +} + +impl<T, F, A: Allocator> DrainFilter<'_, T, F, A> +where + F: FnMut(&mut T) -> bool, +{ + /// Returns a reference to the underlying allocator. + #[unstable(feature = "allocator_api", issue = "32838")] + #[inline] + pub fn allocator(&self) -> &A { + self.vec.allocator() + } +} + +#[unstable(feature = "drain_filter", reason = "recently added", issue = "43244")] +impl<T, F, A: Allocator> Iterator for DrainFilter<'_, T, F, A> +where + F: FnMut(&mut T) -> bool, +{ + type Item = T; + + fn next(&mut self) -> Option<T> { + unsafe { + while self.idx < self.old_len { + let i = self.idx; + let v = slice::from_raw_parts_mut(self.vec.as_mut_ptr(), self.old_len); + self.panic_flag = true; + let drained = (self.pred)(&mut v[i]); + self.panic_flag = false; + // Update the index *after* the predicate is called. If the index + // is updated prior and the predicate panics, the element at this + // index would be leaked. + self.idx += 1; + if drained { + self.del += 1; + return Some(ptr::read(&v[i])); + } else if self.del > 0 { + let del = self.del; + let src: *const T = &v[i]; + let dst: *mut T = &mut v[i - del]; + ptr::copy_nonoverlapping(src, dst, 1); + } + } + None + } + } + + fn size_hint(&self) -> (usize, Option<usize>) { + (0, Some(self.old_len - self.idx)) + } +} + +#[unstable(feature = "drain_filter", reason = "recently added", issue = "43244")] +impl<T, F, A: Allocator> Drop for DrainFilter<'_, T, F, A> +where + F: FnMut(&mut T) -> bool, +{ + fn drop(&mut self) { + struct BackshiftOnDrop<'a, 'b, T, F, A: Allocator> + where + F: FnMut(&mut T) -> bool, + { + drain: &'b mut DrainFilter<'a, T, F, A>, + } + + impl<'a, 'b, T, F, A: Allocator> Drop for BackshiftOnDrop<'a, 'b, T, F, A> + where + F: FnMut(&mut T) -> bool, + { + fn drop(&mut self) { + unsafe { + if self.drain.idx < self.drain.old_len && self.drain.del > 0 { + // This is a pretty messed up state, and there isn't really an + // obviously right thing to do. We don't want to keep trying + // to execute `pred`, so we just backshift all the unprocessed + // elements and tell the vec that they still exist. The backshift + // is required to prevent a double-drop of the last successfully + // drained item prior to a panic in the predicate. + let ptr = self.drain.vec.as_mut_ptr(); + let src = ptr.add(self.drain.idx); + let dst = src.sub(self.drain.del); + let tail_len = self.drain.old_len - self.drain.idx; + src.copy_to(dst, tail_len); + } + self.drain.vec.set_len(self.drain.old_len - self.drain.del); + } + } + } + + let backshift = BackshiftOnDrop { drain: self }; + + // Attempt to consume any remaining elements if the filter predicate + // has not yet panicked. We'll backshift any remaining elements + // whether we've already panicked or if the consumption here panics. + if !backshift.drain.panic_flag { + backshift.drain.for_each(drop); + } + } +} diff --git a/rust/alloc/vec/into_iter.rs b/rust/alloc/vec/into_iter.rs new file mode 100644 index 000000000000..f7a50e76691e --- /dev/null +++ b/rust/alloc/vec/into_iter.rs @@ -0,0 +1,366 @@ +// SPDX-License-Identifier: Apache-2.0 OR MIT + +#[cfg(not(no_global_oom_handling))] +use super::AsVecIntoIter; +use crate::alloc::{Allocator, Global}; +use crate::raw_vec::RawVec; +use core::fmt; +use core::intrinsics::arith_offset; +use core::iter::{ + FusedIterator, InPlaceIterable, SourceIter, TrustedLen, TrustedRandomAccessNoCoerce, +}; +use core::marker::PhantomData; +use core::mem::{self, ManuallyDrop}; +#[cfg(not(no_global_oom_handling))] +use core::ops::Deref; +use core::ptr::{self, NonNull}; +use core::slice::{self}; + +/// An iterator that moves out of a vector. +/// +/// This `struct` is created by the `into_iter` method on [`Vec`](super::Vec) +/// (provided by the [`IntoIterator`] trait). +/// +/// # Example +/// +/// ``` +/// let v = vec![0, 1, 2]; +/// let iter: std::vec::IntoIter<_> = v.into_iter(); +/// ``` +#[stable(feature = "rust1", since = "1.0.0")] +#[rustc_insignificant_dtor] +pub struct IntoIter< + T, + #[unstable(feature = "allocator_api", issue = "32838")] A: Allocator = Global, +> { + pub(super) buf: NonNull<T>, + pub(super) phantom: PhantomData<T>, + pub(super) cap: usize, + // the drop impl reconstructs a RawVec from buf, cap and alloc + // to avoid dropping the allocator twice we need to wrap it into ManuallyDrop + pub(super) alloc: ManuallyDrop<A>, + pub(super) ptr: *const T, + pub(super) end: *const T, +} + +#[stable(feature = "vec_intoiter_debug", since = "1.13.0")] +impl<T: fmt::Debug, A: Allocator> fmt::Debug for IntoIter<T, A> { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + f.debug_tuple("IntoIter").field(&self.as_slice()).finish() + } +} + +impl<T, A: Allocator> IntoIter<T, A> { + /// Returns the remaining items of this iterator as a slice. + /// + /// # Examples + /// + /// ``` + /// let vec = vec!['a', 'b', 'c']; + /// let mut into_iter = vec.into_iter(); + /// assert_eq!(into_iter.as_slice(), &['a', 'b', 'c']); + /// let _ = into_iter.next().unwrap(); + /// assert_eq!(into_iter.as_slice(), &['b', 'c']); + /// ``` + #[stable(feature = "vec_into_iter_as_slice", since = "1.15.0")] + pub fn as_slice(&self) -> &[T] { + unsafe { slice::from_raw_parts(self.ptr, self.len()) } + } + + /// Returns the remaining items of this iterator as a mutable slice. + /// + /// # Examples + /// + /// ``` + /// let vec = vec!['a', 'b', 'c']; + /// let mut into_iter = vec.into_iter(); + /// assert_eq!(into_iter.as_slice(), &['a', 'b', 'c']); + /// into_iter.as_mut_slice()[2] = 'z'; + /// assert_eq!(into_iter.next().unwrap(), 'a'); + /// assert_eq!(into_iter.next().unwrap(), 'b'); + /// assert_eq!(into_iter.next().unwrap(), 'z'); + /// ``` + #[stable(feature = "vec_into_iter_as_slice", since = "1.15.0")] + pub fn as_mut_slice(&mut self) -> &mut [T] { + unsafe { &mut *self.as_raw_mut_slice() } + } + + /// Returns a reference to the underlying allocator. + #[unstable(feature = "allocator_api", issue = "32838")] + #[inline] + pub fn allocator(&self) -> &A { + &self.alloc + } + + fn as_raw_mut_slice(&mut self) -> *mut [T] { + ptr::slice_from_raw_parts_mut(self.ptr as *mut T, self.len()) + } + + /// Drops remaining elements and relinquishes the backing allocation. + /// + /// This is roughly equivalent to the following, but more efficient + /// + /// ``` + /// # let mut into_iter = Vec::<u8>::with_capacity(10).into_iter(); + /// (&mut into_iter).for_each(core::mem::drop); + /// unsafe { core::ptr::write(&mut into_iter, Vec::new().into_iter()); } + /// ``` + /// + /// This method is used by in-place iteration, refer to the vec::in_place_collect + /// documentation for an overview. + #[cfg(not(no_global_oom_handling))] + pub(super) fn forget_allocation_drop_remaining(&mut self) { + let remaining = self.as_raw_mut_slice(); + + // overwrite the individual fields instead of creating a new + // struct and then overwriting &mut self. + // this creates less assembly + self.cap = 0; + self.buf = unsafe { NonNull::new_unchecked(RawVec::NEW.ptr()) }; + self.ptr = self.buf.as_ptr(); + self.end = self.buf.as_ptr(); + + unsafe { + ptr::drop_in_place(remaining); + } + } + + /// Forgets to Drop the remaining elements while still allowing the backing allocation to be freed. + #[allow(dead_code)] + pub(crate) fn forget_remaining_elements(&mut self) { + self.ptr = self.end; + } +} + +#[stable(feature = "vec_intoiter_as_ref", since = "1.46.0")] +impl<T, A: Allocator> AsRef<[T]> for IntoIter<T, A> { + fn as_ref(&self) -> &[T] { + self.as_slice() + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +unsafe impl<T: Send, A: Allocator + Send> Send for IntoIter<T, A> {} +#[stable(feature = "rust1", since = "1.0.0")] +unsafe impl<T: Sync, A: Allocator + Sync> Sync for IntoIter<T, A> {} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<T, A: Allocator> Iterator for IntoIter<T, A> { + type Item = T; + + #[inline] + fn next(&mut self) -> Option<T> { + if self.ptr as *const _ == self.end { + None + } else if mem::size_of::<T>() == 0 { + // purposefully don't use 'ptr.offset' because for + // vectors with 0-size elements this would return the + // same pointer. + self.ptr = unsafe { arith_offset(self.ptr as *const i8, 1) as *mut T }; + + // Make up a value of this ZST. + Some(unsafe { mem::zeroed() }) + } else { + let old = self.ptr; + self.ptr = unsafe { self.ptr.offset(1) }; + + Some(unsafe { ptr::read(old) }) + } + } + + #[inline] + fn size_hint(&self) -> (usize, Option<usize>) { + let exact = if mem::size_of::<T>() == 0 { + self.end.addr().wrapping_sub(self.ptr.addr()) + } else { + unsafe { self.end.sub_ptr(self.ptr) } + }; + (exact, Some(exact)) + } + + #[inline] + fn advance_by(&mut self, n: usize) -> Result<(), usize> { + let step_size = self.len().min(n); + let to_drop = ptr::slice_from_raw_parts_mut(self.ptr as *mut T, step_size); + if mem::size_of::<T>() == 0 { + // SAFETY: due to unchecked casts of unsigned amounts to signed offsets the wraparound + // effectively results in unsigned pointers representing positions 0..usize::MAX, + // which is valid for ZSTs. + self.ptr = unsafe { arith_offset(self.ptr as *const i8, step_size as isize) as *mut T } + } else { + // SAFETY: the min() above ensures that step_size is in bounds + self.ptr = unsafe { self.ptr.add(step_size) }; + } + // SAFETY: the min() above ensures that step_size is in bounds + unsafe { + ptr::drop_in_place(to_drop); + } + if step_size < n { + return Err(step_size); + } + Ok(()) + } + + #[inline] + fn count(self) -> usize { + self.len() + } + + unsafe fn __iterator_get_unchecked(&mut self, i: usize) -> Self::Item + where + Self: TrustedRandomAccessNoCoerce, + { + // SAFETY: the caller must guarantee that `i` is in bounds of the + // `Vec<T>`, so `i` cannot overflow an `isize`, and the `self.ptr.add(i)` + // is guaranteed to pointer to an element of the `Vec<T>` and + // thus guaranteed to be valid to dereference. + // + // Also note the implementation of `Self: TrustedRandomAccess` requires + // that `T: Copy` so reading elements from the buffer doesn't invalidate + // them for `Drop`. + unsafe { + if mem::size_of::<T>() == 0 { mem::zeroed() } else { ptr::read(self.ptr.add(i)) } + } + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<T, A: Allocator> DoubleEndedIterator for IntoIter<T, A> { + #[inline] + fn next_back(&mut self) -> Option<T> { + if self.end == self.ptr { + None + } else if mem::size_of::<T>() == 0 { + // See above for why 'ptr.offset' isn't used + self.end = unsafe { arith_offset(self.end as *const i8, -1) as *mut T }; + + // Make up a value of this ZST. + Some(unsafe { mem::zeroed() }) + } else { + self.end = unsafe { self.end.offset(-1) }; + + Some(unsafe { ptr::read(self.end) }) + } + } + + #[inline] + fn advance_back_by(&mut self, n: usize) -> Result<(), usize> { + let step_size = self.len().min(n); + if mem::size_of::<T>() == 0 { + // SAFETY: same as for advance_by() + self.end = unsafe { + arith_offset(self.end as *const i8, step_size.wrapping_neg() as isize) as *mut T + } + } else { + // SAFETY: same as for advance_by() + self.end = unsafe { self.end.offset(step_size.wrapping_neg() as isize) }; + } + let to_drop = ptr::slice_from_raw_parts_mut(self.end as *mut T, step_size); + // SAFETY: same as for advance_by() + unsafe { + ptr::drop_in_place(to_drop); + } + if step_size < n { + return Err(step_size); + } + Ok(()) + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<T, A: Allocator> ExactSizeIterator for IntoIter<T, A> { + fn is_empty(&self) -> bool { + self.ptr == self.end + } +} + +#[stable(feature = "fused", since = "1.26.0")] +impl<T, A: Allocator> FusedIterator for IntoIter<T, A> {} + +#[unstable(feature = "trusted_len", issue = "37572")] +unsafe impl<T, A: Allocator> TrustedLen for IntoIter<T, A> {} + +#[doc(hidden)] +#[unstable(issue = "none", feature = "std_internals")] +#[rustc_unsafe_specialization_marker] +pub trait NonDrop {} + +// T: Copy as approximation for !Drop since get_unchecked does not advance self.ptr +// and thus we can't implement drop-handling +#[unstable(issue = "none", feature = "std_internals")] +impl<T: Copy> NonDrop for T {} + +#[doc(hidden)] +#[unstable(issue = "none", feature = "std_internals")] +// TrustedRandomAccess (without NoCoerce) must not be implemented because +// subtypes/supertypes of `T` might not be `NonDrop` +unsafe impl<T, A: Allocator> TrustedRandomAccessNoCoerce for IntoIter<T, A> +where + T: NonDrop, +{ + const MAY_HAVE_SIDE_EFFECT: bool = false; +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "vec_into_iter_clone", since = "1.8.0")] +impl<T: Clone, A: Allocator + Clone> Clone for IntoIter<T, A> { + #[cfg(not(test))] + fn clone(&self) -> Self { + self.as_slice().to_vec_in(self.alloc.deref().clone()).into_iter() + } + #[cfg(test)] + fn clone(&self) -> Self { + crate::slice::to_vec(self.as_slice(), self.alloc.deref().clone()).into_iter() + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +unsafe impl<#[may_dangle] T, A: Allocator> Drop for IntoIter<T, A> { + fn drop(&mut self) { + struct DropGuard<'a, T, A: Allocator>(&'a mut IntoIter<T, A>); + + impl<T, A: Allocator> Drop for DropGuard<'_, T, A> { + fn drop(&mut self) { + unsafe { + // `IntoIter::alloc` is not used anymore after this and will be dropped by RawVec + let alloc = ManuallyDrop::take(&mut self.0.alloc); + // RawVec handles deallocation + let _ = RawVec::from_raw_parts_in(self.0.buf.as_ptr(), self.0.cap, alloc); + } + } + } + + let guard = DropGuard(self); + // destroy the remaining elements + unsafe { + ptr::drop_in_place(guard.0.as_raw_mut_slice()); + } + // now `guard` will be dropped and do the rest + } +} + +// In addition to the SAFETY invariants of the following three unsafe traits +// also refer to the vec::in_place_collect module documentation to get an overview +#[unstable(issue = "none", feature = "inplace_iteration")] +#[doc(hidden)] +unsafe impl<T, A: Allocator> InPlaceIterable for IntoIter<T, A> {} + +#[unstable(issue = "none", feature = "inplace_iteration")] +#[doc(hidden)] +unsafe impl<T, A: Allocator> SourceIter for IntoIter<T, A> { + type Source = Self; + + #[inline] + unsafe fn as_inner(&mut self) -> &mut Self::Source { + self + } +} + +#[cfg(not(no_global_oom_handling))] +unsafe impl<T> AsVecIntoIter for IntoIter<T> { + type Item = T; + + fn as_into_iter(&mut self) -> &mut IntoIter<Self::Item> { + self + } +} diff --git a/rust/alloc/vec/is_zero.rs b/rust/alloc/vec/is_zero.rs new file mode 100644 index 000000000000..377f3d172777 --- /dev/null +++ b/rust/alloc/vec/is_zero.rs @@ -0,0 +1,120 @@ +// SPDX-License-Identifier: Apache-2.0 OR MIT + +use crate::boxed::Box; + +#[rustc_specialization_trait] +pub(super) unsafe trait IsZero { + /// Whether this value's representation is all zeros + fn is_zero(&self) -> bool; +} + +macro_rules! impl_is_zero { + ($t:ty, $is_zero:expr) => { + unsafe impl IsZero for $t { + #[inline] + fn is_zero(&self) -> bool { + $is_zero(*self) + } + } + }; +} + +impl_is_zero!(i16, |x| x == 0); +impl_is_zero!(i32, |x| x == 0); +impl_is_zero!(i64, |x| x == 0); +impl_is_zero!(i128, |x| x == 0); +impl_is_zero!(isize, |x| x == 0); + +impl_is_zero!(u16, |x| x == 0); +impl_is_zero!(u32, |x| x == 0); +impl_is_zero!(u64, |x| x == 0); +impl_is_zero!(u128, |x| x == 0); +impl_is_zero!(usize, |x| x == 0); + +impl_is_zero!(bool, |x| x == false); +impl_is_zero!(char, |x| x == '\0'); + +impl_is_zero!(f32, |x: f32| x.to_bits() == 0); +impl_is_zero!(f64, |x: f64| x.to_bits() == 0); + +unsafe impl<T> IsZero for *const T { + #[inline] + fn is_zero(&self) -> bool { + (*self).is_null() + } +} + +unsafe impl<T> IsZero for *mut T { + #[inline] + fn is_zero(&self) -> bool { + (*self).is_null() + } +} + +unsafe impl<T: IsZero, const N: usize> IsZero for [T; N] { + #[inline] + fn is_zero(&self) -> bool { + // Because this is generated as a runtime check, it's not obvious that + // it's worth doing if the array is really long. The threshold here + // is largely arbitrary, but was picked because as of 2022-05-01 LLVM + // can const-fold the check in `vec![[0; 32]; n]` but not in + // `vec![[0; 64]; n]`: https://godbolt.org/z/WTzjzfs5b + // Feel free to tweak if you have better evidence. + + N <= 32 && self.iter().all(IsZero::is_zero) + } +} + +// `Option<&T>` and `Option<Box<T>>` are guaranteed to represent `None` as null. +// For fat pointers, the bytes that would be the pointer metadata in the `Some` +// variant are padding in the `None` variant, so ignoring them and +// zero-initializing instead is ok. +// `Option<&mut T>` never implements `Clone`, so there's no need for an impl of +// `SpecFromElem`. + +unsafe impl<T: ?Sized> IsZero for Option<&T> { + #[inline] + fn is_zero(&self) -> bool { + self.is_none() + } +} + +unsafe impl<T: ?Sized> IsZero for Option<Box<T>> { + #[inline] + fn is_zero(&self) -> bool { + self.is_none() + } +} + +// `Option<num::NonZeroU32>` and similar have a representation guarantee that +// they're the same size as the corresponding `u32` type, as well as a guarantee +// that transmuting between `NonZeroU32` and `Option<num::NonZeroU32>` works. +// While the documentation officially makes it UB to transmute from `None`, +// we're the standard library so we can make extra inferences, and we know that +// the only niche available to represent `None` is the one that's all zeros. + +macro_rules! impl_is_zero_option_of_nonzero { + ($($t:ident,)+) => {$( + unsafe impl IsZero for Option<core::num::$t> { + #[inline] + fn is_zero(&self) -> bool { + self.is_none() + } + } + )+}; +} + +impl_is_zero_option_of_nonzero!( + NonZeroU8, + NonZeroU16, + NonZeroU32, + NonZeroU64, + NonZeroU128, + NonZeroI8, + NonZeroI16, + NonZeroI32, + NonZeroI64, + NonZeroI128, + NonZeroUsize, + NonZeroIsize, +); diff --git a/rust/alloc/vec/mod.rs b/rust/alloc/vec/mod.rs new file mode 100644 index 000000000000..540787804cc2 --- /dev/null +++ b/rust/alloc/vec/mod.rs @@ -0,0 +1,3140 @@ +// SPDX-License-Identifier: Apache-2.0 OR MIT + +//! A contiguous growable array type with heap-allocated contents, written +//! `Vec<T>`. +//! +//! Vectors have *O*(1) indexing, amortized *O*(1) push (to the end) and +//! *O*(1) pop (from the end). +//! +//! Vectors ensure they never allocate more than `isize::MAX` bytes. +//! +//! # Examples +//! +//! You can explicitly create a [`Vec`] with [`Vec::new`]: +//! +//! ``` +//! let v: Vec<i32> = Vec::new(); +//! ``` +//! +//! ...or by using the [`vec!`] macro: +//! +//! ``` +//! let v: Vec<i32> = vec![]; +//! +//! let v = vec![1, 2, 3, 4, 5]; +//! +//! let v = vec![0; 10]; // ten zeroes +//! ``` +//! +//! You can [`push`] values onto the end of a vector (which will grow the vector +//! as needed): +//! +//! ``` +//! let mut v = vec![1, 2]; +//! +//! v.push(3); +//! ``` +//! +//! Popping values works in much the same way: +//! +//! ``` +//! let mut v = vec![1, 2]; +//! +//! let two = v.pop(); +//! ``` +//! +//! Vectors also support indexing (through the [`Index`] and [`IndexMut`] traits): +//! +//! ``` +//! let mut v = vec![1, 2, 3]; +//! let three = v[2]; +//! v[1] = v[1] + 5; +//! ``` +//! +//! [`push`]: Vec::push + +#![stable(feature = "rust1", since = "1.0.0")] + +#[cfg(not(no_global_oom_handling))] +use core::cmp; +use core::cmp::Ordering; +use core::convert::TryFrom; +use core::fmt; +use core::hash::{Hash, Hasher}; +use core::intrinsics::{arith_offset, assume}; +use core::iter; +#[cfg(not(no_global_oom_handling))] +use core::iter::FromIterator; +use core::marker::PhantomData; +use core::mem::{self, ManuallyDrop, MaybeUninit}; +use core::ops::{self, Index, IndexMut, Range, RangeBounds}; +use core::ptr::{self, NonNull}; +use core::slice::{self, SliceIndex}; + +use crate::alloc::{Allocator, Global}; +use crate::borrow::{Cow, ToOwned}; +use crate::boxed::Box; +use crate::collections::TryReserveError; +use crate::raw_vec::RawVec; + +#[unstable(feature = "drain_filter", reason = "recently added", issue = "43244")] +pub use self::drain_filter::DrainFilter; + +mod drain_filter; + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "vec_splice", since = "1.21.0")] +pub use self::splice::Splice; + +#[cfg(not(no_global_oom_handling))] +mod splice; + +#[stable(feature = "drain", since = "1.6.0")] +pub use self::drain::Drain; + +mod drain; + +#[cfg(not(no_global_oom_handling))] +mod cow; + +#[cfg(not(no_global_oom_handling))] +pub(crate) use self::in_place_collect::AsVecIntoIter; +#[stable(feature = "rust1", since = "1.0.0")] +pub use self::into_iter::IntoIter; + +mod into_iter; + +#[cfg(not(no_global_oom_handling))] +use self::is_zero::IsZero; + +mod is_zero; + +#[cfg(not(no_global_oom_handling))] +mod in_place_collect; + +mod partial_eq; + +#[cfg(not(no_global_oom_handling))] +use self::spec_from_elem::SpecFromElem; + +#[cfg(not(no_global_oom_handling))] +mod spec_from_elem; + +#[cfg(not(no_global_oom_handling))] +use self::set_len_on_drop::SetLenOnDrop; + +#[cfg(not(no_global_oom_handling))] +mod set_len_on_drop; + +#[cfg(not(no_global_oom_handling))] +use self::in_place_drop::InPlaceDrop; + +#[cfg(not(no_global_oom_handling))] +mod in_place_drop; + +#[cfg(not(no_global_oom_handling))] +use self::spec_from_iter_nested::SpecFromIterNested; + +#[cfg(not(no_global_oom_handling))] +mod spec_from_iter_nested; + +#[cfg(not(no_global_oom_handling))] +use self::spec_from_iter::SpecFromIter; + +#[cfg(not(no_global_oom_handling))] +mod spec_from_iter; + +#[cfg(not(no_global_oom_handling))] +use self::spec_extend::SpecExtend; + +#[cfg(not(no_global_oom_handling))] +mod spec_extend; + +/// A contiguous growable array type, written as `Vec<T>`, short for 'vector'. +/// +/// # Examples +/// +/// ``` +/// let mut vec = Vec::new(); +/// vec.push(1); +/// vec.push(2); +/// +/// assert_eq!(vec.len(), 2); +/// assert_eq!(vec[0], 1); +/// +/// assert_eq!(vec.pop(), Some(2)); +/// assert_eq!(vec.len(), 1); +/// +/// vec[0] = 7; +/// assert_eq!(vec[0], 7); +/// +/// vec.extend([1, 2, 3].iter().copied()); +/// +/// for x in &vec { +/// println!("{x}"); +/// } +/// assert_eq!(vec, [7, 1, 2, 3]); +/// ``` +/// +/// The [`vec!`] macro is provided for convenient initialization: +/// +/// ``` +/// let mut vec1 = vec![1, 2, 3]; +/// vec1.push(4); +/// let vec2 = Vec::from([1, 2, 3, 4]); +/// assert_eq!(vec1, vec2); +/// ``` +/// +/// It can also initialize each element of a `Vec<T>` with a given value. +/// This may be more efficient than performing allocation and initialization +/// in separate steps, especially when initializing a vector of zeros: +/// +/// ``` +/// let vec = vec![0; 5]; +/// assert_eq!(vec, [0, 0, 0, 0, 0]); +/// +/// // The following is equivalent, but potentially slower: +/// let mut vec = Vec::with_capacity(5); +/// vec.resize(5, 0); +/// assert_eq!(vec, [0, 0, 0, 0, 0]); +/// ``` +/// +/// For more information, see +/// [Capacity and Reallocation](#capacity-and-reallocation). +/// +/// Use a `Vec<T>` as an efficient stack: +/// +/// ``` +/// let mut stack = Vec::new(); +/// +/// stack.push(1); +/// stack.push(2); +/// stack.push(3); +/// +/// while let Some(top) = stack.pop() { +/// // Prints 3, 2, 1 +/// println!("{top}"); +/// } +/// ``` +/// +/// # Indexing +/// +/// The `Vec` type allows to access values by index, because it implements the +/// [`Index`] trait. An example will be more explicit: +/// +/// ``` +/// let v = vec![0, 2, 4, 6]; +/// println!("{}", v[1]); // it will display '2' +/// ``` +/// +/// However be careful: if you try to access an index which isn't in the `Vec`, +/// your software will panic! You cannot do this: +/// +/// ```should_panic +/// let v = vec![0, 2, 4, 6]; +/// println!("{}", v[6]); // it will panic! +/// ``` +/// +/// Use [`get`] and [`get_mut`] if you want to check whether the index is in +/// the `Vec`. +/// +/// # Slicing +/// +/// A `Vec` can be mutable. On the other hand, slices are read-only objects. +/// To get a [slice][prim@slice], use [`&`]. Example: +/// +/// ``` +/// fn read_slice(slice: &[usize]) { +/// // ... +/// } +/// +/// let v = vec![0, 1]; +/// read_slice(&v); +/// +/// // ... and that's all! +/// // you can also do it like this: +/// let u: &[usize] = &v; +/// // or like this: +/// let u: &[_] = &v; +/// ``` +/// +/// In Rust, it's more common to pass slices as arguments rather than vectors +/// when you just want to provide read access. The same goes for [`String`] and +/// [`&str`]. +/// +/// # Capacity and reallocation +/// +/// The capacity of a vector is the amount of space allocated for any future +/// elements that will be added onto the vector. This is not to be confused with +/// the *length* of a vector, which specifies the number of actual elements +/// within the vector. If a vector's length exceeds its capacity, its capacity +/// will automatically be increased, but its elements will have to be +/// reallocated. +/// +/// For example, a vector with capacity 10 and length 0 would be an empty vector +/// with space for 10 more elements. Pushing 10 or fewer elements onto the +/// vector will not change its capacity or cause reallocation to occur. However, +/// if the vector's length is increased to 11, it will have to reallocate, which +/// can be slow. For this reason, it is recommended to use [`Vec::with_capacity`] +/// whenever possible to specify how big the vector is expected to get. +/// +/// # Guarantees +/// +/// Due to its incredibly fundamental nature, `Vec` makes a lot of guarantees +/// about its design. This ensures that it's as low-overhead as possible in +/// the general case, and can be correctly manipulated in primitive ways +/// by unsafe code. Note that these guarantees refer to an unqualified `Vec<T>`. +/// If additional type parameters are added (e.g., to support custom allocators), +/// overriding their defaults may change the behavior. +/// +/// Most fundamentally, `Vec` is and always will be a (pointer, capacity, length) +/// triplet. No more, no less. The order of these fields is completely +/// unspecified, and you should use the appropriate methods to modify these. +/// The pointer will never be null, so this type is null-pointer-optimized. +/// +/// However, the pointer might not actually point to allocated memory. In particular, +/// if you construct a `Vec` with capacity 0 via [`Vec::new`], [`vec![]`][`vec!`], +/// [`Vec::with_capacity(0)`][`Vec::with_capacity`], or by calling [`shrink_to_fit`] +/// on an empty Vec, it will not allocate memory. Similarly, if you store zero-sized +/// types inside a `Vec`, it will not allocate space for them. *Note that in this case +/// the `Vec` might not report a [`capacity`] of 0*. `Vec` will allocate if and only +/// if <code>[mem::size_of::\<T>]\() * [capacity]\() > 0</code>. In general, `Vec`'s allocation +/// details are very subtle --- if you intend to allocate memory using a `Vec` +/// and use it for something else (either to pass to unsafe code, or to build your +/// own memory-backed collection), be sure to deallocate this memory by using +/// `from_raw_parts` to recover the `Vec` and then dropping it. +/// +/// If a `Vec` *has* allocated memory, then the memory it points to is on the heap +/// (as defined by the allocator Rust is configured to use by default), and its +/// pointer points to [`len`] initialized, contiguous elements in order (what +/// you would see if you coerced it to a slice), followed by <code>[capacity] - [len]</code> +/// logically uninitialized, contiguous elements. +/// +/// A vector containing the elements `'a'` and `'b'` with capacity 4 can be +/// visualized as below. The top part is the `Vec` struct, it contains a +/// pointer to the head of the allocation in the heap, length and capacity. +/// The bottom part is the allocation on the heap, a contiguous memory block. +/// +/// ```text +/// ptr len capacity +/// +--------+--------+--------+ +/// | 0x0123 | 2 | 4 | +/// +--------+--------+--------+ +/// | +/// v +/// Heap +--------+--------+--------+--------+ +/// | 'a' | 'b' | uninit | uninit | +/// +--------+--------+--------+--------+ +/// ``` +/// +/// - **uninit** represents memory that is not initialized, see [`MaybeUninit`]. +/// - Note: the ABI is not stable and `Vec` makes no guarantees about its memory +/// layout (including the order of fields). +/// +/// `Vec` will never perform a "small optimization" where elements are actually +/// stored on the stack for two reasons: +/// +/// * It would make it more difficult for unsafe code to correctly manipulate +/// a `Vec`. The contents of a `Vec` wouldn't have a stable address if it were +/// only moved, and it would be more difficult to determine if a `Vec` had +/// actually allocated memory. +/// +/// * It would penalize the general case, incurring an additional branch +/// on every access. +/// +/// `Vec` will never automatically shrink itself, even if completely empty. This +/// ensures no unnecessary allocations or deallocations occur. Emptying a `Vec` +/// and then filling it back up to the same [`len`] should incur no calls to +/// the allocator. If you wish to free up unused memory, use +/// [`shrink_to_fit`] or [`shrink_to`]. +/// +/// [`push`] and [`insert`] will never (re)allocate if the reported capacity is +/// sufficient. [`push`] and [`insert`] *will* (re)allocate if +/// <code>[len] == [capacity]</code>. That is, the reported capacity is completely +/// accurate, and can be relied on. It can even be used to manually free the memory +/// allocated by a `Vec` if desired. Bulk insertion methods *may* reallocate, even +/// when not necessary. +/// +/// `Vec` does not guarantee any particular growth strategy when reallocating +/// when full, nor when [`reserve`] is called. The current strategy is basic +/// and it may prove desirable to use a non-constant growth factor. Whatever +/// strategy is used will of course guarantee *O*(1) amortized [`push`]. +/// +/// `vec![x; n]`, `vec![a, b, c, d]`, and +/// [`Vec::with_capacity(n)`][`Vec::with_capacity`], will all produce a `Vec` +/// with exactly the requested capacity. If <code>[len] == [capacity]</code>, +/// (as is the case for the [`vec!`] macro), then a `Vec<T>` can be converted to +/// and from a [`Box<[T]>`][owned slice] without reallocating or moving the elements. +/// +/// `Vec` will not specifically overwrite any data that is removed from it, +/// but also won't specifically preserve it. Its uninitialized memory is +/// scratch space that it may use however it wants. It will generally just do +/// whatever is most efficient or otherwise easy to implement. Do not rely on +/// removed data to be erased for security purposes. Even if you drop a `Vec`, its +/// buffer may simply be reused by another allocation. Even if you zero a `Vec`'s memory +/// first, that might not actually happen because the optimizer does not consider +/// this a side-effect that must be preserved. There is one case which we will +/// not break, however: using `unsafe` code to write to the excess capacity, +/// and then increasing the length to match, is always valid. +/// +/// Currently, `Vec` does not guarantee the order in which elements are dropped. +/// The order has changed in the past and may change again. +/// +/// [`get`]: ../../std/vec/struct.Vec.html#method.get +/// [`get_mut`]: ../../std/vec/struct.Vec.html#method.get_mut +/// [`String`]: crate::string::String +/// [`&str`]: type@str +/// [`shrink_to_fit`]: Vec::shrink_to_fit +/// [`shrink_to`]: Vec::shrink_to +/// [capacity]: Vec::capacity +/// [`capacity`]: Vec::capacity +/// [mem::size_of::\<T>]: core::mem::size_of +/// [len]: Vec::len +/// [`len`]: Vec::len +/// [`push`]: Vec::push +/// [`insert`]: Vec::insert +/// [`reserve`]: Vec::reserve +/// [`MaybeUninit`]: core::mem::MaybeUninit +/// [owned slice]: Box +#[stable(feature = "rust1", since = "1.0.0")] +#[cfg_attr(not(test), rustc_diagnostic_item = "Vec")] +#[rustc_insignificant_dtor] +pub struct Vec<T, #[unstable(feature = "allocator_api", issue = "32838")] A: Allocator = Global> { + buf: RawVec<T, A>, + len: usize, +} + +//////////////////////////////////////////////////////////////////////////////// +// Inherent methods +//////////////////////////////////////////////////////////////////////////////// + +impl<T> Vec<T> { + /// Constructs a new, empty `Vec<T>`. + /// + /// The vector will not allocate until elements are pushed onto it. + /// + /// # Examples + /// + /// ``` + /// # #![allow(unused_mut)] + /// let mut vec: Vec<i32> = Vec::new(); + /// ``` + #[inline] + #[rustc_const_stable(feature = "const_vec_new", since = "1.39.0")] + #[stable(feature = "rust1", since = "1.0.0")] + #[must_use] + pub const fn new() -> Self { + Vec { buf: RawVec::NEW, len: 0 } + } + + /// Constructs a new, empty `Vec<T>` with the specified capacity. + /// + /// The vector will be able to hold exactly `capacity` elements without + /// reallocating. If `capacity` is 0, the vector will not allocate. + /// + /// It is important to note that although the returned vector has the + /// *capacity* specified, the vector will have a zero *length*. For an + /// explanation of the difference between length and capacity, see + /// *[Capacity and reallocation]*. + /// + /// [Capacity and reallocation]: #capacity-and-reallocation + /// + /// # Panics + /// + /// Panics if the new capacity exceeds `isize::MAX` bytes. + /// + /// # Examples + /// + /// ``` + /// let mut vec = Vec::with_capacity(10); + /// + /// // The vector contains no items, even though it has capacity for more + /// assert_eq!(vec.len(), 0); + /// assert_eq!(vec.capacity(), 10); + /// + /// // These are all done without reallocating... + /// for i in 0..10 { + /// vec.push(i); + /// } + /// assert_eq!(vec.len(), 10); + /// assert_eq!(vec.capacity(), 10); + /// + /// // ...but this may make the vector reallocate + /// vec.push(11); + /// assert_eq!(vec.len(), 11); + /// assert!(vec.capacity() >= 11); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[inline] + #[stable(feature = "rust1", since = "1.0.0")] + #[must_use] + pub fn with_capacity(capacity: usize) -> Self { + Self::with_capacity_in(capacity, Global) + } + + /// Creates a `Vec<T>` directly from the raw components of another vector. + /// + /// # Safety + /// + /// This is highly unsafe, due to the number of invariants that aren't + /// checked: + /// + /// * `ptr` needs to have been previously allocated via [`String`]/`Vec<T>` + /// (at least, it's highly likely to be incorrect if it wasn't). + /// * `T` needs to have the same alignment as what `ptr` was allocated with. + /// (`T` having a less strict alignment is not sufficient, the alignment really + /// needs to be equal to satisfy the [`dealloc`] requirement that memory must be + /// allocated and deallocated with the same layout.) + /// * The size of `T` times the `capacity` (ie. the allocated size in bytes) needs + /// to be the same size as the pointer was allocated with. (Because similar to + /// alignment, [`dealloc`] must be called with the same layout `size`.) + /// * `length` needs to be less than or equal to `capacity`. + /// + /// Violating these may cause problems like corrupting the allocator's + /// internal data structures. For example it is normally **not** safe + /// to build a `Vec<u8>` from a pointer to a C `char` array with length + /// `size_t`, doing so is only safe if the array was initially allocated by + /// a `Vec` or `String`. + /// It's also not safe to build one from a `Vec<u16>` and its length, because + /// the allocator cares about the alignment, and these two types have different + /// alignments. The buffer was allocated with alignment 2 (for `u16`), but after + /// turning it into a `Vec<u8>` it'll be deallocated with alignment 1. To avoid + /// these issues, it is often preferable to do casting/transmuting using + /// [`slice::from_raw_parts`] instead. + /// + /// The ownership of `ptr` is effectively transferred to the + /// `Vec<T>` which may then deallocate, reallocate or change the + /// contents of memory pointed to by the pointer at will. Ensure + /// that nothing else uses the pointer after calling this + /// function. + /// + /// [`String`]: crate::string::String + /// [`dealloc`]: crate::alloc::GlobalAlloc::dealloc + /// + /// # Examples + /// + /// ``` + /// use std::ptr; + /// use std::mem; + /// + /// let v = vec![1, 2, 3]; + /// + // FIXME Update this when vec_into_raw_parts is stabilized + /// // Prevent running `v`'s destructor so we are in complete control + /// // of the allocation. + /// let mut v = mem::ManuallyDrop::new(v); + /// + /// // Pull out the various important pieces of information about `v` + /// let p = v.as_mut_ptr(); + /// let len = v.len(); + /// let cap = v.capacity(); + /// + /// unsafe { + /// // Overwrite memory with 4, 5, 6 + /// for i in 0..len as isize { + /// ptr::write(p.offset(i), 4 + i); + /// } + /// + /// // Put everything back together into a Vec + /// let rebuilt = Vec::from_raw_parts(p, len, cap); + /// assert_eq!(rebuilt, [4, 5, 6]); + /// } + /// ``` + #[inline] + #[stable(feature = "rust1", since = "1.0.0")] + pub unsafe fn from_raw_parts(ptr: *mut T, length: usize, capacity: usize) -> Self { + unsafe { Self::from_raw_parts_in(ptr, length, capacity, Global) } + } +} + +impl<T, A: Allocator> Vec<T, A> { + /// Constructs a new, empty `Vec<T, A>`. + /// + /// The vector will not allocate until elements are pushed onto it. + /// + /// # Examples + /// + /// ``` + /// #![feature(allocator_api)] + /// + /// use std::alloc::System; + /// + /// # #[allow(unused_mut)] + /// let mut vec: Vec<i32, _> = Vec::new_in(System); + /// ``` + #[inline] + #[unstable(feature = "allocator_api", issue = "32838")] + pub const fn new_in(alloc: A) -> Self { + Vec { buf: RawVec::new_in(alloc), len: 0 } + } + + /// Constructs a new, empty `Vec<T, A>` with the specified capacity with the provided + /// allocator. + /// + /// The vector will be able to hold exactly `capacity` elements without + /// reallocating. If `capacity` is 0, the vector will not allocate. + /// + /// It is important to note that although the returned vector has the + /// *capacity* specified, the vector will have a zero *length*. For an + /// explanation of the difference between length and capacity, see + /// *[Capacity and reallocation]*. + /// + /// [Capacity and reallocation]: #capacity-and-reallocation + /// + /// # Panics + /// + /// Panics if the new capacity exceeds `isize::MAX` bytes. + /// + /// # Examples + /// + /// ``` + /// #![feature(allocator_api)] + /// + /// use std::alloc::System; + /// + /// let mut vec = Vec::with_capacity_in(10, System); + /// + /// // The vector contains no items, even though it has capacity for more + /// assert_eq!(vec.len(), 0); + /// assert_eq!(vec.capacity(), 10); + /// + /// // These are all done without reallocating... + /// for i in 0..10 { + /// vec.push(i); + /// } + /// assert_eq!(vec.len(), 10); + /// assert_eq!(vec.capacity(), 10); + /// + /// // ...but this may make the vector reallocate + /// vec.push(11); + /// assert_eq!(vec.len(), 11); + /// assert!(vec.capacity() >= 11); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[inline] + #[unstable(feature = "allocator_api", issue = "32838")] + pub fn with_capacity_in(capacity: usize, alloc: A) -> Self { + Vec { buf: RawVec::with_capacity_in(capacity, alloc), len: 0 } + } + + /// Creates a `Vec<T, A>` directly from the raw components of another vector. + /// + /// # Safety + /// + /// This is highly unsafe, due to the number of invariants that aren't + /// checked: + /// + /// * `ptr` needs to have been previously allocated via [`String`]/`Vec<T>` + /// (at least, it's highly likely to be incorrect if it wasn't). + /// * `T` needs to have the same size and alignment as what `ptr` was allocated with. + /// (`T` having a less strict alignment is not sufficient, the alignment really + /// needs to be equal to satisfy the [`dealloc`] requirement that memory must be + /// allocated and deallocated with the same layout.) + /// * `length` needs to be less than or equal to `capacity`. + /// * `capacity` needs to be the capacity that the pointer was allocated with. + /// + /// Violating these may cause problems like corrupting the allocator's + /// internal data structures. For example it is **not** safe + /// to build a `Vec<u8>` from a pointer to a C `char` array with length `size_t`. + /// It's also not safe to build one from a `Vec<u16>` and its length, because + /// the allocator cares about the alignment, and these two types have different + /// alignments. The buffer was allocated with alignment 2 (for `u16`), but after + /// turning it into a `Vec<u8>` it'll be deallocated with alignment 1. + /// + /// The ownership of `ptr` is effectively transferred to the + /// `Vec<T>` which may then deallocate, reallocate or change the + /// contents of memory pointed to by the pointer at will. Ensure + /// that nothing else uses the pointer after calling this + /// function. + /// + /// [`String`]: crate::string::String + /// [`dealloc`]: crate::alloc::GlobalAlloc::dealloc + /// + /// # Examples + /// + /// ``` + /// #![feature(allocator_api)] + /// + /// use std::alloc::System; + /// + /// use std::ptr; + /// use std::mem; + /// + /// let mut v = Vec::with_capacity_in(3, System); + /// v.push(1); + /// v.push(2); + /// v.push(3); + /// + // FIXME Update this when vec_into_raw_parts is stabilized + /// // Prevent running `v`'s destructor so we are in complete control + /// // of the allocation. + /// let mut v = mem::ManuallyDrop::new(v); + /// + /// // Pull out the various important pieces of information about `v` + /// let p = v.as_mut_ptr(); + /// let len = v.len(); + /// let cap = v.capacity(); + /// let alloc = v.allocator(); + /// + /// unsafe { + /// // Overwrite memory with 4, 5, 6 + /// for i in 0..len as isize { + /// ptr::write(p.offset(i), 4 + i); + /// } + /// + /// // Put everything back together into a Vec + /// let rebuilt = Vec::from_raw_parts_in(p, len, cap, alloc.clone()); + /// assert_eq!(rebuilt, [4, 5, 6]); + /// } + /// ``` + #[inline] + #[unstable(feature = "allocator_api", issue = "32838")] + pub unsafe fn from_raw_parts_in(ptr: *mut T, length: usize, capacity: usize, alloc: A) -> Self { + unsafe { Vec { buf: RawVec::from_raw_parts_in(ptr, capacity, alloc), len: length } } + } + + /// Decomposes a `Vec<T>` into its raw components. + /// + /// Returns the raw pointer to the underlying data, the length of + /// the vector (in elements), and the allocated capacity of the + /// data (in elements). These are the same arguments in the same + /// order as the arguments to [`from_raw_parts`]. + /// + /// After calling this function, the caller is responsible for the + /// memory previously managed by the `Vec`. The only way to do + /// this is to convert the raw pointer, length, and capacity back + /// into a `Vec` with the [`from_raw_parts`] function, allowing + /// the destructor to perform the cleanup. + /// + /// [`from_raw_parts`]: Vec::from_raw_parts + /// + /// # Examples + /// + /// ``` + /// #![feature(vec_into_raw_parts)] + /// let v: Vec<i32> = vec![-1, 0, 1]; + /// + /// let (ptr, len, cap) = v.into_raw_parts(); + /// + /// let rebuilt = unsafe { + /// // We can now make changes to the components, such as + /// // transmuting the raw pointer to a compatible type. + /// let ptr = ptr as *mut u32; + /// + /// Vec::from_raw_parts(ptr, len, cap) + /// }; + /// assert_eq!(rebuilt, [4294967295, 0, 1]); + /// ``` + #[unstable(feature = "vec_into_raw_parts", reason = "new API", issue = "65816")] + pub fn into_raw_parts(self) -> (*mut T, usize, usize) { + let mut me = ManuallyDrop::new(self); + (me.as_mut_ptr(), me.len(), me.capacity()) + } + + /// Decomposes a `Vec<T>` into its raw components. + /// + /// Returns the raw pointer to the underlying data, the length of the vector (in elements), + /// the allocated capacity of the data (in elements), and the allocator. These are the same + /// arguments in the same order as the arguments to [`from_raw_parts_in`]. + /// + /// After calling this function, the caller is responsible for the + /// memory previously managed by the `Vec`. The only way to do + /// this is to convert the raw pointer, length, and capacity back + /// into a `Vec` with the [`from_raw_parts_in`] function, allowing + /// the destructor to perform the cleanup. + /// + /// [`from_raw_parts_in`]: Vec::from_raw_parts_in + /// + /// # Examples + /// + /// ``` + /// #![feature(allocator_api, vec_into_raw_parts)] + /// + /// use std::alloc::System; + /// + /// let mut v: Vec<i32, System> = Vec::new_in(System); + /// v.push(-1); + /// v.push(0); + /// v.push(1); + /// + /// let (ptr, len, cap, alloc) = v.into_raw_parts_with_alloc(); + /// + /// let rebuilt = unsafe { + /// // We can now make changes to the components, such as + /// // transmuting the raw pointer to a compatible type. + /// let ptr = ptr as *mut u32; + /// + /// Vec::from_raw_parts_in(ptr, len, cap, alloc) + /// }; + /// assert_eq!(rebuilt, [4294967295, 0, 1]); + /// ``` + #[unstable(feature = "allocator_api", issue = "32838")] + // #[unstable(feature = "vec_into_raw_parts", reason = "new API", issue = "65816")] + pub fn into_raw_parts_with_alloc(self) -> (*mut T, usize, usize, A) { + let mut me = ManuallyDrop::new(self); + let len = me.len(); + let capacity = me.capacity(); + let ptr = me.as_mut_ptr(); + let alloc = unsafe { ptr::read(me.allocator()) }; + (ptr, len, capacity, alloc) + } + + /// Returns the number of elements the vector can hold without + /// reallocating. + /// + /// # Examples + /// + /// ``` + /// let vec: Vec<i32> = Vec::with_capacity(10); + /// assert_eq!(vec.capacity(), 10); + /// ``` + #[inline] + #[stable(feature = "rust1", since = "1.0.0")] + pub fn capacity(&self) -> usize { + self.buf.capacity() + } + + /// Reserves capacity for at least `additional` more elements to be inserted + /// in the given `Vec<T>`. The collection may reserve more space to avoid + /// frequent reallocations. After calling `reserve`, capacity will be + /// greater than or equal to `self.len() + additional`. Does nothing if + /// capacity is already sufficient. + /// + /// # Panics + /// + /// Panics if the new capacity exceeds `isize::MAX` bytes. + /// + /// # Examples + /// + /// ``` + /// let mut vec = vec![1]; + /// vec.reserve(10); + /// assert!(vec.capacity() >= 11); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[stable(feature = "rust1", since = "1.0.0")] + pub fn reserve(&mut self, additional: usize) { + self.buf.reserve(self.len, additional); + } + + /// Reserves the minimum capacity for exactly `additional` more elements to + /// be inserted in the given `Vec<T>`. After calling `reserve_exact`, + /// capacity will be greater than or equal to `self.len() + additional`. + /// Does nothing if the capacity is already sufficient. + /// + /// Note that the allocator may give the collection more space than it + /// requests. Therefore, capacity can not be relied upon to be precisely + /// minimal. Prefer [`reserve`] if future insertions are expected. + /// + /// [`reserve`]: Vec::reserve + /// + /// # Panics + /// + /// Panics if the new capacity exceeds `isize::MAX` bytes. + /// + /// # Examples + /// + /// ``` + /// let mut vec = vec![1]; + /// vec.reserve_exact(10); + /// assert!(vec.capacity() >= 11); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[stable(feature = "rust1", since = "1.0.0")] + pub fn reserve_exact(&mut self, additional: usize) { + self.buf.reserve_exact(self.len, additional); + } + + /// Tries to reserve capacity for at least `additional` more elements to be inserted + /// in the given `Vec<T>`. The collection may reserve more space to avoid + /// frequent reallocations. After calling `try_reserve`, capacity will be + /// greater than or equal to `self.len() + additional`. Does nothing if + /// capacity is already sufficient. + /// + /// # Errors + /// + /// If the capacity overflows, or the allocator reports a failure, then an error + /// is returned. + /// + /// # Examples + /// + /// ``` + /// use std::collections::TryReserveError; + /// + /// fn process_data(data: &[u32]) -> Result<Vec<u32>, TryReserveError> { + /// let mut output = Vec::new(); + /// + /// // Pre-reserve the memory, exiting if we can't + /// output.try_reserve(data.len())?; + /// + /// // Now we know this can't OOM in the middle of our complex work + /// output.extend(data.iter().map(|&val| { + /// val * 2 + 5 // very complicated + /// })); + /// + /// Ok(output) + /// } + /// # process_data(&[1, 2, 3]).expect("why is the test harness OOMing on 12 bytes?"); + /// ``` + #[stable(feature = "try_reserve", since = "1.57.0")] + pub fn try_reserve(&mut self, additional: usize) -> Result<(), TryReserveError> { + self.buf.try_reserve(self.len, additional) + } + + /// Tries to reserve the minimum capacity for exactly `additional` + /// elements to be inserted in the given `Vec<T>`. After calling + /// `try_reserve_exact`, capacity will be greater than or equal to + /// `self.len() + additional` if it returns `Ok(())`. + /// Does nothing if the capacity is already sufficient. + /// + /// Note that the allocator may give the collection more space than it + /// requests. Therefore, capacity can not be relied upon to be precisely + /// minimal. Prefer [`try_reserve`] if future insertions are expected. + /// + /// [`try_reserve`]: Vec::try_reserve + /// + /// # Errors + /// + /// If the capacity overflows, or the allocator reports a failure, then an error + /// is returned. + /// + /// # Examples + /// + /// ``` + /// use std::collections::TryReserveError; + /// + /// fn process_data(data: &[u32]) -> Result<Vec<u32>, TryReserveError> { + /// let mut output = Vec::new(); + /// + /// // Pre-reserve the memory, exiting if we can't + /// output.try_reserve_exact(data.len())?; + /// + /// // Now we know this can't OOM in the middle of our complex work + /// output.extend(data.iter().map(|&val| { + /// val * 2 + 5 // very complicated + /// })); + /// + /// Ok(output) + /// } + /// # process_data(&[1, 2, 3]).expect("why is the test harness OOMing on 12 bytes?"); + /// ``` + #[stable(feature = "try_reserve", since = "1.57.0")] + pub fn try_reserve_exact(&mut self, additional: usize) -> Result<(), TryReserveError> { + self.buf.try_reserve_exact(self.len, additional) + } + + /// Shrinks the capacity of the vector as much as possible. + /// + /// It will drop down as close as possible to the length but the allocator + /// may still inform the vector that there is space for a few more elements. + /// + /// # Examples + /// + /// ``` + /// let mut vec = Vec::with_capacity(10); + /// vec.extend([1, 2, 3]); + /// assert_eq!(vec.capacity(), 10); + /// vec.shrink_to_fit(); + /// assert!(vec.capacity() >= 3); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[stable(feature = "rust1", since = "1.0.0")] + pub fn shrink_to_fit(&mut self) { + // The capacity is never less than the length, and there's nothing to do when + // they are equal, so we can avoid the panic case in `RawVec::shrink_to_fit` + // by only calling it with a greater capacity. + if self.capacity() > self.len { + self.buf.shrink_to_fit(self.len); + } + } + + /// Shrinks the capacity of the vector with a lower bound. + /// + /// The capacity will remain at least as large as both the length + /// and the supplied value. + /// + /// If the current capacity is less than the lower limit, this is a no-op. + /// + /// # Examples + /// + /// ``` + /// let mut vec = Vec::with_capacity(10); + /// vec.extend([1, 2, 3]); + /// assert_eq!(vec.capacity(), 10); + /// vec.shrink_to(4); + /// assert!(vec.capacity() >= 4); + /// vec.shrink_to(0); + /// assert!(vec.capacity() >= 3); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[stable(feature = "shrink_to", since = "1.56.0")] + pub fn shrink_to(&mut self, min_capacity: usize) { + if self.capacity() > min_capacity { + self.buf.shrink_to_fit(cmp::max(self.len, min_capacity)); + } + } + + /// Converts the vector into [`Box<[T]>`][owned slice]. + /// + /// Note that this will drop any excess capacity. + /// + /// [owned slice]: Box + /// + /// # Examples + /// + /// ``` + /// let v = vec![1, 2, 3]; + /// + /// let slice = v.into_boxed_slice(); + /// ``` + /// + /// Any excess capacity is removed: + /// + /// ``` + /// let mut vec = Vec::with_capacity(10); + /// vec.extend([1, 2, 3]); + /// + /// assert_eq!(vec.capacity(), 10); + /// let slice = vec.into_boxed_slice(); + /// assert_eq!(slice.into_vec().capacity(), 3); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[stable(feature = "rust1", since = "1.0.0")] + pub fn into_boxed_slice(mut self) -> Box<[T], A> { + unsafe { + self.shrink_to_fit(); + let me = ManuallyDrop::new(self); + let buf = ptr::read(&me.buf); + let len = me.len(); + buf.into_box(len).assume_init() + } + } + + /// Shortens the vector, keeping the first `len` elements and dropping + /// the rest. + /// + /// If `len` is greater than the vector's current length, this has no + /// effect. + /// + /// The [`drain`] method can emulate `truncate`, but causes the excess + /// elements to be returned instead of dropped. + /// + /// Note that this method has no effect on the allocated capacity + /// of the vector. + /// + /// # Examples + /// + /// Truncating a five element vector to two elements: + /// + /// ``` + /// let mut vec = vec![1, 2, 3, 4, 5]; + /// vec.truncate(2); + /// assert_eq!(vec, [1, 2]); + /// ``` + /// + /// No truncation occurs when `len` is greater than the vector's current + /// length: + /// + /// ``` + /// let mut vec = vec![1, 2, 3]; + /// vec.truncate(8); + /// assert_eq!(vec, [1, 2, 3]); + /// ``` + /// + /// Truncating when `len == 0` is equivalent to calling the [`clear`] + /// method. + /// + /// ``` + /// let mut vec = vec![1, 2, 3]; + /// vec.truncate(0); + /// assert_eq!(vec, []); + /// ``` + /// + /// [`clear`]: Vec::clear + /// [`drain`]: Vec::drain + #[stable(feature = "rust1", since = "1.0.0")] + pub fn truncate(&mut self, len: usize) { + // This is safe because: + // + // * the slice passed to `drop_in_place` is valid; the `len > self.len` + // case avoids creating an invalid slice, and + // * the `len` of the vector is shrunk before calling `drop_in_place`, + // such that no value will be dropped twice in case `drop_in_place` + // were to panic once (if it panics twice, the program aborts). + unsafe { + // Note: It's intentional that this is `>` and not `>=`. + // Changing it to `>=` has negative performance + // implications in some cases. See #78884 for more. + if len > self.len { + return; + } + let remaining_len = self.len - len; + let s = ptr::slice_from_raw_parts_mut(self.as_mut_ptr().add(len), remaining_len); + self.len = len; + ptr::drop_in_place(s); + } + } + + /// Extracts a slice containing the entire vector. + /// + /// Equivalent to `&s[..]`. + /// + /// # Examples + /// + /// ``` + /// use std::io::{self, Write}; + /// let buffer = vec![1, 2, 3, 5, 8]; + /// io::sink().write(buffer.as_slice()).unwrap(); + /// ``` + #[inline] + #[stable(feature = "vec_as_slice", since = "1.7.0")] + pub fn as_slice(&self) -> &[T] { + self + } + + /// Extracts a mutable slice of the entire vector. + /// + /// Equivalent to `&mut s[..]`. + /// + /// # Examples + /// + /// ``` + /// use std::io::{self, Read}; + /// let mut buffer = vec![0; 3]; + /// io::repeat(0b101).read_exact(buffer.as_mut_slice()).unwrap(); + /// ``` + #[inline] + #[stable(feature = "vec_as_slice", since = "1.7.0")] + pub fn as_mut_slice(&mut self) -> &mut [T] { + self + } + + /// Returns a raw pointer to the vector's buffer. + /// + /// The caller must ensure that the vector outlives the pointer this + /// function returns, or else it will end up pointing to garbage. + /// Modifying the vector may cause its buffer to be reallocated, + /// which would also make any pointers to it invalid. + /// + /// The caller must also ensure that the memory the pointer (non-transitively) points to + /// is never written to (except inside an `UnsafeCell`) using this pointer or any pointer + /// derived from it. If you need to mutate the contents of the slice, use [`as_mut_ptr`]. + /// + /// # Examples + /// + /// ``` + /// let x = vec![1, 2, 4]; + /// let x_ptr = x.as_ptr(); + /// + /// unsafe { + /// for i in 0..x.len() { + /// assert_eq!(*x_ptr.add(i), 1 << i); + /// } + /// } + /// ``` + /// + /// [`as_mut_ptr`]: Vec::as_mut_ptr + #[stable(feature = "vec_as_ptr", since = "1.37.0")] + #[inline] + pub fn as_ptr(&self) -> *const T { + // We shadow the slice method of the same name to avoid going through + // `deref`, which creates an intermediate reference. + let ptr = self.buf.ptr(); + unsafe { + assume(!ptr.is_null()); + } + ptr + } + + /// Returns an unsafe mutable pointer to the vector's buffer. + /// + /// The caller must ensure that the vector outlives the pointer this + /// function returns, or else it will end up pointing to garbage. + /// Modifying the vector may cause its buffer to be reallocated, + /// which would also make any pointers to it invalid. + /// + /// # Examples + /// + /// ``` + /// // Allocate vector big enough for 4 elements. + /// let size = 4; + /// let mut x: Vec<i32> = Vec::with_capacity(size); + /// let x_ptr = x.as_mut_ptr(); + /// + /// // Initialize elements via raw pointer writes, then set length. + /// unsafe { + /// for i in 0..size { + /// *x_ptr.add(i) = i as i32; + /// } + /// x.set_len(size); + /// } + /// assert_eq!(&*x, &[0, 1, 2, 3]); + /// ``` + #[stable(feature = "vec_as_ptr", since = "1.37.0")] + #[inline] + pub fn as_mut_ptr(&mut self) -> *mut T { + // We shadow the slice method of the same name to avoid going through + // `deref_mut`, which creates an intermediate reference. + let ptr = self.buf.ptr(); + unsafe { + assume(!ptr.is_null()); + } + ptr + } + + /// Returns a reference to the underlying allocator. + #[unstable(feature = "allocator_api", issue = "32838")] + #[inline] + pub fn allocator(&self) -> &A { + self.buf.allocator() + } + + /// Forces the length of the vector to `new_len`. + /// + /// This is a low-level operation that maintains none of the normal + /// invariants of the type. Normally changing the length of a vector + /// is done using one of the safe operations instead, such as + /// [`truncate`], [`resize`], [`extend`], or [`clear`]. + /// + /// [`truncate`]: Vec::truncate + /// [`resize`]: Vec::resize + /// [`extend`]: Extend::extend + /// [`clear`]: Vec::clear + /// + /// # Safety + /// + /// - `new_len` must be less than or equal to [`capacity()`]. + /// - The elements at `old_len..new_len` must be initialized. + /// + /// [`capacity()`]: Vec::capacity + /// + /// # Examples + /// + /// This method can be useful for situations in which the vector + /// is serving as a buffer for other code, particularly over FFI: + /// + /// ```no_run + /// # #![allow(dead_code)] + /// # // This is just a minimal skeleton for the doc example; + /// # // don't use this as a starting point for a real library. + /// # pub struct StreamWrapper { strm: *mut std::ffi::c_void } + /// # const Z_OK: i32 = 0; + /// # extern "C" { + /// # fn deflateGetDictionary( + /// # strm: *mut std::ffi::c_void, + /// # dictionary: *mut u8, + /// # dictLength: *mut usize, + /// # ) -> i32; + /// # } + /// # impl StreamWrapper { + /// pub fn get_dictionary(&self) -> Option<Vec<u8>> { + /// // Per the FFI method's docs, "32768 bytes is always enough". + /// let mut dict = Vec::with_capacity(32_768); + /// let mut dict_length = 0; + /// // SAFETY: When `deflateGetDictionary` returns `Z_OK`, it holds that: + /// // 1. `dict_length` elements were initialized. + /// // 2. `dict_length` <= the capacity (32_768) + /// // which makes `set_len` safe to call. + /// unsafe { + /// // Make the FFI call... + /// let r = deflateGetDictionary(self.strm, dict.as_mut_ptr(), &mut dict_length); + /// if r == Z_OK { + /// // ...and update the length to what was initialized. + /// dict.set_len(dict_length); + /// Some(dict) + /// } else { + /// None + /// } + /// } + /// } + /// # } + /// ``` + /// + /// While the following example is sound, there is a memory leak since + /// the inner vectors were not freed prior to the `set_len` call: + /// + /// ``` + /// let mut vec = vec![vec![1, 0, 0], + /// vec![0, 1, 0], + /// vec![0, 0, 1]]; + /// // SAFETY: + /// // 1. `old_len..0` is empty so no elements need to be initialized. + /// // 2. `0 <= capacity` always holds whatever `capacity` is. + /// unsafe { + /// vec.set_len(0); + /// } + /// ``` + /// + /// Normally, here, one would use [`clear`] instead to correctly drop + /// the contents and thus not leak memory. + #[inline] + #[stable(feature = "rust1", since = "1.0.0")] + pub unsafe fn set_len(&mut self, new_len: usize) { + debug_assert!(new_len <= self.capacity()); + + self.len = new_len; + } + + /// Removes an element from the vector and returns it. + /// + /// The removed element is replaced by the last element of the vector. + /// + /// This does not preserve ordering, but is *O*(1). + /// If you need to preserve the element order, use [`remove`] instead. + /// + /// [`remove`]: Vec::remove + /// + /// # Panics + /// + /// Panics if `index` is out of bounds. + /// + /// # Examples + /// + /// ``` + /// let mut v = vec!["foo", "bar", "baz", "qux"]; + /// + /// assert_eq!(v.swap_remove(1), "bar"); + /// assert_eq!(v, ["foo", "qux", "baz"]); + /// + /// assert_eq!(v.swap_remove(0), "foo"); + /// assert_eq!(v, ["baz", "qux"]); + /// ``` + #[inline] + #[stable(feature = "rust1", since = "1.0.0")] + pub fn swap_remove(&mut self, index: usize) -> T { + #[cold] + #[inline(never)] + fn assert_failed(index: usize, len: usize) -> ! { + panic!("swap_remove index (is {index}) should be < len (is {len})"); + } + + let len = self.len(); + if index >= len { + assert_failed(index, len); + } + unsafe { + // We replace self[index] with the last element. Note that if the + // bounds check above succeeds there must be a last element (which + // can be self[index] itself). + let value = ptr::read(self.as_ptr().add(index)); + let base_ptr = self.as_mut_ptr(); + ptr::copy(base_ptr.add(len - 1), base_ptr.add(index), 1); + self.set_len(len - 1); + value + } + } + + /// Inserts an element at position `index` within the vector, shifting all + /// elements after it to the right. + /// + /// # Panics + /// + /// Panics if `index > len`. + /// + /// # Examples + /// + /// ``` + /// let mut vec = vec![1, 2, 3]; + /// vec.insert(1, 4); + /// assert_eq!(vec, [1, 4, 2, 3]); + /// vec.insert(4, 5); + /// assert_eq!(vec, [1, 4, 2, 3, 5]); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[stable(feature = "rust1", since = "1.0.0")] + pub fn insert(&mut self, index: usize, element: T) { + #[cold] + #[inline(never)] + fn assert_failed(index: usize, len: usize) -> ! { + panic!("insertion index (is {index}) should be <= len (is {len})"); + } + + let len = self.len(); + if index > len { + assert_failed(index, len); + } + + // space for the new element + if len == self.buf.capacity() { + self.reserve(1); + } + + unsafe { + // infallible + // The spot to put the new value + { + let p = self.as_mut_ptr().add(index); + // Shift everything over to make space. (Duplicating the + // `index`th element into two consecutive places.) + ptr::copy(p, p.offset(1), len - index); + // Write it in, overwriting the first copy of the `index`th + // element. + ptr::write(p, element); + } + self.set_len(len + 1); + } + } + + /// Removes and returns the element at position `index` within the vector, + /// shifting all elements after it to the left. + /// + /// Note: Because this shifts over the remaining elements, it has a + /// worst-case performance of *O*(*n*). If you don't need the order of elements + /// to be preserved, use [`swap_remove`] instead. If you'd like to remove + /// elements from the beginning of the `Vec`, consider using + /// [`VecDeque::pop_front`] instead. + /// + /// [`swap_remove`]: Vec::swap_remove + /// [`VecDeque::pop_front`]: crate::collections::VecDeque::pop_front + /// + /// # Panics + /// + /// Panics if `index` is out of bounds. + /// + /// # Examples + /// + /// ``` + /// let mut v = vec![1, 2, 3]; + /// assert_eq!(v.remove(1), 2); + /// assert_eq!(v, [1, 3]); + /// ``` + #[stable(feature = "rust1", since = "1.0.0")] + #[track_caller] + pub fn remove(&mut self, index: usize) -> T { + #[cold] + #[inline(never)] + #[track_caller] + fn assert_failed(index: usize, len: usize) -> ! { + panic!("removal index (is {index}) should be < len (is {len})"); + } + + let len = self.len(); + if index >= len { + assert_failed(index, len); + } + unsafe { + // infallible + let ret; + { + // the place we are taking from. + let ptr = self.as_mut_ptr().add(index); + // copy it out, unsafely having a copy of the value on + // the stack and in the vector at the same time. + ret = ptr::read(ptr); + + // Shift everything down to fill in that spot. + ptr::copy(ptr.offset(1), ptr, len - index - 1); + } + self.set_len(len - 1); + ret + } + } + + /// Retains only the elements specified by the predicate. + /// + /// In other words, remove all elements `e` for which `f(&e)` returns `false`. + /// This method operates in place, visiting each element exactly once in the + /// original order, and preserves the order of the retained elements. + /// + /// # Examples + /// + /// ``` + /// let mut vec = vec![1, 2, 3, 4]; + /// vec.retain(|&x| x % 2 == 0); + /// assert_eq!(vec, [2, 4]); + /// ``` + /// + /// Because the elements are visited exactly once in the original order, + /// external state may be used to decide which elements to keep. + /// + /// ``` + /// let mut vec = vec![1, 2, 3, 4, 5]; + /// let keep = [false, true, true, false, true]; + /// let mut iter = keep.iter(); + /// vec.retain(|_| *iter.next().unwrap()); + /// assert_eq!(vec, [2, 3, 5]); + /// ``` + #[stable(feature = "rust1", since = "1.0.0")] + pub fn retain<F>(&mut self, mut f: F) + where + F: FnMut(&T) -> bool, + { + self.retain_mut(|elem| f(elem)); + } + + /// Retains only the elements specified by the predicate, passing a mutable reference to it. + /// + /// In other words, remove all elements `e` such that `f(&mut e)` returns `false`. + /// This method operates in place, visiting each element exactly once in the + /// original order, and preserves the order of the retained elements. + /// + /// # Examples + /// + /// ``` + /// let mut vec = vec![1, 2, 3, 4]; + /// vec.retain_mut(|x| if *x > 3 { + /// false + /// } else { + /// *x += 1; + /// true + /// }); + /// assert_eq!(vec, [2, 3, 4]); + /// ``` + #[stable(feature = "vec_retain_mut", since = "1.61.0")] + pub fn retain_mut<F>(&mut self, mut f: F) + where + F: FnMut(&mut T) -> bool, + { + let original_len = self.len(); + // Avoid double drop if the drop guard is not executed, + // since we may make some holes during the process. + unsafe { self.set_len(0) }; + + // Vec: [Kept, Kept, Hole, Hole, Hole, Hole, Unchecked, Unchecked] + // |<- processed len ->| ^- next to check + // |<- deleted cnt ->| + // |<- original_len ->| + // Kept: Elements which predicate returns true on. + // Hole: Moved or dropped element slot. + // Unchecked: Unchecked valid elements. + // + // This drop guard will be invoked when predicate or `drop` of element panicked. + // It shifts unchecked elements to cover holes and `set_len` to the correct length. + // In cases when predicate and `drop` never panick, it will be optimized out. + struct BackshiftOnDrop<'a, T, A: Allocator> { + v: &'a mut Vec<T, A>, + processed_len: usize, + deleted_cnt: usize, + original_len: usize, + } + + impl<T, A: Allocator> Drop for BackshiftOnDrop<'_, T, A> { + fn drop(&mut self) { + if self.deleted_cnt > 0 { + // SAFETY: Trailing unchecked items must be valid since we never touch them. + unsafe { + ptr::copy( + self.v.as_ptr().add(self.processed_len), + self.v.as_mut_ptr().add(self.processed_len - self.deleted_cnt), + self.original_len - self.processed_len, + ); + } + } + // SAFETY: After filling holes, all items are in contiguous memory. + unsafe { + self.v.set_len(self.original_len - self.deleted_cnt); + } + } + } + + let mut g = BackshiftOnDrop { v: self, processed_len: 0, deleted_cnt: 0, original_len }; + + fn process_loop<F, T, A: Allocator, const DELETED: bool>( + original_len: usize, + f: &mut F, + g: &mut BackshiftOnDrop<'_, T, A>, + ) where + F: FnMut(&mut T) -> bool, + { + while g.processed_len != original_len { + // SAFETY: Unchecked element must be valid. + let cur = unsafe { &mut *g.v.as_mut_ptr().add(g.processed_len) }; + if !f(cur) { + // Advance early to avoid double drop if `drop_in_place` panicked. + g.processed_len += 1; + g.deleted_cnt += 1; + // SAFETY: We never touch this element again after dropped. + unsafe { ptr::drop_in_place(cur) }; + // We already advanced the counter. + if DELETED { + continue; + } else { + break; + } + } + if DELETED { + // SAFETY: `deleted_cnt` > 0, so the hole slot must not overlap with current element. + // We use copy for move, and never touch this element again. + unsafe { + let hole_slot = g.v.as_mut_ptr().add(g.processed_len - g.deleted_cnt); + ptr::copy_nonoverlapping(cur, hole_slot, 1); + } + } + g.processed_len += 1; + } + } + + // Stage 1: Nothing was deleted. + process_loop::<F, T, A, false>(original_len, &mut f, &mut g); + + // Stage 2: Some elements were deleted. + process_loop::<F, T, A, true>(original_len, &mut f, &mut g); + + // All item are processed. This can be optimized to `set_len` by LLVM. + drop(g); + } + + /// Removes all but the first of consecutive elements in the vector that resolve to the same + /// key. + /// + /// If the vector is sorted, this removes all duplicates. + /// + /// # Examples + /// + /// ``` + /// let mut vec = vec![10, 20, 21, 30, 20]; + /// + /// vec.dedup_by_key(|i| *i / 10); + /// + /// assert_eq!(vec, [10, 20, 30, 20]); + /// ``` + #[stable(feature = "dedup_by", since = "1.16.0")] + #[inline] + pub fn dedup_by_key<F, K>(&mut self, mut key: F) + where + F: FnMut(&mut T) -> K, + K: PartialEq, + { + self.dedup_by(|a, b| key(a) == key(b)) + } + + /// Removes all but the first of consecutive elements in the vector satisfying a given equality + /// relation. + /// + /// The `same_bucket` function is passed references to two elements from the vector and + /// must determine if the elements compare equal. The elements are passed in opposite order + /// from their order in the slice, so if `same_bucket(a, b)` returns `true`, `a` is removed. + /// + /// If the vector is sorted, this removes all duplicates. + /// + /// # Examples + /// + /// ``` + /// let mut vec = vec!["foo", "bar", "Bar", "baz", "bar"]; + /// + /// vec.dedup_by(|a, b| a.eq_ignore_ascii_case(b)); + /// + /// assert_eq!(vec, ["foo", "bar", "baz", "bar"]); + /// ``` + #[stable(feature = "dedup_by", since = "1.16.0")] + pub fn dedup_by<F>(&mut self, mut same_bucket: F) + where + F: FnMut(&mut T, &mut T) -> bool, + { + let len = self.len(); + if len <= 1 { + return; + } + + /* INVARIANT: vec.len() > read >= write > write-1 >= 0 */ + struct FillGapOnDrop<'a, T, A: core::alloc::Allocator> { + /* Offset of the element we want to check if it is duplicate */ + read: usize, + + /* Offset of the place where we want to place the non-duplicate + * when we find it. */ + write: usize, + + /* The Vec that would need correction if `same_bucket` panicked */ + vec: &'a mut Vec<T, A>, + } + + impl<'a, T, A: core::alloc::Allocator> Drop for FillGapOnDrop<'a, T, A> { + fn drop(&mut self) { + /* This code gets executed when `same_bucket` panics */ + + /* SAFETY: invariant guarantees that `read - write` + * and `len - read` never overflow and that the copy is always + * in-bounds. */ + unsafe { + let ptr = self.vec.as_mut_ptr(); + let len = self.vec.len(); + + /* How many items were left when `same_bucket` panicked. + * Basically vec[read..].len() */ + let items_left = len.wrapping_sub(self.read); + + /* Pointer to first item in vec[write..write+items_left] slice */ + let dropped_ptr = ptr.add(self.write); + /* Pointer to first item in vec[read..] slice */ + let valid_ptr = ptr.add(self.read); + + /* Copy `vec[read..]` to `vec[write..write+items_left]`. + * The slices can overlap, so `copy_nonoverlapping` cannot be used */ + ptr::copy(valid_ptr, dropped_ptr, items_left); + + /* How many items have been already dropped + * Basically vec[read..write].len() */ + let dropped = self.read.wrapping_sub(self.write); + + self.vec.set_len(len - dropped); + } + } + } + + let mut gap = FillGapOnDrop { read: 1, write: 1, vec: self }; + let ptr = gap.vec.as_mut_ptr(); + + /* Drop items while going through Vec, it should be more efficient than + * doing slice partition_dedup + truncate */ + + /* SAFETY: Because of the invariant, read_ptr, prev_ptr and write_ptr + * are always in-bounds and read_ptr never aliases prev_ptr */ + unsafe { + while gap.read < len { + let read_ptr = ptr.add(gap.read); + let prev_ptr = ptr.add(gap.write.wrapping_sub(1)); + + if same_bucket(&mut *read_ptr, &mut *prev_ptr) { + // Increase `gap.read` now since the drop may panic. + gap.read += 1; + /* We have found duplicate, drop it in-place */ + ptr::drop_in_place(read_ptr); + } else { + let write_ptr = ptr.add(gap.write); + + /* Because `read_ptr` can be equal to `write_ptr`, we either + * have to use `copy` or conditional `copy_nonoverlapping`. + * Looks like the first option is faster. */ + ptr::copy(read_ptr, write_ptr, 1); + + /* We have filled that place, so go further */ + gap.write += 1; + gap.read += 1; + } + } + + /* Technically we could let `gap` clean up with its Drop, but + * when `same_bucket` is guaranteed to not panic, this bloats a little + * the codegen, so we just do it manually */ + gap.vec.set_len(gap.write); + mem::forget(gap); + } + } + + /// Appends an element to the back of a collection. + /// + /// # Panics + /// + /// Panics if the new capacity exceeds `isize::MAX` bytes. + /// + /// # Examples + /// + /// ``` + /// let mut vec = vec![1, 2]; + /// vec.push(3); + /// assert_eq!(vec, [1, 2, 3]); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[inline] + #[stable(feature = "rust1", since = "1.0.0")] + pub fn push(&mut self, value: T) { + // This will panic or abort if we would allocate > isize::MAX bytes + // or if the length increment would overflow for zero-sized types. + if self.len == self.buf.capacity() { + self.buf.reserve_for_push(self.len); + } + unsafe { + let end = self.as_mut_ptr().add(self.len); + ptr::write(end, value); + self.len += 1; + } + } + + /// Tries to append an element to the back of a collection. + /// + /// # Examples + /// + /// ``` + /// let mut vec = vec![1, 2]; + /// vec.try_push(3).unwrap(); + /// assert_eq!(vec, [1, 2, 3]); + /// ``` + #[inline] + #[stable(feature = "kernel", since = "1.0.0")] + pub fn try_push(&mut self, value: T) -> Result<(), TryReserveError> { + if self.len == self.buf.capacity() { + self.buf.try_reserve_for_push(self.len)?; + } + unsafe { + let end = self.as_mut_ptr().add(self.len); + ptr::write(end, value); + self.len += 1; + } + Ok(()) + } + + /// Removes the last element from a vector and returns it, or [`None`] if it + /// is empty. + /// + /// If you'd like to pop the first element, consider using + /// [`VecDeque::pop_front`] instead. + /// + /// [`VecDeque::pop_front`]: crate::collections::VecDeque::pop_front + /// + /// # Examples + /// + /// ``` + /// let mut vec = vec![1, 2, 3]; + /// assert_eq!(vec.pop(), Some(3)); + /// assert_eq!(vec, [1, 2]); + /// ``` + #[inline] + #[stable(feature = "rust1", since = "1.0.0")] + pub fn pop(&mut self) -> Option<T> { + if self.len == 0 { + None + } else { + unsafe { + self.len -= 1; + Some(ptr::read(self.as_ptr().add(self.len()))) + } + } + } + + /// Moves all the elements of `other` into `self`, leaving `other` empty. + /// + /// # Panics + /// + /// Panics if the number of elements in the vector overflows a `usize`. + /// + /// # Examples + /// + /// ``` + /// let mut vec = vec![1, 2, 3]; + /// let mut vec2 = vec![4, 5, 6]; + /// vec.append(&mut vec2); + /// assert_eq!(vec, [1, 2, 3, 4, 5, 6]); + /// assert_eq!(vec2, []); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[inline] + #[stable(feature = "append", since = "1.4.0")] + pub fn append(&mut self, other: &mut Self) { + unsafe { + self.append_elements(other.as_slice() as _); + other.set_len(0); + } + } + + /// Appends elements to `self` from other buffer. + #[cfg(not(no_global_oom_handling))] + #[inline] + unsafe fn append_elements(&mut self, other: *const [T]) { + let count = unsafe { (*other).len() }; + self.reserve(count); + let len = self.len(); + unsafe { ptr::copy_nonoverlapping(other as *const T, self.as_mut_ptr().add(len), count) }; + self.len += count; + } + + /// Removes the specified range from the vector in bulk, returning all + /// removed elements as an iterator. If the iterator is dropped before + /// being fully consumed, it drops the remaining removed elements. + /// + /// The returned iterator keeps a mutable borrow on the vector to optimize + /// its implementation. + /// + /// # Panics + /// + /// Panics if the starting point is greater than the end point or if + /// the end point is greater than the length of the vector. + /// + /// # Leaking + /// + /// If the returned iterator goes out of scope without being dropped (due to + /// [`mem::forget`], for example), the vector may have lost and leaked + /// elements arbitrarily, including elements outside the range. + /// + /// # Examples + /// + /// ``` + /// let mut v = vec![1, 2, 3]; + /// let u: Vec<_> = v.drain(1..).collect(); + /// assert_eq!(v, &[1]); + /// assert_eq!(u, &[2, 3]); + /// + /// // A full range clears the vector, like `clear()` does + /// v.drain(..); + /// assert_eq!(v, &[]); + /// ``` + #[stable(feature = "drain", since = "1.6.0")] + pub fn drain<R>(&mut self, range: R) -> Drain<'_, T, A> + where + R: RangeBounds<usize>, + { + // Memory safety + // + // When the Drain is first created, it shortens the length of + // the source vector to make sure no uninitialized or moved-from elements + // are accessible at all if the Drain's destructor never gets to run. + // + // Drain will ptr::read out the values to remove. + // When finished, remaining tail of the vec is copied back to cover + // the hole, and the vector length is restored to the new length. + // + let len = self.len(); + let Range { start, end } = slice::range(range, ..len); + + unsafe { + // set self.vec length's to start, to be safe in case Drain is leaked + self.set_len(start); + // Use the borrow in the IterMut to indicate borrowing behavior of the + // whole Drain iterator (like &mut T). + let range_slice = slice::from_raw_parts_mut(self.as_mut_ptr().add(start), end - start); + Drain { + tail_start: end, + tail_len: len - end, + iter: range_slice.iter(), + vec: NonNull::from(self), + } + } + } + + /// Clears the vector, removing all values. + /// + /// Note that this method has no effect on the allocated capacity + /// of the vector. + /// + /// # Examples + /// + /// ``` + /// let mut v = vec![1, 2, 3]; + /// + /// v.clear(); + /// + /// assert!(v.is_empty()); + /// ``` + #[inline] + #[stable(feature = "rust1", since = "1.0.0")] + pub fn clear(&mut self) { + let elems: *mut [T] = self.as_mut_slice(); + + // SAFETY: + // - `elems` comes directly from `as_mut_slice` and is therefore valid. + // - Setting `self.len` before calling `drop_in_place` means that, + // if an element's `Drop` impl panics, the vector's `Drop` impl will + // do nothing (leaking the rest of the elements) instead of dropping + // some twice. + unsafe { + self.len = 0; + ptr::drop_in_place(elems); + } + } + + /// Returns the number of elements in the vector, also referred to + /// as its 'length'. + /// + /// # Examples + /// + /// ``` + /// let a = vec![1, 2, 3]; + /// assert_eq!(a.len(), 3); + /// ``` + #[inline] + #[stable(feature = "rust1", since = "1.0.0")] + pub fn len(&self) -> usize { + self.len + } + + /// Returns `true` if the vector contains no elements. + /// + /// # Examples + /// + /// ``` + /// let mut v = Vec::new(); + /// assert!(v.is_empty()); + /// + /// v.push(1); + /// assert!(!v.is_empty()); + /// ``` + #[stable(feature = "rust1", since = "1.0.0")] + pub fn is_empty(&self) -> bool { + self.len() == 0 + } + + /// Splits the collection into two at the given index. + /// + /// Returns a newly allocated vector containing the elements in the range + /// `[at, len)`. After the call, the original vector will be left containing + /// the elements `[0, at)` with its previous capacity unchanged. + /// + /// # Panics + /// + /// Panics if `at > len`. + /// + /// # Examples + /// + /// ``` + /// let mut vec = vec![1, 2, 3]; + /// let vec2 = vec.split_off(1); + /// assert_eq!(vec, [1]); + /// assert_eq!(vec2, [2, 3]); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[inline] + #[must_use = "use `.truncate()` if you don't need the other half"] + #[stable(feature = "split_off", since = "1.4.0")] + pub fn split_off(&mut self, at: usize) -> Self + where + A: Clone, + { + #[cold] + #[inline(never)] + fn assert_failed(at: usize, len: usize) -> ! { + panic!("`at` split index (is {at}) should be <= len (is {len})"); + } + + if at > self.len() { + assert_failed(at, self.len()); + } + + if at == 0 { + // the new vector can take over the original buffer and avoid the copy + return mem::replace( + self, + Vec::with_capacity_in(self.capacity(), self.allocator().clone()), + ); + } + + let other_len = self.len - at; + let mut other = Vec::with_capacity_in(other_len, self.allocator().clone()); + + // Unsafely `set_len` and copy items to `other`. + unsafe { + self.set_len(at); + other.set_len(other_len); + + ptr::copy_nonoverlapping(self.as_ptr().add(at), other.as_mut_ptr(), other.len()); + } + other + } + + /// Resizes the `Vec` in-place so that `len` is equal to `new_len`. + /// + /// If `new_len` is greater than `len`, the `Vec` is extended by the + /// difference, with each additional slot filled with the result of + /// calling the closure `f`. The return values from `f` will end up + /// in the `Vec` in the order they have been generated. + /// + /// If `new_len` is less than `len`, the `Vec` is simply truncated. + /// + /// This method uses a closure to create new values on every push. If + /// you'd rather [`Clone`] a given value, use [`Vec::resize`]. If you + /// want to use the [`Default`] trait to generate values, you can + /// pass [`Default::default`] as the second argument. + /// + /// # Examples + /// + /// ``` + /// let mut vec = vec![1, 2, 3]; + /// vec.resize_with(5, Default::default); + /// assert_eq!(vec, [1, 2, 3, 0, 0]); + /// + /// let mut vec = vec![]; + /// let mut p = 1; + /// vec.resize_with(4, || { p *= 2; p }); + /// assert_eq!(vec, [2, 4, 8, 16]); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[stable(feature = "vec_resize_with", since = "1.33.0")] + pub fn resize_with<F>(&mut self, new_len: usize, f: F) + where + F: FnMut() -> T, + { + let len = self.len(); + if new_len > len { + self.extend_with(new_len - len, ExtendFunc(f)); + } else { + self.truncate(new_len); + } + } + + /// Consumes and leaks the `Vec`, returning a mutable reference to the contents, + /// `&'a mut [T]`. Note that the type `T` must outlive the chosen lifetime + /// `'a`. If the type has only static references, or none at all, then this + /// may be chosen to be `'static`. + /// + /// As of Rust 1.57, this method does not reallocate or shrink the `Vec`, + /// so the leaked allocation may include unused capacity that is not part + /// of the returned slice. + /// + /// This function is mainly useful for data that lives for the remainder of + /// the program's life. Dropping the returned reference will cause a memory + /// leak. + /// + /// # Examples + /// + /// Simple usage: + /// + /// ``` + /// let x = vec![1, 2, 3]; + /// let static_ref: &'static mut [usize] = x.leak(); + /// static_ref[0] += 1; + /// assert_eq!(static_ref, &[2, 2, 3]); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[stable(feature = "vec_leak", since = "1.47.0")] + #[inline] + pub fn leak<'a>(self) -> &'a mut [T] + where + A: 'a, + { + let mut me = ManuallyDrop::new(self); + unsafe { slice::from_raw_parts_mut(me.as_mut_ptr(), me.len) } + } + + /// Returns the remaining spare capacity of the vector as a slice of + /// `MaybeUninit<T>`. + /// + /// The returned slice can be used to fill the vector with data (e.g. by + /// reading from a file) before marking the data as initialized using the + /// [`set_len`] method. + /// + /// [`set_len`]: Vec::set_len + /// + /// # Examples + /// + /// ``` + /// // Allocate vector big enough for 10 elements. + /// let mut v = Vec::with_capacity(10); + /// + /// // Fill in the first 3 elements. + /// let uninit = v.spare_capacity_mut(); + /// uninit[0].write(0); + /// uninit[1].write(1); + /// uninit[2].write(2); + /// + /// // Mark the first 3 elements of the vector as being initialized. + /// unsafe { + /// v.set_len(3); + /// } + /// + /// assert_eq!(&v, &[0, 1, 2]); + /// ``` + #[stable(feature = "vec_spare_capacity", since = "1.60.0")] + #[inline] + pub fn spare_capacity_mut(&mut self) -> &mut [MaybeUninit<T>] { + // Note: + // This method is not implemented in terms of `split_at_spare_mut`, + // to prevent invalidation of pointers to the buffer. + unsafe { + slice::from_raw_parts_mut( + self.as_mut_ptr().add(self.len) as *mut MaybeUninit<T>, + self.buf.capacity() - self.len, + ) + } + } + + /// Returns vector content as a slice of `T`, along with the remaining spare + /// capacity of the vector as a slice of `MaybeUninit<T>`. + /// + /// The returned spare capacity slice can be used to fill the vector with data + /// (e.g. by reading from a file) before marking the data as initialized using + /// the [`set_len`] method. + /// + /// [`set_len`]: Vec::set_len + /// + /// Note that this is a low-level API, which should be used with care for + /// optimization purposes. If you need to append data to a `Vec` + /// you can use [`push`], [`extend`], [`extend_from_slice`], + /// [`extend_from_within`], [`insert`], [`append`], [`resize`] or + /// [`resize_with`], depending on your exact needs. + /// + /// [`push`]: Vec::push + /// [`extend`]: Vec::extend + /// [`extend_from_slice`]: Vec::extend_from_slice + /// [`extend_from_within`]: Vec::extend_from_within + /// [`insert`]: Vec::insert + /// [`append`]: Vec::append + /// [`resize`]: Vec::resize + /// [`resize_with`]: Vec::resize_with + /// + /// # Examples + /// + /// ``` + /// #![feature(vec_split_at_spare)] + /// + /// let mut v = vec![1, 1, 2]; + /// + /// // Reserve additional space big enough for 10 elements. + /// v.reserve(10); + /// + /// let (init, uninit) = v.split_at_spare_mut(); + /// let sum = init.iter().copied().sum::<u32>(); + /// + /// // Fill in the next 4 elements. + /// uninit[0].write(sum); + /// uninit[1].write(sum * 2); + /// uninit[2].write(sum * 3); + /// uninit[3].write(sum * 4); + /// + /// // Mark the 4 elements of the vector as being initialized. + /// unsafe { + /// let len = v.len(); + /// v.set_len(len + 4); + /// } + /// + /// assert_eq!(&v, &[1, 1, 2, 4, 8, 12, 16]); + /// ``` + #[unstable(feature = "vec_split_at_spare", issue = "81944")] + #[inline] + pub fn split_at_spare_mut(&mut self) -> (&mut [T], &mut [MaybeUninit<T>]) { + // SAFETY: + // - len is ignored and so never changed + let (init, spare, _) = unsafe { self.split_at_spare_mut_with_len() }; + (init, spare) + } + + /// Safety: changing returned .2 (&mut usize) is considered the same as calling `.set_len(_)`. + /// + /// This method provides unique access to all vec parts at once in `extend_from_within`. + unsafe fn split_at_spare_mut_with_len( + &mut self, + ) -> (&mut [T], &mut [MaybeUninit<T>], &mut usize) { + let ptr = self.as_mut_ptr(); + // SAFETY: + // - `ptr` is guaranteed to be valid for `self.len` elements + // - but the allocation extends out to `self.buf.capacity()` elements, possibly + // uninitialized + let spare_ptr = unsafe { ptr.add(self.len) }; + let spare_ptr = spare_ptr.cast::<MaybeUninit<T>>(); + let spare_len = self.buf.capacity() - self.len; + + // SAFETY: + // - `ptr` is guaranteed to be valid for `self.len` elements + // - `spare_ptr` is pointing one element past the buffer, so it doesn't overlap with `initialized` + unsafe { + let initialized = slice::from_raw_parts_mut(ptr, self.len); + let spare = slice::from_raw_parts_mut(spare_ptr, spare_len); + + (initialized, spare, &mut self.len) + } + } +} + +impl<T: Clone, A: Allocator> Vec<T, A> { + /// Resizes the `Vec` in-place so that `len` is equal to `new_len`. + /// + /// If `new_len` is greater than `len`, the `Vec` is extended by the + /// difference, with each additional slot filled with `value`. + /// If `new_len` is less than `len`, the `Vec` is simply truncated. + /// + /// This method requires `T` to implement [`Clone`], + /// in order to be able to clone the passed value. + /// If you need more flexibility (or want to rely on [`Default`] instead of + /// [`Clone`]), use [`Vec::resize_with`]. + /// If you only need to resize to a smaller size, use [`Vec::truncate`]. + /// + /// # Examples + /// + /// ``` + /// let mut vec = vec!["hello"]; + /// vec.resize(3, "world"); + /// assert_eq!(vec, ["hello", "world", "world"]); + /// + /// let mut vec = vec![1, 2, 3, 4]; + /// vec.resize(2, 0); + /// assert_eq!(vec, [1, 2]); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[stable(feature = "vec_resize", since = "1.5.0")] + pub fn resize(&mut self, new_len: usize, value: T) { + let len = self.len(); + + if new_len > len { + self.extend_with(new_len - len, ExtendElement(value)) + } else { + self.truncate(new_len); + } + } + + /// Clones and appends all elements in a slice to the `Vec`. + /// + /// Iterates over the slice `other`, clones each element, and then appends + /// it to this `Vec`. The `other` slice is traversed in-order. + /// + /// Note that this function is same as [`extend`] except that it is + /// specialized to work with slices instead. If and when Rust gets + /// specialization this function will likely be deprecated (but still + /// available). + /// + /// # Examples + /// + /// ``` + /// let mut vec = vec![1]; + /// vec.extend_from_slice(&[2, 3, 4]); + /// assert_eq!(vec, [1, 2, 3, 4]); + /// ``` + /// + /// [`extend`]: Vec::extend + #[cfg(not(no_global_oom_handling))] + #[stable(feature = "vec_extend_from_slice", since = "1.6.0")] + pub fn extend_from_slice(&mut self, other: &[T]) { + self.spec_extend(other.iter()) + } + + /// Copies elements from `src` range to the end of the vector. + /// + /// # Panics + /// + /// Panics if the starting point is greater than the end point or if + /// the end point is greater than the length of the vector. + /// + /// # Examples + /// + /// ``` + /// let mut vec = vec![0, 1, 2, 3, 4]; + /// + /// vec.extend_from_within(2..); + /// assert_eq!(vec, [0, 1, 2, 3, 4, 2, 3, 4]); + /// + /// vec.extend_from_within(..2); + /// assert_eq!(vec, [0, 1, 2, 3, 4, 2, 3, 4, 0, 1]); + /// + /// vec.extend_from_within(4..8); + /// assert_eq!(vec, [0, 1, 2, 3, 4, 2, 3, 4, 0, 1, 4, 2, 3, 4]); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[stable(feature = "vec_extend_from_within", since = "1.53.0")] + pub fn extend_from_within<R>(&mut self, src: R) + where + R: RangeBounds<usize>, + { + let range = slice::range(src, ..self.len()); + self.reserve(range.len()); + + // SAFETY: + // - `slice::range` guarantees that the given range is valid for indexing self + unsafe { + self.spec_extend_from_within(range); + } + } +} + +impl<T, A: Allocator, const N: usize> Vec<[T; N], A> { + /// Takes a `Vec<[T; N]>` and flattens it into a `Vec<T>`. + /// + /// # Panics + /// + /// Panics if the length of the resulting vector would overflow a `usize`. + /// + /// This is only possible when flattening a vector of arrays of zero-sized + /// types, and thus tends to be irrelevant in practice. If + /// `size_of::<T>() > 0`, this will never panic. + /// + /// # Examples + /// + /// ``` + /// #![feature(slice_flatten)] + /// + /// let mut vec = vec![[1, 2, 3], [4, 5, 6], [7, 8, 9]]; + /// assert_eq!(vec.pop(), Some([7, 8, 9])); + /// + /// let mut flattened = vec.into_flattened(); + /// assert_eq!(flattened.pop(), Some(6)); + /// ``` + #[unstable(feature = "slice_flatten", issue = "95629")] + pub fn into_flattened(self) -> Vec<T, A> { + let (ptr, len, cap, alloc) = self.into_raw_parts_with_alloc(); + let (new_len, new_cap) = if mem::size_of::<T>() == 0 { + (len.checked_mul(N).expect("vec len overflow"), usize::MAX) + } else { + // SAFETY: + // - `cap * N` cannot overflow because the allocation is already in + // the address space. + // - Each `[T; N]` has `N` valid elements, so there are `len * N` + // valid elements in the allocation. + unsafe { (len.unchecked_mul(N), cap.unchecked_mul(N)) } + }; + // SAFETY: + // - `ptr` was allocated by `self` + // - `ptr` is well-aligned because `[T; N]` has the same alignment as `T`. + // - `new_cap` refers to the same sized allocation as `cap` because + // `new_cap * size_of::<T>()` == `cap * size_of::<[T; N]>()` + // - `len` <= `cap`, so `len * N` <= `cap * N`. + unsafe { Vec::<T, A>::from_raw_parts_in(ptr.cast(), new_len, new_cap, alloc) } + } +} + +// This code generalizes `extend_with_{element,default}`. +trait ExtendWith<T> { + fn next(&mut self) -> T; + fn last(self) -> T; +} + +struct ExtendElement<T>(T); +impl<T: Clone> ExtendWith<T> for ExtendElement<T> { + fn next(&mut self) -> T { + self.0.clone() + } + fn last(self) -> T { + self.0 + } +} + +struct ExtendFunc<F>(F); +impl<T, F: FnMut() -> T> ExtendWith<T> for ExtendFunc<F> { + fn next(&mut self) -> T { + (self.0)() + } + fn last(mut self) -> T { + (self.0)() + } +} + +impl<T, A: Allocator> Vec<T, A> { + #[cfg(not(no_global_oom_handling))] + /// Extend the vector by `n` values, using the given generator. + fn extend_with<E: ExtendWith<T>>(&mut self, n: usize, mut value: E) { + self.reserve(n); + + unsafe { + let mut ptr = self.as_mut_ptr().add(self.len()); + // Use SetLenOnDrop to work around bug where compiler + // might not realize the store through `ptr` through self.set_len() + // don't alias. + let mut local_len = SetLenOnDrop::new(&mut self.len); + + // Write all elements except the last one + for _ in 1..n { + ptr::write(ptr, value.next()); + ptr = ptr.offset(1); + // Increment the length in every step in case next() panics + local_len.increment_len(1); + } + + if n > 0 { + // We can write the last element directly without cloning needlessly + ptr::write(ptr, value.last()); + local_len.increment_len(1); + } + + // len set by scope guard + } + } +} + +impl<T: PartialEq, A: Allocator> Vec<T, A> { + /// Removes consecutive repeated elements in the vector according to the + /// [`PartialEq`] trait implementation. + /// + /// If the vector is sorted, this removes all duplicates. + /// + /// # Examples + /// + /// ``` + /// let mut vec = vec![1, 2, 2, 3, 2]; + /// + /// vec.dedup(); + /// + /// assert_eq!(vec, [1, 2, 3, 2]); + /// ``` + #[stable(feature = "rust1", since = "1.0.0")] + #[inline] + pub fn dedup(&mut self) { + self.dedup_by(|a, b| a == b) + } +} + +//////////////////////////////////////////////////////////////////////////////// +// Internal methods and functions +//////////////////////////////////////////////////////////////////////////////// + +#[doc(hidden)] +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "rust1", since = "1.0.0")] +pub fn from_elem<T: Clone>(elem: T, n: usize) -> Vec<T> { + <T as SpecFromElem>::from_elem(elem, n, Global) +} + +#[doc(hidden)] +#[cfg(not(no_global_oom_handling))] +#[unstable(feature = "allocator_api", issue = "32838")] +pub fn from_elem_in<T: Clone, A: Allocator>(elem: T, n: usize, alloc: A) -> Vec<T, A> { + <T as SpecFromElem>::from_elem(elem, n, alloc) +} + +trait ExtendFromWithinSpec { + /// # Safety + /// + /// - `src` needs to be valid index + /// - `self.capacity() - self.len()` must be `>= src.len()` + unsafe fn spec_extend_from_within(&mut self, src: Range<usize>); +} + +impl<T: Clone, A: Allocator> ExtendFromWithinSpec for Vec<T, A> { + default unsafe fn spec_extend_from_within(&mut self, src: Range<usize>) { + // SAFETY: + // - len is increased only after initializing elements + let (this, spare, len) = unsafe { self.split_at_spare_mut_with_len() }; + + // SAFETY: + // - caller guaratees that src is a valid index + let to_clone = unsafe { this.get_unchecked(src) }; + + iter::zip(to_clone, spare) + .map(|(src, dst)| dst.write(src.clone())) + // Note: + // - Element was just initialized with `MaybeUninit::write`, so it's ok to increase len + // - len is increased after each element to prevent leaks (see issue #82533) + .for_each(|_| *len += 1); + } +} + +impl<T: Copy, A: Allocator> ExtendFromWithinSpec for Vec<T, A> { + unsafe fn spec_extend_from_within(&mut self, src: Range<usize>) { + let count = src.len(); + { + let (init, spare) = self.split_at_spare_mut(); + + // SAFETY: + // - caller guaratees that `src` is a valid index + let source = unsafe { init.get_unchecked(src) }; + + // SAFETY: + // - Both pointers are created from unique slice references (`&mut [_]`) + // so they are valid and do not overlap. + // - Elements are :Copy so it's OK to to copy them, without doing + // anything with the original values + // - `count` is equal to the len of `source`, so source is valid for + // `count` reads + // - `.reserve(count)` guarantees that `spare.len() >= count` so spare + // is valid for `count` writes + unsafe { ptr::copy_nonoverlapping(source.as_ptr(), spare.as_mut_ptr() as _, count) }; + } + + // SAFETY: + // - The elements were just initialized by `copy_nonoverlapping` + self.len += count; + } +} + +//////////////////////////////////////////////////////////////////////////////// +// Common trait implementations for Vec +//////////////////////////////////////////////////////////////////////////////// + +#[stable(feature = "rust1", since = "1.0.0")] +impl<T, A: Allocator> ops::Deref for Vec<T, A> { + type Target = [T]; + + fn deref(&self) -> &[T] { + unsafe { slice::from_raw_parts(self.as_ptr(), self.len) } + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<T, A: Allocator> ops::DerefMut for Vec<T, A> { + fn deref_mut(&mut self) -> &mut [T] { + unsafe { slice::from_raw_parts_mut(self.as_mut_ptr(), self.len) } + } +} + +#[cfg(not(no_global_oom_handling))] +trait SpecCloneFrom { + fn clone_from(this: &mut Self, other: &Self); +} + +#[cfg(not(no_global_oom_handling))] +impl<T: Clone, A: Allocator> SpecCloneFrom for Vec<T, A> { + default fn clone_from(this: &mut Self, other: &Self) { + // drop anything that will not be overwritten + this.truncate(other.len()); + + // self.len <= other.len due to the truncate above, so the + // slices here are always in-bounds. + let (init, tail) = other.split_at(this.len()); + + // reuse the contained values' allocations/resources. + this.clone_from_slice(init); + this.extend_from_slice(tail); + } +} + +#[cfg(not(no_global_oom_handling))] +impl<T: Copy, A: Allocator> SpecCloneFrom for Vec<T, A> { + fn clone_from(this: &mut Self, other: &Self) { + this.clear(); + this.extend_from_slice(other); + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "rust1", since = "1.0.0")] +impl<T: Clone, A: Allocator + Clone> Clone for Vec<T, A> { + #[cfg(not(test))] + fn clone(&self) -> Self { + let alloc = self.allocator().clone(); + <[T]>::to_vec_in(&**self, alloc) + } + + // HACK(japaric): with cfg(test) the inherent `[T]::to_vec` method, which is + // required for this method definition, is not available. Instead use the + // `slice::to_vec` function which is only available with cfg(test) + // NB see the slice::hack module in slice.rs for more information + #[cfg(test)] + fn clone(&self) -> Self { + let alloc = self.allocator().clone(); + crate::slice::to_vec(&**self, alloc) + } + + fn clone_from(&mut self, other: &Self) { + SpecCloneFrom::clone_from(self, other) + } +} + +/// The hash of a vector is the same as that of the corresponding slice, +/// as required by the `core::borrow::Borrow` implementation. +/// +/// ``` +/// #![feature(build_hasher_simple_hash_one)] +/// use std::hash::BuildHasher; +/// +/// let b = std::collections::hash_map::RandomState::new(); +/// let v: Vec<u8> = vec![0xa8, 0x3c, 0x09]; +/// let s: &[u8] = &[0xa8, 0x3c, 0x09]; +/// assert_eq!(b.hash_one(v), b.hash_one(s)); +/// ``` +#[stable(feature = "rust1", since = "1.0.0")] +impl<T: Hash, A: Allocator> Hash for Vec<T, A> { + #[inline] + fn hash<H: Hasher>(&self, state: &mut H) { + Hash::hash(&**self, state) + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +#[rustc_on_unimplemented( + message = "vector indices are of type `usize` or ranges of `usize`", + label = "vector indices are of type `usize` or ranges of `usize`" +)] +impl<T, I: SliceIndex<[T]>, A: Allocator> Index<I> for Vec<T, A> { + type Output = I::Output; + + #[inline] + fn index(&self, index: I) -> &Self::Output { + Index::index(&**self, index) + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +#[rustc_on_unimplemented( + message = "vector indices are of type `usize` or ranges of `usize`", + label = "vector indices are of type `usize` or ranges of `usize`" +)] +impl<T, I: SliceIndex<[T]>, A: Allocator> IndexMut<I> for Vec<T, A> { + #[inline] + fn index_mut(&mut self, index: I) -> &mut Self::Output { + IndexMut::index_mut(&mut **self, index) + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "rust1", since = "1.0.0")] +impl<T> FromIterator<T> for Vec<T> { + #[inline] + fn from_iter<I: IntoIterator<Item = T>>(iter: I) -> Vec<T> { + <Self as SpecFromIter<T, I::IntoIter>>::from_iter(iter.into_iter()) + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<T, A: Allocator> IntoIterator for Vec<T, A> { + type Item = T; + type IntoIter = IntoIter<T, A>; + + /// Creates a consuming iterator, that is, one that moves each value out of + /// the vector (from start to end). The vector cannot be used after calling + /// this. + /// + /// # Examples + /// + /// ``` + /// let v = vec!["a".to_string(), "b".to_string()]; + /// for s in v.into_iter() { + /// // s has type String, not &String + /// println!("{s}"); + /// } + /// ``` + #[inline] + fn into_iter(self) -> IntoIter<T, A> { + unsafe { + let mut me = ManuallyDrop::new(self); + let alloc = ManuallyDrop::new(ptr::read(me.allocator())); + let begin = me.as_mut_ptr(); + let end = if mem::size_of::<T>() == 0 { + arith_offset(begin as *const i8, me.len() as isize) as *const T + } else { + begin.add(me.len()) as *const T + }; + let cap = me.buf.capacity(); + IntoIter { + buf: NonNull::new_unchecked(begin), + phantom: PhantomData, + cap, + alloc, + ptr: begin, + end, + } + } + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<'a, T, A: Allocator> IntoIterator for &'a Vec<T, A> { + type Item = &'a T; + type IntoIter = slice::Iter<'a, T>; + + fn into_iter(self) -> slice::Iter<'a, T> { + self.iter() + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<'a, T, A: Allocator> IntoIterator for &'a mut Vec<T, A> { + type Item = &'a mut T; + type IntoIter = slice::IterMut<'a, T>; + + fn into_iter(self) -> slice::IterMut<'a, T> { + self.iter_mut() + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "rust1", since = "1.0.0")] +impl<T, A: Allocator> Extend<T> for Vec<T, A> { + #[inline] + fn extend<I: IntoIterator<Item = T>>(&mut self, iter: I) { + <Self as SpecExtend<T, I::IntoIter>>::spec_extend(self, iter.into_iter()) + } + + #[inline] + fn extend_one(&mut self, item: T) { + self.push(item); + } + + #[inline] + fn extend_reserve(&mut self, additional: usize) { + self.reserve(additional); + } +} + +impl<T, A: Allocator> Vec<T, A> { + // leaf method to which various SpecFrom/SpecExtend implementations delegate when + // they have no further optimizations to apply + #[cfg(not(no_global_oom_handling))] + fn extend_desugared<I: Iterator<Item = T>>(&mut self, mut iterator: I) { + // This is the case for a general iterator. + // + // This function should be the moral equivalent of: + // + // for item in iterator { + // self.push(item); + // } + while let Some(element) = iterator.next() { + let len = self.len(); + if len == self.capacity() { + let (lower, _) = iterator.size_hint(); + self.reserve(lower.saturating_add(1)); + } + unsafe { + ptr::write(self.as_mut_ptr().add(len), element); + // Since next() executes user code which can panic we have to bump the length + // after each step. + // NB can't overflow since we would have had to alloc the address space + self.set_len(len + 1); + } + } + } + + /// Creates a splicing iterator that replaces the specified range in the vector + /// with the given `replace_with` iterator and yields the removed items. + /// `replace_with` does not need to be the same length as `range`. + /// + /// `range` is removed even if the iterator is not consumed until the end. + /// + /// It is unspecified how many elements are removed from the vector + /// if the `Splice` value is leaked. + /// + /// The input iterator `replace_with` is only consumed when the `Splice` value is dropped. + /// + /// This is optimal if: + /// + /// * The tail (elements in the vector after `range`) is empty, + /// * or `replace_with` yields fewer or equal elements than `range`’s length + /// * or the lower bound of its `size_hint()` is exact. + /// + /// Otherwise, a temporary vector is allocated and the tail is moved twice. + /// + /// # Panics + /// + /// Panics if the starting point is greater than the end point or if + /// the end point is greater than the length of the vector. + /// + /// # Examples + /// + /// ``` + /// let mut v = vec![1, 2, 3, 4]; + /// let new = [7, 8, 9]; + /// let u: Vec<_> = v.splice(1..3, new).collect(); + /// assert_eq!(v, &[1, 7, 8, 9, 4]); + /// assert_eq!(u, &[2, 3]); + /// ``` + #[cfg(not(no_global_oom_handling))] + #[inline] + #[stable(feature = "vec_splice", since = "1.21.0")] + pub fn splice<R, I>(&mut self, range: R, replace_with: I) -> Splice<'_, I::IntoIter, A> + where + R: RangeBounds<usize>, + I: IntoIterator<Item = T>, + { + Splice { drain: self.drain(range), replace_with: replace_with.into_iter() } + } + + /// Creates an iterator which uses a closure to determine if an element should be removed. + /// + /// If the closure returns true, then the element is removed and yielded. + /// If the closure returns false, the element will remain in the vector and will not be yielded + /// by the iterator. + /// + /// Using this method is equivalent to the following code: + /// + /// ``` + /// # let some_predicate = |x: &mut i32| { *x == 2 || *x == 3 || *x == 6 }; + /// # let mut vec = vec![1, 2, 3, 4, 5, 6]; + /// let mut i = 0; + /// while i < vec.len() { + /// if some_predicate(&mut vec[i]) { + /// let val = vec.remove(i); + /// // your code here + /// } else { + /// i += 1; + /// } + /// } + /// + /// # assert_eq!(vec, vec![1, 4, 5]); + /// ``` + /// + /// But `drain_filter` is easier to use. `drain_filter` is also more efficient, + /// because it can backshift the elements of the array in bulk. + /// + /// Note that `drain_filter` also lets you mutate every element in the filter closure, + /// regardless of whether you choose to keep or remove it. + /// + /// # Examples + /// + /// Splitting an array into evens and odds, reusing the original allocation: + /// + /// ``` + /// #![feature(drain_filter)] + /// let mut numbers = vec![1, 2, 3, 4, 5, 6, 8, 9, 11, 13, 14, 15]; + /// + /// let evens = numbers.drain_filter(|x| *x % 2 == 0).collect::<Vec<_>>(); + /// let odds = numbers; + /// + /// assert_eq!(evens, vec![2, 4, 6, 8, 14]); + /// assert_eq!(odds, vec![1, 3, 5, 9, 11, 13, 15]); + /// ``` + #[unstable(feature = "drain_filter", reason = "recently added", issue = "43244")] + pub fn drain_filter<F>(&mut self, filter: F) -> DrainFilter<'_, T, F, A> + where + F: FnMut(&mut T) -> bool, + { + let old_len = self.len(); + + // Guard against us getting leaked (leak amplification) + unsafe { + self.set_len(0); + } + + DrainFilter { vec: self, idx: 0, del: 0, old_len, pred: filter, panic_flag: false } + } +} + +/// Extend implementation that copies elements out of references before pushing them onto the Vec. +/// +/// This implementation is specialized for slice iterators, where it uses [`copy_from_slice`] to +/// append the entire slice at once. +/// +/// [`copy_from_slice`]: slice::copy_from_slice +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "extend_ref", since = "1.2.0")] +impl<'a, T: Copy + 'a, A: Allocator + 'a> Extend<&'a T> for Vec<T, A> { + fn extend<I: IntoIterator<Item = &'a T>>(&mut self, iter: I) { + self.spec_extend(iter.into_iter()) + } + + #[inline] + fn extend_one(&mut self, &item: &'a T) { + self.push(item); + } + + #[inline] + fn extend_reserve(&mut self, additional: usize) { + self.reserve(additional); + } +} + +/// Implements comparison of vectors, [lexicographically](core::cmp::Ord#lexicographical-comparison). +#[stable(feature = "rust1", since = "1.0.0")] +impl<T: PartialOrd, A: Allocator> PartialOrd for Vec<T, A> { + #[inline] + fn partial_cmp(&self, other: &Self) -> Option<Ordering> { + PartialOrd::partial_cmp(&**self, &**other) + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<T: Eq, A: Allocator> Eq for Vec<T, A> {} + +/// Implements ordering of vectors, [lexicographically](core::cmp::Ord#lexicographical-comparison). +#[stable(feature = "rust1", since = "1.0.0")] +impl<T: Ord, A: Allocator> Ord for Vec<T, A> { + #[inline] + fn cmp(&self, other: &Self) -> Ordering { + Ord::cmp(&**self, &**other) + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +unsafe impl<#[may_dangle] T, A: Allocator> Drop for Vec<T, A> { + fn drop(&mut self) { + unsafe { + // use drop for [T] + // use a raw slice to refer to the elements of the vector as weakest necessary type; + // could avoid questions of validity in certain cases + ptr::drop_in_place(ptr::slice_from_raw_parts_mut(self.as_mut_ptr(), self.len)) + } + // RawVec handles deallocation + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +#[rustc_const_unstable(feature = "const_default_impls", issue = "87864")] +impl<T> const Default for Vec<T> { + /// Creates an empty `Vec<T>`. + fn default() -> Vec<T> { + Vec::new() + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<T: fmt::Debug, A: Allocator> fmt::Debug for Vec<T, A> { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + fmt::Debug::fmt(&**self, f) + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<T, A: Allocator> AsRef<Vec<T, A>> for Vec<T, A> { + fn as_ref(&self) -> &Vec<T, A> { + self + } +} + +#[stable(feature = "vec_as_mut", since = "1.5.0")] +impl<T, A: Allocator> AsMut<Vec<T, A>> for Vec<T, A> { + fn as_mut(&mut self) -> &mut Vec<T, A> { + self + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl<T, A: Allocator> AsRef<[T]> for Vec<T, A> { + fn as_ref(&self) -> &[T] { + self + } +} + +#[stable(feature = "vec_as_mut", since = "1.5.0")] +impl<T, A: Allocator> AsMut<[T]> for Vec<T, A> { + fn as_mut(&mut self) -> &mut [T] { + self + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "rust1", since = "1.0.0")] +impl<T: Clone> From<&[T]> for Vec<T> { + /// Allocate a `Vec<T>` and fill it by cloning `s`'s items. + /// + /// # Examples + /// + /// ``` + /// assert_eq!(Vec::from(&[1, 2, 3][..]), vec![1, 2, 3]); + /// ``` + #[cfg(not(test))] + fn from(s: &[T]) -> Vec<T> { + s.to_vec() + } + #[cfg(test)] + fn from(s: &[T]) -> Vec<T> { + crate::slice::to_vec(s, Global) + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "vec_from_mut", since = "1.19.0")] +impl<T: Clone> From<&mut [T]> for Vec<T> { + /// Allocate a `Vec<T>` and fill it by cloning `s`'s items. + /// + /// # Examples + /// + /// ``` + /// assert_eq!(Vec::from(&mut [1, 2, 3][..]), vec![1, 2, 3]); + /// ``` + #[cfg(not(test))] + fn from(s: &mut [T]) -> Vec<T> { + s.to_vec() + } + #[cfg(test)] + fn from(s: &mut [T]) -> Vec<T> { + crate::slice::to_vec(s, Global) + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "vec_from_array", since = "1.44.0")] +impl<T, const N: usize> From<[T; N]> for Vec<T> { + /// Allocate a `Vec<T>` and move `s`'s items into it. + /// + /// # Examples + /// + /// ``` + /// assert_eq!(Vec::from([1, 2, 3]), vec![1, 2, 3]); + /// ``` + #[cfg(not(test))] + fn from(s: [T; N]) -> Vec<T> { + <[T]>::into_vec(box s) + } + + #[cfg(test)] + fn from(s: [T; N]) -> Vec<T> { + crate::slice::into_vec(box s) + } +} + +#[stable(feature = "vec_from_cow_slice", since = "1.14.0")] +impl<'a, T> From<Cow<'a, [T]>> for Vec<T> +where + [T]: ToOwned<Owned = Vec<T>>, +{ + /// Convert a clone-on-write slice into a vector. + /// + /// If `s` already owns a `Vec<T>`, it will be returned directly. + /// If `s` is borrowing a slice, a new `Vec<T>` will be allocated and + /// filled by cloning `s`'s items into it. + /// + /// # Examples + /// + /// ``` + /// # use std::borrow::Cow; + /// let o: Cow<[i32]> = Cow::Owned(vec![1, 2, 3]); + /// let b: Cow<[i32]> = Cow::Borrowed(&[1, 2, 3]); + /// assert_eq!(Vec::from(o), Vec::from(b)); + /// ``` + fn from(s: Cow<'a, [T]>) -> Vec<T> { + s.into_owned() + } +} + +// note: test pulls in libstd, which causes errors here +#[cfg(not(test))] +#[stable(feature = "vec_from_box", since = "1.18.0")] +impl<T, A: Allocator> From<Box<[T], A>> for Vec<T, A> { + /// Convert a boxed slice into a vector by transferring ownership of + /// the existing heap allocation. + /// + /// # Examples + /// + /// ``` + /// let b: Box<[i32]> = vec![1, 2, 3].into_boxed_slice(); + /// assert_eq!(Vec::from(b), vec![1, 2, 3]); + /// ``` + fn from(s: Box<[T], A>) -> Self { + s.into_vec() + } +} + +// note: test pulls in libstd, which causes errors here +#[cfg(not(no_global_oom_handling))] +#[cfg(not(test))] +#[stable(feature = "box_from_vec", since = "1.20.0")] +impl<T, A: Allocator> From<Vec<T, A>> for Box<[T], A> { + /// Convert a vector into a boxed slice. + /// + /// If `v` has excess capacity, its items will be moved into a + /// newly-allocated buffer with exactly the right capacity. + /// + /// # Examples + /// + /// ``` + /// assert_eq!(Box::from(vec![1, 2, 3]), vec![1, 2, 3].into_boxed_slice()); + /// ``` + fn from(v: Vec<T, A>) -> Self { + v.into_boxed_slice() + } +} + +#[cfg(not(no_global_oom_handling))] +#[stable(feature = "rust1", since = "1.0.0")] +impl From<&str> for Vec<u8> { + /// Allocate a `Vec<u8>` and fill it with a UTF-8 string. + /// + /// # Examples + /// + /// ``` + /// assert_eq!(Vec::from("123"), vec![b'1', b'2', b'3']); + /// ``` + fn from(s: &str) -> Vec<u8> { + From::from(s.as_bytes()) + } +} + +#[stable(feature = "array_try_from_vec", since = "1.48.0")] +impl<T, A: Allocator, const N: usize> TryFrom<Vec<T, A>> for [T; N] { + type Error = Vec<T, A>; + + /// Gets the entire contents of the `Vec<T>` as an array, + /// if its size exactly matches that of the requested array. + /// + /// # Examples + /// + /// ``` + /// assert_eq!(vec![1, 2, 3].try_into(), Ok([1, 2, 3])); + /// assert_eq!(<Vec<i32>>::new().try_into(), Ok([])); + /// ``` + /// + /// If the length doesn't match, the input comes back in `Err`: + /// ``` + /// let r: Result<[i32; 4], _> = (0..10).collect::<Vec<_>>().try_into(); + /// assert_eq!(r, Err(vec![0, 1, 2, 3, 4, 5, 6, 7, 8, 9])); + /// ``` + /// + /// If you're fine with just getting a prefix of the `Vec<T>`, + /// you can call [`.truncate(N)`](Vec::truncate) first. + /// ``` + /// let mut v = String::from("hello world").into_bytes(); + /// v.sort(); + /// v.truncate(2); + /// let [a, b]: [_; 2] = v.try_into().unwrap(); + /// assert_eq!(a, b' '); + /// assert_eq!(b, b'd'); + /// ``` + fn try_from(mut vec: Vec<T, A>) -> Result<[T; N], Vec<T, A>> { + if vec.len() != N { + return Err(vec); + } + + // SAFETY: `.set_len(0)` is always sound. + unsafe { vec.set_len(0) }; + + // SAFETY: A `Vec`'s pointer is always aligned properly, and + // the alignment the array needs is the same as the items. + // We checked earlier that we have sufficient items. + // The items will not double-drop as the `set_len` + // tells the `Vec` not to also drop them. + let array = unsafe { ptr::read(vec.as_ptr() as *const [T; N]) }; + Ok(array) + } +} diff --git a/rust/alloc/vec/partial_eq.rs b/rust/alloc/vec/partial_eq.rs new file mode 100644 index 000000000000..10ad4e492287 --- /dev/null +++ b/rust/alloc/vec/partial_eq.rs @@ -0,0 +1,49 @@ +// SPDX-License-Identifier: Apache-2.0 OR MIT + +use crate::alloc::Allocator; +#[cfg(not(no_global_oom_handling))] +use crate::borrow::Cow; + +use super::Vec; + +macro_rules! __impl_slice_eq1 { + ([$($vars:tt)*] $lhs:ty, $rhs:ty $(where $ty:ty: $bound:ident)?, #[$stability:meta]) => { + #[$stability] + impl<T, U, $($vars)*> PartialEq<$rhs> for $lhs + where + T: PartialEq<U>, + $($ty: $bound)? + { + #[inline] + fn eq(&self, other: &$rhs) -> bool { self[..] == other[..] } + #[inline] + fn ne(&self, other: &$rhs) -> bool { self[..] != other[..] } + } + } +} + +__impl_slice_eq1! { [A1: Allocator, A2: Allocator] Vec<T, A1>, Vec<U, A2>, #[stable(feature = "rust1", since = "1.0.0")] } +__impl_slice_eq1! { [A: Allocator] Vec<T, A>, &[U], #[stable(feature = "rust1", since = "1.0.0")] } +__impl_slice_eq1! { [A: Allocator] Vec<T, A>, &mut [U], #[stable(feature = "rust1", since = "1.0.0")] } +__impl_slice_eq1! { [A: Allocator] &[T], Vec<U, A>, #[stable(feature = "partialeq_vec_for_ref_slice", since = "1.46.0")] } +__impl_slice_eq1! { [A: Allocator] &mut [T], Vec<U, A>, #[stable(feature = "partialeq_vec_for_ref_slice", since = "1.46.0")] } +__impl_slice_eq1! { [A: Allocator] Vec<T, A>, [U], #[stable(feature = "partialeq_vec_for_slice", since = "1.48.0")] } +__impl_slice_eq1! { [A: Allocator] [T], Vec<U, A>, #[stable(feature = "partialeq_vec_for_slice", since = "1.48.0")] } +#[cfg(not(no_global_oom_handling))] +__impl_slice_eq1! { [A: Allocator] Cow<'_, [T]>, Vec<U, A> where T: Clone, #[stable(feature = "rust1", since = "1.0.0")] } +#[cfg(not(no_global_oom_handling))] +__impl_slice_eq1! { [] Cow<'_, [T]>, &[U] where T: Clone, #[stable(feature = "rust1", since = "1.0.0")] } +#[cfg(not(no_global_oom_handling))] +__impl_slice_eq1! { [] Cow<'_, [T]>, &mut [U] where T: Clone, #[stable(feature = "rust1", since = "1.0.0")] } +__impl_slice_eq1! { [A: Allocator, const N: usize] Vec<T, A>, [U; N], #[stable(feature = "rust1", since = "1.0.0")] } +__impl_slice_eq1! { [A: Allocator, const N: usize] Vec<T, A>, &[U; N], #[stable(feature = "rust1", since = "1.0.0")] } + +// NOTE: some less important impls are omitted to reduce code bloat +// FIXME(Centril): Reconsider this? +//__impl_slice_eq1! { [const N: usize] Vec<A>, &mut [B; N], } +//__impl_slice_eq1! { [const N: usize] [A; N], Vec<B>, } +//__impl_slice_eq1! { [const N: usize] &[A; N], Vec<B>, } +//__impl_slice_eq1! { [const N: usize] &mut [A; N], Vec<B>, } +//__impl_slice_eq1! { [const N: usize] Cow<'a, [A]>, [B; N], } +//__impl_slice_eq1! { [const N: usize] Cow<'a, [A]>, &[B; N], } +//__impl_slice_eq1! { [const N: usize] Cow<'a, [A]>, &mut [B; N], } diff --git a/rust/bindgen_parameters b/rust/bindgen_parameters new file mode 100644 index 000000000000..be4963bf7203 --- /dev/null +++ b/rust/bindgen_parameters @@ -0,0 +1,21 @@ +# SPDX-License-Identifier: GPL-2.0 + +--opaque-type xregs_state +--opaque-type desc_struct +--opaque-type arch_lbr_state +--opaque-type local_apic + +# Packed type cannot transitively contain a `#[repr(align)]` type. +--opaque-type x86_msi_data +--opaque-type x86_msi_addr_lo + +# `try` is a reserved keyword since Rust 2018; solved in `bindgen` v0.59.2, +# commit 2aed6b021680 ("context: Escape the try keyword properly"). +--opaque-type kunit_try_catch + +# If SMP is disabled, `arch_spinlock_t` is defined as a ZST which triggers a Rust +# warning. We don't need to peek into it anyway. +--opaque-type spinlock + +# `seccomp`'s comment gets understood as a doctest +--no-doc-comments diff --git a/rust/bindings/bindings_helper.h b/rust/bindings/bindings_helper.h new file mode 100644 index 000000000000..c48bc284214a --- /dev/null +++ b/rust/bindings/bindings_helper.h @@ -0,0 +1,13 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Header that contains the code (mostly headers) for which Rust bindings + * will be automatically generated by `bindgen`. + * + * Sorted alphabetically. + */ + +#include <linux/slab.h> + +/* `bindgen` gets confused at certain things. */ +const gfp_t BINDINGS_GFP_KERNEL = GFP_KERNEL; +const gfp_t BINDINGS___GFP_ZERO = __GFP_ZERO; diff --git a/rust/bindings/lib.rs b/rust/bindings/lib.rs new file mode 100644 index 000000000000..6c50ee62c56b --- /dev/null +++ b/rust/bindings/lib.rs @@ -0,0 +1,53 @@ +// SPDX-License-Identifier: GPL-2.0 + +//! Bindings. +//! +//! Imports the generated bindings by `bindgen`. +//! +//! This crate may not be directly used. If you need a kernel C API that is +//! not ported or wrapped in the `kernel` crate, then do so first instead of +//! using this crate. + +#![no_std] +#![feature(core_ffi_c)] +// See <https://github.com/rust-lang/rust-bindgen/issues/1651>. +#![cfg_attr(test, allow(deref_nullptr))] +#![cfg_attr(test, allow(unaligned_references))] +#![cfg_attr(test, allow(unsafe_op_in_unsafe_fn))] +#![allow( + clippy::all, + missing_docs, + non_camel_case_types, + non_upper_case_globals, + non_snake_case, + improper_ctypes, + unreachable_pub, + unsafe_op_in_unsafe_fn +)] + +mod bindings_raw { + // Use glob import here to expose all helpers. + // Symbols defined within the module will take precedence to the glob import. + pub use super::bindings_helper::*; + include!(concat!( + env!("OBJTREE"), + "/rust/bindings/bindings_generated.rs" + )); +} + +// When both a directly exposed symbol and a helper exists for the same function, +// the directly exposed symbol is preferred and the helper becomes dead code, so +// ignore the warning here. +#[allow(dead_code)] +mod bindings_helper { + // Import the generated bindings for types. + include!(concat!( + env!("OBJTREE"), + "/rust/bindings/bindings_helpers_generated.rs" + )); +} + +pub use bindings_raw::*; + +pub const GFP_KERNEL: gfp_t = BINDINGS_GFP_KERNEL; +pub const __GFP_ZERO: gfp_t = BINDINGS___GFP_ZERO; diff --git a/rust/compiler_builtins.rs b/rust/compiler_builtins.rs new file mode 100644 index 000000000000..f8f39a3e6855 --- /dev/null +++ b/rust/compiler_builtins.rs @@ -0,0 +1,63 @@ +// SPDX-License-Identifier: GPL-2.0 + +//! Our own `compiler_builtins`. +//! +//! Rust provides [`compiler_builtins`] as a port of LLVM's [`compiler-rt`]. +//! Since we do not need the vast majority of them, we avoid the dependency +//! by providing this file. +//! +//! At the moment, some builtins are required that should not be. For instance, +//! [`core`] has 128-bit integers functionality which we should not be compiling +//! in. We will work with upstream [`core`] to provide feature flags to disable +//! the parts we do not need. For the moment, we define them to [`panic!`] at +//! runtime for simplicity to catch mistakes, instead of performing surgery +//! on `core.o`. +//! +//! In any case, all these symbols are weakened to ensure we do not override +//! those that may be provided by the rest of the kernel. +//! +//! [`compiler_builtins`]: https://github.com/rust-lang/compiler-builtins +//! [`compiler-rt`]: https://compiler-rt.llvm.org/ + +#![feature(compiler_builtins)] +#![compiler_builtins] +#![no_builtins] +#![no_std] + +macro_rules! define_panicking_intrinsics( + ($reason: tt, { $($ident: ident, )* }) => { + $( + #[doc(hidden)] + #[no_mangle] + pub extern "C" fn $ident() { + panic!($reason); + } + )* + } +); + +define_panicking_intrinsics!("`f32` should not be used", { + __eqsf2, + __gesf2, + __lesf2, + __nesf2, + __unordsf2, +}); + +define_panicking_intrinsics!("`f64` should not be used", { + __unorddf2, +}); + +define_panicking_intrinsics!("`i128` should not be used", { + __ashrti3, + __muloti4, + __multi3, +}); + +define_panicking_intrinsics!("`u128` should not be used", { + __ashlti3, + __lshrti3, + __udivmodti4, + __udivti3, + __umodti3, +}); diff --git a/rust/exports.c b/rust/exports.c new file mode 100644 index 000000000000..bb7cc64cecd0 --- /dev/null +++ b/rust/exports.c @@ -0,0 +1,21 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * A hack to export Rust symbols for loadable modules without having to redo + * the entire `include/linux/export.h` logic in Rust. + * + * This requires the Rust's new/future `v0` mangling scheme because the default + * one ("legacy") uses invalid characters for C identifiers (thus we cannot use + * the `EXPORT_SYMBOL_*` macros). + * + * All symbols are exported as GPL-only to guarantee no GPL-only feature is + * accidentally exposed. + */ + +#include <linux/module.h> + +#define EXPORT_SYMBOL_RUST_GPL(sym) extern int sym; EXPORT_SYMBOL_GPL(sym) + +#include "exports_core_generated.h" +#include "exports_alloc_generated.h" +#include "exports_bindings_generated.h" +#include "exports_kernel_generated.h" diff --git a/rust/helpers.c b/rust/helpers.c new file mode 100644 index 000000000000..b4f15eee2ffd --- /dev/null +++ b/rust/helpers.c @@ -0,0 +1,51 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Non-trivial C macros cannot be used in Rust. Similarly, inlined C functions + * cannot be called either. This file explicitly creates functions ("helpers") + * that wrap those so that they can be called from Rust. + * + * Even though Rust kernel modules should never use directly the bindings, some + * of these helpers need to be exported because Rust generics and inlined + * functions may not get their code generated in the crate where they are + * defined. Other helpers, called from non-inline functions, may not be + * exported, in principle. However, in general, the Rust compiler does not + * guarantee codegen will be performed for a non-inline function either. + * Therefore, this file exports all the helpers. In the future, this may be + * revisited to reduce the number of exports after the compiler is informed + * about the places codegen is required. + * + * All symbols are exported as GPL-only to guarantee no GPL-only feature is + * accidentally exposed. + */ + +#include <linux/bug.h> +#include <linux/build_bug.h> + +__noreturn void rust_helper_BUG(void) +{ + BUG(); +} +EXPORT_SYMBOL_GPL(rust_helper_BUG); + +/* + * We use `bindgen`'s `--size_t-is-usize` option to bind the C `size_t` type + * as the Rust `usize` type, so we can use it in contexts where Rust + * expects a `usize` like slice (array) indices. `usize` is defined to be + * the same as C's `uintptr_t` type (can hold any pointer) but not + * necessarily the same as `size_t` (can hold the size of any single + * object). Most modern platforms use the same concrete integer type for + * both of them, but in case we find ourselves on a platform where + * that's not true, fail early instead of risking ABI or + * integer-overflow issues. + * + * If your platform fails this assertion, it means that you are in + * danger of integer-overflow bugs (even if you attempt to remove + * `--size_t-is-usize`). It may be easiest to change the kernel ABI on + * your platform such that `size_t` matches `uintptr_t` (i.e., to increase + * `size_t`, because `uintptr_t` has to be at least as big as `size_t`). + */ +static_assert( + sizeof(size_t) == sizeof(uintptr_t) && + __alignof__(size_t) == __alignof__(uintptr_t), + "Rust code expects C `size_t` to match Rust `usize`" +); diff --git a/rust/kernel/allocator.rs b/rust/kernel/allocator.rs new file mode 100644 index 000000000000..397a3dd57a9b --- /dev/null +++ b/rust/kernel/allocator.rs @@ -0,0 +1,64 @@ +// SPDX-License-Identifier: GPL-2.0 + +//! Allocator support. + +use core::alloc::{GlobalAlloc, Layout}; +use core::ptr; + +use crate::bindings; + +struct KernelAllocator; + +unsafe impl GlobalAlloc for KernelAllocator { + unsafe fn alloc(&self, layout: Layout) -> *mut u8 { + // `krealloc()` is used instead of `kmalloc()` because the latter is + // an inline function and cannot be bound to as a result. + unsafe { bindings::krealloc(ptr::null(), layout.size(), bindings::GFP_KERNEL) as *mut u8 } + } + + unsafe fn dealloc(&self, ptr: *mut u8, _layout: Layout) { + unsafe { + bindings::kfree(ptr as *const core::ffi::c_void); + } + } +} + +#[global_allocator] +static ALLOCATOR: KernelAllocator = KernelAllocator; + +// `rustc` only generates these for some crate types. Even then, we would need +// to extract the object file that has them from the archive. For the moment, +// let's generate them ourselves instead. +// +// Note that `#[no_mangle]` implies exported too, nowadays. +#[no_mangle] +fn __rust_alloc(size: usize, _align: usize) -> *mut u8 { + unsafe { bindings::krealloc(core::ptr::null(), size, bindings::GFP_KERNEL) as *mut u8 } +} + +#[no_mangle] +fn __rust_dealloc(ptr: *mut u8, _size: usize, _align: usize) { + unsafe { bindings::kfree(ptr as *const core::ffi::c_void) }; +} + +#[no_mangle] +fn __rust_realloc(ptr: *mut u8, _old_size: usize, _align: usize, new_size: usize) -> *mut u8 { + unsafe { + bindings::krealloc( + ptr as *const core::ffi::c_void, + new_size, + bindings::GFP_KERNEL, + ) as *mut u8 + } +} + +#[no_mangle] +fn __rust_alloc_zeroed(size: usize, _align: usize) -> *mut u8 { + unsafe { + bindings::krealloc( + core::ptr::null(), + size, + bindings::GFP_KERNEL | bindings::__GFP_ZERO, + ) as *mut u8 + } +} diff --git a/rust/kernel/error.rs b/rust/kernel/error.rs new file mode 100644 index 000000000000..466b2a8fe569 --- /dev/null +++ b/rust/kernel/error.rs @@ -0,0 +1,59 @@ +// SPDX-License-Identifier: GPL-2.0 + +//! Kernel errors. +//! +//! C header: [`include/uapi/asm-generic/errno-base.h`](../../../include/uapi/asm-generic/errno-base.h) + +use alloc::collections::TryReserveError; + +/// Contains the C-compatible error codes. +pub mod code { + /// Out of memory. + pub const ENOMEM: super::Error = super::Error(-(crate::bindings::ENOMEM as i32)); +} + +/// Generic integer kernel error. +/// +/// The kernel defines a set of integer generic error codes based on C and +/// POSIX ones. These codes may have a more specific meaning in some contexts. +/// +/// # Invariants +/// +/// The value is a valid `errno` (i.e. `>= -MAX_ERRNO && < 0`). +#[derive(Clone, Copy, PartialEq, Eq)] +pub struct Error(core::ffi::c_int); + +impl Error { + /// Returns the kernel error code. + pub fn to_kernel_errno(self) -> core::ffi::c_int { + self.0 + } +} + +impl From<TryReserveError> for Error { + fn from(_: TryReserveError) -> Error { + code::ENOMEM + } +} + +/// A [`Result`] with an [`Error`] error type. +/// +/// To be used as the return type for functions that may fail. +/// +/// # Error codes in C and Rust +/// +/// In C, it is common that functions indicate success or failure through +/// their return value; modifying or returning extra data through non-`const` +/// pointer parameters. In particular, in the kernel, functions that may fail +/// typically return an `int` that represents a generic error code. We model +/// those as [`Error`]. +/// +/// In Rust, it is idiomatic to model functions that may fail as returning +/// a [`Result`]. Since in the kernel many functions return an error code, +/// [`Result`] is a type alias for a [`core::result::Result`] that uses +/// [`Error`] as its error type. +/// +/// Note that even if a function does not return anything when it succeeds, +/// it should still be modeled as returning a `Result` rather than +/// just an [`Error`]. +pub type Result<T = ()> = core::result::Result<T, Error>; diff --git a/rust/kernel/lib.rs b/rust/kernel/lib.rs new file mode 100644 index 000000000000..abd46261d385 --- /dev/null +++ b/rust/kernel/lib.rs @@ -0,0 +1,78 @@ +// SPDX-License-Identifier: GPL-2.0 + +//! The `kernel` crate. +//! +//! This crate contains the kernel APIs that have been ported or wrapped for +//! usage by Rust code in the kernel and is shared by all of them. +//! +//! In other words, all the rest of the Rust code in the kernel (e.g. kernel +//! modules written in Rust) depends on [`core`], [`alloc`] and this crate. +//! +//! If you need a kernel C API that is not ported or wrapped yet here, then +//! do so first instead of bypassing this crate. + +#![no_std] +#![feature(core_ffi_c)] + +// Ensure conditional compilation based on the kernel configuration works; +// otherwise we may silently break things like initcall handling. +#[cfg(not(CONFIG_RUST))] +compile_error!("Missing kernel configuration for conditional compilation"); + +#[cfg(not(test))] +#[cfg(not(testlib))] +mod allocator; +pub mod error; +pub mod prelude; +pub mod print; +pub mod str; + +#[doc(hidden)] +pub use bindings; +pub use macros; + +/// Prefix to appear before log messages printed from within the `kernel` crate. +const __LOG_PREFIX: &[u8] = b"rust_kernel\0"; + +/// The top level entrypoint to implementing a kernel module. +/// +/// For any teardown or cleanup operations, your type may implement [`Drop`]. +pub trait Module: Sized + Sync { + /// Called at module initialization time. + /// + /// Use this method to perform whatever setup or registration your module + /// should do. + /// + /// Equivalent to the `module_init` macro in the C API. + fn init(module: &'static ThisModule) -> error::Result<Self>; +} + +/// Equivalent to `THIS_MODULE` in the C API. +/// +/// C header: `include/linux/export.h` +pub struct ThisModule(*mut bindings::module); + +// SAFETY: `THIS_MODULE` may be used from all threads within a module. +unsafe impl Sync for ThisModule {} + +impl ThisModule { + /// Creates a [`ThisModule`] given the `THIS_MODULE` pointer. + /// + /// # Safety + /// + /// The pointer must be equal to the right `THIS_MODULE`. + pub const unsafe fn from_ptr(ptr: *mut bindings::module) -> ThisModule { + ThisModule(ptr) + } +} + +#[cfg(not(any(testlib, test)))] +#[panic_handler] +fn panic(info: &core::panic::PanicInfo<'_>) -> ! { + pr_emerg!("{}\n", info); + // SAFETY: FFI call. + unsafe { bindings::BUG() }; + // Bindgen currently does not recognize `__noreturn` so `BUG` returns `()` + // instead of `!`. See <https://github.com/rust-lang/rust-bindgen/issues/2094>. + loop {} +} diff --git a/rust/kernel/prelude.rs b/rust/kernel/prelude.rs new file mode 100644 index 000000000000..495e22250726 --- /dev/null +++ b/rust/kernel/prelude.rs @@ -0,0 +1,20 @@ +// SPDX-License-Identifier: GPL-2.0 + +//! The `kernel` prelude. +//! +//! These are the most common items used by Rust code in the kernel, +//! intended to be imported by all Rust code, for convenience. +//! +//! # Examples +//! +//! ``` +//! use kernel::prelude::*; +//! ``` + +pub use super::{ + error::{Error, Result}, + pr_emerg, pr_info, ThisModule, +}; +pub use alloc::{boxed::Box, vec::Vec}; +pub use core::pin::Pin; +pub use macros::module; diff --git a/rust/kernel/print.rs b/rust/kernel/print.rs new file mode 100644 index 000000000000..55db5a1ba752 --- /dev/null +++ b/rust/kernel/print.rs @@ -0,0 +1,198 @@ +// SPDX-License-Identifier: GPL-2.0 + +//! Printing facilities. +//! +//! C header: [`include/linux/printk.h`](../../../../include/linux/printk.h) +//! +//! Reference: <https://www.kernel.org/doc/html/latest/core-api/printk-basics.html> + +use core::{ + ffi::{c_char, c_void}, + fmt, +}; + +use crate::str::RawFormatter; + +#[cfg(CONFIG_PRINTK)] +use crate::bindings; + +// Called from `vsprintf` with format specifier `%pA`. +#[no_mangle] +unsafe fn rust_fmt_argument(buf: *mut c_char, end: *mut c_char, ptr: *const c_void) -> *mut c_char { + use fmt::Write; + // SAFETY: The C contract guarantees that `buf` is valid if it's less than `end`. + let mut w = unsafe { RawFormatter::from_ptrs(buf.cast(), end.cast()) }; + let _ = w.write_fmt(unsafe { *(ptr as *const fmt::Arguments<'_>) }); + w.pos().cast() +} + +/// Format strings. +/// +/// Public but hidden since it should only be used from public macros. +#[doc(hidden)] +pub mod format_strings { + use crate::bindings; + + /// The length we copy from the `KERN_*` kernel prefixes. + const LENGTH_PREFIX: usize = 2; + + /// The length of the fixed format strings. + pub const LENGTH: usize = 10; + + /// Generates a fixed format string for the kernel's [`_printk`]. + /// + /// The format string is always the same for a given level, i.e. for a + /// given `prefix`, which are the kernel's `KERN_*` constants. + /// + /// [`_printk`]: ../../../../include/linux/printk.h + const fn generate(is_cont: bool, prefix: &[u8; 3]) -> [u8; LENGTH] { + // Ensure the `KERN_*` macros are what we expect. + assert!(prefix[0] == b'\x01'); + if is_cont { + assert!(prefix[1] == b'c'); + } else { + assert!(prefix[1] >= b'0' && prefix[1] <= b'7'); + } + assert!(prefix[2] == b'\x00'); + + let suffix: &[u8; LENGTH - LENGTH_PREFIX] = if is_cont { + b"%pA\0\0\0\0\0" + } else { + b"%s: %pA\0" + }; + + [ + prefix[0], prefix[1], suffix[0], suffix[1], suffix[2], suffix[3], suffix[4], suffix[5], + suffix[6], suffix[7], + ] + } + + // Generate the format strings at compile-time. + // + // This avoids the compiler generating the contents on the fly in the stack. + // + // Furthermore, `static` instead of `const` is used to share the strings + // for all the kernel. + pub static EMERG: [u8; LENGTH] = generate(false, bindings::KERN_EMERG); + pub static INFO: [u8; LENGTH] = generate(false, bindings::KERN_INFO); +} + +/// Prints a message via the kernel's [`_printk`]. +/// +/// Public but hidden since it should only be used from public macros. +/// +/// # Safety +/// +/// The format string must be one of the ones in [`format_strings`], and +/// the module name must be null-terminated. +/// +/// [`_printk`]: ../../../../include/linux/_printk.h +#[doc(hidden)] +#[cfg_attr(not(CONFIG_PRINTK), allow(unused_variables))] +pub unsafe fn call_printk( + format_string: &[u8; format_strings::LENGTH], + module_name: &[u8], + args: fmt::Arguments<'_>, +) { + // `_printk` does not seem to fail in any path. + #[cfg(CONFIG_PRINTK)] + unsafe { + bindings::_printk( + format_string.as_ptr() as _, + module_name.as_ptr(), + &args as *const _ as *const c_void, + ); + } +} + +/// Performs formatting and forwards the string to [`call_printk`]. +/// +/// Public but hidden since it should only be used from public macros. +#[doc(hidden)] +#[cfg(not(testlib))] +#[macro_export] +#[allow(clippy::crate_in_macro_def)] +macro_rules! print_macro ( + // The non-continuation cases (most of them, e.g. `INFO`). + ($format_string:path, $($arg:tt)+) => ( + // SAFETY: This hidden macro should only be called by the documented + // printing macros which ensure the format string is one of the fixed + // ones. All `__LOG_PREFIX`s are null-terminated as they are generated + // by the `module!` proc macro or fixed values defined in a kernel + // crate. + unsafe { + $crate::print::call_printk( + &$format_string, + crate::__LOG_PREFIX, + format_args!($($arg)+), + ); + } + ); +); + +/// Stub for doctests +#[cfg(testlib)] +#[macro_export] +macro_rules! print_macro ( + ($format_string:path, $e:expr, $($arg:tt)+) => ( + () + ); +); + +// We could use a macro to generate these macros. However, doing so ends +// up being a bit ugly: it requires the dollar token trick to escape `$` as +// well as playing with the `doc` attribute. Furthermore, they cannot be easily +// imported in the prelude due to [1]. So, for the moment, we just write them +// manually, like in the C side; while keeping most of the logic in another +// macro, i.e. [`print_macro`]. +// +// [1]: https://github.com/rust-lang/rust/issues/52234 + +/// Prints an emergency-level message (level 0). +/// +/// Use this level if the system is unusable. +/// +/// Equivalent to the kernel's [`pr_emerg`] macro. +/// +/// Mimics the interface of [`std::print!`]. See [`core::fmt`] and +/// `alloc::format!` for information about the formatting syntax. +/// +/// [`pr_emerg`]: https://www.kernel.org/doc/html/latest/core-api/printk-basics.html#c.pr_emerg +/// [`std::print!`]: https://doc.rust-lang.org/std/macro.print.html +/// +/// # Examples +/// +/// ``` +/// pr_emerg!("hello {}\n", "there"); +/// ``` +#[macro_export] +macro_rules! pr_emerg ( + ($($arg:tt)*) => ( + $crate::print_macro!($crate::print::format_strings::EMERG, $($arg)*) + ) +); + +/// Prints an info-level message (level 6). +/// +/// Use this level for informational messages. +/// +/// Equivalent to the kernel's [`pr_info`] macro. +/// +/// Mimics the interface of [`std::print!`]. See [`core::fmt`] and +/// `alloc::format!` for information about the formatting syntax. +/// +/// [`pr_info`]: https://www.kernel.org/doc/html/latest/core-api/printk-basics.html#c.pr_info +/// [`std::print!`]: https://doc.rust-lang.org/std/macro.print.html +/// +/// # Examples +/// +/// ``` +/// pr_info!("hello {}\n", "there"); +/// ``` +#[macro_export] +#[doc(alias = "print")] +macro_rules! pr_info ( + ($($arg:tt)*) => ( + $crate::print_macro!($crate::print::format_strings::INFO, $($arg)*) + ) +); diff --git a/rust/kernel/str.rs b/rust/kernel/str.rs new file mode 100644 index 000000000000..e45ff220ae50 --- /dev/null +++ b/rust/kernel/str.rs @@ -0,0 +1,72 @@ +// SPDX-License-Identifier: GPL-2.0 + +//! String representations. + +use core::fmt; + +/// Allows formatting of [`fmt::Arguments`] into a raw buffer. +/// +/// It does not fail if callers write past the end of the buffer so that they can calculate the +/// size required to fit everything. +/// +/// # Invariants +/// +/// The memory region between `pos` (inclusive) and `end` (exclusive) is valid for writes if `pos` +/// is less than `end`. +pub(crate) struct RawFormatter { + // Use `usize` to use `saturating_*` functions. + #[allow(dead_code)] + beg: usize, + pos: usize, + end: usize, +} + +impl RawFormatter { + /// Creates a new instance of [`RawFormatter`] with the given buffer pointers. + /// + /// # Safety + /// + /// If `pos` is less than `end`, then the region between `pos` (inclusive) and `end` + /// (exclusive) must be valid for writes for the lifetime of the returned [`RawFormatter`]. + pub(crate) unsafe fn from_ptrs(pos: *mut u8, end: *mut u8) -> Self { + // INVARIANT: The safety requierments guarantee the type invariants. + Self { + beg: pos as _, + pos: pos as _, + end: end as _, + } + } + + /// Returns the current insert position. + /// + /// N.B. It may point to invalid memory. + pub(crate) fn pos(&self) -> *mut u8 { + self.pos as _ + } +} + +impl fmt::Write for RawFormatter { + fn write_str(&mut self, s: &str) -> fmt::Result { + // `pos` value after writing `len` bytes. This does not have to be bounded by `end`, but we + // don't want it to wrap around to 0. + let pos_new = self.pos.saturating_add(s.len()); + + // Amount that we can copy. `saturating_sub` ensures we get 0 if `pos` goes past `end`. + let len_to_copy = core::cmp::min(pos_new, self.end).saturating_sub(self.pos); + + if len_to_copy > 0 { + // SAFETY: If `len_to_copy` is non-zero, then we know `pos` has not gone past `end` + // yet, so it is valid for write per the type invariants. + unsafe { + core::ptr::copy_nonoverlapping( + s.as_bytes().as_ptr(), + self.pos as *mut u8, + len_to_copy, + ) + }; + } + + self.pos = pos_new; + Ok(()) + } +} diff --git a/rust/macros/helpers.rs b/rust/macros/helpers.rs new file mode 100644 index 000000000000..cdc7dc6135d2 --- /dev/null +++ b/rust/macros/helpers.rs @@ -0,0 +1,51 @@ +// SPDX-License-Identifier: GPL-2.0 + +use proc_macro::{token_stream, TokenTree}; + +pub(crate) fn try_ident(it: &mut token_stream::IntoIter) -> Option<String> { + if let Some(TokenTree::Ident(ident)) = it.next() { + Some(ident.to_string()) + } else { + None + } +} + +pub(crate) fn try_literal(it: &mut token_stream::IntoIter) -> Option<String> { + if let Some(TokenTree::Literal(literal)) = it.next() { + Some(literal.to_string()) + } else { + None + } +} + +pub(crate) fn try_byte_string(it: &mut token_stream::IntoIter) -> Option<String> { + try_literal(it).and_then(|byte_string| { + if byte_string.starts_with("b\"") && byte_string.ends_with('\"') { + Some(byte_string[2..byte_string.len() - 1].to_string()) + } else { + None + } + }) +} + +pub(crate) fn expect_ident(it: &mut token_stream::IntoIter) -> String { + try_ident(it).expect("Expected Ident") +} + +pub(crate) fn expect_punct(it: &mut token_stream::IntoIter) -> char { + if let TokenTree::Punct(punct) = it.next().expect("Reached end of token stream for Punct") { + punct.as_char() + } else { + panic!("Expected Punct"); + } +} + +pub(crate) fn expect_byte_string(it: &mut token_stream::IntoIter) -> String { + try_byte_string(it).expect("Expected byte string") +} + +pub(crate) fn expect_end(it: &mut token_stream::IntoIter) { + if it.next().is_some() { + panic!("Expected end"); + } +} diff --git a/rust/macros/lib.rs b/rust/macros/lib.rs new file mode 100644 index 000000000000..91764bfb1f89 --- /dev/null +++ b/rust/macros/lib.rs @@ -0,0 +1,72 @@ +// SPDX-License-Identifier: GPL-2.0 + +//! Crate for all kernel procedural macros. + +mod helpers; +mod module; + +use proc_macro::TokenStream; + +/// Declares a kernel module. +/// +/// The `type` argument should be a type which implements the [`Module`] +/// trait. Also accepts various forms of kernel metadata. +/// +/// C header: [`include/linux/moduleparam.h`](../../../include/linux/moduleparam.h) +/// +/// [`Module`]: ../kernel/trait.Module.html +/// +/// # Examples +/// +/// ```ignore +/// use kernel::prelude::*; +/// +/// module!{ +/// type: MyModule, +/// name: b"my_kernel_module", +/// author: b"Rust for Linux Contributors", +/// description: b"My very own kernel module!", +/// license: b"GPL", +/// params: { +/// my_i32: i32 { +/// default: 42, +/// permissions: 0o000, +/// description: b"Example of i32", +/// }, +/// writeable_i32: i32 { +/// default: 42, +/// permissions: 0o644, +/// description: b"Example of i32", +/// }, +/// }, +/// } +/// +/// struct MyModule; +/// +/// impl kernel::Module for MyModule { +/// fn init() -> Result<Self> { +/// // If the parameter is writeable, then the kparam lock must be +/// // taken to read the parameter: +/// { +/// let lock = THIS_MODULE.kernel_param_lock(); +/// pr_info!("i32 param is: {}\n", writeable_i32.read(&lock)); +/// } +/// // If the parameter is read only, it can be read without locking +/// // the kernel parameters: +/// pr_info!("i32 param is: {}\n", my_i32.read()); +/// Ok(Self) +/// } +/// } +/// ``` +/// +/// # Supported argument types +/// - `type`: type which implements the [`Module`] trait (required). +/// - `name`: byte array of the name of the kernel module (required). +/// - `author`: byte array of the author of the kernel module. +/// - `description`: byte array of the description of the kernel module. +/// - `license`: byte array of the license of the kernel module (required). +/// - `alias`: byte array of alias name of the kernel module. +#[proc_macro] +pub fn module(ts: TokenStream) -> TokenStream { + module::module(ts) +} diff --git a/rust/macros/module.rs b/rust/macros/module.rs new file mode 100644 index 000000000000..186a5b8be23c --- /dev/null +++ b/rust/macros/module.rs @@ -0,0 +1,282 @@ +// SPDX-License-Identifier: GPL-2.0 + +use crate::helpers::*; +use proc_macro::{token_stream, Literal, TokenStream, TokenTree}; +use std::fmt::Write; + +struct ModInfoBuilder<'a> { + module: &'a str, + counter: usize, + buffer: String, +} + +impl<'a> ModInfoBuilder<'a> { + fn new(module: &'a str) -> Self { + ModInfoBuilder { + module, + counter: 0, + buffer: String::new(), + } + } + + fn emit_base(&mut self, field: &str, content: &str, builtin: bool) { + let string = if builtin { + // Built-in modules prefix their modinfo strings by `module.`. + format!( + "{module}.{field}={content}\0", + module = self.module, + field = field, + content = content + ) + } else { + // Loadable modules' modinfo strings go as-is. + format!("{field}={content}\0", field = field, content = content) + }; + + write!( + &mut self.buffer, + " + {cfg} + #[doc(hidden)] + #[link_section = \".modinfo\"] + #[used] + pub static __{module}_{counter}: [u8; {length}] = *{string}; + ", + cfg = if builtin { + "#[cfg(not(MODULE))]" + } else { + "#[cfg(MODULE)]" + }, + module = self.module.to_uppercase(), + counter = self.counter, + length = string.len(), + string = Literal::byte_string(string.as_bytes()), + ) + .unwrap(); + + self.counter += 1; + } + + fn emit_only_builtin(&mut self, field: &str, content: &str) { + self.emit_base(field, content, true) + } + + fn emit_only_loadable(&mut self, field: &str, content: &str) { + self.emit_base(field, content, false) + } + + fn emit(&mut self, field: &str, content: &str) { + self.emit_only_builtin(field, content); + self.emit_only_loadable(field, content); + } +} + +#[derive(Debug, Default)] +struct ModuleInfo { + type_: String, + license: String, + name: String, + author: Option<String>, + description: Option<String>, + alias: Option<String>, +} + +impl ModuleInfo { + fn parse(it: &mut token_stream::IntoIter) -> Self { + let mut info = ModuleInfo::default(); + + const EXPECTED_KEYS: &[&str] = + &["type", "name", "author", "description", "license", "alias"]; + const REQUIRED_KEYS: &[&str] = &["type", "name", "license"]; + let mut seen_keys = Vec::new(); + + loop { + let key = match it.next() { + Some(TokenTree::Ident(ident)) => ident.to_string(), + Some(_) => panic!("Expected Ident or end"), + None => break, + }; + + if seen_keys.contains(&key) { + panic!( + "Duplicated key \"{}\". Keys can only be specified once.", + key + ); + } + + assert_eq!(expect_punct(it), ':'); + + match key.as_str() { + "type" => info.type_ = expect_ident(it), + "name" => info.name = expect_byte_string(it), + "author" => info.author = Some(expect_byte_string(it)), + "description" => info.description = Some(expect_byte_string(it)), + "license" => info.license = expect_byte_string(it), + "alias" => info.alias = Some(expect_byte_string(it)), + _ => panic!( + "Unknown key \"{}\". Valid keys are: {:?}.", + key, EXPECTED_KEYS + ), + } + + assert_eq!(expect_punct(it), ','); + + seen_keys.push(key); + } + + expect_end(it); + + for key in REQUIRED_KEYS { + if !seen_keys.iter().any(|e| e == key) { + panic!("Missing required key \"{}\".", key); + } + } + + let mut ordered_keys: Vec<&str> = Vec::new(); + for key in EXPECTED_KEYS { + if seen_keys.iter().any(|e| e == key) { + ordered_keys.push(key); + } + } + + if seen_keys != ordered_keys { + panic!( + "Keys are not ordered as expected. Order them like: {:?}.", + ordered_keys + ); + } + + info + } +} + +pub(crate) fn module(ts: TokenStream) -> TokenStream { + let mut it = ts.into_iter(); + + let info = ModuleInfo::parse(&mut it); + + let mut modinfo = ModInfoBuilder::new(info.name.as_ref()); + if let Some(author) = info.author { + modinfo.emit("author", &author); + } + if let Some(description) = info.description { + modinfo.emit("description", &description); + } + modinfo.emit("license", &info.license); + if let Some(alias) = info.alias { + modinfo.emit("alias", &alias); + } + + // Built-in modules also export the `file` modinfo string. + let file = + std::env::var("RUST_MODFILE").expect("Unable to fetch RUST_MODFILE environmental variable"); + modinfo.emit_only_builtin("file", &file); + + format!( + " + /// The module name. + /// + /// Used by the printing macros, e.g. [`info!`]. + const __LOG_PREFIX: &[u8] = b\"{name}\\0\"; + + /// The \"Rust loadable module\" mark, for `scripts/is_rust_module.sh`. + // + // This may be best done another way later on, e.g. as a new modinfo + // key or a new section. For the moment, keep it simple. + #[cfg(MODULE)] + #[doc(hidden)] + #[used] + static __IS_RUST_MODULE: () = (); + + static mut __MOD: Option<{type_}> = None; + + // SAFETY: `__this_module` is constructed by the kernel at load time and will not be + // freed until the module is unloaded. + #[cfg(MODULE)] + static THIS_MODULE: kernel::ThisModule = unsafe {{ + kernel::ThisModule::from_ptr(&kernel::bindings::__this_module as *const _ as *mut _) + }}; + #[cfg(not(MODULE))] + static THIS_MODULE: kernel::ThisModule = unsafe {{ + kernel::ThisModule::from_ptr(core::ptr::null_mut()) + }}; + + // Loadable modules need to export the `{{init,cleanup}}_module` identifiers. + #[cfg(MODULE)] + #[doc(hidden)] + #[no_mangle] + pub extern \"C\" fn init_module() -> core::ffi::c_int {{ + __init() + }} + + #[cfg(MODULE)] + #[doc(hidden)] + #[no_mangle] + pub extern \"C\" fn cleanup_module() {{ + __exit() + }} + + // Built-in modules are initialized through an initcall pointer + // and the identifiers need to be unique. + #[cfg(not(MODULE))] + #[cfg(not(CONFIG_HAVE_ARCH_PREL32_RELOCATIONS))] + #[doc(hidden)] + #[link_section = \"{initcall_section}\"] + #[used] + pub static __{name}_initcall: extern \"C\" fn() -> core::ffi::c_int = __{name}_init; + + #[cfg(not(MODULE))] + #[cfg(CONFIG_HAVE_ARCH_PREL32_RELOCATIONS)] + core::arch::global_asm!( + r#\".section \"{initcall_section}\", \"a\" + __{name}_initcall: + .long __{name}_init - . + .previous + \"# + ); + + #[cfg(not(MODULE))] + #[doc(hidden)] + #[no_mangle] + pub extern \"C\" fn __{name}_init() -> core::ffi::c_int {{ + __init() + }} + + #[cfg(not(MODULE))] + #[doc(hidden)] + #[no_mangle] + pub extern \"C\" fn __{name}_exit() {{ + __exit() + }} + + fn __init() -> core::ffi::c_int {{ + match <{type_} as kernel::Module>::init(&THIS_MODULE) {{ + Ok(m) => {{ + unsafe {{ + __MOD = Some(m); + }} + return 0; + }} + Err(e) => {{ + return e.to_kernel_errno(); + }} + }} + }} + + fn __exit() {{ + unsafe {{ + // Invokes `drop()` on `__MOD`, which should be used for cleanup. + __MOD = None; + }} + }} + + {modinfo} + ", + type_ = info.type_, + name = info.name, + modinfo = modinfo.buffer, + initcall_section = ".initcall6.init" + ) + .parse() + .expect("Error parsing formatted string into token stream.") +} diff --git a/samples/Kconfig b/samples/Kconfig index 470ee3baf2e1..0d81c00289ee 100644 --- a/samples/Kconfig +++ b/samples/Kconfig @@ -263,6 +263,8 @@ config SAMPLE_CORESIGHT_SYSCFG This demonstrates how a user may create their own CoreSight configurations and easily load them into the system at runtime. +source "samples/rust/Kconfig" + endif # SAMPLES config HAVE_SAMPLE_FTRACE_DIRECT diff --git a/samples/Makefile b/samples/Makefile index 701e912ab5af..9832ef3f8fcb 100644 --- a/samples/Makefile +++ b/samples/Makefile @@ -35,3 +35,4 @@ subdir-$(CONFIG_SAMPLE_WATCH_QUEUE) += watch_queue obj-$(CONFIG_DEBUG_KMEMLEAK_TEST) += kmemleak/ obj-$(CONFIG_SAMPLE_CORESIGHT_SYSCFG) += coresight/ obj-$(CONFIG_SAMPLE_FPROBE) += fprobe/ +obj-$(CONFIG_SAMPLES_RUST) += rust/ diff --git a/samples/rust/Kconfig b/samples/rust/Kconfig new file mode 100644 index 000000000000..841e0906e943 --- /dev/null +++ b/samples/rust/Kconfig @@ -0,0 +1,30 @@ +# SPDX-License-Identifier: GPL-2.0 + +menuconfig SAMPLES_RUST + bool "Rust samples" + depends on RUST + help + You can build sample Rust kernel code here. + + If unsure, say N. + +if SAMPLES_RUST + +config SAMPLE_RUST_MINIMAL + tristate "Minimal" + help + This option builds the Rust minimal module sample. + + To compile this as a module, choose M here: + the module will be called rust_minimal. + + If unsure, say N. + +config SAMPLE_RUST_HOSTPROGS + bool "Host programs" + help + This option builds the Rust host program samples. + + If unsure, say N. + +endif # SAMPLES_RUST diff --git a/samples/rust/Makefile b/samples/rust/Makefile new file mode 100644 index 000000000000..1daba5f8658a --- /dev/null +++ b/samples/rust/Makefile @@ -0,0 +1,5 @@ +# SPDX-License-Identifier: GPL-2.0 + +obj-$(CONFIG_SAMPLE_RUST_MINIMAL) += rust_minimal.o + +subdir-$(CONFIG_SAMPLE_RUST_HOSTPROGS) += hostprogs diff --git a/samples/rust/hostprogs/.gitignore b/samples/rust/hostprogs/.gitignore new file mode 100644 index 000000000000..a6c173da5048 --- /dev/null +++ b/samples/rust/hostprogs/.gitignore @@ -0,0 +1,3 @@ +# SPDX-License-Identifier: GPL-2.0 + +single diff --git a/samples/rust/hostprogs/Makefile b/samples/rust/hostprogs/Makefile new file mode 100644 index 000000000000..8ddcbd7416db --- /dev/null +++ b/samples/rust/hostprogs/Makefile @@ -0,0 +1,5 @@ +# SPDX-License-Identifier: GPL-2.0 + +hostprogs-always-y := single + +single-rust := y diff --git a/samples/rust/hostprogs/a.rs b/samples/rust/hostprogs/a.rs new file mode 100644 index 000000000000..f7a4a3d0f4e0 --- /dev/null +++ b/samples/rust/hostprogs/a.rs @@ -0,0 +1,7 @@ +// SPDX-License-Identifier: GPL-2.0 + +//! Rust single host program sample: module `a`. + +pub(crate) fn f(x: i32) { + println!("The number is {}.", x); +} diff --git a/samples/rust/hostprogs/b.rs b/samples/rust/hostprogs/b.rs new file mode 100644 index 000000000000..c1675890648f --- /dev/null +++ b/samples/rust/hostprogs/b.rs @@ -0,0 +1,5 @@ +// SPDX-License-Identifier: GPL-2.0 + +//! Rust single host program sample: module `b`. + +pub(crate) const CONSTANT: i32 = 42; diff --git a/samples/rust/hostprogs/single.rs b/samples/rust/hostprogs/single.rs new file mode 100644 index 000000000000..8c48a119339a --- /dev/null +++ b/samples/rust/hostprogs/single.rs @@ -0,0 +1,12 @@ +// SPDX-License-Identifier: GPL-2.0 + +//! Rust single host program sample. + +mod a; +mod b; + +fn main() { + println!("Hello world!"); + + a::f(b::CONSTANT); +} diff --git a/samples/rust/rust_minimal.rs b/samples/rust/rust_minimal.rs new file mode 100644 index 000000000000..54ad17685742 --- /dev/null +++ b/samples/rust/rust_minimal.rs @@ -0,0 +1,38 @@ +// SPDX-License-Identifier: GPL-2.0 + +//! Rust minimal sample. + +use kernel::prelude::*; + +module! { + type: RustMinimal, + name: b"rust_minimal", + author: b"Rust for Linux Contributors", + description: b"Rust minimal sample", + license: b"GPL", +} + +struct RustMinimal { + numbers: Vec<i32>, +} + +impl kernel::Module for RustMinimal { + fn init(_module: &'static ThisModule) -> Result<Self> { + pr_info!("Rust minimal sample (init)\n"); + pr_info!("Am I built-in? {}\n", !cfg!(MODULE)); + + let mut numbers = Vec::new(); + numbers.try_push(72)?; + numbers.try_push(108)?; + numbers.try_push(200)?; + + Ok(RustMinimal { numbers }) + } +} + +impl Drop for RustMinimal { + fn drop(&mut self) { + pr_info!("My numbers are {:?}\n", self.numbers); + pr_info!("Rust minimal sample (exit)\n"); + } +} diff --git a/scripts/.gitignore b/scripts/.gitignore index eed308bef604..b7aec8eb1bd4 100644 --- a/scripts/.gitignore +++ b/scripts/.gitignore @@ -1,6 +1,7 @@ # SPDX-License-Identifier: GPL-2.0-only /asn1_compiler /bin2c +/generate_rust_target /insert-sys-cert /kallsyms /module.lds diff --git a/scripts/Kconfig.include b/scripts/Kconfig.include index a0ccceb22cf8..274125307ebd 100644 --- a/scripts/Kconfig.include +++ b/scripts/Kconfig.include @@ -36,12 +36,12 @@ ld-option = $(success,$(LD) -v $(1)) as-instr = $(success,printf "%b\n" "$(1)" | $(CC) $(CLANG_FLAGS) -c -x assembler -o /dev/null -) # check if $(CC) and $(LD) exist -$(error-if,$(failure,command -v $(CC)),compiler '$(CC)' not found) +$(error-if,$(failure,command -v $(CC)),C compiler '$(CC)' not found) $(error-if,$(failure,command -v $(LD)),linker '$(LD)' not found) -# Get the compiler name, version, and error out if it is not supported. +# Get the C compiler name, version, and error out if it is not supported. cc-info := $(shell,$(srctree)/scripts/cc-version.sh $(CC)) -$(error-if,$(success,test -z "$(cc-info)"),Sorry$(comma) this compiler is not supported.) +$(error-if,$(success,test -z "$(cc-info)"),Sorry$(comma) this C compiler is not supported.) cc-name := $(shell,set -- $(cc-info) && echo $1) cc-version := $(shell,set -- $(cc-info) && echo $2) diff --git a/scripts/Makefile b/scripts/Makefile index f084f08ed176..1575af84d557 100644 --- a/scripts/Makefile +++ b/scripts/Makefile @@ -10,6 +10,9 @@ hostprogs-always-$(CONFIG_BUILDTIME_TABLE_SORT) += sorttable hostprogs-always-$(CONFIG_ASN1) += asn1_compiler hostprogs-always-$(CONFIG_MODULE_SIG_FORMAT) += sign-file hostprogs-always-$(CONFIG_SYSTEM_EXTRA_CERTIFICATE) += insert-sys-cert +hostprogs-always-$(CONFIG_RUST) += generate_rust_target + +generate_rust_target-rust := y HOSTCFLAGS_sorttable.o = -I$(srctree)/tools/include HOSTLDLIBS_sorttable = -lpthread diff --git a/scripts/Makefile.build b/scripts/Makefile.build index 784f46d41959..27be77c0d6d8 100644 --- a/scripts/Makefile.build +++ b/scripts/Makefile.build @@ -26,6 +26,7 @@ EXTRA_CPPFLAGS := EXTRA_LDFLAGS := asflags-y := ccflags-y := +rustflags-y := cppflags-y := ldflags-y := @@ -271,6 +272,65 @@ quiet_cmd_cc_lst_c = MKLST $@ $(obj)/%.lst: $(src)/%.c FORCE $(call if_changed_dep,cc_lst_c) +# Compile Rust sources (.rs) +# --------------------------------------------------------------------------- + +rust_allowed_features := core_ffi_c + +rust_common_cmd = \ + RUST_MODFILE=$(modfile) $(RUSTC_OR_CLIPPY) $(rust_flags) \ + -Zallow-features=$(rust_allowed_features) \ + -Zcrate-attr=no_std \ + -Zcrate-attr='feature($(rust_allowed_features))' \ + --extern alloc --extern kernel \ + --crate-type rlib --out-dir $(obj) -L $(objtree)/rust/ \ + --crate-name $(basename $(notdir $@)) + +rust_handle_depfile = \ + mv $(obj)/$(basename $(notdir $@)).d $(depfile); \ + sed -i '/^\#/d' $(depfile) + +# `--emit=obj`, `--emit=asm` and `--emit=llvm-ir` imply a single codegen unit +# will be used. We explicitly request `-Ccodegen-units=1` in any case, and +# the compiler shows a warning if it is not 1. However, if we ever stop +# requesting it explicitly and we start using some other `--emit` that does not +# imply it (and for which codegen is performed), then we would be out of sync, +# i.e. the outputs we would get for the different single targets (e.g. `.ll`) +# would not match each other. + +quiet_cmd_rustc_o_rs = $(RUSTC_OR_CLIPPY_QUIET) $(quiet_modtag) $@ + cmd_rustc_o_rs = \ + $(rust_common_cmd) --emit=dep-info,obj $<; \ + $(rust_handle_depfile) + +$(obj)/%.o: $(src)/%.rs FORCE + $(call if_changed_dep,rustc_o_rs) + +quiet_cmd_rustc_rsi_rs = $(RUSTC_OR_CLIPPY_QUIET) $(quiet_modtag) $@ + cmd_rustc_rsi_rs = \ + $(rust_common_cmd) --emit=dep-info -Zunpretty=expanded $< >$@; \ + command -v $(RUSTFMT) >/dev/null && $(RUSTFMT) $@; \ + $(rust_handle_depfile) + +$(obj)/%.rsi: $(src)/%.rs FORCE + $(call if_changed_dep,rustc_rsi_rs) + +quiet_cmd_rustc_s_rs = $(RUSTC_OR_CLIPPY_QUIET) $(quiet_modtag) $@ + cmd_rustc_s_rs = \ + $(rust_common_cmd) --emit=dep-info,asm $<; \ + $(rust_handle_depfile) + +$(obj)/%.s: $(src)/%.rs FORCE + $(call if_changed_dep,rustc_s_rs) + +quiet_cmd_rustc_ll_rs = $(RUSTC_OR_CLIPPY_QUIET) $(quiet_modtag) $@ + cmd_rustc_ll_rs = \ + $(rust_common_cmd) --emit=dep-info,llvm-ir $<; \ + $(rust_handle_depfile) + +$(obj)/%.ll: $(src)/%.rs FORCE + $(call if_changed_dep,rustc_ll_rs) + # Compile assembler sources (.S) # --------------------------------------------------------------------------- diff --git a/scripts/Makefile.debug b/scripts/Makefile.debug index 8cf1cb22dd93..332c486f705f 100644 --- a/scripts/Makefile.debug +++ b/scripts/Makefile.debug @@ -1,4 +1,6 @@ DEBUG_CFLAGS := +DEBUG_RUSTFLAGS := + debug-flags-y := -g ifdef CONFIG_DEBUG_INFO_SPLIT @@ -17,9 +19,12 @@ KBUILD_AFLAGS += $(debug-flags-y) ifdef CONFIG_DEBUG_INFO_REDUCED DEBUG_CFLAGS += -fno-var-tracking +DEBUG_RUSTFLAGS += -Cdebuginfo=1 ifdef CONFIG_CC_IS_GCC DEBUG_CFLAGS += -femit-struct-debug-baseonly endif +else +DEBUG_RUSTFLAGS += -Cdebuginfo=2 endif ifdef CONFIG_DEBUG_INFO_COMPRESSED @@ -30,3 +35,6 @@ endif KBUILD_CFLAGS += $(DEBUG_CFLAGS) export DEBUG_CFLAGS + +KBUILD_RUSTFLAGS += $(DEBUG_RUSTFLAGS) +export DEBUG_RUSTFLAGS diff --git a/scripts/Makefile.host b/scripts/Makefile.host index 278b4d6ac945..da133780b751 100644 --- a/scripts/Makefile.host +++ b/scripts/Makefile.host @@ -22,6 +22,8 @@ $(obj)/%.tab.c $(obj)/%.tab.h: $(src)/%.y FORCE # to preprocess a data file. # # Both C and C++ are supported, but preferred language is C for such utilities. +# Rust is also supported, but it may only be used in scenarios where a Rust +# toolchain is required to be available (e.g. when `CONFIG_RUST` is enabled). # # Sample syntax (see Documentation/kbuild/makefiles.rst for reference) # hostprogs := bin2hex @@ -37,15 +39,20 @@ $(obj)/%.tab.c $(obj)/%.tab.h: $(src)/%.y FORCE # qconf-objs := menu.o # Will compile qconf as a C++ program, and menu as a C program. # They are linked as C++ code to the executable qconf +# +# hostprogs := target +# target-rust := y +# Will compile `target` as a Rust program, using `target.rs` as the crate root. +# The crate may consist of several source files. # C code # Executables compiled from a single .c file host-csingle := $(foreach m,$(hostprogs), \ - $(if $($(m)-objs)$($(m)-cxxobjs),,$(m))) + $(if $($(m)-objs)$($(m)-cxxobjs)$($(m)-rust),,$(m))) # C executables linked based on several .o files host-cmulti := $(foreach m,$(hostprogs),\ - $(if $($(m)-cxxobjs),,$(if $($(m)-objs),$(m)))) + $(if $($(m)-cxxobjs)$($(m)-rust),,$(if $($(m)-objs),$(m)))) # Object (.o) files compiled from .c files host-cobjs := $(sort $(foreach m,$(hostprogs),$($(m)-objs))) @@ -58,11 +65,17 @@ host-cxxmulti := $(foreach m,$(hostprogs),$(if $($(m)-cxxobjs),$(m))) # C++ Object (.o) files compiled from .cc files host-cxxobjs := $(sort $(foreach m,$(host-cxxmulti),$($(m)-cxxobjs))) +# Rust code +# Executables compiled from a single Rust crate (which may consist of +# one or more .rs files) +host-rust := $(foreach m,$(hostprogs),$(if $($(m)-rust),$(m))) + host-csingle := $(addprefix $(obj)/,$(host-csingle)) host-cmulti := $(addprefix $(obj)/,$(host-cmulti)) host-cobjs := $(addprefix $(obj)/,$(host-cobjs)) host-cxxmulti := $(addprefix $(obj)/,$(host-cxxmulti)) host-cxxobjs := $(addprefix $(obj)/,$(host-cxxobjs)) +host-rust := $(addprefix $(obj)/,$(host-rust)) ##### # Handle options to gcc. Support building with separate output directory @@ -71,6 +84,8 @@ _hostc_flags = $(KBUILD_HOSTCFLAGS) $(HOST_EXTRACFLAGS) \ $(HOSTCFLAGS_$(target-stem).o) _hostcxx_flags = $(KBUILD_HOSTCXXFLAGS) $(HOST_EXTRACXXFLAGS) \ $(HOSTCXXFLAGS_$(target-stem).o) +_hostrust_flags = $(KBUILD_HOSTRUSTFLAGS) $(HOST_EXTRARUSTFLAGS) \ + $(HOSTRUSTFLAGS_$(target-stem)) # $(objtree)/$(obj) for including generated headers from checkin source files ifeq ($(KBUILD_EXTMOD),) @@ -82,6 +97,7 @@ endif hostc_flags = -Wp,-MMD,$(depfile) $(_hostc_flags) hostcxx_flags = -Wp,-MMD,$(depfile) $(_hostcxx_flags) +hostrust_flags = $(_hostrust_flags) ##### # Compile programs on the host @@ -128,5 +144,17 @@ quiet_cmd_host-cxxobjs = HOSTCXX $@ $(host-cxxobjs): $(obj)/%.o: $(src)/%.cc FORCE $(call if_changed_dep,host-cxxobjs) +# Create executable from a single Rust crate (which may consist of +# one or more `.rs` files) +# host-rust -> Executable +quiet_cmd_host-rust = HOSTRUSTC $@ + cmd_host-rust = \ + $(HOSTRUSTC) $(hostrust_flags) --emit=dep-info,link \ + --out-dir=$(obj)/ $<; \ + mv $(obj)/$(target-stem).d $(depfile); \ + sed -i '/^\#/d' $(depfile) +$(host-rust): $(obj)/%: $(src)/%.rs FORCE + $(call if_changed_dep,host-rust) + targets += $(host-csingle) $(host-cmulti) $(host-cobjs) \ - $(host-cxxmulti) $(host-cxxobjs) + $(host-cxxmulti) $(host-cxxobjs) $(host-rust) diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib index 3fb6a99e78c4..c88b98b5dc44 100644 --- a/scripts/Makefile.lib +++ b/scripts/Makefile.lib @@ -8,6 +8,7 @@ ldflags-y += $(EXTRA_LDFLAGS) # flags that take effect in current and sub directories KBUILD_AFLAGS += $(subdir-asflags-y) KBUILD_CFLAGS += $(subdir-ccflags-y) +KBUILD_RUSTFLAGS += $(subdir-rustflags-y) # Figure out what we need to build from the various variables # =========================================================================== @@ -128,6 +129,10 @@ _c_flags = $(filter-out $(CFLAGS_REMOVE_$(target-stem).o), \ $(filter-out $(ccflags-remove-y), \ $(KBUILD_CPPFLAGS) $(KBUILD_CFLAGS) $(ccflags-y)) \ $(CFLAGS_$(target-stem).o)) +_rust_flags = $(filter-out $(RUSTFLAGS_REMOVE_$(target-stem).o), \ + $(filter-out $(rustflags-remove-y), \ + $(KBUILD_RUSTFLAGS) $(rustflags-y)) \ + $(RUSTFLAGS_$(target-stem).o)) _a_flags = $(filter-out $(AFLAGS_REMOVE_$(target-stem).o), \ $(filter-out $(asflags-remove-y), \ $(KBUILD_CPPFLAGS) $(KBUILD_AFLAGS) $(asflags-y)) \ @@ -202,6 +207,11 @@ modkern_cflags = \ $(KBUILD_CFLAGS_MODULE) $(CFLAGS_MODULE), \ $(KBUILD_CFLAGS_KERNEL) $(CFLAGS_KERNEL) $(modfile_flags)) +modkern_rustflags = \ + $(if $(part-of-module), \ + $(KBUILD_RUSTFLAGS_MODULE) $(RUSTFLAGS_MODULE), \ + $(KBUILD_RUSTFLAGS_KERNEL) $(RUSTFLAGS_KERNEL)) + modkern_aflags = $(if $(part-of-module), \ $(KBUILD_AFLAGS_MODULE) $(AFLAGS_MODULE), \ $(KBUILD_AFLAGS_KERNEL) $(AFLAGS_KERNEL)) @@ -211,6 +221,8 @@ c_flags = -Wp,-MMD,$(depfile) $(NOSTDINC_FLAGS) $(LINUXINCLUDE) \ $(_c_flags) $(modkern_cflags) \ $(basename_flags) $(modname_flags) +rust_flags = $(_rust_flags) $(modkern_rustflags) @$(objtree)/include/generated/rustc_cfg + a_flags = -Wp,-MMD,$(depfile) $(NOSTDINC_FLAGS) $(LINUXINCLUDE) \ $(_a_flags) $(modkern_aflags) diff --git a/scripts/Makefile.modfinal b/scripts/Makefile.modfinal index 35100e981f4a..9a1fa6aa30fe 100644 --- a/scripts/Makefile.modfinal +++ b/scripts/Makefile.modfinal @@ -39,11 +39,13 @@ quiet_cmd_ld_ko_o = LD [M] $@ quiet_cmd_btf_ko = BTF [M] $@ cmd_btf_ko = \ - if [ -f vmlinux ]; then \ + if [ ! -f vmlinux ]; then \ + printf "Skipping BTF generation for %s due to unavailability of vmlinux\n" $@ 1>&2; \ + elif [ -n "$(CONFIG_RUST)" ] && $(srctree)/scripts/is_rust_module.sh $@; then \ + printf "Skipping BTF generation for %s because it's a Rust module\n" $@ 1>&2; \ + else \ LLVM_OBJCOPY="$(OBJCOPY)" $(PAHOLE) -J $(PAHOLE_FLAGS) --btf_base vmlinux $@; \ $(RESOLVE_BTFIDS) -b vmlinux $@; \ - else \ - printf "Skipping BTF generation for %s due to unavailability of vmlinux\n" $@ 1>&2; \ fi; # Same as newer-prereqs, but allows to exclude specified extra dependencies diff --git a/scripts/cc-version.sh b/scripts/cc-version.sh index f1952c522466..2401c86fcf53 100755 --- a/scripts/cc-version.sh +++ b/scripts/cc-version.sh @@ -1,13 +1,13 @@ #!/bin/sh # SPDX-License-Identifier: GPL-2.0 # -# Print the compiler name and its version in a 5 or 6-digit form. +# Print the C compiler name and its version in a 5 or 6-digit form. # Also, perform the minimum version check. set -e -# Print the compiler name and some version components. -get_compiler_info() +# Print the C compiler name and some version components. +get_c_compiler_info() { cat <<- EOF | "$@" -E -P -x c - 2>/dev/null #if defined(__clang__) @@ -32,7 +32,7 @@ get_canonical_version() # $@ instead of $1 because multiple words might be given, e.g. CC="ccache gcc". orig_args="$@" -set -- $(get_compiler_info "$@") +set -- $(get_c_compiler_info "$@") name=$1 @@ -52,7 +52,7 @@ ICC) min_version=$($min_tool_version icc) ;; *) - echo "$orig_args: unknown compiler" >&2 + echo "$orig_args: unknown C compiler" >&2 exit 1 ;; esac @@ -62,7 +62,7 @@ min_cversion=$(get_canonical_version $min_version) if [ "$cversion" -lt "$min_cversion" ]; then echo >&2 "***" - echo >&2 "*** Compiler is too old." + echo >&2 "*** C compiler is too old." echo >&2 "*** Your $name version: $version" echo >&2 "*** Minimum $name version: $min_version" echo >&2 "***" diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl index 4aa09e0cb86a..c8a616a9d034 100755 --- a/scripts/checkpatch.pl +++ b/scripts/checkpatch.pl @@ -3616,7 +3616,7 @@ sub process { my $comment = ""; if ($realfile =~ /\.(h|s|S)$/) { $comment = '/*'; - } elsif ($realfile =~ /\.(c|dts|dtsi)$/) { + } elsif ($realfile =~ /\.(c|rs|dts|dtsi)$/) { $comment = '//'; } elsif (($checklicenseline == 2) || $realfile =~ /\.(sh|pl|py|awk|tc|yaml)$/) { $comment = '#'; @@ -3664,7 +3664,7 @@ sub process { } # check we are in a valid source file if not then ignore this hunk - next if ($realfile !~ /\.(h|c|s|S|sh|dtsi|dts)$/); + next if ($realfile !~ /\.(h|c|rs|s|S|sh|dtsi|dts)$/); # check for using SPDX-License-Identifier on the wrong line number if ($realline != $checklicenseline && @@ -6783,15 +6783,19 @@ sub process { } if ($bad_specifier ne "") { my $stat_real = get_stat_real($linenr, $lc); + my $msg_level = \&WARN; my $ext_type = "Invalid"; my $use = ""; if ($bad_specifier =~ /p[Ff]/) { $use = " - use %pS instead"; $use =~ s/pS/ps/ if ($bad_specifier =~ /pf/); + } elsif ($bad_specifier =~ /pA/) { + $use = " - '%pA' is only intended to be used from Rust code"; + $msg_level = \&ERROR; } - WARN("VSPRINTF_POINTER_EXTENSION", - "$ext_type vsprintf pointer extension '$bad_specifier'$use\n" . "$here\n$stat_real\n"); + &{$msg_level}("VSPRINTF_POINTER_EXTENSION", + "$ext_type vsprintf pointer extension '$bad_specifier'$use\n" . "$here\n$stat_real\n"); } } } diff --git a/scripts/decode_stacktrace.sh b/scripts/decode_stacktrace.sh index 7075e26ab2c4..564c5632e1a2 100755 --- a/scripts/decode_stacktrace.sh +++ b/scripts/decode_stacktrace.sh @@ -8,6 +8,14 @@ usage() { echo " $0 -r <release> | <vmlinux> [<base path>|auto] [<modules path>]" } +# Try to find a Rust demangler +if type llvm-cxxfilt >/dev/null 2>&1 ; then + cppfilt=llvm-cxxfilt +elif type c++filt >/dev/null 2>&1 ; then + cppfilt=c++filt + cppfilt_opts=-i +fi + if [[ $1 == "-r" ]] ; then vmlinux="" basepath="auto" @@ -180,6 +188,12 @@ parse_symbol() { # In the case of inlines, move everything to same line code=${code//$'\n'/' '} + # Demangle if the name looks like a Rust symbol and if + # we got a Rust demangler + if [[ $name =~ ^_R && $cppfilt != "" ]] ; then + name=$("$cppfilt" "$cppfilt_opts" "$name") + fi + # Replace old address with pretty line numbers symbol="$segment$name ($code)" } diff --git a/scripts/generate_rust_analyzer.py b/scripts/generate_rust_analyzer.py new file mode 100755 index 000000000000..75bb611bd751 --- /dev/null +++ b/scripts/generate_rust_analyzer.py @@ -0,0 +1,135 @@ +#!/usr/bin/env python3 +# SPDX-License-Identifier: GPL-2.0 +"""generate_rust_analyzer - Generates the `rust-project.json` file for `rust-analyzer`. +""" + +import argparse +import json +import logging +import pathlib +import sys + +def generate_crates(srctree, objtree, sysroot_src): + # Generate the configuration list. + cfg = [] + with open(objtree / "include" / "generated" / "rustc_cfg") as fd: + for line in fd: + line = line.replace("--cfg=", "") + line = line.replace("\n", "") + cfg.append(line) + + # Now fill the crates list -- dependencies need to come first. + # + # Avoid O(n^2) iterations by keeping a map of indexes. + crates = [] + crates_indexes = {} + + def append_crate(display_name, root_module, deps, cfg=[], is_workspace_member=True, is_proc_macro=False): + crates_indexes[display_name] = len(crates) + crates.append({ + "display_name": display_name, + "root_module": str(root_module), + "is_workspace_member": is_workspace_member, + "is_proc_macro": is_proc_macro, + "deps": [{"crate": crates_indexes[dep], "name": dep} for dep in deps], + "cfg": cfg, + "edition": "2021", + "env": { + "RUST_MODFILE": "This is only for rust-analyzer" + } + }) + + # First, the ones in `rust/` since they are a bit special. + append_crate( + "core", + sysroot_src / "core" / "src" / "lib.rs", + [], + is_workspace_member=False, + ) + + append_crate( + "compiler_builtins", + srctree / "rust" / "compiler_builtins.rs", + [], + ) + + append_crate( + "alloc", + srctree / "rust" / "alloc" / "lib.rs", + ["core", "compiler_builtins"], + ) + + append_crate( + "macros", + srctree / "rust" / "macros" / "lib.rs", + [], + is_proc_macro=True, + ) + crates[-1]["proc_macro_dylib_path"] = "rust/libmacros.so" + + append_crate( + "bindings", + srctree / "rust"/ "bindings" / "lib.rs", + ["core"], + cfg=cfg, + ) + crates[-1]["env"]["OBJTREE"] = str(objtree.resolve(True)) + + append_crate( + "kernel", + srctree / "rust" / "kernel" / "lib.rs", + ["core", "alloc", "macros", "bindings"], + cfg=cfg, + ) + crates[-1]["source"] = { + "include_dirs": [ + str(srctree / "rust" / "kernel"), + str(objtree / "rust") + ], + "exclude_dirs": [], + } + + # Then, the rest outside of `rust/`. + # + # We explicitly mention the top-level folders we want to cover. + for folder in ("samples", "drivers"): + for path in (srctree / folder).rglob("*.rs"): + logging.info("Checking %s", path) + name = path.name.replace(".rs", "") + + # Skip those that are not crate roots. + if f"{name}.o" not in open(path.parent / "Makefile").read(): + continue + + logging.info("Adding %s", name) + append_crate( + name, + path, + ["core", "alloc", "kernel"], + cfg=cfg, + ) + + return crates + +def main(): + parser = argparse.ArgumentParser() + parser.add_argument('--verbose', '-v', action='store_true') + parser.add_argument("srctree", type=pathlib.Path) + parser.add_argument("objtree", type=pathlib.Path) + parser.add_argument("sysroot_src", type=pathlib.Path) + args = parser.parse_args() + + logging.basicConfig( + format="[%(asctime)s] [%(levelname)s] %(message)s", + level=logging.INFO if args.verbose else logging.WARNING + ) + + rust_project = { + "crates": generate_crates(args.srctree, args.objtree, args.sysroot_src), + "sysroot_src": str(args.sysroot_src), + } + + json.dump(rust_project, sys.stdout, sort_keys=True, indent=4) + +if __name__ == "__main__": + main() diff --git a/scripts/generate_rust_target.rs b/scripts/generate_rust_target.rs new file mode 100644 index 000000000000..3c6cbe2b278d --- /dev/null +++ b/scripts/generate_rust_target.rs @@ -0,0 +1,182 @@ +// SPDX-License-Identifier: GPL-2.0 + +//! The custom target specification file generator for `rustc`. +//! +//! To configure a target from scratch, a JSON-encoded file has to be passed +//! to `rustc` (introduced in [RFC 131]). These options and the file itself are +//! unstable. Eventually, `rustc` should provide a way to do this in a stable +//! manner. For instance, via command-line arguments. Therefore, this file +//! should avoid using keys which can be set via `-C` or `-Z` options. +//! +//! [RFC 131]: https://rust-lang.github.io/rfcs/0131-target-specification.html + +use std::{ + collections::HashMap, + fmt::{Display, Formatter, Result}, + io::BufRead, +}; + +enum Value { + Boolean(bool), + Number(i32), + String(String), + Object(Object), +} + +type Object = Vec<(String, Value)>; + +/// Minimal "almost JSON" generator (e.g. no `null`s, no arrays, no escaping), +/// enough for this purpose. +impl Display for Value { + fn fmt(&self, formatter: &mut Formatter<'_>) -> Result { + match self { + Value::Boolean(boolean) => write!(formatter, "{}", boolean), + Value::Number(number) => write!(formatter, "{}", number), + Value::String(string) => write!(formatter, "\"{}\"", string), + Value::Object(object) => { + formatter.write_str("{")?; + if let [ref rest @ .., ref last] = object[..] { + for (key, value) in rest { + write!(formatter, "\"{}\": {},", key, value)?; + } + write!(formatter, "\"{}\": {}", last.0, last.1)?; + } + formatter.write_str("}") + } + } + } +} + +struct TargetSpec(Object); + +impl TargetSpec { + fn new() -> TargetSpec { + TargetSpec(Vec::new()) + } +} + +trait Push<T> { + fn push(&mut self, key: &str, value: T); +} + +impl Push<bool> for TargetSpec { + fn push(&mut self, key: &str, value: bool) { + self.0.push((key.to_string(), Value::Boolean(value))); + } +} + +impl Push<i32> for TargetSpec { + fn push(&mut self, key: &str, value: i32) { + self.0.push((key.to_string(), Value::Number(value))); + } +} + +impl Push<String> for TargetSpec { + fn push(&mut self, key: &str, value: String) { + self.0.push((key.to_string(), Value::String(value))); + } +} + +impl Push<&str> for TargetSpec { + fn push(&mut self, key: &str, value: &str) { + self.push(key, value.to_string()); + } +} + +impl Push<Object> for TargetSpec { + fn push(&mut self, key: &str, value: Object) { + self.0.push((key.to_string(), Value::Object(value))); + } +} + +impl Display for TargetSpec { + fn fmt(&self, formatter: &mut Formatter<'_>) -> Result { + // We add some newlines for clarity. + formatter.write_str("{\n")?; + if let [ref rest @ .., ref last] = self.0[..] { + for (key, value) in rest { + write!(formatter, " \"{}\": {},\n", key, value)?; + } + write!(formatter, " \"{}\": {}\n", last.0, last.1)?; + } + formatter.write_str("}") + } +} + +struct KernelConfig(HashMap<String, String>); + +impl KernelConfig { + /// Parses `include/config/auto.conf` from `stdin`. + fn from_stdin() -> KernelConfig { + let mut result = HashMap::new(); + + let stdin = std::io::stdin(); + let mut handle = stdin.lock(); + let mut line = String::new(); + + loop { + line.clear(); + + if handle.read_line(&mut line).unwrap() == 0 { + break; + } + + if line.starts_with('#') { + continue; + } + + let (key, value) = line.split_once('=').expect("Missing `=` in line."); + result.insert(key.to_string(), value.trim_end_matches('\n').to_string()); + } + + KernelConfig(result) + } + + /// Does the option exist in the configuration (any value)? + /// + /// The argument must be passed without the `CONFIG_` prefix. + /// This avoids repetition and it also avoids `fixdep` making us + /// depend on it. + fn has(&self, option: &str) -> bool { + let option = "CONFIG_".to_owned() + option; + self.0.contains_key(&option) + } +} + +fn main() { + let cfg = KernelConfig::from_stdin(); + let mut ts = TargetSpec::new(); + + // `llvm-target`s are taken from `scripts/Makefile.clang`. + if cfg.has("X86_64") { + ts.push("arch", "x86_64"); + ts.push( + "data-layout", + "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", + ); + let mut features = "-3dnow,-3dnowa,-mmx,+soft-float".to_string(); + if cfg.has("RETPOLINE") { + features += ",+retpoline-external-thunk"; + } + ts.push("features", features); + ts.push("llvm-target", "x86_64-linux-gnu"); + ts.push("target-pointer-width", "64"); + } else { + panic!("Unsupported architecture"); + } + + ts.push("emit-debug-gdb-scripts", false); + ts.push("frame-pointer", "may-omit"); + ts.push( + "stack-probes", + vec![("kind".to_string(), Value::String("none".to_string()))], + ); + + // Everything else is LE, whether `CPU_LITTLE_ENDIAN` is declared or not + // (e.g. x86). It is also `rustc`'s default. + if cfg.has("CPU_BIG_ENDIAN") { + ts.push("target-endian", "big"); + } + + println!("{}", ts); +} diff --git a/scripts/is_rust_module.sh b/scripts/is_rust_module.sh new file mode 100755 index 000000000000..28b3831a7593 --- /dev/null +++ b/scripts/is_rust_module.sh @@ -0,0 +1,16 @@ +#!/bin/sh +# SPDX-License-Identifier: GPL-2.0 +# +# is_rust_module.sh module.ko +# +# Returns `0` if `module.ko` is a Rust module, `1` otherwise. + +set -e + +# Using the `16_` prefix ensures other symbols with the same substring +# are not picked up (even if it would be unlikely). The last part is +# used just in case LLVM decides to use the `.` suffix. +# +# In the future, checking for the `.comment` section may be another +# option, see https://github.com/rust-lang/rust/pull/97550. +${NM} "$*" | grep -qE '^[0-9a-fA-F]+ r _R[^[:space:]]+16___IS_RUST_MODULE[^[:space:]]*$' diff --git a/scripts/kallsyms.c b/scripts/kallsyms.c index f18e6dfc68c5..c4793301a27e 100644 --- a/scripts/kallsyms.c +++ b/scripts/kallsyms.c @@ -27,7 +27,23 @@ #define ARRAY_SIZE(arr) (sizeof(arr) / sizeof(arr[0])) -#define KSYM_NAME_LEN 128 +#define _stringify_1(x) #x +#define _stringify(x) _stringify_1(x) + +#define KSYM_NAME_LEN 512 + +/* + * A substantially bigger size than the current maximum. + * + * It cannot be defined as an expression because it gets stringified + * for the fscanf() format string. Therefore, a _Static_assert() is + * used instead to maintain the relationship with KSYM_NAME_LEN. + */ +#define KSYM_NAME_LEN_BUFFER 2048 +_Static_assert( + KSYM_NAME_LEN_BUFFER == KSYM_NAME_LEN * 4, + "Please keep KSYM_NAME_LEN_BUFFER in sync with KSYM_NAME_LEN" +); struct sym_entry { unsigned long long addr; @@ -198,15 +214,15 @@ static void check_symbol_range(const char *sym, unsigned long long addr, static struct sym_entry *read_symbol(FILE *in) { - char name[500], type; + char name[KSYM_NAME_LEN_BUFFER+1], type; unsigned long long addr; unsigned int len; struct sym_entry *sym; int rc; - rc = fscanf(in, "%llx %c %499s\n", &addr, &type, name); + rc = fscanf(in, "%llx %c %" _stringify(KSYM_NAME_LEN_BUFFER) "s\n", &addr, &type, name); if (rc != 3) { - if (rc != EOF && fgets(name, 500, in) == NULL) + if (rc != EOF && fgets(name, ARRAY_SIZE(name), in) == NULL) fprintf(stderr, "Read error or end of file.\n"); return NULL; } @@ -471,12 +487,35 @@ static void write_src(void) if ((i & 0xFF) == 0) markers[i >> 8] = off; - printf("\t.byte 0x%02x", table[i]->len); + /* There cannot be any symbol of length zero. */ + if (table[i]->len == 0) { + fprintf(stderr, "kallsyms failure: " + "unexpected zero symbol length\n"); + exit(EXIT_FAILURE); + } + + /* Only lengths that fit in up-to-two-byte ULEB128 are supported. */ + if (table[i]->len > 0x3FFF) { + fprintf(stderr, "kallsyms failure: " + "unexpected huge symbol length\n"); + exit(EXIT_FAILURE); + } + + /* Encode length with ULEB128. */ + if (table[i]->len <= 0x7F) { + /* Most symbols use a single byte for the length. */ + printf("\t.byte 0x%02x", table[i]->len); + off += table[i]->len + 1; + } else { + /* "Big" symbols use two bytes. */ + printf("\t.byte 0x%02x, 0x%02x", + (table[i]->len & 0x7F) | 0x80, + (table[i]->len >> 7) & 0x7F); + off += table[i]->len + 2; + } for (k = 0; k < table[i]->len; k++) printf(", 0x%02x", table[i]->sym[k]); printf("\n"); - - off += table[i]->len + 1; } printf("\n"); diff --git a/scripts/kconfig/confdata.c b/scripts/kconfig/confdata.c index c4340c90e172..b7c9f1dd5e42 100644 --- a/scripts/kconfig/confdata.c +++ b/scripts/kconfig/confdata.c @@ -216,6 +216,13 @@ static const char *conf_get_autoheader_name(void) return name ? name : "include/generated/autoconf.h"; } +static const char *conf_get_rustccfg_name(void) +{ + char *name = getenv("KCONFIG_RUSTCCFG"); + + return name ? name : "include/generated/rustc_cfg"; +} + static int conf_set_sym_val(struct symbol *sym, int def, int def_flags, char *p) { char *p2; @@ -605,6 +612,9 @@ static const struct comment_style comment_style_c = { static void conf_write_heading(FILE *fp, const struct comment_style *cs) { + if (!cs) + return; + fprintf(fp, "%s\n", cs->prefix); fprintf(fp, "%s Automatically generated file; DO NOT EDIT.\n", @@ -745,6 +755,65 @@ static void print_symbol_for_c(FILE *fp, struct symbol *sym) free(escaped); } +static void print_symbol_for_rustccfg(FILE *fp, struct symbol *sym) +{ + const char *val; + const char *val_prefix = ""; + char *val_prefixed = NULL; + size_t val_prefixed_len; + char *escaped = NULL; + + if (sym->type == S_UNKNOWN) + return; + + val = sym_get_string_value(sym); + + switch (sym->type) { + case S_BOOLEAN: + case S_TRISTATE: + /* + * We do not care about disabled ones, i.e. no need for + * what otherwise are "comments" in other printers. + */ + if (*val == 'n') + return; + + /* + * To have similar functionality to the C macro `IS_ENABLED()` + * we provide an empty `--cfg CONFIG_X` here in both `y` + * and `m` cases. + * + * Then, the common `fprintf()` below will also give us + * a `--cfg CONFIG_X="y"` or `--cfg CONFIG_X="m"`, which can + * be used as the equivalent of `IS_BUILTIN()`/`IS_MODULE()`. + */ + fprintf(fp, "--cfg=%s%s\n", CONFIG_, sym->name); + break; + case S_HEX: + if (val[0] != '0' || (val[1] != 'x' && val[1] != 'X')) + val_prefix = "0x"; + break; + default: + break; + } + + if (strlen(val_prefix) > 0) { + val_prefixed_len = strlen(val) + strlen(val_prefix) + 1; + val_prefixed = xmalloc(val_prefixed_len); + snprintf(val_prefixed, val_prefixed_len, "%s%s", val_prefix, val); + val = val_prefixed; + } + + /* All values get escaped: the `--cfg` option only takes strings */ + escaped = escape_string_value(val); + val = escaped; + + fprintf(fp, "--cfg=%s%s=%s\n", CONFIG_, sym->name, val); + + free(escaped); + free(val_prefixed); +} + /* * Write out a minimal config. * All values that has default values are skipped as this is redundant. @@ -1132,6 +1201,12 @@ int conf_write_autoconf(int overwrite) if (ret) return ret; + ret = __conf_write_autoconf(conf_get_rustccfg_name(), + print_symbol_for_rustccfg, + NULL); + if (ret) + return ret; + /* * Create include/config/auto.conf. This must be the last step because * Kbuild has a dependency on auto.conf and this marks the successful diff --git a/scripts/min-tool-version.sh b/scripts/min-tool-version.sh index 250925aab101..b6593eac5003 100755 --- a/scripts/min-tool-version.sh +++ b/scripts/min-tool-version.sh @@ -30,6 +30,12 @@ llvm) echo 11.0.0 fi ;; +rustc) + echo 1.62.0 + ;; +bindgen) + echo 0.56.0 + ;; *) echo "$1: unknown tool" >&2 exit 1 diff --git a/scripts/rust_is_available.sh b/scripts/rust_is_available.sh new file mode 100755 index 000000000000..aebbf1913970 --- /dev/null +++ b/scripts/rust_is_available.sh @@ -0,0 +1,160 @@ +#!/bin/sh +# SPDX-License-Identifier: GPL-2.0 +# +# Tests whether a suitable Rust toolchain is available. +# +# Pass `-v` for human output and more checks (as warnings). + +set -e + +min_tool_version=$(dirname $0)/min-tool-version.sh + +# Convert the version string x.y.z to a canonical up-to-7-digits form. +# +# Note that this function uses one more digit (compared to other +# instances in other version scripts) to give a bit more space to +# `rustc` since it will reach 1.100.0 in late 2026. +get_canonical_version() +{ + IFS=. + set -- $1 + echo $((100000 * $1 + 100 * $2 + $3)) +} + +# Check that the Rust compiler exists. +if ! command -v "$RUSTC" >/dev/null; then + if [ "$1" = -v ]; then + echo >&2 "***" + echo >&2 "*** Rust compiler '$RUSTC' could not be found." + echo >&2 "***" + fi + exit 1 +fi + +# Check that the Rust bindings generator exists. +if ! command -v "$BINDGEN" >/dev/null; then + if [ "$1" = -v ]; then + echo >&2 "***" + echo >&2 "*** Rust bindings generator '$BINDGEN' could not be found." + echo >&2 "***" + fi + exit 1 +fi + +# Check that the Rust compiler version is suitable. +# +# Non-stable and distributions' versions may have a version suffix, e.g. `-dev`. +rust_compiler_version=$( \ + LC_ALL=C "$RUSTC" --version 2>/dev/null \ + | head -n 1 \ + | grep -oE '[0-9]+\.[0-9]+\.[0-9]+' \ +) +rust_compiler_min_version=$($min_tool_version rustc) +rust_compiler_cversion=$(get_canonical_version $rust_compiler_version) +rust_compiler_min_cversion=$(get_canonical_version $rust_compiler_min_version) +if [ "$rust_compiler_cversion" -lt "$rust_compiler_min_cversion" ]; then + if [ "$1" = -v ]; then + echo >&2 "***" + echo >&2 "*** Rust compiler '$RUSTC' is too old." + echo >&2 "*** Your version: $rust_compiler_version" + echo >&2 "*** Minimum version: $rust_compiler_min_version" + echo >&2 "***" + fi + exit 1 +fi +if [ "$1" = -v ] && [ "$rust_compiler_cversion" -gt "$rust_compiler_min_cversion" ]; then + echo >&2 "***" + echo >&2 "*** Rust compiler '$RUSTC' is too new. This may or may not work." + echo >&2 "*** Your version: $rust_compiler_version" + echo >&2 "*** Expected version: $rust_compiler_min_version" + echo >&2 "***" +fi + +# Check that the Rust bindings generator is suitable. +# +# Non-stable and distributions' versions may have a version suffix, e.g. `-dev`. +rust_bindings_generator_version=$( \ + LC_ALL=C "$BINDGEN" --version 2>/dev/null \ + | head -n 1 \ + | grep -oE '[0-9]+\.[0-9]+\.[0-9]+' \ +) +rust_bindings_generator_min_version=$($min_tool_version bindgen) +rust_bindings_generator_cversion=$(get_canonical_version $rust_bindings_generator_version) +rust_bindings_generator_min_cversion=$(get_canonical_version $rust_bindings_generator_min_version) +if [ "$rust_bindings_generator_cversion" -lt "$rust_bindings_generator_min_cversion" ]; then + if [ "$1" = -v ]; then + echo >&2 "***" + echo >&2 "*** Rust bindings generator '$BINDGEN' is too old." + echo >&2 "*** Your version: $rust_bindings_generator_version" + echo >&2 "*** Minimum version: $rust_bindings_generator_min_version" + echo >&2 "***" + fi + exit 1 +fi +if [ "$1" = -v ] && [ "$rust_bindings_generator_cversion" -gt "$rust_bindings_generator_min_cversion" ]; then + echo >&2 "***" + echo >&2 "*** Rust bindings generator '$BINDGEN' is too new. This may or may not work." + echo >&2 "*** Your version: $rust_bindings_generator_version" + echo >&2 "*** Expected version: $rust_bindings_generator_min_version" + echo >&2 "***" +fi + +# Check that the `libclang` used by the Rust bindings generator is suitable. +bindgen_libclang_version=$( \ + LC_ALL=C "$BINDGEN" $(dirname $0)/rust_is_available_bindgen_libclang.h 2>&1 >/dev/null \ + | grep -F 'clang version ' \ + | grep -oE '[0-9]+\.[0-9]+\.[0-9]+' \ + | head -n 1 \ +) +bindgen_libclang_min_version=$($min_tool_version llvm) +bindgen_libclang_cversion=$(get_canonical_version $bindgen_libclang_version) +bindgen_libclang_min_cversion=$(get_canonical_version $bindgen_libclang_min_version) +if [ "$bindgen_libclang_cversion" -lt "$bindgen_libclang_min_cversion" ]; then + if [ "$1" = -v ]; then + echo >&2 "***" + echo >&2 "*** libclang (used by the Rust bindings generator '$BINDGEN') is too old." + echo >&2 "*** Your version: $bindgen_libclang_version" + echo >&2 "*** Minimum version: $bindgen_libclang_min_version" + echo >&2 "***" + fi + exit 1 +fi + +# If the C compiler is Clang, then we can also check whether its version +# matches the `libclang` version used by the Rust bindings generator. +# +# In the future, we might be able to perform a full version check, see +# https://github.com/rust-lang/rust-bindgen/issues/2138. +if [ "$1" = -v ]; then + cc_name=$($(dirname $0)/cc-version.sh "$CC" | cut -f1 -d' ') + if [ "$cc_name" = Clang ]; then + clang_version=$( \ + LC_ALL=C "$CC" --version 2>/dev/null \ + | sed -nE '1s:.*version ([0-9]+\.[0-9]+\.[0-9]+).*:\1:p' + ) + if [ "$clang_version" != "$bindgen_libclang_version" ]; then + echo >&2 "***" + echo >&2 "*** libclang (used by the Rust bindings generator '$BINDGEN')" + echo >&2 "*** version does not match Clang's. This may be a problem." + echo >&2 "*** libclang version: $bindgen_libclang_version" + echo >&2 "*** Clang version: $clang_version" + echo >&2 "***" + fi + fi +fi + +# Check that the source code for the `core` standard library exists. +# +# `$KRUSTFLAGS` is passed in case the user added `--sysroot`. +rustc_sysroot=$("$RUSTC" $KRUSTFLAGS --print sysroot) +rustc_src=${RUST_LIB_SRC:-"$rustc_sysroot/lib/rustlib/src/rust/library"} +rustc_src_core="$rustc_src/core/src/lib.rs" +if [ ! -e "$rustc_src_core" ]; then + if [ "$1" = -v ]; then + echo >&2 "***" + echo >&2 "*** Source code for the 'core' standard library could not be found" + echo >&2 "*** at '$rustc_src_core'." + echo >&2 "***" + fi + exit 1 +fi diff --git a/scripts/rust_is_available_bindgen_libclang.h b/scripts/rust_is_available_bindgen_libclang.h new file mode 100644 index 000000000000..0ef6db10d674 --- /dev/null +++ b/scripts/rust_is_available_bindgen_libclang.h @@ -0,0 +1,2 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#pragma message("clang version " __clang_version__) diff --git a/tools/include/linux/kallsyms.h b/tools/include/linux/kallsyms.h index efb6c3f5f2a9..5a37ccbec54f 100644 --- a/tools/include/linux/kallsyms.h +++ b/tools/include/linux/kallsyms.h @@ -6,7 +6,7 @@ #include <stdio.h> #include <unistd.h> -#define KSYM_NAME_LEN 128 +#define KSYM_NAME_LEN 512 struct module; diff --git a/tools/lib/perf/include/perf/event.h b/tools/lib/perf/include/perf/event.h index 93bf93a59c99..d8ae4e944467 100644 --- a/tools/lib/perf/include/perf/event.h +++ b/tools/lib/perf/include/perf/event.h @@ -97,7 +97,7 @@ struct perf_record_throttle { }; #ifndef KSYM_NAME_LEN -#define KSYM_NAME_LEN 256 +#define KSYM_NAME_LEN 512 #endif struct perf_record_ksymbol { diff --git a/tools/lib/symbol/kallsyms.h b/tools/lib/symbol/kallsyms.h index 72ab9870454b..542f9b059c3b 100644 --- a/tools/lib/symbol/kallsyms.h +++ b/tools/lib/symbol/kallsyms.h @@ -7,7 +7,7 @@ #include <linux/types.h> #ifndef KSYM_NAME_LEN -#define KSYM_NAME_LEN 256 +#define KSYM_NAME_LEN 512 #endif static inline u8 kallsyms2elf_binding(char type) |