RustInQemu: Difference between revisions

From QEMU
 
(57 intermediate revisions by 3 users not shown)
Line 1: Line 1:
For the old RustInQemu page, see [[RustInQemu/2022]]
For the old RustInQemu page, see [[RustInQemu/2022]]


== Active efforts in 2024 ==
== Active effort ==


* <code>Subject: [RFC 0/6] scripts: Rewrite simpletrace printer in Rust</code><br /><code>Date: Mon, 27 May 2024 16:14:15 +0800</code><br />[https://lore.kernel.org/qemu-devel/20240527081421.2258624-1-zhao1.liu@intel.com/ RFC v1]
* [https://lore.kernel.org/qemu-devel/d3d1944e-2482-4aa7-b621-596246a08107@gnu.org/ Current state]
* ARM PL011 UART device model in Rust<br /><code>Subject: [RFC PATCH v1 0/6] Implement ARM PL011 in Rust</code><br /><code>Date: Mon, 10 Jun 2024 21:22:35 +0300</code><br />[https://lore.kernel.org/qemu-devel/cover.rust-pl011-rfc-v1.git.manos.pitsidianakis@linaro.org/ RFC v1] [https://lore.kernel.org/qemu-devel/rust-pl011-rfc-v6.git.manos.pitsidianakis@linaro.org/ v6]
* [https://www.qemu.org/docs/master/devel/rust.html Documentation]
** Meson and configure integration
* Past updates: [https://lore.kernel.org/qemu-devel/17ad81c3-98fc-44c2-8f65-f5e2cc07030b@gnu.org/ January 2025] [https://lore.kernel.org/all/cc40943e-dec1-4890-a1d9-579350ce296f@pbonzini.local/ November 2024]
** Bindings generation
 
After some discussion we decided to use rustc directly without going through cargo and each external crate becomes a Meson subproject. Meson 1.5.0+'s support for <tt>method = cargo</tt> is pretty good for host-compiled packages, but it doesn't work for dependencies of build-compiled packages (i.e. dependencies of procedural macros). This is tracked by https://github.com/mesonbuild/meson/issues/11121. Until then, meson.build files for procedural macros and their dependencies have to be written by hand. (Currently _all_ dependencies use hand-written meson.build).
 
While cargo would be fine as an interim solution, using Meson to compile the Rust code removes some moving parts. The extra cost of writing meson.build files for dependencies is expected to be small and will mostly happen upfront.


== Past efforts ==
== Past efforts ==


* <code>[RFC v3 00/32] Rust binding for QAPI and qemu-ga QMP handler examples</code><br />[https://patchew.org/QEMU/20210907121943.3498701-1-marcandre.lureau@redhat.com/ on patchew] [https://lore.kernel.org/qemu-devel/20210907121943.3498701-1-marcandre.lureau@redhat.com/ on lore]
* <code>[RFC v3 00/32] Rust binding for QAPI and qemu-ga QMP handler examples</code><br />[https://patchew.org/QEMU/20210907121943.3498701-1-marcandre.lureau@redhat.com/ on patchew] [https://lore.kernel.org/qemu-devel/20210907121943.3498701-1-marcandre.lureau@redhat.com/ on lore]
* <code>Subject: [RFC 0/6] scripts: Rewrite simpletrace printer in Rust</code><br /><code>Date: Mon, 27 May 2024 16:14:15 +0800</code><br />[https://lore.kernel.org/qemu-devel/20240527081421.2258624-1-zhao1.liu@intel.com/ RFC v1]


== Minimum supported versions ==
== Work in progress ==
1.59.0 has const <tt>CStr::from_bytes_with_nul_unchecked</tt>, not really possible to go below that.
=== Paolo ===
* Waiting for Meson 1.8.1: "objects:", rustdoc,
* Later: pinned_init... maybe use MaybeUninit as a first step?


Bindgen 0.60.x has --allowlist-file. Ubuntu 22.04 has 0.59.x which is a problem.
=== Manos ===
=== Zhao ===


In the future 1.83.0 would allow nested offset_of.
=== Kevin ===
* Executing async code, block driver implementations
** https://lore.kernel.org/qemu-devel/20250211214328.640374-1-kwolf@redhat.com/


== Done ==
== Bugs ==
* only generate bindings.rs.inc once
* Add BQL abstraction for tests - avoid running them with "--test-threads 1"
* disabling the cast_ptr_alignment lint is too broad
* [https://github.com/mesonbuild/meson/issues/8828 Distros would prefer to link libstd dynamically]; Rust-enabled QEMU binaries are huge
* update bundled meson to 1.5.0
* add support for --rustc and RUSTC environment variables
* remove --no-include-path-detection from bindgen invocation
* rename subprojects to use Meson 1.5.0+ convention and add meson.override_dependency() call
* add support for cross compilation of Rust subprojects (<tt>native: true</tt> for deps of procedural macro crates)
* fix licenses for Rust subprojects


== Posted ==
== Possible project targets ==
* Enable use of rustc 1.63.0
** fix cfgs for proc-macro2
** core::ffi (stable in 1.64.0)
** "if let" usage in bilge-impl needs to be patched out (stable in 1.65.0)
** std::sync::OnceLock (stable in 1.70.0)
** MaybeUninit::zeroed() (stable in 1.75.0 in const context)
** c"" literals (stable in 1.77.0)
** offset_of! (stable in 1.77.0)
* Enable use of older bindgen
** What to do about Ubuntu 22.04? Bindgen 0.59.x does not have --allowlist-file
* Install rustc/bindgen in all containers


* feature parity for pl011
=== Nice things to have ===
** [https://lore.kernel.org/qemu-devel/CAFEAcA9X5iBFMEw3dxsgBPZo7FBuNYp3dK5gf7v+cEyVa+pErA@mail.gmail.com/T/ support pl011-luminary]
** migration
== To be posted ==
* qemu_irq (https://gitlab.com/bonzini/qemu/-/commit/b04e20648568f7b73de8af2bc5c6b0dadca3dbaf)
* QOM (https://gitlab.com/bonzini/qemu/-/commits/rust-qom (defining QOM classes entirely without macros, QOM casts and reference counting)
** also adds documentation for public APIs in qemu-api and qemu-api-macros
== TODO ==
* Remove need for manual "meson subprojects update --reset" when updating packagefiles/
* Remove need for manual "meson subprojects update --reset" when updating packagefiles/
* Remove remaining uses of Cargo, or clean up. Either all this should be driven entirely by Meson (for example via compile_commands.json, similar to scripts/check_sparse.py), or it should use a Cargo workspace so that a single invocation of "cargo clippy" will handle all crates.
* Improve tool integration
** Perform clippy checks ("ninja clippy"? Maybe "meson test --suite codecheck"?)
** Run code checks a la "meson test --suite codecheck"? (clippy and rustfmt)
** Reformat code ("ninja rustfmt"? Maybe also include it in the same suite as clippy?)
* Place rustdoc output for master somewhere?
** Generate documentation ("ninja rustdoc"? Maybe check how Linux runs doctests?)
* split qemu_api into multiple crates
* More CI
* distros need dynamically linked libstd
** Rustfmt checks?
** rewrite all main()s in Rust and use [https://github.com/mesonbuild/meson/issues/14224 -C prefer-dynamic]?
** Run nightly clippy
** or should Meson be able to link C programs with rlibs?
** Rustdoc
 
=== New code to write===
* more QOM procedural macros
** generate qdev properties
** generate parts of TypeInfo?
** monomorphization of <tt>T: IsA&lt;Class&gt;</tt> to <tt>&T</tt> (like the momo crate)
* feature parity for pl011
* feature parity for pl011
** check for new commits
* safe object creation with [https://docs.rs/pinned-init/ pinned_init, originating from Linux])
* bindings
* Generic Rust<->C interop, Error (https://github.com/bonzini/rust-qemu)
** MemoryRegion (callbacks)
* DMA (investigate vm-memory?)
** Chardev (callbacks, functions)
* Trace/log (Summer of Code)
** DMA (investigate vm-memory?)
* other experiments: https://github.com/bonzini/rust-qemu (Generic Rust<->C interop, Error)
** safe object creation: use MaybeUninit in instance_init? (long-term investigate [https://docs.rs/pinned-init/ pinned_init crate, originating from Linux])
* more QOM procedural macros (currently <tt>#[derive(Object)]</tt>)
** generate qdev properties?
** generate parts of TypeInfo?


== Ideas for lints without breaking CI ==
=== Improved clippy coverage in CI? ===
See https://github.com/bonzini/rust-qemu/commit/95b25f7c5f4e2694a85a5503050cc98da7562c7c
* run clippy as part of "make check"
* run clippy as part of "make check", possibly only if Rust is newer than some version (1.74.0 so that clippy can be configured in Cargo.toml?)
* deny many individual lints, do not deny groups (complexity, perf, style, suspicious) on regular builds. allow unknown_lints.
* add to CI a fallible job that runs on nightly clippy with <tt>-Dclippy::complexity -Dclippy::perf -Dclippy::suspicious -Dclippy::style -Dunknown_lints</tt>. the job should generally pass, and if a new lint triggers it probably should be added to Cargo.toml as either "allow" (rare) or "deny" (possibly after adding #[allow()] to the source).
* add to CI a fallible job that runs on nightly clippy with <tt>-Dclippy::complexity -Dclippy::perf -Dclippy::suspicious -Dclippy::style -Dunknown_lints</tt>. the job should generally pass, and if a new lint triggers it probably should be added to Cargo.toml as either "allow" (rare) or "deny" (possibly after adding #[allow()] to the source).


== Possible project targets ==
=== Miscellanea  ===
 
* <code>qemu-bridge-helper.c</code> Re-write SUID C executable with useful features.


=== Miscellanea  ===
=== Bindings ===


- <code>qemu-bridge-helper.c</code> Re-write SUID C executable with useful features.
* visitors/QOM properties, could be interesting for <code>hw/sensor/tmp105.c</code>...
* complete vmstate bindings (needs const_refs_static in Rust 1.83.0, Zhao+Paolo have a prototype)


=== Devices  ===
=== Devices  ===
Line 92: Line 69:
** needs block bindings
** needs block bindings
* <code>hw/timer/i8254.c</code> (claimed, WIP)
* <code>hw/timer/i8254.c</code> (claimed, WIP)
** needs timer bindgings
* <code>hw/mem/nvdimm.c</code> (suggested by [[User:Manos|Manos]])
* <code>hw/mem/nvdimm.c</code> (suggested by [[User:Manos|Manos]])


[[Category:RustInQemu]]
[[Category:RustInQemu]]

Latest revision as of 12:50, 2 May 2025

For the old RustInQemu page, see RustInQemu/2022

Active effort

Past efforts

  • [RFC v3 00/32] Rust binding for QAPI and qemu-ga QMP handler examples
    on patchew on lore
  • Subject: [RFC 0/6] scripts: Rewrite simpletrace printer in Rust
    Date: Mon, 27 May 2024 16:14:15 +0800
    RFC v1

Work in progress

Paolo

  • Waiting for Meson 1.8.1: "objects:", rustdoc,
  • Later: pinned_init... maybe use MaybeUninit as a first step?

Manos

Zhao

Kevin

Bugs

Possible project targets

Nice things to have

  • Remove need for manual "meson subprojects update --reset" when updating packagefiles/
  • Improve tool integration
    • Run code checks a la "meson test --suite codecheck"? (clippy and rustfmt)
  • Place rustdoc output for master somewhere?
  • split qemu_api into multiple crates
  • distros need dynamically linked libstd
    • rewrite all main()s in Rust and use -C prefer-dynamic?
    • or should Meson be able to link C programs with rlibs?

New code to write

  • more QOM procedural macros
    • generate qdev properties
    • generate parts of TypeInfo?
    • monomorphization of T: IsA<Class> to &T (like the momo crate)
  • feature parity for pl011
  • safe object creation with pinned_init, originating from Linux)
  • Generic Rust<->C interop, Error (https://github.com/bonzini/rust-qemu)
  • DMA (investigate vm-memory?)
  • Trace/log (Summer of Code)

Improved clippy coverage in CI?

  • run clippy as part of "make check"
  • add to CI a fallible job that runs on nightly clippy with -Dclippy::complexity -Dclippy::perf -Dclippy::suspicious -Dclippy::style -Dunknown_lints. the job should generally pass, and if a new lint triggers it probably should be added to Cargo.toml as either "allow" (rare) or "deny" (possibly after adding #[allow()] to the source).

Miscellanea

  • qemu-bridge-helper.c Re-write SUID C executable with useful features.

Bindings

  • visitors/QOM properties, could be interesting for hw/sensor/tmp105.c...
  • complete vmstate bindings (needs const_refs_static in Rust 1.83.0, Zhao+Paolo have a prototype)

Devices

  • hw/block/pflash_cfi01.c, hw/block/pflash_cfi02.c (claimed, WIP)
    • needs block bindings
  • hw/timer/i8254.c (claimed, WIP)
  • hw/mem/nvdimm.c (suggested by Manos)