News › linux
By The Linux Camp News, October 20, 2026
Ojeda's October 18 submission targets Rust infrastructure, testing and kernel APIs. The large library import is mostly mechanical; Binder logging and bitmap bounds are more consequential changes to review.
Miguel Ojeda submitted the next round of Rust support for Linux 7.4 on October 18, 2026. This is a pull request covering shared Rust infrastructure and kernel APIs, not a release announcement. The supplied sources, retrieved October 20, do not establish mainline acceptance. 2 1
Much of the volume comes from updating the bundled Zerocopy library to v0.8.62, advancing eight releases. Zerocopy supplies traits and macros for converting between raw bytes and structured types. Ojeda reports almost no required kernel-side adaptation, so the import's size should not be read as equivalent growth in kernel functionality. 2 1
Rust doctests, tests derived from documentation examples, would generate KUnit suites through the existing #[kunit_tests] and #[test] attributes rather than C foreign-function interfaces. Reusing that Rust test machinery removes the generated C glue file: the change simplifies how tests connect to the kernel test infrastructure. 2
For reproducible-build work, check the rustdoc version and whether the build uses O=. The submission trims generated doctest function names specifically for rustdoc older than 1.95.0 with O= builds, where those names were not reproducible. That stated condition is narrower than a general Rust build reproducibility fix. 2
Rust Binder would adopt new rate-limited printing macros built on a wrapper around the kernel's ratelimit_state. This gives Rust code a way to limit repeated messages, including printing triggered by userspace. For operators reviewing Binder logs, the relevant change is message rate limiting, not a demonstrated system-wide performance gain. 2
The bitmap API also caps lengths at i32::MAX. Ojeda says this fixes truncating u32 casts in copy_and_extend that could otherwise cause out-of-bounds writes. That is a maintainer-described correctness fix; the supplied account does not establish affected releases, exploitability or stable-backport plans. 2
Integration remains a separate check. Ojeda reported at least two linux-next rounds and a conflict-free test merge, but expected conflicts with char-misc, DRM and Kbuild. Those checks support integration readiness, not proof that the submitted changes were accepted unchanged. 2