IBM s390 patch series moves Rust support into Linux kernel review

IBM’s s390 is now the latest proving ground for Rust-for-Linux, as Jan Polensky posted the first patch set needed to build Rust code on IBM Z hardware. The series landed on LKML on Tue, 12 May 2026 at 12:59:20 +0200 under the subject “[PATCH 4/4] s390: enable Rust support,” putting a long-lived mainframe architecture into the kernel’s Rust review queue.

That matters because this is not a finished driver rollout. The series is architecture glue, the kind of groundwork the Linux kernel needs before a new target can fully participate in the Rust infrastructure. It wires s390 in as a Rust-capable 64-bit architecture, adds the assembly interfaces Rust needs for WARN and BUG reporting, makes room for static branch support, and adjusts bindgen parameters so Rust can cope with the packed and aligned layouts common in s390 code.

For Rust in the kernel, that kind of plumbing is the real milestone. The work is less about abstract language support than about proving that Rust’s safety model can coexist with the Linux kernel’s low-level rules on a conservative platform with strict layout and calling-convention expectations. LWN.net also noted that the current s390 setup still requires a nightly rustc because of -Zpacked-stack, while the minimum Rust tool version gating is being adjusted for the architecture.

The patch series arrives against a backdrop that already shows how far Rust-for-Linux has come. The Linux kernel’s Rust documentation says support was merged into mainline in v6.1 as an experiment to determine whether Rust was worth the tradeoffs. The current architecture-support documentation lists arm, arm64, loongarch, riscv, um, and x86 as working targets, with x86 limited to x86_64 and riscv limited to riscv64. s390 would expand that list further, giving kernel developers another major platform where Rust code can eventually compile and run.

That is why this patch set resonates beyond one architecture checkbox. IBM Z and s390 carry hardware and software lineage back to the 1990s, and getting Rust support into review there signals that Rust in the kernel is moving from experiment toward infrastructure. The series may be small, but on a platform defined by reliability, longevity, and exacting compatibility, even the arch glue is a meaningful vote of confidence.