NEWSDESK
Jon Masters summarises the latest happenings in the Linux kernel, so that you don’t have to.
Linus Torvalds has announced the release of Linux 6.9, noting L that “the whole release has felt pretty normal”. The new kernel adds support for pidfds by individual threads (not just processes) – simplifying race-free process management in userspace – the new FRED (Fast Return and Event Delivery) mechanism present on future Intel processors (intended to simplify and speed up interrupt delivery), host support for AMD’s Confidential Compute feature SEV-SNP (Secure Nested Paging), along with a large number of other features. As usual, KernelNewbies has an excellent summary of the new features, which you can find here: https://kernelnewbies.org/Linux_6.9.
HANGING AROUND
With the release of Linux 6.9 came the opening of the merge window (period of time during which disruptive changes are allowed to be made to the kernel code) for what will be Linux 6.10 in a couple of months. The merge window was open for the customary two weeks, and was then closed with the release of Linux 6.10-rc1. In addition to removing support for older Alpha chips, 6.10 adds a lot more Rust language infrastructure (including initial support for Rust on RISC-V), initial – but not yet enabled – support for the ntsync Windows NT (aka modern Windows) sychronisation primitives (similar to futexes but different enough to require a special subsystem, which should speed up games running under Windows emulation in Wine by as much as 150%), and the new mseal system call that effectively locks the address space of an application against future changes.
A few regressions in 6.9 were highlighted by Thorsten Leemhuis, including an AB-BA (competing spinlocks acquired by two processes in the opposite order) deadlock in some code used to manage LEDs on network cards that was apparently causing a bunch of users to experience unexpected hangs when booting 6.9.
