The past few weeks we have eagerly been following a Linux kernel patch series after AI found some major bottlenecks in the Linux kernel build code. With AI helping to address these bottlenecks, Linux kernel build times are substantially reduced. These optimizations might be ready in time for Linux 7.4 for reducing incremental kernel build times as much as around 70%. This week the latest iteration of these patches were posted by Arm engineer Lorenzo Stoakes.

The third revision is a set of 20 patches sent out by Lorenzo Stoakes for significantly speeding up kernel build times. The patches are rebased against the latest upstream Linux Git, there is addressing of Sashiko-reported warnings, and other mostly minor changes.

We've talked a lot about the very exciting gains found on AMD EPYC and Threadripper with these patches while the Apple Silicon with the M2 SoC also shows very positive gains too. These patches are a major win for the Linux kernel at large.

With the Apple M2 tests on Linux, a full kernel build with Clang was around 10% faster, an incremental kernel build 55% faster with Clang or 45% faster with GCC, and a no-op build (make on an already compiled kernel tree) was 74~76% faster. Great numbers and in a similar ballpark to the AMD x86_64 numbers.

The v3 patches for those interested can be found on the LKML.

Very meaningful time savings for those compiling the Linux kernel from source. Hopefully these efficiency improvements will still be on track to be ready for landing in Linux 7.4. With the time it used to take to compile the Linux kernel often provided a good enough window to get a cup of coffee or other beverage. With today's processors those kernel build times are much lower in the past while these Kbuild improvements being worked on by Lorenzo Stoakes further closes that coffee window.