Linux 4.17 has been released
Summary: This release adds support for AMD Radeon Vega 12 and it enables the "display code" by default in supported AMD Radeon GPUs; it also adds a kernel TLS receive path; a more efficient idle loop that prevent CPUs from spending too much time in shallow idle states; eight unmaintained architectures have been removed and another, the Andes NDS32 architecture, has been added; XFS has gained lazytime support; the CPU load estimations have been modified; Intel Cannonlake GPUs are supported and a kernel memory consistency model has been added; and many new drivers and other improvements.
AMD Radeon Vega 12 support + "Display Code" enabled by default
This release adds support for AMD Radeon Vega 12.
It also enables by default the display code added in Linux 4.15
Code: commit
Kernel TLS receive path
Linux 4.13 added
Code: commit
A more efficient idle loop
This release reworks the idle loop in order to prevent CPUs from spending too much time in shallow idle states. It reduces idle power on some systems by 10% or more and may improve performance of workloads in which the idle loop overhead matters.
Code: commit
Removal of unused architectures
Eight architectures have been removed due to lack of maintenance: score, blackfin, CRIS, frv, m32r, metag, mn10300 and tile
Recommended LWN article: Shedding old architectures and compilers in the kernel
Code: commit
New NDS32 architecture
While eight unmaintained architectures are being removed, one has been added: NDS32. The Andes nds32 architecture
Code: (including interrupt controller driver and timer driver). It is able to boot to shell and passes most LTP-2017 testsuites in nds32 AE3XX platform Y n g:|Y r T:|Y ǧ:|Y Z:|Y n 8:|Y W ]:|Y 8Y :|Y _ :|Y h Hm:|Y &:|Y ~ ٫:|Y F wX:|Y F:|Y%Code: [[https://git.kernel.org/linus/7f65ea42eb00bc902f1c37a71e984e4f4064cfa9|commit
XFS lazytime support
XFS adds support for the lazytime, a feature that helps to improve the performance impact caused by atime updates.
Code: commit
Better load estimation
This release changes the algorithms used to estimate the CPU load to make it a better fit for the scheduling of tasks common in embedded mobile use-cases, without affecting other classes of workloads.
Recommended LWN article: Toward better CPU load estimation
Code: commit
Intel Cannonlake support
Support for Intel Cannonlake GPUs is enabled in this release.
Code: commit
Latest protections against CPU bugs
This release adds support for SSBD mitigation
Code: commit
Provide a Linux-kernel memory model
This release adds the Linux Kernel Memory Consistency Model (LKMM) subsystem, which is an an array of tools in tools/memory-model/ that formally describe the Linux memory coherency model (a.k.a. Documentation/memory-barriers.txt), and also produce 'litmus tests' in form of kernel code which can be directly executed and tested.
Recommended LWN article: A formal kernel memory-ordering model (part 1)
Code: commit