Linux Kernel News Archives

The crypto subsystem updates for the Linux 6.7 kernel includes the usual churn like various crypto acceleration updates for different SoCs and other routine changes plus is also limiting the role of some insecure and/or obsolete crypto hashing algorithms.

One of the last major blockers before the remaining real-time "PREEMPT_RT" patches can be upstreamed is sorting out threaded / atomic console printing. With the in-development Linux 6.7 kernel, there's been more work upstreamed in that endeavor.

Overnight the mainline Linux kernel has retired support for Intel Itanium (IA-64) processors.

Since Linux 6.6 we've been seeing work upstreamed for sysctl working to remove its sentinel, the final empty element on each sysctl array. This will cut-down on around 64 bytes of bloat per array, help with kernel build times, and an all-around improvement. With Linux 6.7 more of the sysctl changes are ready and hopefully for the v6.8 kernel cycle next year the effort will be completed.

Sent out this morning was the Direct Rendering Manager (DRM) feature pull request of updated graphics/display drivers for the in-development Linux 6.7 kernel. Notable this round is Intel Meteor Lake integrated graphics now being considered stable / enabled out-of-the-box, Intel Lunar Lake graphics support has started to get underway, and AMD continues working on their upcoming hardware platforms.

Linux's FSCRYPT file-system encryption framework allows for native file encryption support on the likes of EXT4, F2FS, and UBIFS. FSCRYPT can make use of inline encryption hardware for accelerating the file-system encryption support and with the Linux 6.7 kernel will work for more scenarios.

As anticipated Linus Torvalds just released the Linux 6.6 kernel as stable.

From the perspective of Linux distributions trying to reduce their attack surface while still making it possible for users to run legacy software without recompiling their kernel, SUSE has spearheaded the effort for boot-time enabling/disabling of x86 32-bit support for whether 32-bit user-space programs and 32-bit system calls can be executed. That code has been submitted for the imminent Linux 6.7 merge window.

One of the many early pull requests sent in for Linux 6.7 were the x86/boot changes that are headlined by a rework to the PE header generation in order to generate a modern, 4K-aligned kernel image view to ultimately aim for better system security.

With Linux 6.6 expected to be released tomorrow as stable, the Linux 6.7 merge window in turn will be opened. Here's a preview of some of the changes expected for this next kernel cycle.

Tomorrow the Linux 6.6 kernel is expected to be released as stable unless Linus Torvalds has last minute reservations and decides to extend the cycle by an extra week. While there were many last minute fixes this week, the changes don't appear to be too scary or invasive. In any event the Linux 6.6 kernel is bringing some exciting features.

The VKMS driver as a virtual kernel mode-setting solution as a software-only implementation of KMS for use with primarily headless Linux systems continues tacking on more features. Maíra Canal and Melissa Wen presented last week at XDC 2023 on the VKMS driver that continues proving very useful for headless platforms as well as for testing purposes.

Linus Torvalds today released the seventh weekly release candidate of Linux 6.6 while is hoping to release the stable kernel version next weekend.

Sent out today were a new batch of drm-misc-next changes intended for Linux 6.7. There's various fixes to the smaller Direct Rendering Manager (DRM) drivers, DRM VM_BIND async documentation, and other small changes. Plus the DRM_GPUVM code has been re-licensed from being GPLv2-only to now GPLv2 or MIT.

Google is proposing a new mseal() memory sealing system call for the Linux kernel. Google intends for this architecture independent system call to be initially used by the Google Chrome web browser on Chrome OS while experiments are underway for use by Glibc in the dynamic linker to seal all non-writable segments at startup.

Linus Torvalds just released Linux 6.6-rc6 and expressed his intention of releasing Linux 6.6 stable in two weeks unless anything wild happens as we approach the end of this kernel cycle.

Queued up this week as part of the HID subsystem fixes ahead of today's Linux 6.6-rc6 kernel test release has been a rather embarrassing bug: unplugging Logitech USB receivers has for the past several months been causing the Linux kernel to crash. After a number of bug reports around this issue from unplugging Logitech keyboard/mice receivers to simply switching away on a USB switch with the device(s) attached, the Linux 6.6-rc6 kernel is carrying the fix and it's also marked for back-porting to existing stable Linux kernel series.

As we approach the Linux 6.6 stable release in a few weeks, Linus Torvalds today released Linux 6.6-rc5 with everything looking "normal" for this week's test release.

A set of patches posted last week can improve the Linux kernel's performance of accounted kernel memory allocations by around 30%.

Currently when hibernating a Linux system LZO compression is used for preserving the memory contents while a new patch series posted today by Qualcomm allow for changing out the compression API used and makes LZ4 a new option during Linux hibernation.

Rust bindings are coming to the Linux kernel's workqueue (WQ) infrastructure that is used for deferring work to a kernel thread for asynchronous process execution.

Linus Torvalds just released Linux 6.6-rc4 as the newest test release for this kernel that is looking to be released as stable around the end of October.

With today's release of kmod 31, Linux's modprobe utility for loading kernel modules can finally allow arbitrary paths to allow loading new kernel modules from anywhere on the file-system.

With roughly just about one month to go until the stable release, Linux 6.6-rc3 was released today as the newest test release of Linux 6.6.

Linus Torvalds released Linux 6.6-rc2 today and it also happens to be 32 years to the day since he introduced the Linux 0.01 kernel version.

Following SLOB's removal and SLAB being deprecated and set for removal, the Linux kernel is all-in on the SLUB allocator. A new patch series posted on Friday is aiming to help prevent the possibility of cross-cache attacks with the SLUB memory allocator in the kernel.

The threaded / atomic console support for the Linux kernel is the last main blocker before the real-time "RT" patches can finally be mainlined.

The Linux kernel currently allows disabling support for 32-bit programs and 32-bit system calls at compile-time, but a new option expected to be introduced with Linux 6.7 this winter will provide a new "ia32_emulation" boot time option to allow unconditionally disabling support for 32-bit programs and system calls.

Still not over the finish line for the mainline Linux kernel are the real-time "RT" patches but it continues inching toward that long-awaited goal. Out today is v6.6-rc1-rt1 as the real-time kernel patch series updated against the initial Linux 6.6 Git state.

During the Linux 6.6 merge window a pull request submitted the new "GenPD" subsystem. While the pull request did land for Linux 6.6, Linus Torvalds took issue with it -- not because of the code but over the lack of clarity on what "GenPD" is for those not domain experts in this area. To help clear things up, GenPD is being renamed to "pmdomain" to provide a bit more clarity.

With the Linux 6.6 merge window the cgroup changes brought one change worth mentioning.

The Linux 6.6 merge window is over as the period by which new features and improvements are added to the kernel. Linux 6.6-rc1 is out the door as the kernel developers and testing community begin stabilizing this next major Linux kernel release. With Linux 6.6 there are many exciting feature additions but also one notable addition that once again didn't make the cut.

In addition to the EEVDF scheduler replacing the CFS code in Linux 6.6, another fundamental and interesting change with Linux 6.6 is on the workqueue (WQ) side with a rework that can benefit systems with multiple L3 caches like modern AMD chiplet-based systems.

In addition to all the open-source kernel graphics/display driver updates for Linux 6.6, merged this afternoon ahead of the Linux 6.6-rc1 tagging is merging of the DRM continuous integration (CI) code to hopefully lead to better testing of DRM subsystem/driver changes.

While last week saw the main set of power management updates for the Linux 6.6 kernel, this week a follow-on pull request was submitted -- and merged -- with a few last minute additions.

Merged overnight were the perf subsystem tooling updates for the Linux 6.6 merge window that is set to close later on Sunday.

The USB/Thunderbolt subsystem updates were merged this past weekend for the ongoing Linux 6.6 kernel merge window.

Masahiro Yamada submitted today all of the Kconfig feature updates targeting the Linux 6.6 merge window. Most notable with the Kconfig work is the "make xconfig" GUI seeing a port for Qt6 toolkit compatibility.

A set of printk clean-ups were sent in today for the Linux 6.6 merge window. These clean-ups are important as they are a stepping stone towards the threaded / atomic console printing and in turn that is the last major blocker before the real-time (PREEMPT_RT) support can finally be upstreamed in the kernel.

As part of the "smp/core" changes that were merged last week for the Linux 6.6 kernel, partial SMT enablement landed for processors that support more than two threads per physical core to allow greater run-time control over just how many threads to enable.

Among the many changes to land during this first week of the Linux 6.6 cycle were the x86 CPU microcode loader updates that now unconditionally makes that support part of the x86/x86_64 kernel builds.

The Direct Rendering Manager (DRM) subsystem updates for the Linux 6.6 kernel to provide new features for the number of kernel graphics drivers and other AI accelerator drivers within the mainline kernel.

A pull request was sent out this week to introduce the GenPD subsystem and add a new drivers/genpd area for the GenPD provider drivers to the kernel... To which Linus Torvalds wonders what the heck is a "GenPD" as probably most of you are also wondering about.

Some code cleaning within the sysctl space of the Linux kernel will eventually eliminate around 64 bytes of bloat per array within the kernel where a sentinel can be removed.

The EEVDF scheduler code has been merged for the in-development Linux 6.6 kernel. EEVDF replaces the existing CFS scheduler code. There is the likelihood of some performance regressions initially though but the developers will be working to address them as they arise. Additionally, this scheduler pull also re-introduces cluster scheduling for Intel Core hybrid processors.

Building off yesterday's release of Linux 6.5, GNU Linux-libre 6.5-gnu is now available for this downstream kernel maintained by the Free Software Foundation Latin America crew that removes support for binary-only kernel modules and stripping out other kernel code that depends on non-free-software microcode/firmware and other elements not deemed in the interests of pure free software.

While at the start of Linux 6.5 cycle Linus Torvalds was concerned this release "may be one of those releases that may drag out", in the end it is releasing today and right on schedule. Linux 6.5 was just published rather than going into overtime without any extra release candidate.

Ahead of the Linux 6.5 stable kernel being released there have already been a few pull requests submitted of new feature material for Linux 6.6. Among the early pulls for v6.6 is one from prominent Linux developer Christian Brauner at Microsoft that introduces the fchmodat2() system call as an evolution of fchmodat().

While there are many exciting features with Linux 6.5 which will hopefully be released as stable later today, ahead for the imminent Linux 6.6 merge window are another hearty batch of new features set to be introduced.

While Linux 6.5 is expected for release tomorrow, the flow of last minute fixes isn't over.