Linux Kernel News Archives

Following Sunday's release of Linux 6.0-rc1, yesterday saw the release of 6.0-rc1-rt1 as the set of patches for providing real-time kernel support atop the upstream code-base. There is just roughly 50 patches to go until the PREEMPT_RT functionality is mainlined!

After Qualcomm announced their Cloud AI 100 Accelerator back in 2019, in 2020 during the early days of the pandemic they posted a Linux driver for this accelerator. That driver didn't get picked up for the mainline Linux kernel and two years later there still is little fanfare around the Qualcomm AI Cloud Accelerator hardware. However, now they have posted a new Linux driver that goes the DRM driver route.

While Linux 6.0 will bring a lot of shiny new features, Multi-Gen LRU (MGLRU) is one of the anticipated changes that isn't going to land now until Linux 6.1. But in the interim, MGLRU v14 was posted today that re-bases the code against 6.0-rc1 to help facilitate more testing of this kernel change that primarily helps Linux systems under memory pressure.

After the two week long merge window, Linus Torvalds this afternoon released the first release candidate of Linux 6.0. Over the next roughly two months the Linux 6.0 kernel will stabilize but already from my early testing on various systems it is in nice shape and the features and performance are looking great.

While support for the loongArch Chinese CPU architecture was merged in Linux 5.19, it wasn't actually enough to yield a booting system due to some driver code not yet being finished and ready for merging in time. LoongArch was allowed to merge that preliminary code in v5.19 so the Glibc support could land and now for Linux 6.0 more of the CPU port is ready to hit the kernel.

A number of new Linux kernel stable releases are out this week with new mitigations around the latest batch of published CPU security vulnerabilities. Linux 5.19.1, 5.18.17, 5.15.60, 5.10.136, 5.4.210, and 4.19.255 are the new releases out today.

Following recent upstream discussions around the -O3 compiler optimizations for the Linux kernel, the Kconfig switch advertising this option is being removed in Linux 6.0.

The EFI changes were merged last week for the Linux 6.0 cycle and contain two notable improvements on the ARM64 side.

Today's busy patch Tuesday for Intel continues with the Linux kernel getting mitigated for EIBRS Post-barrier Return Stack Buffer (PBRSB). This PBRSB is the latest handling on the "CPU vulnerability nightmares front", the pull request calls it.

A patch coming about earlier this year allows setting the system's hostname before user-space starts by way of the hostname= kernel parameter. That patch has now landed as part of Andrew Morton's accumulated changes for Linux 6.0.

The "char/misc" changes were merged a few days back for the Linux 6.0 kernel with this pull being the rather "random catch-all" area of the kernel for drivers not fitting within other subsystems. Most notable with the char/misc updates for Linux 6.0 is introducing support for Intel's Habana Labs Gaudi2.

The IBM Power CPU architecture updates have been merged for the in-development Linux 6.0 kernel.

Along with his various other pull requests for areas of the kernel he oversees, Greg Kroah-Hartman submitted the Linux 6.0 staging changes this week.

BUS1 started out as a Linux kernel IPC module following the failure of KDBUS and while there still are occasional commits to that out-of-tree BUS1 kernel module, the involved (Red Hat) developers have been primarily working on Dbus-Broker as the high performance, user-space D-Bus implementation that delivers greater speed and reliability over the reference D-Bus code. Now also popping up under the BUS1 umbrella is "r-linux" as a Rust-written, capability-based Linux runtime.

Earlier this year a developer stepped up willing to maintain Linux's FBDEV subsystem for frame-buffer device drivers since it fell into an unmaintained state in 2016 but even prior to that had been on the decline in the era of more proper DRM/KMS drivers. Helge Deller continues that work overseeing the frame-buffer device "FBDEV" subsystem and this week sent in the new patches for Linux 6.0.

Support for old NEC VR4100 CPUs based on the MIPS R4000 core is being removed with the Linux 6.0 kernel leading to devices like the old IBM WorkPad Z50 no longer being supported.

While OpenRISC has been around a decade longer than RISC-V and its original support in the Linux kernel dates back to the v3.1 days, on the hardware side OpenRISC hasn't enjoyed nearly as much success as RISC-V and its kernel support not advancing nearly as rapidly. Now with Linux 6.0, OpenRISC is finally exposing PCI bus support.

Earlier this week saw the Rust for Linux v8 patches posted that introduced a number of new abstractions and expanding the Rust programming language integration to more areas of the kernel. Those patches amounted to 43.6k lines of new code while "Rust for Linux v9" was posted today and comes in at just 12.5k lines of new code.

A lot of great features are landing for Linux 6.0 but two more are now confirmed to partake in this next major kernel version: Multi-Gen LRU (MGLRU) and the Maple Leaf data structure.

Another big ticket feature has made it for the Linux 6.0 kernel: the Runtime Verification infrastructure for running Linux on safety-critical systems.

Direct Rendering Manager (DRM) subsystem maintainer David Airlie has submitted all of the open-source GPU/display driver updates for the Linux 6.0 (nee 5.20) merge window. Much of this work is centered around bringing up the Intel Arc Graphics DG2/Alchemist discrete graphics cards and on the AMD side continued work around enabling RDNA3 and Instinct MI300 graphics processors.

As part of getting PREEMPT_RT support ready to be mainlined at long last for real-time kernel builds, a number of Linux 6.0 (nee 5.20) pull requests have revolved around additional RT preparatory changes. With the printk changes for Linux 6.0, the console drivers will now be skipped in RT mode.

Ingo Molnar today submitted the main set of kernel scheduler updates for the in-development Linux 6.0 (nee 5.20). The scheduler updates contain some notable changes that will be interesting to benchmark in the days ahead.

Following yesterday's release of Linux 5.19 stable the FSFLA folks maintaining the GNU Linux-libre kernel released their downstream version that strips out support for loading proprietary kernel modules as well as stripping out drivers/support that requires non-free/closed-source microcode/firmware files.

In case you missed it in yesterday's Linux 5.19 announcement and to avoid reader questions/confusion in the days ahead, just making it loud and clear here: what was referred to as the Linux 5.20 kernel in development will most likely be called Linux 6.0.

Linus Torvalds just released Linux 5.19 as stable for the newest version of the Linux kernel. He also mentioned this is the first time he released the new Linux kernel from an ARM64 laptop in the form of an Apple MacBook running an AArch64 Apple M2 SoC.

You may recall a month ago the lone developer still working on open-source VIA x86 graphics support for Linux hoped to finally mainline this "OpenChrome" DRM/KMS driver for the Linux 5.20 cycle. Well, Linux 5.19 is being released today and that opens up the Linux 5.20 merge window but still the OpenChrome DRM driver isn't ready to go yet.

Submitted early ahead of tomorrow's Linux 5.19 stable kernel release are the SoC changes destined for the Linux 5.20 merge window. There are more than one thousand SoC patches for Linux 5.20 cycle adding and updating many SoCs and board/platform coverage. One of several notable additions this cycle is introducing the Qualcomm Snapdragon 8cx Gen3 support for high-end Arm laptops.

The Linux 5.19 kernel is to be released this weekend and in turn will mark the start of the Linux 5.20 merge window. Based on various Git "-next" queues and mailing list indications, here is a look at some of the changes expected for the Linux 5.20 kernel.

Linux 5.19 is set to be released this weekend, so here is a reminder about some of the big ticket items to find in this summer 2022 kernel release.

After years in the works, the "PREEMPT_RT" support for building a real-time Linux kernel might finally be mainlined for the upcoming Linux 5.20 cycle if the last few remaining bits are reviewed/signed-off on in time for next week's merge window.

ByteDance as the Chinese company behind TikTok has been working on a number of Linux kernel optimizations the past few years and their most recent work is for faster Kexec rebooting of the kernel. With their massive fleet of servers powering TikTok and other apps, they will do whatever they can to shave milliseconds off the boot/reboot time of their servers and that is what most of their Linux optimizations have been about -- including this newest patch series for faster Kexec reboots.

Samsung has posted a set of Linux kernel patches for enabling their Trinity neural processing unit (NPU) hardware as accelerators found within some of their embedded systems. Samsung Trinity NPUs have been in use since 2018 and their newer Trinity "TRIV2" accelerator is found within the latest Samsung TVs for AI purposes. Samsung hopes to upstream this new Linux driver simply named "trinity" into the mainline kernel.

As was expected given all the recent kernel activity, Linux 5.19-rc8 was released today rather than going straight to the Linux 5.19 stable release, which has been diverted now to next Sunday. Linux 5.19-rc8 ships with many last minute fixes for this summer 2022 kernel update.

While normally big CPU security mitigation work done behind closed-doors is in good shape for the vulnerability embargo date, Retbleed has been an exception. Nearly two weeks since Retbleed was made public, the Linux kernel patches around it continue with more now sent in today ahead of Linux 5.19-rc8 to address fallout from the mitigation handling.

Toshiba's Visconti SoC provides an optimized image recognition processor and geared for advanced driver assistance system (ADAS) solutions for automobiles and similar modern use-cases. Toshiba engineers are now publishing patches for their DNN image processing accelerator driver with hopes of getting the code upstreamed into the mainline Linux kernel.

Nadav Amit who previously spearheaded work on reducing unnecessary TLB flushes, concurrent TLB flushes, and other low level optimizations over the years. The latest work is now on "relaxed" TLB flushes as another low-level performance improvement.

Stemming from my article last week noting how Linux 5.19 Git broke Intel Alder Lake P graphics support due to requiring new firmware while not retaining backwards compatibility with the existing Intel GuC firmware, a solution is still being worked on prior to Linux 5.19 final whether it be a revert or the proposed patch working on GuC v69/70 firmware compatibility. Linux firmware guidelines are also being proposed to ensure kernel developers in the future don't try to break firmware support guarantees.

Linus Torvalds just released Linux 5.19-rc7 following a busy week due to the Retbleed security mitigation and not only the CPU overhead performance impact it puts on users but the mess it has on kernel development especially when it comes to embargoed issues that make the patches difficult to review/test well prior to embargo lift.

The AMDGPU kernel graphics driver had been preparing to make use of the buddy allocator started by the Intel Linux graphics driver. But now with today's batch of DRM fixes that AMDGPU support is being reverted for Linux 5.19 due to it causing garbled screens for some users.

Merged last year for Linux 5.16 was the FUTEX2 system call for improving the FUTEX interface on Linux and with the original motivation for those improvements to wait on multiple futexes so Wine/Proton can better match Microsoft Windows' behavior for a more optimal Linux gaming experience. Beyond the performance/efficiency benefit of FUTEX2, other improvements have also been talked about while we are now beginning to see action on one of them: NUMA awareness.

Following the upstream discussions over -O3'ing the Linux kernel last month I ran some fresh benchmarks of the Linux kernel built with -O2 versus -O3. After the -O3 optimized kernel build results weren't too impressive, a number of Phoronix readers were virtually shouting that "-O3 -march=native" is where it's at for fun and performance... To appease those even though in the past it hasn't proven worthwhile and upstream kernel developers are against it, here are those numbers.

Rob Clark as the lead developer of the MSM DRM kernel driver and the Freedreno/TURNIP Mesa drivers for open-source Qualcomm Adreno graphics driver support has submitted the Direct Rendering Manager driver changes for the upcoming Linux 5.20 merge window.

Linus Torvalds just announced the availability of Linux 5.19-rc6 as the latest routine test release for the upcoming Linux 5.19.

A set of proposed patches promise to make the Linux kernel's memchr() implementation faster for locating a character within a block of memory. In tests carried out by the developer, the new implementation can be nearly four times faster on large searches.

In addition to the OpenChrome DRM/KMS driver hoping to be finally mainlined in 2022 for supporting aging VIA graphics hardware from the long-ago days of their x86 chipsets, separately there is a DRM/KMS kernel driver in the works for something even older... A Linux DRM graphics driver for the Atari Falcon from the early 90's.

The newest version of the MGLRU patches were posted and I've also finished up some more benchmarks of this Multi-Gen LRU kernel feature.

MGLRU as the "Multi-Gen LRU" for reworking the Linux kernel's page reclamation code to be less taxing on the CPU and making better choices continues to look very good for the future of Linux performance.

A patch from TUXEDO Computers that is now queued into the input subsystem's "next" branch for Linux 5.20 adds a number of quirks to fix several Clevo / TUXEDO laptops from touchpad and keyboard issues after suspending the system.

Being worked on the past several years by Google engineers and others has been the KernelMemorySanitizer (KMSAN) that has already found more than 300 kernel bugs even prior to being mainlined. Sent out prior to the US holiday weekend as the fourth iteration of these patches, building off the "request for comments" sent out in 2020.