Phoronix

Google's AI team has announced "Lyra" as a very low bit-rate codec for speech compression designed for use-cases like WebRTC and other video chats... With a bit rate so low that when combined with the likes of the AV1 video codec could potentially allow video chats over 56kbps Internet connections...

Mageia 8 is out today as a significant and long overdue update to this Linux distribution long ago derived from Mandriva/Mandrake lineage...

Wine 6.3 is out today as the newest bi-weekly development snapshot of this free software solution for running Windows games and applications under Linux and macOS...

Linus Torvalds just merged a set of patches that includes KFence. Short for the Kernel Electric Fence, KFence is a low-overhead memory safety error detector/validator that is suitable for use in production kernel builds...

The Linux 5.12 merge window is drawing to a close this weekend while being sent in this morning were the RISC-V updates that tend to excite the free software enthusiasts...

As a quick PSA for those that may be eager to test out early Git builds of the Linux 5.12 kernel, I've been hitting a very nasty issue on multiple systems leading to corruption / data loss...

Last week saw the main set of ACPI and power management updates for Linux 5.12 while for the second week of the merge window has been the follow-up work with Intel Simple Firmware Interface removal and also an additional ACPI update...

Last week the main set of DRM subsystem updates were sent in for the Linux 5.12 merge window. That pull included exciting additions like Radeon RX 6000 series OverDrive and Intel Xe VRR. Mistakenly left out of that pull request last week were the open-source Qualcomm Adreno driver improvements for the "MSM" kernel driver while now that code has landed...

The newest GNU project seeing its first release is GNU Poke, which is being inaugurated at v1.0 after being in development for the past three years...

Genode OS 21.02 is out as the latest feature release to this open-source operating system framework...

The new OpenGL renderer work for GTK 4 as a post-4.0 improvement is shaping up well and should really help push along the open-source toolkit on macOS...

Blender 2.92 is out today as the latest feature release for this award-winning, open-source 3D graphics/modeling software...

Intel's "Simple Firmware Interface" that dates back to the company's early Atom-powered mobile days is being eliminated with the in-development Linux 5.12 kernel...

The VFIO changes to the Linux 5.12 kernel include an optimization worth mentioning...

Last month Red Hat announced no-cost Red Hat Enterprise Linux for small production environments while now they have extended their gratis RHEL offering to also include "open-source infrastructure" being entitled to no-cost usage...

NVIDIA has updated their 460 series Linux driver to provide launch-day support for the GeForce RTX 3060...

New restrictions will be in place beginning with Ubuntu 20.04.3 LTS this summer to try to enforce better quality releases with less regressions by enforcing better quality control...

Lavapipe as Mesa's CPU-based Vulkan driver implementation akin to LLVMpipe for OpenGL can now run on Microsoft Windows...

Most of you probably haven't heard of DFI much in nearly two decades since the days of their colorful "LanParty" motherboards that were well known at the time, but these days they are focused on the industrial computer industry and have now teamed up with Canonical to partake in the Ubuntu IoT Hardware Certification Partner Program...

AMD has released AMDVLK 2021.Q1.4 as their newest snapshot of this open-source Radeon Vulkan driver for Linux systems...

The beta of GNOME 40 is out today, one month ahead of the stable release of this big open-source desktop environment update...

Last week I noted "GFX90A" appearing in the AMD LLVM back-end and now the AMDGPU Linux kernel driver patches have appeared for "Aldebaran" that appear to be the codename for the next-generation CDNA part making use of GFX90A...

Google is announcing today in cooperation with The Linux Foundation that they are providing funding for two full-time developers to focus solely on security issues...

Immediately following the publishing of the

After recently looking at the early LLVM Clang 12 compiler performance on the AMD Ryzen 9 5950X, in today's benchmarking is a look at how the GCC 11 compiler performance is looking in its near final state compared to GCC 10 under a variety of build CFLAGS/CXXFLAGS configurations on the AMD Zen 3 desktop.

Ampere Computing with their 80-core Ampere Altra processors and Mount Jade reference platform continue making progress on open-source firmware support for interested parties...

Linux 5.12 is bringing the initial infrastructure around ACPI Platform Profile support and with this kernel it's implemented for newer Lenovo ThinkPad and IdeaPad laptops. The support allow for altering the system's power/performance characteristics depending upon your desire for a speedy, quiet, or cool experience. With Linux 5.13 it looks like HP laptops with this capability will begin to see working Platform Profile support too...

Debian is the latest major Linux distribution deploying a Debuginfod web server so that ELF/DWARF/source-code information can be supplied via HTTP to clients on-demand when debugging...

An Oracle engineer has proposed introducing a new "-ftrivial-auto-var-init=" option for the GCC compiler that would allowing initializing automatic variables with either a pattern or zeroes in the name of security...

Pop open the champagne as the in-development Linux 5.12 kernel will be able to support link-time optimizations (LTO) in conjunction with the LLVM Clang compiler on not only AArch64 (64-bit ARM) but also x86_64...

After a two year hiatus on the patches, VMware's Nadav Amit has gotten back to working on current TLB flushing support for the Linux kernel in yielding a small but measurable performance improvement...

In yesterday's Linux distribution benchmark comparison on a Ryzen 9 5900X, you may have noticed some of the Ubuntu 21.04 development benchmarks coming in slower than Ubuntu 20.10 but overall a tight race. That is something I've seen now on other systems too -- such as these results to pass along today from an Intel notebook with the latest Ubuntu 20.10 vs. 21.04 development tests...

Thanks to Google engineers the Linux 5.12 kernel is providing punctual support for eMMC inline encryption that is being ratified with a forthcoming specification update and already being found within some mobile hardware...

Firefox 86.0 is out today as the latest monthly update to this open-source web browser that continues to work on ramping up its security offerings...

The open-source Mesa Radeon Vulkan driver "RADV" has now enabled displayable DCC (Delta Color Compression) support that should yield some performance benefit while there still is more work to be completed...

The HID subsystem updates were sent in this morning for the ongoing Linux 5.12 merge window...

GNOME 40 beta is coming together with the GNOME Shell and Mutter components having seen new releases on Monday...

Libinput 1.17 is out as the newest version of this Wayland/X.Org input handling library for the Linux desktop...

Released yesterday was dav1d 0.8.2 as a fairly significant update to this AV1 CPU-based decoder. For those wondering what this update means for performance, here are some initial benchmarks...

The first release candidate of systemd 248 is now available with a number of improvements ranging from a new "system extensions images" concept to the out-of-memory daemon (OOMD) being declared stable...

One of many interesting and original open-source projects to be started in 2020 was ZLUDA, an open-spurce drop-in CUDA implementation for Intel graphics. ZLUDA - developed independent of Intel and NVIDIA - is built atop Intel's oneAPI Level Zero interface (hence the name, ZLUDA) and allows for unmodified CUDA applications to run on Intel UHD/Xe Graphics hardware with near-native performance. Well, that's the goal at least but with the initial ZLUDA release were a number of support limitations...

The changes within the remote direct memory access (RDMA) subsystem for Linux 5.12 are deemed "quite small" but there is one interesting addition courtesy of Intel...

Given the recent release of Arch Linux based EndeavourOS and a Phoronix Premium supporter recently pointing out SalientOS as another interesting Arch-based Linux distribution, here are benchmarks showing how these easy/quick to deploy Arch based operating systems with sane defaults compare to that of Ubuntu, Fedora Workstation, and Intel's own Clear Linux. This round of February 2021 Linux benchmarking was carried out on an AMD Ryzen 9 5900X desktop stemming from premium member feedback.

Microsoft engineers continue increasing their contributions to the Linux kernel where it makes business sense for them, such as in the case of securing the Azure cloud given that around 50% or more of the instances run Linux. With Linux 5.12 there are integrity subsystem improvements coming from Microsoft...

Sent in last week were some AMDGPU "fixes" for Linux 5.12. While there are some fixes as part of the series, there are some new (minor) features enabled...

Dav1d is already the most performant and leading AV1 software decoder we have seen while out today is v0.8.2 that should speed-up the video decode process even more on modern x86/x86_64 and ARM hardware...

Greg Kroah-Hartman sent in the staging/IIO updates for the Linux 5.12 kernel and this time around are lighter than normal but still with a few interesting items worth mentioning...

Following the recent major rewrite to KDE's KWin compositor code there are more exciting improvements likely to come for KWin in improving its Wayland compositor support...

The block subsystem and related storage changes were merged today for the in-development Linux 5.12 kernel...

Earlier this month at the virtual FOSDEM 2021 conference was an interesting presentation on how European developers are preparing for AMD-powered supercomputers and beginning to figure out the best approaches for converting existing NVIDIA CUDA GPU code to run on Radeon GPUs as well as whether writing new GPU-focused code with OpenMP device offload is worthwhile...