AMD has issued a nice end-of-year update to the AMD Optimizing C/C++ Compiler (AOCC) that also includes Fortran support as well as a new release of their AMD Optimizing CPU Libraries (AOCL)...
GCC 12 is nearing release in a few months time as the annual feature update to the GNU Compiler Collection for this Free Software Foundation backed code compiler. On top of new C/C++ language features and various optimization improvements, there is updated tuning for Intel's new Alder Lake processors. Here are some early GCC 11.2 vs. GCC 12 development benchmarks looking at the performance on a Core i5 12600K.
Fwupd 1.7.3 is out today as the newest version of this open-source software that integrates with the Linux Vendor Firmware Service (LVFS) for offering streamlined system and device firmware updates under Linux...
Queued up as part of the x86/core changes intended for the Linux 5.17 cycle is dropping of the AMD 3DNow! code within the kernel. While 3DNow! brings back fond memories from the days of AMD's K6 and early Athlon processors, AMD deprecated the instructions a decade ago and no longer found in newer processors. Removing of the 3DNow! kernel code is being done as part of some code improvements...
Ubisoft issued a new job posting for a Linux developer, which has many Linux gamers excited especially as it's mentioned for an "unannounced project." Unfortunately, contrary to all the emailed tips in overnight about the job posting, it ultimately will likely prove to be of little interest to the gaming community...
It looks like EXT4 come Linux 5.17 could be making use of the kernel's new mount API. Queued up into EXT4's "dev" branch is transitioning the EXT4 file-system driver to using the kernel's modern mount API...
AMD's "RDNA" class GPUs support wavefront sizes of 32 and 64 compared to older GCN GPUs at 64 threads. Going back to 2019 RadeonSI began making use of Wave32 for some shaders but now for Mesa 22.0 next quarter there are greater Wave32 improvements that have landed...
Linus Torvalds just released Linux 5.16-rc5 and while things are looking normal at this stage, he announced that this 5.16 cycle will drag on longer due to the Christmas / New Year's holidays...
One of the Intel patch series we have been waiting to see mainlined since all the way back in 2018 is around per-client GPU metrics reporting for being able to show various GPU engine activity on a per-process basis. Every once in a while the patches have been revived but have yet to reach mainline. They recently were revved once again, leaving us hope that in 2022 we might finally see this standardized per-client/process GPU statistics reporting land in the mainline kernel...
This year there has been a lot of Linux kernel work around improving the handling of tiered memory servers, namely those with traditional system RAM augmented by Intel Optane DC persistent memory. There has been work to demote pages during reclamation to the slower persistent memory, improving NUMA balancing around such systems to optimize memory placement, transparent page placement and related work around tiered memory Linux servers...
Coming with future Intel CPUs is Trust Domain Extensions (TDX) to further enhance the security of virtual machines (VMs) and it's sounded a lot like AMD's Secure Encrypted Virtualization (SEV) in many regards and in fact now for the Linux kernel Intel is looking at leveraging some of that SEV code to allow for more code sharing between these CPU features to improve virtualization security...
Aside from the separate work around experimental Vulkan Video decode support, thanks to Intel recently there have been a number of Vulkan improvements to the FFmpeg code around new accelerated filters...
One year after Apple introduced the M1 SoC and the effort began to bring-up this Apple Silicon under Linux, the effort remains ongoing and more code is inching closer to the mainline kernel...
The recent activity around x86 (x86_64 included) straight-line speculation mitigation handling is set to culminate with this security feature being set for mainline with the upcoming Linux 5.17 cycle...
Ubuntu 22.04 LTS isn't expected to run on aging IBM POWER8 hardware as Canonical is shifting its PPC64EL architecture baseline to POWER9 for building packages...
Even with the holidays quickly approaching, KDE developers remain very busy in landing fixes -- especially crash fixes -- and fixing up Plasma's Wayland session for ensuring it is very polished for 2022...
Following last week's Wine 6.23 development release, Wine 7.0-rc1 was just declared in marking the end of feature development and beginning preparations for issuing Wine 7.0.0 stable in January...
The open-source Radeon Linux graphics driver work covered yesterday about a big optimization by leveraging NIR and going through that intermediate representation and relying on common NIR optimizations has now been merged into Mesa 22.0. This is a step-up from the existing open-source OpenGL driver support for old Radeon 9500 through Radeon X1000 series (R500) graphics processors. A similar conversion is also planned for the old Nouveau driver handling NVIDIA "NV30" era graphics processors too...
While it won't make it for the upcoming Wine 7.0, the Wayland driver for natively supporting this X11 successor continues maturing and in the not too distant future will hopefully begin receiving more widespread testing via Wine-Staging...
It's been a while since last having a hearty BSD benchmark comparison on Phoronix in part due to the latest hardware platforms generally lagging behind with how well supported they are by the various BSDs. But stemming from a Phoronix Premium supporter recently requesting some fresh BSD benchmarks, here is a look at how DragonFlyBSD 6.0.1, FreeBSD 13.0, NetBSD 9.2, and OpenBSD 7.0 are competing against various Linux distributions like CentOS, Clear Linux, and Ubuntu.
With Intel's Atom x6000E "Elkhart Lake" SoCs there is a new block called the Programmable Services Engine (PSE) that is an Arm Cortex-M7 companion core that handles various tasks. Unfortunately, with the PSE it means a new binary-only firmware module. With the Programmable Services Engine likely to come with future Intel platforms too, Coreboot developers and open-source enthusiasts are calling on Intel to now open-source this firmware to avoid having this extra binary blob and further complicating future open-source firmware efforts...
Back in 2018 for the Linux 4.18 kernel was introducing the Restartable Sequences system call for allowing faster user-space operations on per-CPU data. By avoiding atomic operations in cases like incrementing per-CPU counters, modifying per-CPU spinlocks, reading/writing to per-CPU ring buffers, and similar, Restartable Sequences can provide a performance advantage. The GNU C Library is landing its revised support for making use of this system call...
Sent to DRM-Next this week for queuing until the Linux 5.17 merge window in January is the latest batch of drm-intel-gt-next updates, which has hang fixes and more preparations for Intel discrete graphics...
It's been over one month since the release of AMDVLK 2021.Q4.1 as the latest open-source AMD Vulkan driver update, which is off the wagon compared to the prior weekly/bi-weekly release cadence. But today thankfully it's been succeeded by AMDVLK 2021.Q4.2 as the newest driver release...
Patches started earlier this year for allowing the parallel bring-up of secondary CPU cores for x86_64 processors have gotten back to being worked on and were sent out on Thursday for review...
While earlier this year AMD dropped pre-Polaris support from their mainline Radeon Software Windows driver, under Linux with open-source software older GPUs can live on much longer with superior driver support... Pending for Mesa 22.0 and as a surprise Christmas gift for those with nearly two decade old GPUs, a big optimization is pending for those with ATI Radeon R300/R400/R500 series graphics cards still in operation...
Systemd 250 is gearing up for release this month and today marked the availability of the first release candidate (and RC2 as a brown paper bag update). Systemd 250 is packing a rather large number of new features and changes across the board for this dominant Linux init system and service manager...
An exciting new development for GNOME 42 is allowing input events to happen at their full input device rate, which is great news for high refresh rate Linux gamers, input tablets, and similar use-cases...
At the end of October came the pleasant surprise of the introduction of the Raspberry Pi Zero 2 W. This drop-in replacement to the original Raspberry Pi Zero features a more powerful 1.0GHz quad-core Cortex-A53 compared to the miniscule 1GHz single-core design of the original Pi Zero while boasting 512MB of LPDDR2 RAM. Here are some initial benchmarks of the Raspberry Pi Zero 2 W for those curious about its performance.
In addition to the LLVM SPIR-V back-end appearing ready for merging, also working through the final steps for being mainlined in the LLVM compiler stack is also Facebook's BOLT project for optimizing the performance of binaries...
Taking place in San Francisco from Monday through yesterday evening was the RISC-V Summit for discussions around this dominant open-source processor ISA. For those that did not make it to the event, many of the slide decks are available...
It is not too often getting to talk about performance optimizations for Mesa's Virgl code that along with in conjunction with related "Virgil" components allows for hardware-accelerated 3D/OpenGL running within virtual machines. Hitting Mesa 22.0 this week though is some Virgl code improvements for allowing lower memory use within virtual machines...
Intel's oneDNN Deep Neural Network Library that is part of their oneAPI toolkit is out with version 2.5 and brings RISC-V CPU support among other updates...
Engineers from Intel and Arm in cooperation with The Khronos Group feel ready now to begin landing their SPIR-V back-end within the upstream LLVM source tree! This SPIR-V back-end for LLVM would ultimately allow LLVM front-ends for different languages to more easily target this industry-standard shader representation so that it could be ingested by Vulkan / OpenCL drivers...
While Ubuntu switched from LZ4 to Zstd for compressing its initramfs, they now are finding they were too aggressive in defaulting to Zstd with the highest compression level of 19. Due to speed and memory consumption concerns, they are looking at lowering their Zstd compression level...
The Cloud-Hypervisor open-source project started by Intel a few years ago as a Rust-written VMM focused on cloud workloads is now going to be hosted by the Linux Foundation...
Microsoft has introduced "Microsoft-Performance-Tools for Linux-Android" as a collection of open-source tools for analyzing system performance on Linux and Android...
When it comes to the state of packaged web browsers for Debian GNU/Linux, unfortunately it leaves a lot to be desired at the moment and for those wanting to be secure and up-to-date it can mean resorting to proprietary or un-packaged browser builds...
Mesa's Radeon Vulkan driver "RADV" is implementing emulated support for ETC2 texture compression to use with newer AMD GPUs to improve compatibility with Google's Android operating system...
Google is looking to upstream their Linux kernel driver for Open Profile for DICE, a secret derivation protocol used currently by some Android devices...
One of many exciting features/changes with upcoming Intel Xeon Scalable "Sapphire Rapids" processors is the introduction of Advanced Matrix Extensions (AMX). While initial AMX support is premiering with Linux 5.16 due out in stable form as the start of the new year, it currently doesn't allow for KVM virtualized guests to make use of the new capabilities...
While all of the software components are out there now for being able to run NVIDIA's proprietary driver stack with modern (GBM-based) Wayland compositors by default, including XWayland support, Fedora Workstation currently defaults to using an X.Org based session with the green binary blob. However, for Fedora 36 next spring they are planning on using the Wayland-based desktop here too...