Servo Web Engine Further Tuning Performance, Screen Reader & Other New Features

The Servo open-source web layout engine continues advancing with its demo Servoshell and continued work around making it suitable for embedding into other software. The Servo project this morning published their latest monthly status update to inform the community what they have been up to the past several weeks.

For those curious what Servo developers have been backing on this summer, some of the highlights they published today on their development blog include:

- Servo landed some additional work towards supporting incremental layout handling.

- Various other performance optimizations.

- Continued work
Continue reading...

Intel QATlib 25.08 Brings Hugepages Support & Other Improvements

Intel engineers yesterday released QATlib 25.08 as the first new update in nearly one year for this QuickAssist Technology library. Intel QuickAssist allows hardware-accelerated offloading of various security authentication and compression operations from the CPU onto dedicated accelerator IP found in recent Xeon processors. Intel's QATlib is the open-source library for enabling that magic to happen from the user-space side.

Intel QATlib 25.08 continues supporting both QuickAssist Gen4 and Gen5 devices -- there are older QAT hardware out

AMD Bringing SmartMux Support To Linux 6.17 For Better Hybrid GPU Laptop Support

For the upcoming Linux 6.17 kernel cycle AMD already queued fixes for GPU compute on some older AMD hardware , improved debugging support for AMDGPU , and other enhancements. Sent out today was a final batch of feature changes for AMDGPU/AMDKFD expected for the upcoming Linux 6.17 merge window. Most notable is AMD SmartMux support coming to Linux.

AMD SmartMux already works with current/recent hardware and is supported under Microsoft Windows drivers it seems. Now though AMD SmartMux is coming to Linux for improving hybrid GPU
Continue reading...

LLVM 21.1-rc1 Released For Testing: Better RISC-V Support, AMD GFX1250 & NVIDIA GB10

The first release candidate of LLVM 21.1 is now available for testing, which under their modern versioning scheme will represent the first stable version of the LLVM 21 compiler stack.

As noted earlier this week when LLVM 21 was branched and LLVM 22 now open for development , the LLVM 21.1 series is going to bring a lot of notable additions and refinements to existing functionality for this half-year compiler update. LLVM 21 introduces the AMD GFX1250 target for what is
Continue reading...

Linux Receiving Fix For AMD Radeon Polaris GPUs Producing Lots Of Log Spam

For those using an AMD Radeon RX 500 "Polaris" graphics card on Linux and routinely suspend/resume your system, going into the Linux 6.16 kernel and then to be back-ported to the stable series is a fix where the AMDGPU driver could end up producing a lot of spam in the kernel log.

Sent out today was this week's drm-fixes-6.16 for the Radeon and AMDGPU open-source kernel graphics drivers. There is a memory leak fix in
Continue reading...

New FFmpeg AVX-512 Optimizations Hit Up To 36x The Performance Of Plain C Code

Some commits merged today to FFmpeg Git provide additional hand-tuned Assembly code for AVX-512 with capable Intel and AMD processors.

Open-source multimedia developer Niklas Haas today upstreamed some additional AVX2 and AVX-512 tuning to FFmpeg, on top of the multimedia library's already vast array of hand-tuned code for leveraging Advanced Vector Extensions.

For FFmpeg's avfilter scene_sad code, there is now an AVX-512 implementation added that comes in at 36.
Continue reading...

Google Continues Working On "Magma" For Mesa Cross-Platform System Call Interface

Mesa 25.2 entered its feature freeze yesterday with many exciting driver improvements with new features and performance optimizations while one feature that wasn't ready for merging in this quarter's release is Magma, which is a recent effort by Google engineers working on a cross-platform system call interface for Mesa. And it's written in Rust.

Magma is a cross-platform GPU system call library in Mesa. Google engineers have been developing it with an eye on Chrome OS use and for possible future use as
Continue reading...

Single RunQueue Proxy Execution Appears Ready For Linux 6.17

The long in development work around proxy execution for the Linux kernel appears to be ready for the upcoming Linux 6.17 merge window with the Single RunQueue Proxy Execution patches queued into a TIP branch after going through 19 rounds of patch review/revisions.

John Stultz of Google has been leading the effort around proxy execution as a kernel mechanism for mutex-owning tasks to inherit the scheduling context of higher priority waiters. Proxy execution for this means of priority inheritance has been sought by Google for Android use to improve the
Continue reading...

Apple Silicon

Linux 6.17 To Upstream Support For The Decade Old Marvell PXA1908 SoC

Launched back in 2014 was the Marvell PXA1908 SoC intended for 4G LTE smartphones and featured four Arm Cortex-A53 cores. Not too impressive for its time and far less so today. Though after a decade of not seeing mainline Linux kernel support and some vendor kernels stuck in the Linux 3.14 era, the upcoming Linux 6.17 cycle is expected to upstream support for this old smartphone SoC.

Queued up ahead of the Linux 6.17 kernel
Continue reading...

Linux 6.17 To Fix AMDGPU Hibernation So It Doesn't Take ~50 Minutes On Large GPU Servers

While late in the Linux 6.16 cycle and hitting the cut-off for when the period to queue new DRM driver feature material for Linux 6.17 ends, an additional drm-misc-next pull request was sent out today with some last minute kernel graphics driver changes for this next kernel cycle. Motivating this extra pull were the recent AMDGPU system hibernation patches.

The headline change with today's drm-misc-next pull is incorporating the AMD patches to reduce system memory requirements for hibernation on
Continue reading...