With vulnerabilities like L1TF and Microarchitectural Data Sampling (MDS) prominently showing the insecurities of Intel Hyper Threading, DigitalOcean and other organizations continue spearheading a core scheduling implementation for Linux that could allow HT to remain enabled but with reducing the security risk...
When the Zink Gallium3D driver running OpenGL over Vulkan was first introduced in 2018 and since one of the main blockers besides the performance overhead has been the limited OpenGL 2/3 support. The GL3/GL4 support has been improving with time for Zink and when making use of the latest out-of-tree patches is even possible to get OpenGL 4.6 running over Vulkan with Zink!..
As part of the virtual SIGGRAPH20, Intel is using the opportunity to talk up their ray-tracing efforts...
Proposed last summer by Valve and Collabora developers were extending the Linux kernel's futex system call to allow for more optimal thread pool synchronization and paired with Wine/Proton work to better match the semantics of Windows. That then spun into creating a new system call, futex2. With the recently closed Linux 5.9 merge window the new futex2 system call didn't land, but the work is still being pursued...
For a number of years the GNU Compiler Collection has shipped experimental support for the DWARF 5 debugging data format while finally for next year's GCC 11 release it might be deemed stable and used by default...
The Mesa Radeon Vulkan driver "RADV" has seen an initial trap handler implementation for helping to more easily catch and diagnose various issues stemming from Vulkan shaders...
The hipSYCL effort has been about supporting the Khronos SYCL single-source language built on C++ across any CPU with OpenMP as well as AMD Radeon GPUs via ROCm and NVIDIA GPUs via CUDA. The hipSYCL effort has a new "Lite" experimental runtime under development...
Pages