Qualcomm has unveiled Linux-Qualcomm distribution Linux 2.0, which provides a software stack optimized for Qualcomm's Dragonwing IoT platforms. The project implements a unified distribution that enables rapid development of solutions for various Qualcomm SoCs, such as the QCS5430, QCS6490, IQ-8275, IQ-9075, and IQ-615. Qualcomm uses the Dragonwing platform as its foundation. Linux The toolkit and meta-packages from the Yocto and OpenEmbedded projects are used, supplemented by patches, drivers, firmware, packages and configuration files to support specific hardware blocks used in Qualcomm chips.
The distribution is developed using an open development model and contributes to upstream projects. Qualcomm's proprietary components are integrated into the distribution as dedicated, modular overlays. The software stack supports all processors, subsystems, and components of the Qualcomm platform, including CPUs, GPUs, and various hardware accelerators for AI, data processing, and multimedia operations (VPUs, DPUs, NPUs, and DSPs). In addition to a Yocto-based user environment, the distribution includes a bootloader with verified boot support, UEFI firmware, and kernel packages maintained by Qualcomm engineers. Linux, including additional drivers and patches.
The platform is designated as fully ready for building production systems based on Qualcomm hardware. Features include virtualization support and mandatory access control using SE.Linux, testing of all components on reference hardware, a long package maintenance cycle for all supported platforms, and the ability to integrate custom BSPs (Board Support Packages) and OEM layers. Updates can be made by flashing a system image or automatically delivering updates over the air. 
Qualcomm Release Linux 2.0 is notable for its transition to the Yocto 6.0 platform and the inclusion of the kernel Linux 6.18 (previously Yocto 5.0 and kernel 6.6 were used). The core distribution now includes capabilities for running applications in real-time mode (the kernel is built with PREEMPT_RT support). Instead of two separately maintained distribution variants, "Base" and "Custom," a unified stack with a single, shared kernel and system environment (rootfs) is offered.
The openness of development processes has been increased: releases are no longer created behind closed doors, but are publicly available on GitHub using an open continuous integration system and a predictable update schedule. The organization of modular overlays with add-ons has been redesigned; they are now created from a single open code base, rather than separated into open and proprietary builds. In addition to the previously supported QCS5430, QCS6490, IQ9075, IQ8275, and IQ615 SoCs, support has been added for the IQ-X platform for industrial computers.
Meta-packages with basic functionality are now being developed in publicly accessible repositories and are open to accepting contributions from third-party contributors, creating forks, or using them without waiting for releases. Such meta-packages include: meta-qcom (a hardware support layer), meta-qcom-distro (a distribution framework, system image build scripts, and configuration), meta-qcom-3rdparty (third-party platform support components), meta-qcom-releases (reproducible release builds), meta-selinux (mandatory access control via SE).Linux), meta-updater (OTA updates based on OSTree), meta-security (additional mechanisms for enhancing database security and auditing), meta-virtualization (Docker, Kubernetes, KVM).
In addition to meta-packages, overlays are also available—separately delivered versioned sets of kernel modules and libraries supporting advanced features. Installing an overlay installs the corresponding optimized kernel modules and libraries on top of the standard base environment without modifying it. Available overlays include:
- working with sound (AudioReach and hardware acceleration of sound processing);
- camera control (camX ISP);
- working with graphics (acceleration of OpenGL ES, OpenCL and Vulkan on Adreno GPU);
- video processing (using the Iris VPU to accelerate video encoding and decoding);
- sensor support (Qualcomm Sensor Hub integration);
- Solving computer vision problems (FastCV acceleration on DSP/Adreno).
Source: opennet.ru
