The Linux kernel is the foundation of modern computing, connecting increasingly complex hardware to demanding workloads in gaming, AI, embedded systems, and cloud infrastructure. Igalia provides specialized kernel engineering services to optimize, extend, and maintain this critical layer.

We combine deep architectural knowledge with active upstream leadership. As maintainers and key contributors, we strive to shape the kernel evolution across many areas, including the GPU DRM stack, BPF-based scheduling, and memory-safe Rust abstractions. Our approach emphasizes long-term maintainability, ensuring that our partners’ innovations align with the upstream ecosystem.

Core Kernel Infrastructure

Our engineers have deep expertise in scheduling, concurrency, and memory management. We design and refine scheduling behavior to balance fairness, throughput, and latency across diverse workloads. We analyze and resolve complex concurrency issues, improve synchronization primitives such as futexes, and eliminate bottlenecks in highly parallel applications. In memory management, we optimize allocation strategies and strengthen system stability under sustained memory pressure, enabling partners to extract maximum performance from their platforms.

Beyond core performance, we enhance the kernel’s diagnostic and debugging tools, including tracing frameworks, pstore, and kdump. Our expertise also extends to power management and virtualization environments such as KVM and Hyper-V for long-term sustainability.

Kernel Graphics Layer

The Linux graphics ecosystem is experiencing an unprecedented surge in innovation. From the explosion of Linux-based gaming and high-end consumer electronics to the critical role of GPGPU acceleration in AI, the graphics layer has become the heart of the modern Linux stack.

Our engineers serve as active contributors and maintainers within the DRM and KMS subsystems. We specialize in implementing cutting-edge features such as HDR, color management, Variable Refresh Rate, and Asynchronous Page Flips. We actively improve DRM scheduling to enhance multitasking performance and develop robust GPU recovery features to ensure system stability under heavy loads. We also remain at the forefront of research, exploring the application of CRIU (Checkpoint/Restore In Userspace) within graphics contexts to enable new possibilities in virtualization and process migration.

Safe Kernel Extension and Specialization

As system demands become increasingly specialized, the ability to extend kernel functionality safely and dynamically has become a critical requirement. At Igalia, we are pioneers in using modern frameworks to tailor the kernel to specific workloads while maintaining the rigorous safety standards required for production environments.

Our engineers are key contributors to the sched_ext framework, which enables the implementation of custom, user-defined thread schedulers. We leverage these capabilities to create workload-optimized solutions, such as specialized Linux gaming scheduling and advanced cache management policies. By implementing kernel-level logic that is both dynamic and verifiably safe, we enable our partners to achieve significant performance gains and lower tail latency. In parallel, we contribute to Rust for Linux by developing safe abstractions for subsystems like DRM. By bridging the gap between C and Rust, we help create a more secure, stable, and maintainable kernel.

File and Storage Technologies

The storage stack has evolved into a performance-critical subsystem driven by NVMe, multi-queue block architectures, and highly virtualized deployments. Igalia has deep expertise across the kernel block layer, including NVMe, (i)SCSI, and software RAID, with a focus on optimizing I/O paths for low latency and high throughput in demanding cloud and enterprise environments.

Our work in FUSE includes improving restartability and crash recovery to ensure that userspace filesystem daemons can be updated or recovered without destabilizing active mounts. We also contribute to improvements in the Linux filesystem layer, including upstream support for case-insensitive operation in tmpfs, and OverlayFS, alongside with improvements to Unicode subsystem. Combined with expertise in distributed filesystems and high-performance interconnects such as InfiniBand, we help design scalable, fault-tolerant storage architectures for modern distributed systems.

Maintainability across Architectures and Versions

Supporting diverse hardware platforms and long-lived deployments requires deep architectural understanding and disciplined upstream collaboration. Igalia has extensive experience across x86, ARM, PowerPC, and RISC-V, addressing architecture-specific challenges in performance, correctness, and hardware enablement.

Our years of specialization in identifying and fixing kernel bugs have honed our expertise in reproducing race conditions, memory corruptions, and other complex issues, collaborating effectively with the upstream community, and delivering reliable patches. This hands-on experience allows us to identify root causes across a wide range of subsystems, resulting in a more resilient and secure kernel that benefits both our clients and the broader Linux ecosystem.

Additionally, we offer comprehensive lifecycle support by backporting security fixes and functional improvements across different kernel versions. By carefully managing versions and aligning with upstream developments, we ensure that production systems remain stable, secure, and maintainable over time.