Walking Through the Linux-Based Graphics Stack
summary
A concrete walk through how a buffer moves from app to panel: allocates memory (sys RAM or VRAM), shares it via dmabuf + EGL or SHM, compositor composites on the GPU, and submits a KMS page flip with GEM-managed buffers. Coverage includes DRM KMS + Atomic vs legacy FBDEV, cache and IOMMU concerns, zero copy via PRIME, and systemd-logind FD passing so compositors run unprivileged. Shows where tearing happens, why double vs triple buffering changes latency, and which kernel/ioctls are used.
AI-written summary · about
chapters (4)
- 0:00Start
- 2:37Graphics Hardware: Memory
- 9:41Graphics Hardware: Rendering
- 20:57Graphics APIs: Summary Diagram
related talks
39mNavigating the Linux Graphics Stack
46mDemystifying Graphics and Media Work Synchronisation
20mLinux Graphics 101
40mDemystifying the Embedded Linux Graphics Stack: An Easy Introduction for Beginners
32mThe Modern Linux Graphics Stack on Embedded Systems
46mGraphic Testing Without Hardware: Discovering the Power of VKMS!
44mA Current Overview of the DRM KMS Driver-Side APIs
52mGetting pixels on screen on Linux: introduction to Kernel Mode Setting