GPU Reset Recovery and Error Handling in Linux

GPU Reset Recovery and Error Handling in Linux

Learn about GPU reset handling in Linux through this 25-minute conference talk from XDC 2023. Explore the complex challenges of managing GPU errors and system recovery when devices get stuck. Discover the inner workings of DRM device resets, understand error propagation through the stack, and examine methods for identifying reset triggers. Despite some audio issues in the first 3 minutes, gain valuable insights into proposed improvements for GPU reset handling and recovery mechanisms in the Linux ecosystem.