Session

Precision timing in Linux networking has largely been limited to control-plane timestamping, where timing is observed but not directly controlled, typically at rates below modern line speeds. Bringing timestamping into the data plane at line rate is challenging because driver behavior introduces variability that can undermine timestamp fidelity even with hardware assistance. We examine a hardware-centric approach that relocates clock synchronization functions from the control plane to the data plane within a \SNIC architecture. The design applies real-time clock corrections using packet-derived timestamps and a lightweight two-bit synchronization protocol.

However, timestamp accuracy is not sufficient on its own. Linux can measure time, but it does not yet reliably enforce timing in the transmit path. We explore how SO_TXTIME shifts timing from observation to control by allowing applications to explicitly schedule packet transmission. While powerful, software-based scheduling struggles to scale to modern link rates and tight jitter bounds. To complete the model, we propose hardware-timed pause semantics can be used for low-jitter transmit control. Together, data-plane timestamping, explicit transmit scheduling, and hardware-backed timing control move Linux networking toward deterministic, time-aware data planes.