Valve's Steam Deck, equipped with SteamOS, offers a unique feature: built-in crash data collection. This feature, along with logging system events like split-lock detection, provides valuable insights for both users and developers. The data collection is entirely opt-in, allowing users to choose when and how much information they share with Steam. However, for those eager to delve deeper into this process, a recent presentation at the Linux Plumbers Conference shed light on the intricacies of Steam Deck data collection.
Guilherme Piccoli, a long-standing partner with Valve from the Linux consulting firm Igalia, presented on the data collection methods employed by the Steam Deck. The focus was on two key areas: addressing software/hardware failures and optimizing game/engine/library performance on the Steam Deck and SteamOS.
The presentation highlighted various breakages, such as crashes, GPU hangs, kernel oops, and out-of-memory events, as well as the importance of split lock detection and other non-critical but useful data. It's worth noting that the data collection process is entirely user-controlled, ensuring privacy and transparency.
For those interested in learning more, the presentation's PDF slide deck is available online. While the video recordings from the Linux Plumbers Conference 2025 are not yet accessible, the slide deck provides a comprehensive overview of the data collection process. This resource is a valuable reference for anyone curious about the technical aspects of the Steam Deck's data collection capabilities.
