lockdep: How to read its cryptic output

Session information has not yet been published for this event.

Scheduled: Friday, September 9, 2011 from 2:20 – 3:10pm in Alexander Valley Ball Room

One Line Summary

Learn to read the output from lockdep when your code can trigger a deadlock


Ever since lockdep was introduced in 2006 most deadlock scenarios have been avoided. But when lockdep does detect a problem with the locking, the output it gives can be very cryptic and hard to understand, especially when the locking issues have to do with interrupts or del_timer_sync().

This talk will explain how to translate the lockdep output so that you can understand the locking problem scenario and confidently fix your code (or someone else’s code) to avoid any deadlocks.


interrupts, debugging, locking, realtime

Presentation Materials



  • Steven Rostedt

    Red Hat


    Steven Rostedt has been coding the Linux kernel since 1998. He started while working on his Master’s in CS. In 2001 he joined TimeSys and started his professional career in Linux kernel hacking. Currently Steven works with Red Hat supporting their Real Time kernel, but spends a lot of his time doing mainline work with tracing, the scheduler and pretty much anything else he wants to tinker with in the kernel.