How does the kernel release a process's locks when it gets killed
 

Hi,
I'm taking an OS class and we just completed an assignment related to adding thread support to a kernel (xv6 OS). One of the cases we *didn't* have to handle was releasing locks held by a process in case it got killed. I was wondering how is this problem generally solved (maybe in the linux kernel)?

I don't know enough to answer your question, but some of the tools in this article might help: http://www.linuxjournal.com/article/3829

Also, the interactive kernel map could be a nice reference: http://www.makelinux.net/kernel_map/