On Fri, Sep 27, 2013 at 10:04:44AM -0600, Bjorn Helgaas wrote: > [+cc Thomas, Russell]
Someone is doing something quite bad in the kernel, and as yet I've not figured out a way to track it down. The issue is this: someone is kfree'ing a kobject before its release function has been called, and the memory is being re-used. The problem is that when the last reference has been dropped with the debug enabled, the kobject is linked into the timer lists for the delayed work. When the timer lists get run, they're found to be corrupted. The obvious solution to this is to move the delayed work out of the kobject into a separately allocated structure. That would work if x86 didn't register kobjects very early in boot, before the memory allocators were up and running. Frankly, I've no idea how to solve this. So I regard x86 as just being difficult and broken. :) If anyone has any ideas, then I'm all ears. http://www.annhuey.com/ed-pix/fa_i-pix/I%27m-All-Ears.jpg -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
