On Thu, 2014-10-23 at 09:58 +0300, Kevin Wilson wrote:

I am getting this message when running a patched kernel: BUG: unable to handle kernel NULL pointer dereference at (null)

That is just what you get when the kernel dereferences NULL, isn't it? (If the kernel dereferences, say, something halfway some struct, you'll see a hex address instead of "(null)".) Assuming x86, see: arch/x86/mm/fault.c:show_fault_oops() and lib/vsprintf.c:pointer().

Is there some kernel configuration item which enables getting , instead "(null)", the proper value (I assume it should be a method name)

Doesn't your BUG also print a backtrace that points at a (possible) culprit? Paul Bolle