Re: LKPK (Live Kernel Patching Kit)

Adam Sulmicki (adam@cfar.UMD.EDU)
Sat, 25 Jul 1998 19:46:34 -0400

Chad C Giffin writes:

->I just starting designing such a thing last week. I made the mistake of
->talking about it on IRC and now suddenly I see this? [1]

I don't think so. I mentioned it on linux-kernel several time. I did came
up with it the first time, when the off-by-one bug surfaced. I never
heard it on IRC.

I just had to finish reading "Intel Architecture Software Dev. Manual"
Vols : 1,2,3, thus the dealy :-).

-> To provide a means by which one may upgrade his kernel image and,
->at the same time, not interrupt services, vm, or running processes by
->allowing the administrator to boot a new kernel image overtop the
->existing image.

I think you are talking about something else. I even did not try to think
about anything like this as it is _extermely_ complicated.

My idea is to design _simple_ "patch" which allows you patch binary,
running kernel so that you can continue to run your machine until the
hardware becomes obsolete without need for a single reboot :-). But it is
as means of fixing bugs, the version of the kernel is to stay the same.

It is also as menas of _supplement_ to IKD (Inegerated Kernel Debugger)
which while is cool, requires you to have prepatched kernel ahead of time
and requires you patch for each and every single kernel. While my metchod
allows you debug and test any kernel.

As I said I have it alrady working for the off-by-one bug, it looks really

I still need to find out solution to changing address of a variable.
Since I don't know gcc sources, I don't know how compiler/program
declares static variables, as in opposite to pointers.

I need something like 'extern' but with ability to specify explict
address. And have gcc trust me that I'm right about the address.

BTW: How about cooperation instead of competition? :P


To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
Please read the FAQ at