On Tue, Dec 10, 2013 at 09:35:40AM -0500, Vivek Goyal wrote: > On Mon, Dec 09, 2013 at 05:06:10PM -0800, Kees Cook wrote: > > On Mon, Dec 9, 2013 at 4:34 PM, H. Peter Anvin <hpa at zytor.com> wrote: > > > On 12/09/2013 04:16 PM, Kees Cook wrote: > > >> For general-purpose (i.e. distro) kernel builds it makes sense to build with > > >> CONFIG_KEXEC to allow end users to choose what kind of things they want to do > > >> with kexec. However, in the face of trying to lock down a system with such > > >> a kernel, there needs to be a way to disable kexec (much like module loading > > >> can be disabled). Without this, it is too easy for the root user to modify > > >> kernel memory even when CONFIG_STRICT_DEVMEM and modules_disabled are set. > > >> > > >> Signed-off-by: Kees Cook <keescook at chromium.org> > > >> Acked-by: Rik van Riel <riel at redhat.com> > > > > > > So the logic is to load a crashkernel and then lock down the machine > > > before services, networking etc. are enabled? > > > > Right, or to just turn it off at boot time if kexec will not be used at all. > > kdump kernel is loaded with the help of kdump service. Different distro's > might have different dependencies for that serivce. But recently in fedora > we wait network to come up before starting that service. (So that nfs > targets can be mounted and checked for valid dump destinations). > > IOW, crash kernel is loaded quite late in the game (quite a few services > have run and possibly networking is up too). To me, practically one will > disable kdump also if you change state of this knob early. Of course it also removes the possibility of re-loading crash kernel after doing some changes to /etc/kdump.conf (like dump destination or kernel command line). I am wondering if it is sufficient to disable kexec jump back functionality only while retaining kexec and kdump. That seems to be easiest way to change kernel's data structures. Thanks Vivek
