Attached is the dmesg from patched kernel.

I'm not an expert at debugging the kernel, obviously.  I tried setting
up a serial console before without much luck as part of this debugging
session.  I'll try again when I have some time, but I couldn't get to
the exact message before -- it scrolled by too quickly (and boot_delay
froze the output).

On Fri, Jan 25, 2019 at 10:58 AM Michal Hocko <mhocko@kernel.org> wrote:
>
> On Fri 25-01-19 10:52:06, robert shteynfeld wrote:
> > The person who pointed to mm/page_alloc.c commits likely causing the
> > issue did not have time to build a patched/reverted kernel to confirm
> > his hypothesis.  When I tried backing out the two separate commits he
> > suggested, the first commit (ie. the one in the subject) was the one
> > that when backed out fixed the boot issue.  Reverting the second one
> > had no effect.
>
> OK, so can you provide the full dmesg (with revert) along wth
> memblock=debug kernel parameter. Also would it be possible to configure
> serial or net console to get the actual panic report please?
> --
> Michal Hocko
> SUSE Labs