[PATCH 0/7] tip-queue 2015-12-14

[Borislav Petkov <bp@alien8.de>]

From: Borislav Petkov <bp@suse.de>

Hi Ingo,

here's a pile which addresses issues which arise when using an NMI
button on certain platforms which provide it, to cause an external NMI
and thus cause a concurrent kdump to fail. For that, we're getting a
new cmdline option called "apic_extnmi" which can control which CPUs
(or none) to receive the external NMI. The default behavior is what the
kernel does currently - only the BSP receives it.

I've made sure that each commit message explains in great detail what
the problem is as this whole area is not trivial.

Please apply,
thanks.

Borislav Petkov (1):
  x86/mm: Align macro defines

Hidehiro Kawai (6):
  panic, x86: Fix re-entrance problem due to panic on NMI
  panic, x86: Allow CPUs to save registers even if looping in NMI context
  kexec: Fix race between panic() and crash_kexec()
  x86/apic: Introduce apic_extnmi command line parameter
  x86/nmi: Save regs in crash dump on external NMI
  Documentation: Document kernel.panic_on_io_nmi sysctl