panic, kexec: make __crash_kexec() NMI safe

Attempting to get a crash dump out of a debug PREEMPT_RT kernel via an NMI
panic() doesn't work.  The cause of that lies in the PREEMPT_RT definition
of mutex_trylock():

	if (IS_ENABLED(CONFIG_DEBUG_RT_MUTEXES) && WARN_ON_ONCE(!in_task()))
		return 0;

This prevents an nmi_panic() from executing the main body of
__crash_kexec() which does the actual kexec into the kdump kernel.  The
warning and return are explained by:

  6ce47fd9 ("rtmutex: Warn if trylock is called from hard/softirq context")
  [...]
  The reasons for this are:

      1) There is a potential deadlock in the slowpath

      2) Another cpu which blocks on the rtmutex will boost the task
	 which allegedly locked the rtmutex, but that cannot work
	 because the hard/softirq context borrows the task context.

Furthermore, grabbing the lock isn't NMI safe, so do away with kexec_mutex
and replace it with an atomic variable.  This is somewhat overzealous as
*some* callsites could keep using a mutex (e.g.  the sysfs-facing ones
like crash_shrink_memory()), but this has the benefit of involving a
single unified lock and preventing any future NMI-related surprises.

Tested by triggering NMI panics via:

  $ echo 1 > /proc/sys/kernel/panic_on_unrecovered_nmi
  $ echo 1 > /proc/sys/kernel/unknown_nmi_panic
  $ echo 1 > /proc/sys/kernel/panic

  $ ipmitool power diag

Link: https://lkml.kernel.org/r/20220630223258.4144112-3-vschneid@redhat.com
Fixes: 6ce47fd9

 ("rtmutex: Warn if trylock is called from hard/softirq context")
Signed-off-by: Valentin Schneider <vschneid@redhat.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Baoquan He <bhe@redhat.com>
Cc: "Eric W . Biederman" <ebiederm@xmission.com>
Cc: Juri Lelli <jlelli@redhat.com>
Cc: Luis Claudio R. Goncalves <lgoncalv@redhat.com>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Petr Mladek <pmladek@suse.com>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>