stop_machine: implement stop_machine_from_inactive_cpu()
authorTejun Heo <tj@kernel.org>
Thu, 23 Jun 2011 18:19:28 +0000 (11:19 -0700)
committerH. Peter Anvin <hpa@linux.intel.com>
Mon, 27 Jun 2011 22:17:08 +0000 (15:17 -0700)
commitf740e6cd0cb5e7468e46831aeb4d9c30e03d5ebc
tree83d81928de5d3ebc3b325f633dddd77a0e6ad662
parentfd7355ba1e936487f5aae6fc058c6cb300e44a64
stop_machine: implement stop_machine_from_inactive_cpu()

Currently, mtrr wants stop_machine functionality while a CPU is being
brought up.  As stop_machine() requires the calling CPU to be active,
mtrr implements its own stop_machine using stop_one_cpu() on each
online CPU.  This doesn't only unnecessarily duplicate complex logic
but also introduces a possibility of deadlock when it races against
the generic stop_machine().

This patch implements stop_machine_from_inactive_cpu() to serve such
use cases.  Its functionality is basically the same as stop_machine();
however, it should be called from a CPU which isn't active and doesn't
depend on working scheduling on the calling CPU.

This is achieved by using busy loops for synchronization and
open-coding stop_cpus queuing and waiting with direct invocation of
fn() for local CPU inbetween.

Signed-off-by: Tejun Heo <tj@kernel.org>
Link: http://lkml.kernel.org/r/20110623182056.982526827@sbsiddha-MOBL3.sc.intel.com
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
include/linux/stop_machine.h
kernel/stop_machine.c