stop_machine: Clean up the usage of the preemption counter in cpu_stopper_thread()
authorOleg Nesterov <oleg@redhat.com>
Sun, 15 Nov 2015 19:33:32 +0000 (20:33 +0100)
committerIngo Molnar <mingo@kernel.org>
Mon, 23 Nov 2015 08:48:20 +0000 (09:48 +0100)
commitaccaf6ea3db6f5fb997f096b6eefd5431d03f7e5
tree3e17684822d4d0e9685ef0d598c2fa55ab88495e
parentdd2e3121e3cb16d03a6e3f2db48f260f046f39c2
stop_machine: Clean up the usage of the preemption counter in cpu_stopper_thread()

1. Change this code to use preempt_count_inc/preempt_count_dec; this way
   it works even if CONFIG_PREEMPT_COUNT=n, and we avoid the unnecessary
   __preempt_schedule() check (stop_sched_class is not preemptible).

   And this makes clear that we only want to make preempt_count() != 0
   for __might_sleep() / schedule_debug().

2. Change WARN_ONCE() to use %pf to print the function name and remove
   kallsyms_lookup/ksym_buf.

3. Move "int ret" into the "if (work)" block, this looks more consistent.

Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Tejun Heo <tj@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Milos Vyletel <milos@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Prarit Bhargava <prarit@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20151115193332.GA8281@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
kernel/stop_machine.c