Doug / smarc-fsl-linux-kernel

Download as
Browse Code ยป

Commit 9398180097e359646d46083c3e079a54e20bee82

Authored by Tejun Heo
Committed by Linus Torvalds
1 parent e13193319d
Exists in master and in 7 other branches imx_3.0.35_4.1.0, imx_3.10.17_1.0.1_ga, imx_3.10.53_1.1.0_ga, imx_3.14.28_1.0.0_ga, smarc-imx_3.10.53_1.1.0_ga, smarc-imx_3.14.28_1.0.0_ga, smarc-l5.0.0_1.0.0-ga

workqueue: fix race condition in schedule_on_each_cpu() 

Commit 65a64464349883891e21e74af16c05d6e1eeb4e9 ("HWPOISON: Allow
schedule_on_each_cpu() from keventd") which allows schedule_on_each_cpu()
to be called from keventd added a race condition.  schedule_on_each_cpu()
may race with cpu hotplug and end up executing the function twice on a
cpu.

Fix it by moving direct execution into the section protected with
get/put_online_cpus().  While at it, update code such that direct
execution is done after works have been scheduled for all other cpus and
drop unnecessary cpu != orig test from flush loop.

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Andi Kleen <ak@linux.intel.com>
Acked-by: Oleg Nesterov <oleg@redhat.com>
Cc: Ingo Molnar <mingo@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Showing 1 changed file with 13 additions and 15 deletions Side-by-side Diff

... ... @@ -692,31 +692,29 @@
692 692 if (!works)
693 693 return -ENOMEM;
694 694  
  695 + get_online_cpus();
  696 +
695 697 /*
696   - * when running in keventd don't schedule a work item on itself.
697   - * Can just call directly because the work queue is already bound.
698   - * This also is faster.
699   - * Make this a generic parameter for other workqueues?
  698 + * When running in keventd don't schedule a work item on
  699 + * itself. Can just call directly because the work queue is
  700 + * already bound. This also is faster.
700 701 */
701   - if (current_is_keventd()) {
  702 + if (current_is_keventd())
702 703 orig = raw_smp_processor_id();
703   - INIT_WORK(per_cpu_ptr(works, orig), func);
704   - func(per_cpu_ptr(works, orig));
705   - }
706 704  
707   - get_online_cpus();
708 705 for_each_online_cpu(cpu) {
709 706 struct work_struct *work = per_cpu_ptr(works, cpu);
710 707  
711   - if (cpu == orig)
712   - continue;
713 708 INIT_WORK(work, func);
714   - schedule_work_on(cpu, work);
715   - }
716   - for_each_online_cpu(cpu) {
717 709 if (cpu != orig)
718   - flush_work(per_cpu_ptr(works, cpu));
  710 + schedule_work_on(cpu, work);
719 711 }
  712 + if (orig >= 0)
  713 + func(per_cpu_ptr(works, orig));
  714 +
  715 + for_each_online_cpu(cpu)
  716 + flush_work(per_cpu_ptr(works, cpu));
  717 +
720 718 put_online_cpus();
721 719 free_percpu(works);
722 720 return 0;