Hello, Linus.

Just one patch to add destroy_work_on_stack() annotations to help
debugobj debugging.  The change is almost trivial but unfortunately
causes the following merge conflict.

  long work_on_cpu(int cpu, long (*fn)(void *), void *arg)
  {
          struct work_for_cpu wfc = { .fn = fn, .arg = arg };

          INIT_WORK_ONSTACK(&wfc.work, work_for_cpu_fn);
          schedule_work_on(cpu, &wfc.work);
  <<<<<<< HEAD
          flush_work(&wfc.work);
  =======

          /*
           * The work item is on-stack and can't lead to deadlock through
           * flushing.  Use __flush_work() to avoid spurious lockdep warnings
           * when work_on_cpu()s are nested.
           */
          __flush_work(&wfc.work);

          destroy_work_on_stack(&wfc.work);
  >>>>>>> 440a11360326044a9addf1c652a0364aad0be90c
          return wfc.ret;
  }

The conflict is between the following two commits.

* master        12997d1a999c ("Revert "workqueue: allow work_on_cpu() to be 
called recursively"")
* wq/for-3.14   440a11360326 ("workqueue: Calling destroy_work_on_stack() to 
pair with INIT_WORK_ONSTACK()")

The former updated work_on_cpu() to drop __flush_work() change to
allow recursive work_on_cpu() usage which no longer was necessary
after PCI update.  The latter added destroy_work_on_stack() call to
the function thus causing context conflict.  It can be easily resolved
by adding destroy_work_on_stack() to the version in master.

  long work_on_cpu(int cpu, long (*fn)(void *), void *arg)
  {
          struct work_for_cpu wfc = { .fn = fn, .arg = arg };

          INIT_WORK_ONSTACK(&wfc.work, work_for_cpu_fn);
          schedule_work_on(cpu, &wfc.work);
          flush_work(&wfc.work);
          destroy_work_on_stack(&wfc.work);
          return wfc.ret;
  }

The merge is also available in the test-merge-3.14 branch of wq tree.

The changes are available in the following git branch.

  git://git.kernel.org/pub/scm/linux/kernel/git/tj/wq.git for-3.14

for you to fetch changes up to 440a11360326044a9addf1c652a0364aad0be90c:

  workqueue: Calling destroy_work_on_stack() to pair with INIT_WORK_ONSTACK() 
(2014-01-11 22:26:33 -0500)

Thanks.
----------------------------------------------------------------
Chuansheng Liu (1):
      workqueue: Calling destroy_work_on_stack() to pair with 
INIT_WORK_ONSTACK()

 kernel/workqueue.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index 987293d..00df112 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -4776,6 +4776,7 @@ static int workqueue_cpu_down_callback(struct 
notifier_block *nfb,
 
                /* wait for per-cpu unbinding to finish */
                flush_work(&unbind_work);
+               destroy_work_on_stack(&unbind_work);
                break;
        }
        return NOTIFY_OK;
@@ -4822,6 +4823,7 @@ long work_on_cpu(int cpu, long (*fn)(void *), void *arg)
         */
        __flush_work(&wfc.work);
 
+       destroy_work_on_stack(&wfc.work);
        return wfc.ret;
 }
 EXPORT_SYMBOL_GPL(work_on_cpu);
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Reply via email to