We hit one rare case below:
T1 calling disable_irq(), but hanging at synchronize_irq()
always;
The corresponding irq thread is in sleeping state;
And all CPUs are in idle state;
After analysis, we found there is one possible scenerio which
causes T1 is waiting there forever:
CPU0 CPU1
synchronize_irq()
wait_event()
spin_lock()
atomic_dec_and_test(&threads_active)
insert the __wait into queue
spin_unlock()
if(waitqueue_active)
atomic_read(&threads_active)
wait_up()
Here after inserted the __wait into queue on CPU0, and before
test if queue is empty on CPU1, there is no barrier, it maybe
cause it is not visible for CPU1 immediately, although CPU0 has
updated the queue list.
It is similar for CPU0 atomic_read() threads_active also.
So we need one smp_mb() before waitqueue_active or something like
that.
Thomas shared one good option that removing waitqueue_active()
judgement directly, it will make things to be simple and clear.
Cc: Thomas Gleixner <[email protected]>
Cc: Xiaoming Wang <[email protected]>
Signed-off-by: Chuansheng Liu <[email protected]>
---
kernel/irq/manage.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c
index 481a13c..d3bf660 100644
--- a/kernel/irq/manage.c
+++ b/kernel/irq/manage.c
@@ -802,8 +802,7 @@ static irqreturn_t irq_thread_fn(struct irq_desc *desc,
static void wake_threads_waitq(struct irq_desc *desc)
{
- if (atomic_dec_and_test(&desc->threads_active) &&
- waitqueue_active(&desc->wait_for_threads))
+ if (atomic_dec_and_test(&desc->threads_active))
wake_up(&desc->wait_for_threads);
}
--
1.9.rc0
Commit-ID: c685689fd24d310343ac33942e9a54a974ae9c43
Gitweb: http://git.kernel.org/tip/c685689fd24d310343ac33942e9a54a974ae9c43
Author: Chuansheng Liu <[email protected]>
AuthorDate: Mon, 24 Feb 2014 11:29:50 +0800
Committer: Thomas Gleixner <[email protected]>
CommitDate: Thu, 27 Feb 2014 10:54:16 +0100
genirq: Remove racy waitqueue_active check
We hit one rare case below:
T1 calling disable_irq(), but hanging at synchronize_irq()
always;
The corresponding irq thread is in sleeping state;
And all CPUs are in idle state;
After analysis, we found there is one possible scenerio which
causes T1 is waiting there forever:
CPU0 CPU1
synchronize_irq()
wait_event()
spin_lock()
atomic_dec_and_test(&threads_active)
insert the __wait into queue
spin_unlock()
if(waitqueue_active)
atomic_read(&threads_active)
wake_up()
Here after inserted the __wait into queue on CPU0, and before
test if queue is empty on CPU1, there is no barrier, it maybe
cause it is not visible for CPU1 immediately, although CPU0 has
updated the queue list.
It is similar for CPU0 atomic_read() threads_active also.
So we'd need one smp_mb() before waitqueue_active.that, but removing
the waitqueue_active() check solves it as wel l and it makes
things simple and clear.
Signed-off-by: Chuansheng Liu <[email protected]>
Cc: Xiaoming Wang <[email protected]>
Link: http://lkml.kernel.org/r/[email protected]
Cc: [email protected]
Signed-off-by: Thomas Gleixner <[email protected]>
---
kernel/irq/manage.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c
index 481a13c..d3bf660 100644
--- a/kernel/irq/manage.c
+++ b/kernel/irq/manage.c
@@ -802,8 +802,7 @@ static irqreturn_t irq_thread_fn(struct irq_desc *desc,
static void wake_threads_waitq(struct irq_desc *desc)
{
- if (atomic_dec_and_test(&desc->threads_active) &&
- waitqueue_active(&desc->wait_for_threads))
+ if (atomic_dec_and_test(&desc->threads_active))
wake_up(&desc->wait_for_threads);
}