Re: [BUG] WARNING: at kernel/trace/ring_buffer.c:3420

From: Steven Rostedt
Date: Mon Mar 08 2010 - 10:03:24 EST


On Mon, 2010-03-08 at 14:50 +0800, Lai Jiangshan wrote:
> Li Zefan wrote:
> > While running a ftrace test, a kernel warning showed up. The bug is not so
> > easy to reproduce, normally it takes 30 mins to 2 hours.
> >
> > After revering 3c05d7482777f15e71bb4cb1ba78dee2800dfec6, I ran the test
> > for 15 hours, and the bug seems disapeared.
> >
>
> This bug exhausted us, we found a issue: a tiny window which is
> one of causes of this bug. This patch is to close this tiny window.
> (bug is not fixed)
>
> From: Lai Jiangshan <laijs@xxxxxxxxxxxxxx>
>
> ringbuffer resizing and reseting will increase the ->record_disabled
> and then wait until a rcu_shced grace period passes.
>
> Contrarily, testing ->record_disabled should be at the same
> preempt disabled critical region as writing into ringbuffer, otherwise
> it will leave a window break ringbuffer resizing or reseting.

So the resizing and the resetting need a synchronize_sched() after the
disabling of the buffers, right?

-- Steve

>
> Signed-off-by: Lai Jiangshan <laijs@xxxxxxxxxxxxxx>
> Reported-by: Li Zefan <lizf@xxxxxxxxxxxxxx>
> ---
> diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c
> index 8c1b2d2..54191d6 100644
> --- a/kernel/trace/ring_buffer.c
> +++ b/kernel/trace/ring_buffer.c
> @@ -2232,12 +2232,12 @@ ring_buffer_lock_reserve(struct ring_buffer *buffer, unsigned long length)
> if (ring_buffer_flags != RB_BUFFERS_ON)
> return NULL;
>
> - if (atomic_read(&buffer->record_disabled))
> - return NULL;
> -
> /* If we are tracing schedule, we don't want to recurse */
> resched = ftrace_preempt_disable();
>
> + if (atomic_read(&buffer->record_disabled))
> + goto out_nocheck;
> +
> if (trace_recursive_lock())
> goto out_nocheck;
>
> @@ -2469,11 +2469,11 @@ int ring_buffer_write(struct ring_buffer *buffer,
> if (ring_buffer_flags != RB_BUFFERS_ON)
> return -EBUSY;
>
> - if (atomic_read(&buffer->record_disabled))
> - return -EBUSY;
> -
> resched = ftrace_preempt_disable();
>
> + if (atomic_read(&buffer->record_disabled))
> + goto out;
> +
> cpu = raw_smp_processor_id();
>
> if (!cpumask_test_cpu(cpu, buffer->cpumask))
>
>


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/