Re: [PATCH v20 07/10] ring-buffer: Skip invalid sub-buffers for iterator

[Masami Hiramatsu (Google) <mhiramat@kernel.org>]

On Wed, 20 May 2026 14:49:45 -0400
Steven Rostedt <rostedt@kernel.org> wrote:

> From: Steven Rostedt <rostedt@goodmis.org>
> 
> On bootup if the persistent ring buffer finds an invalid sub-buffer, it
> only invalidates the invalid sub-buffer and continues. Several sub-buffers
> may be invalid and this can cause the iterator to loop more than 3 times
> looking for a new event. If that happens, then it returns NULL. Having a
> NULL return early can confuse the iterator looking for the next event, and
> may show events out of order.
> 
> Have the same logic for the consuming read for the iterator that will
> allow the loop to find the next event to happen the number of sub-buffers
> and not just 3.
> 
> Fixes: **TBD** ring-buffer: Skip invalid sub-buffers when validating persistent ring buffer
> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>

Should we merge this into the original one?

Anyway, this looks good to me.

Thanks,

> ---
>  kernel/trace/ring_buffer.c | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
> 
> diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c
> index c6c2f92bfc24..bda53a2d2159 100644
> --- a/kernel/trace/ring_buffer.c
> +++ b/kernel/trace/ring_buffer.c
> @@ -6103,12 +6103,14 @@ rb_iter_peek(struct ring_buffer_iter *iter, u64 *ts)
>  	struct ring_buffer_per_cpu *cpu_buffer;
>  	struct ring_buffer_event *event;
>  	int nr_loops = 0;
> +	int max_loops;
>  
>  	if (ts)
>  		*ts = 0;
>  
>  	cpu_buffer = iter->cpu_buffer;
>  	buffer = cpu_buffer->buffer;
> +	max_loops = cpu_buffer->ring_meta ? cpu_buffer->nr_pages : 3;
>  
>  	/*
>  	 * Check if someone performed a consuming read to the buffer
> @@ -6131,7 +6133,7 @@ rb_iter_peek(struct ring_buffer_iter *iter, u64 *ts)
>  	 * the ring buffer with an active write as the consumer is.
>  	 * Do not warn if the three failures is reached.
>  	 */
> -	if (++nr_loops > 3)
> +	if (++nr_loops > max_loops)
>  		return NULL;
>  
>  	if (rb_per_cpu_empty(cpu_buffer))
> -- 
> 2.53.0
> 
> 


-- 
Masami Hiramatsu (Google) <mhiramat@kernel.org>