Re: [PATCH v20 09/10] ring-buffer: Show persistent buffer dropped events in trace file

[Google]

On Wed, 20 May 2026 14:49:47 -0400
Steven Rostedt <rostedt@xxxxxxxxxx> wrote:

> From: Steven Rostedt <rostedt@xxxxxxxxxxx>
> 
> When the persistent ring buffer is validated on boot up, if a subbuffer is
> deemed invalid, it resets the buffer and continues. Currently, these lost
> events are not shown in the trace file output.
> 
> Have the trace iterator look for subbuffers that have the RB_MISSED_EVENTS
> set and set the iter->missed_events flag when it is detected. This will
> then have the trace file shows "LOST EVENTS" when it reads across a
> subbuffer that was corrupted and invalidated.

I think it is good to show the dropped events. BTW, is it better to
comment out the line, just for parser?
For example, add a '#' at the like.

# CPU:5 [LOST EVENTS]

Ah, but it is already done...

Thanks,

> 
> For example:
> 
>  <...>-1016    [005] ...1.  6230.660403: preempt_disable: caller=__mod_memcg_state+0x1c8/0x200 parent=__mod_memcg_state+0x1c8/0x200
> CPU:5 [LOST EVENTS]
>  <...>-1016    [005] .....  6230.660673: kmem_cache_alloc: call_site=__anon_vma_prepare+0x1ad/0x1e0 ptr=000000006e40294c name=anon_vma bytes_req=200 bytes_alloc=208 gfp_flags=GFP_KERNEL node=-1 accounted=true
> 
> Signed-off-by: Steven Rostedt <rostedt@xxxxxxxxxxx>
> ---
>  kernel/trace/ring_buffer.c | 14 +++++++++++---
>  1 file changed, 11 insertions(+), 3 deletions(-)
> 
> diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c
> index ae5c645b59c9..9cdbee171cdc 100644
> --- a/kernel/trace/ring_buffer.c
> +++ b/kernel/trace/ring_buffer.c
> @@ -3518,6 +3518,9 @@ static void rb_inc_iter(struct ring_buffer_iter *iter)
>  	else
>  		rb_inc_page(&iter->head_page);
>  
> +	if (rb_page_commit(iter->head_page) & RB_MISSED_EVENTS)
> +		iter->missed_events = -1;
> +
>  	iter->page_stamp = iter->read_stamp = iter->head_page->page->time_stamp;
>  	iter->head = 0;
>  	iter->next_event = 0;
> @@ -5579,6 +5582,7 @@ static void rb_iter_reset(struct ring_buffer_iter *iter)
>  	iter->head_page = cpu_buffer->reader_page;
>  	iter->head = cpu_buffer->reader_page->read;
>  	iter->next_event = iter->head;
> +	iter->missed_events = 0;
>  
>  	iter->cache_reader_page = iter->head_page;
>  	iter->cache_read = cpu_buffer->read;
> @@ -7053,7 +7057,7 @@ int ring_buffer_read_page(struct trace_buffer *buffer,
>  	struct ring_buffer_event *event;
>  	struct buffer_data_page *dpage;
>  	struct buffer_page *reader;
> -	unsigned long missed_events;
> +	long missed_events;
>  	unsigned int commit;
>  	unsigned int read;
>  	u64 save_timestamp;
> @@ -7179,6 +7183,8 @@ int ring_buffer_read_page(struct trace_buffer *buffer,
>  		local_set(&reader->entries, 0);
>  		reader->read = 0;
>  		data_page->data = dpage;
> +		if (!missed_events && rb_data_page_commit(dpage) & RB_MISSED_EVENTS)
> +			missed_events = -1;
>  
>  		/*
>  		 * Use the real_end for the data size,
> @@ -7196,10 +7202,12 @@ int ring_buffer_read_page(struct trace_buffer *buffer,
>  	 * Set a flag in the commit field if we lost events
>  	 */
>  	if (missed_events) {
> -		/* If there is room at the end of the page to save the
> +		/*
> +		 * If there is room at the end of the page to save the
>  		 * missed events, then record it there.
>  		 */
> -		if (buffer->subbuf_size - commit >= sizeof(missed_events)) {
> +		if (missed_events > 0 &&
> +		    buffer->subbuf_size - commit >= sizeof(missed_events)) {
>  			memcpy(&dpage->data[commit], &missed_events,
>  			       sizeof(missed_events));
>  			local_add(RB_MISSED_STORED, &dpage->commit);
> -- 
> 2.53.0
> 
> 


-- 
Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx>