Re: [PATCH v5] ring-buffer: Ensure proper resetting of atomic variables in ring_buffer_reset_online_cpus

[Steven Rostedt]

For some reason, this email did not make it to
linux-trace-kernel@xxxxxxxxxxxxxxx, and therefore did not make it into
patchwork?

John?

-- Steve


On Wed, 26 Apr 2023 09:04:44 +0800
Tze-nan.Wu <Tze-nan.Wu@xxxxxxxxxxxx> wrote:

> From: "Tze-nan Wu" <Tze-nan.Wu@xxxxxxxxxxxx>
> 
> In ring_buffer_reset_online_cpus, the buffer_size_kb write operation
> may permanently fail if the cpu_online_mask changes between two
> for_each_online_buffer_cpu loops. The number of increases and decreases
> on both cpu_buffer->resize_disabled and cpu_buffer->record_disabled may be
> inconsistent, causing some CPUs to have non-zero values for these atomic
> variables after the function returns.
> 
> This issue can be reproduced by "echo 0 > trace" while hotplugging cpu.
> After reproducing success, we can find out buffer_size_kb will not be
> functional anymore.
> 
> To prevent leaving 'resize_disabled' and 'record_disabled' non-zero after
> ring_buffer_reset_online_cpus returns, we ensure that each atomic variable
> has been set up before atomic_sub() to it.
> 
> Cc: stable@xxxxxxxxxxxxxxx
> Cc: npiggin@xxxxxxxxx
> Fixes: b23d7a5f4a07 ("ring-buffer: speed up buffer resets by avoiding synchronize_rcu for each CPU")
> Reviewed-by: Cheng-Jui Wang <cheng-jui.wang@xxxxxxxxxxxx>
> Signed-off-by: Tze-nan Wu <Tze-nan.Wu@xxxxxxxxxxxx>
> ---
> Changes from v4 to v5: https://lore.kernel.org/lkml/20230412112401.25081-1-Tze-nan.Wu@xxxxxxxxxxxx/
>   - Move the define before the function
> ---
>  kernel/trace/ring_buffer.c | 16 +++++++++++++---
>  1 file changed, 13 insertions(+), 3 deletions(-)
> 
> diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c
> index 76a2d91eecad..253ef85a9ec3 100644
> --- a/kernel/trace/ring_buffer.c
> +++ b/kernel/trace/ring_buffer.c
> @@ -5345,6 +5345,9 @@ void ring_buffer_reset_cpu(struct trace_buffer *buffer, int cpu)
>  }
>  EXPORT_SYMBOL_GPL(ring_buffer_reset_cpu);
>  
> +/* Flag to ensure proper resetting of atomic variables */
> +#define RESET_BIT	(1 << 30)
> +
>  /**
>   * ring_buffer_reset_online_cpus - reset a ring buffer per CPU buffer
>   * @buffer: The ring buffer to reset a per cpu buffer of
> @@ -5361,20 +5364,27 @@ void ring_buffer_reset_online_cpus(struct trace_buffer *buffer)
>  	for_each_online_buffer_cpu(buffer, cpu) {
>  		cpu_buffer = buffer->buffers[cpu];
>  
> -		atomic_inc(&cpu_buffer->resize_disabled);
> +		atomic_add(RESET_BIT, &cpu_buffer->resize_disabled);
>  		atomic_inc(&cpu_buffer->record_disabled);
>  	}
>  
>  	/* Make sure all commits have finished */
>  	synchronize_rcu();
>  
> -	for_each_online_buffer_cpu(buffer, cpu) {
> +	for_each_buffer_cpu(buffer, cpu) {
>  		cpu_buffer = buffer->buffers[cpu];
>  
> +		/*
> +		 * If a CPU came online during the synchronize_rcu(), then
> +		 * ignore it.
> +		 */
> +		if (!(atomic_read(&cpu_buffer->resize_disabled) & RESET_BIT))
> +			continue;
> +
>  		reset_disabled_cpu_buffer(cpu_buffer);
>  
>  		atomic_dec(&cpu_buffer->record_disabled);
> -		atomic_dec(&cpu_buffer->resize_disabled);
> +		atomic_sub(RESET_BIT, &cpu_buffer->resize_disabled);
>  	}
>  
>  	mutex_unlock(&buffer->mutex);