Re: [PATCH net-next 04/37] can: proc: reset pkg_stats atomics individually

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] can: dropped statistics updates due to unprotected atomic_long_set()
--

commit a887271f48442196a8a536b23ac554282d14a9f4
Author: Runyu Xiao <runyu.xiao@seu.edu.cn>

can: proc: reset pkg_stats atomics individually

This commit replaces a whole-struct memset() in can_init_stats() with a
helper that explicitly resets individual fields. The memset() caused
non-atomic writes to atomic_long_t members when called concurrently with
hot-path statistics updates.

Link: https://patch.msgid.link/20260512133937.21957-1-runyu.xiao@seu.edu.cn

> diff --git a/net/can/proc.c b/net/can/proc.c
> index de4d05ae34597..64b3bdc2fa7e0 100644
> --- a/net/can/proc.c
> +++ b/net/can/proc.c
> @@ -76,16 +76,39 @@ static const char rx_list_name[][8] = {
>   * af_can statistics stuff
>   */
>  
> +static void can_reset_pkg_stats(struct can_pkg_stats *pkg_stats)
> +{
> +	atomic_long_set(&pkg_stats->rx_frames, 0);
> +	atomic_long_set(&pkg_stats->tx_frames, 0);
> +	atomic_long_set(&pkg_stats->matches, 0);
> +
> +	pkg_stats->total_rx_rate = 0;
> +	pkg_stats->total_tx_rate = 0;
> +	pkg_stats->total_rx_match_ratio = 0;
> +
> +	pkg_stats->current_rx_rate = 0;
> +	pkg_stats->current_tx_rate = 0;
> +	pkg_stats->current_rx_match_ratio = 0;
> +
> +	pkg_stats->max_rx_rate = 0;
> +	pkg_stats->max_tx_rate = 0;
> +	pkg_stats->max_rx_match_ratio = 0;
> +
> +	atomic_long_set(&pkg_stats->rx_frames_delta, 0);
> +	atomic_long_set(&pkg_stats->tx_frames_delta, 0);
> +	atomic_long_set(&pkg_stats->matches_delta, 0);

[Severity: Medium]
Does unconditionally clearing the delta counters here drop concurrent
increments?

If a packet is received or transmitted concurrently during a statistics reset
or during the can_stat_update() calculation window in net/can/proc.c,
can_receive() in net/can/af_can.c will increment the delta counter via
atomic_long_inc(&pkg_stats->rx_frames_delta).

If atomic_long_set(..., 0) runs right after that increment, won't it
unconditionally zero the counter, dropping the increment entirely from both the
current and next intervals?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261015015336.816543-1-mkl@pengutronix.de?part=4