Re: [PATCH net v2] net: dsa: mt7530: fix .get_stats64 sleeping in atomic context

[Paolo Abeni <pabeni@redhat.com>]

On 4/19/26 12:43 PM, Daniel Golle wrote:
> The .get_stats64 callback runs in atomic context, but on
> MDIO-connected switches every register read acquires the MDIO bus
> mutex, which can sleep:
> [   12.645973] BUG: sleeping function called from invalid context at kernel/locking/mutex.c:609
> [   12.654442] in_atomic(): 0, irqs_disabled(): 0, non_block: 0, pid: 759, name: grep
> [   12.663377] preempt_count: 0, expected: 0
> [   12.667410] RCU nest depth: 1, expected: 0
> [   12.671511] INFO: lockdep is turned off.
> [   12.675441] CPU: 0 UID: 0 PID: 759 Comm: grep Tainted: G S      W           7.0.0+ #0 PREEMPT
> [   12.675453] Tainted: [S]=CPU_OUT_OF_SPEC, [W]=WARN
> [   12.675456] Hardware name: Bananapi BPI-R64 (DT)
> [   12.675459] Call trace:
> [   12.675462]  show_stack+0x14/0x1c (C)
> [   12.675477]  dump_stack_lvl+0x68/0x8c
> [   12.675487]  dump_stack+0x14/0x1c
> [   12.675495]  __might_resched+0x14c/0x220
> [   12.675504]  __might_sleep+0x44/0x80
> [   12.675511]  __mutex_lock+0x50/0xb10
> [   12.675523]  mutex_lock_nested+0x20/0x30
> [   12.675532]  mt7530_get_stats64+0x40/0x2ac
> [   12.675542]  dsa_user_get_stats64+0x2c/0x40
> [   12.675553]  dev_get_stats+0x44/0x1e0
> [   12.675564]  dev_seq_printf_stats+0x24/0xe0
> [   12.675575]  dev_seq_show+0x14/0x3c
> [   12.675583]  seq_read_iter+0x37c/0x480
> [   12.675595]  seq_read+0xd0/0xec
> [   12.675605]  proc_reg_read+0x94/0xe4
> [   12.675615]  vfs_read+0x98/0x29c
> [   12.675625]  ksys_read+0x54/0xdc
> [   12.675633]  __arm64_sys_read+0x18/0x20
> [   12.675642]  invoke_syscall.constprop.0+0x54/0xec
> [   12.675653]  do_el0_svc+0x3c/0xb4
> [   12.675662]  el0_svc+0x38/0x200
> [   12.675670]  el0t_64_sync_handler+0x98/0xdc
> [   12.675679]  el0t_64_sync+0x158/0x15c
> 
> For MDIO-connected switches, poll MIB counters asynchronously using a
> delayed workqueue every second and let .get_stats64 return the cached
> values under a spinlock. A mod_delayed_work() call on each read
> triggers an immediate refresh so counters stay responsive when queried
> more frequently.
> 
> MMIO-connected switches (MT7988, EN7581, AN7583) are not affected
> because their regmap does not sleep, so they continue to read MIB
> counters directly in .get_stats64.
> 
> Fixes: 88c810f35ed5 ("net: dsa: mt7530: implement .get_stats64")
> Signed-off-by: Daniel Golle <daniel@makrotopia.org>
> Acked-by: Chester A. Unal <chester.a.unal@arinc9.com>
> Reviewed-by: Andrew Lunn <andrew@lunn.ch>
> ---
> v2:
>  * use spin_lock_bh()/spin_unlock_bh() to prevent potential deadlock
>  * rate-limit mod_delayed_work() refresh to at most once per 100ms
>  * move cancel_delayed_work_sync() after dsa_unregister_switch()
>  * add mt753x_teardown() callback to cancel the stats work
>  * fix commit message
> 
>  drivers/net/dsa/mt7530.c | 66 ++++++++++++++++++++++++++++++++++++++--
>  drivers/net/dsa/mt7530.h |  8 +++++
>  2 files changed, 71 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/net/dsa/mt7530.c b/drivers/net/dsa/mt7530.c
> index b9423389c2ef0..8c1186ba2279b 100644
> --- a/drivers/net/dsa/mt7530.c
> +++ b/drivers/net/dsa/mt7530.c
> @@ -25,6 +25,9 @@
>  
>  #include "mt7530.h"
>  
> +#define MT7530_STATS_POLL_INTERVAL	(1 * HZ)
> +#define MT7530_STATS_RATE_LIMIT		(HZ / 10)
> +
>  static struct mt753x_pcs *pcs_to_mt753x_pcs(struct phylink_pcs *pcs)
>  {
>  	return container_of(pcs, struct mt753x_pcs, pcs);
> @@ -906,10 +909,9 @@ static void mt7530_get_rmon_stats(struct dsa_switch *ds, int port,
>  	*ranges = mt7530_rmon_ranges;
>  }
>  
> -static void mt7530_get_stats64(struct dsa_switch *ds, int port,
> -			       struct rtnl_link_stats64 *storage)
> +static void mt7530_read_port_stats64(struct mt7530_priv *priv, int port,
> +				     struct rtnl_link_stats64 *storage)
>  {
> -	struct mt7530_priv *priv = ds->priv;
>  	uint64_t data;
>  
>  	/* MIB counter doesn't provide a FramesTransmittedOK but instead
> @@ -951,6 +953,45 @@ static void mt7530_get_stats64(struct dsa_switch *ds, int port,
>  			       &storage->rx_crc_errors);
>  }
>  
> +static void mt7530_stats_poll(struct work_struct *work)
> +{
> +	struct mt7530_priv *priv = container_of(work, struct mt7530_priv,
> +						stats_work.work);
> +	struct rtnl_link_stats64 stats = {};
> +	struct dsa_port *dp;
> +	int port;
> +
> +	dsa_switch_for_each_user_port(dp, priv->ds) {
> +		port = dp->index;
> +
> +		mt7530_read_port_stats64(priv, port, &stats);
> +
> +		spin_lock_bh(&priv->stats_lock);
> +		priv->ports[port].stats = stats;
> +		spin_unlock_bh(&priv->stats_lock);
> +	}
> +
> +	priv->stats_last = jiffies;
> +	schedule_delayed_work(&priv->stats_work,
> +			      MT7530_STATS_POLL_INTERVAL);
> +}
> +
> +static void mt7530_get_stats64(struct dsa_switch *ds, int port,
> +			       struct rtnl_link_stats64 *storage)
> +{
> +	struct mt7530_priv *priv = ds->priv;
> +
> +	if (priv->bus) {
> +		spin_lock_bh(&priv->stats_lock);
> +		*storage = priv->ports[port].stats;
> +		spin_unlock_bh(&priv->stats_lock);
> +		if (time_after(jiffies, priv->stats_last + MT7530_STATS_RATE_LIMIT))

Since both the `stats_last` access and read are lockless, it looks like
they may race leading wrong/unexpected delay. I think it would be better
move both under the spinlock (yes, the write will happen multiple times
per stats update, I don't think it will matter).

/P