[PATCH v2 3/5] ath10k: drain tx before restarting hw

[Kalle Valo]

Michal Kazior <michal.kazior at tieto.com> writes:

> This makes sure no further tx requests are
> submitted to HTT before driver teardown.
>
> This should prevent invalid pointer/NULL
> dereference on htt tx pool in ath10k_htt_tx() in
> some cases of heavy traffic.
>
> Reported-By: Ben Greear <greearb at candelatech.com>
> Signed-off-by: Michal Kazior <michal.kazior at tieto.com>
> ---
>  drivers/net/wireless/ath/ath10k/mac.c | 30 ++++++++++++++++++++++++------
>  1 file changed, 24 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/net/wireless/ath/ath10k/mac.c b/drivers/net/wireless/ath/ath10k/mac.c
> index b5d75b8..577a3a5 100644
> --- a/drivers/net/wireless/ath/ath10k/mac.c
> +++ b/drivers/net/wireless/ath/ath10k/mac.c
> @@ -2289,6 +2289,21 @@ static void ath10k_tx(struct ieee80211_hw *hw,
>  /*
>   * Initialize various parameters with default vaules.
>   */
> +static void ath10k_drain_tx(struct ath10k *ar)
> +{
> +	/* workers can hold conf_mutex -- avoid deadlock */
> +	WARN_ON(lockdep_is_held(&ar->conf_mutex));

Based on the discussion I modified the function to this now. Are you ok
with that?

/* Must not be called with conf_mutex held as workers can use that also. */
static void ath10k_drain_tx(struct ath10k *ar)
{
	/* make sure rcu-protected mac80211 tx path itself is drained */
	synchronize_net();

	ath10k_offchan_tx_purge(ar);
	ath10k_mgmt_over_wmi_tx_purge(ar);

	cancel_work_sync(&ar->offchan_tx_work);
	cancel_work_sync(&ar->wmi_mgmt_tx_work);
}

Full patch:

https://github.com/kvalo/ath/commit/e3c513cd03802e75802a6ce5efbadaa0dbf04b27

-- 
Kalle Valo