On 10/03/2015 06:19 PM, Nicholas Krause wrote:
> This fixes the locking region in the function brcmf_sdio_sendfromq
> to properly lock around the call to the function brcmrf_sdio_txpkt
> in order to avoid concurrent access issues when calling this
> function as stated in the function's comments about the caller
> being required to lock around the call to this particular function.
>
> Signed-off-by: Nicholas Krause <xerofoify@gmail.com>
> ---
>   drivers/net/wireless/brcm80211/brcmfmac/sdio.c | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/net/wireless/brcm80211/brcmfmac/sdio.c b/drivers/net/wireless/brcm80211/brcmfmac/sdio.c
> index f990e3d..da59683 100644
> --- a/drivers/net/wireless/brcm80211/brcmfmac/sdio.c
> +++ b/drivers/net/wireless/brcm80211/brcmfmac/sdio.c
> @@ -2388,11 +2388,11 @@ static uint brcmf_sdio_sendfromq(struct brcmf_sdio *bus, uint maxframes)
>   				break;
>   			__skb_queue_tail(&pktq, pkt);
>   		}
> -		spin_unlock_bh(&bus->txq_lock);
>   		if (i == 0)
>   			break;

So what if we break here. Seems you need to do unlock if i == 0 here.

Regards,
Arend

>
>   		ret = brcmf_sdio_txpkt(bus, &pktq, SDPCM_DATA_CHANNEL);
> +		spin_unlock_bh(&bus->txq_lock);
>
>   		cnt += i;
>
>