Some firmware revisions don't seem to deilver
management frames with FCS error via WMI so narrow
down the HTT rule to not drop corrupted management
frames.
This basically increases number of frames ath10k
reports while sniffing.
Signed-off-by: Michal Kazior <[email protected]>
---
drivers/net/wireless/ath/ath10k/htt_rx.c | 13 +++++++++++--
drivers/net/wireless/ath/ath10k/wmi.c | 7 +++++--
2 files changed, 16 insertions(+), 4 deletions(-)
diff --git a/drivers/net/wireless/ath/ath10k/htt_rx.c b/drivers/net/wireless/ath/ath10k/htt_rx.c
index 75b2293..85910b2 100644
--- a/drivers/net/wireless/ath/ath10k/htt_rx.c
+++ b/drivers/net/wireless/ath/ath10k/htt_rx.c
@@ -1381,6 +1381,8 @@ static bool ath10k_htt_rx_amsdu_allowed(struct ath10k *ar,
{
struct sk_buff *msdu;
struct htt_rx_desc *rxd;
+ bool is_mgmt;
+ bool has_fcs_err;
msdu = skb_peek(amsdu);
rxd = (void *)msdu->data - sizeof(*rxd);
@@ -1394,12 +1396,19 @@ static bool ath10k_htt_rx_amsdu_allowed(struct ath10k *ar,
return false;
}
+ is_mgmt = !!(rxd->attention.flags &
+ __cpu_to_le32(RX_ATTENTION_FLAGS_MGMT_TYPE));
+ has_fcs_err = !!(rxd->attention.flags &
+ __cpu_to_le32(RX_ATTENTION_FLAGS_FCS_ERR));
+
/* Management frames are handled via WMI events. The pros of such
* approach is that channel is explicitly provided in WMI events
* whereas HTT doesn't provide channel information for Rxed frames.
+ *
+ * However some firmware revisions don't report corrupted frames via
+ * WMI so don't drop them.
*/
- if (rxd->attention.flags &
- __cpu_to_le32(RX_ATTENTION_FLAGS_MGMT_TYPE)) {
+ if (is_mgmt && !has_fcs_err) {
ath10k_dbg(ar, ATH10K_DBG_HTT, "htt rx mgmt ctrl\n");
return false;
}
diff --git a/drivers/net/wireless/ath/ath10k/wmi.c b/drivers/net/wireless/ath/ath10k/wmi.c
index c2bc828..df948a4 100644
--- a/drivers/net/wireless/ath/ath10k/wmi.c
+++ b/drivers/net/wireless/ath/ath10k/wmi.c
@@ -1166,8 +1166,11 @@ static int ath10k_wmi_event_mgmt_rx(struct ath10k *ar, struct sk_buff *skb)
return 0;
}
- if (rx_status & WMI_RX_STATUS_ERR_CRC)
- status->flag |= RX_FLAG_FAILED_FCS_CRC;
+ if (rx_status & WMI_RX_STATUS_ERR_CRC) {
+ dev_kfree_skb(skb);
+ return 0;
+ }
+
if (rx_status & WMI_RX_STATUS_ERR_MIC)
status->flag |= RX_FLAG_MMIC_ERROR;
--
1.8.5.3
Michal Kazior <[email protected]> writes:
> Some firmware revisions don't seem to deilver
> management frames with FCS error via WMI so narrow
> down the HTT rule to not drop corrupted management
> frames.
>
> This basically increases number of frames ath10k
> reports while sniffing.
>
> Signed-off-by: Michal Kazior <[email protected]>
Thanks, applied.
--
Kalle Valo
Michal Kazior <[email protected]> writes:
> Some firmware revisions don't seem to deilver
> management frames with FCS error via WMI so narrow
> down the HTT rule to not drop corrupted management
> frames.
>
> This basically increases number of frames ath10k
> reports while sniffing.
>
> Signed-off-by: Michal Kazior <[email protected]>
Thanks, applied.
--
Kalle Valo
Michal Kazior <[email protected]> writes:
> On 25 November 2014 at 10:04, Kalle Valo <[email protected]> wrote:
>> Michal Kazior <[email protected]> writes:
> [...]
>>> @@ -1394,12 +1396,19 @@ static bool ath10k_htt_rx_amsdu_allowed(struct ath10k *ar,
>>> return false;
>>> }
>>>
>>> + is_mgmt = !!(rxd->attention.flags &
>>> + __cpu_to_le32(RX_ATTENTION_FLAGS_MGMT_TYPE));
>>> + has_fcs_err = !!(rxd->attention.flags &
>>> + __cpu_to_le32(RX_ATTENTION_FLAGS_FCS_ERR));
>>
>> I think I asked this before in some other patch, but isn't '!!' operator
>> useless here? is_mgmt is a boolean so the compiler should convert it
>> correctly without '!!' anyway, right?
>
> If I remove `!!` I get:
>
> warning: incorrect type in assignment (different base types)
> expected bool [unsigned] [usertype] is_mgmt
> got restricted __le32
Oh. Sorry, my bad.
> I could do a temporary `u32 attention` to get rid of the `!!`. You
> want me to update it?
Please don't, this is fine.
--
Kalle Valo
Michal Kazior <[email protected]> writes:
> --- a/drivers/net/wireless/ath/ath10k/htt_rx.c
> +++ b/drivers/net/wireless/ath/ath10k/htt_rx.c
> @@ -1381,6 +1381,8 @@ static bool ath10k_htt_rx_amsdu_allowed(struct ath10k *ar,
> {
> struct sk_buff *msdu;
> struct htt_rx_desc *rxd;
> + bool is_mgmt;
> + bool has_fcs_err;
>
> msdu = skb_peek(amsdu);
> rxd = (void *)msdu->data - sizeof(*rxd);
> @@ -1394,12 +1396,19 @@ static bool ath10k_htt_rx_amsdu_allowed(struct ath10k *ar,
> return false;
> }
>
> + is_mgmt = !!(rxd->attention.flags &
> + __cpu_to_le32(RX_ATTENTION_FLAGS_MGMT_TYPE));
> + has_fcs_err = !!(rxd->attention.flags &
> + __cpu_to_le32(RX_ATTENTION_FLAGS_FCS_ERR));
I think I asked this before in some other patch, but isn't '!!' operator
useless here? is_mgmt is a boolean so the compiler should convert it
correctly without '!!' anyway, right?
--
Kalle Valo
On 25 November 2014 at 10:04, Kalle Valo <[email protected]> wrote:
> Michal Kazior <[email protected]> writes:
[...]
>> @@ -1394,12 +1396,19 @@ static bool ath10k_htt_rx_amsdu_allowed(struct ath10k *ar,
>> return false;
>> }
>>
>> + is_mgmt = !!(rxd->attention.flags &
>> + __cpu_to_le32(RX_ATTENTION_FLAGS_MGMT_TYPE));
>> + has_fcs_err = !!(rxd->attention.flags &
>> + __cpu_to_le32(RX_ATTENTION_FLAGS_FCS_ERR));
>
> I think I asked this before in some other patch, but isn't '!!' operator
> useless here? is_mgmt is a boolean so the compiler should convert it
> correctly without '!!' anyway, right?
If I remove `!!` I get:
warning: incorrect type in assignment (different base types)
expected bool [unsigned] [usertype] is_mgmt
got restricted __le32
I could do a temporary `u32 attention` to get rid of the `!!`. You
want me to update it?
Michał