[PATCH 2/3] wcn36xx: Clean up wcn36xx_smd_send_beacon

[Eugene Krasnikov]

size_of(struct wcn36xx_hal_send_beacon_req_msg) is different with your
patch and that means PREPARE_HAL_BUF will copy more bytes to the
hal_buf buffer. So in the end wcn36xx_smd_send_and_wait will send to
the hardware all the bytes that are in the hal_buf buffer and that
means extra 4 bytes will be sent to HW. Or do i miss something?

2014-02-12 19:50 GMT+00:00 Pontus Fuchs <pontus.fuchs at gmail.com>:
> On 2014-02-12 17:10, Eugene Krasnikov wrote:
>>
>> Hm... I thought if you change any of the
>> wcn36xx_hal_send_beacon_req_msg structures you will change the
>> protocol between the host and the chip. So by adding beacon_length6;
>> you increases the size of the message body on 4 bytes. Are you sure
>> there are no functional changes in this patch?
>
>
> Yes I am pretty sure. See below why.
>
>
>> 2014-02-11 15:53 GMT+00:00 Pontus Fuchs <pontus.fuchs at gmail.com>:
>>>
>>> No functional changes.
>>>
>>> Signed-off-by: Pontus Fuchs <pontus.fuchs at gmail.com>
>>> ---
>>>   hal.h |  7 +++++--
>>>   smd.c | 12 +++++-------
>>>   2 files changed, 10 insertions(+), 9 deletions(-)
>>>
>>> diff --git a/hal.h b/hal.h
>>> index c5e7473..0a4d3fb 100644
>>> --- a/hal.h
>>> +++ b/hal.h
>>> @@ -52,8 +52,8 @@
>>>   #define WALN_HAL_STA_INVALID_IDX 0xFF
>>>   #define WCN36XX_HAL_BSS_INVALID_IDX 0xFF
>>>
>>> -/* Default Beacon template size */
>>> -#define BEACON_TEMPLATE_SIZE 0x180
>>> +/* Default Beacon template size. */
>>> +#define BEACON_TEMPLATE_SIZE 0x17C
>>>
>>>   /* Param Change Bitmap sent to HAL */
>>>   #define PARAM_BCN_INTERVAL_CHANGED                      (1 << 0)
>>> @@ -2883,6 +2883,9 @@ struct update_beacon_rsp_msg {
>>>   struct wcn36xx_hal_send_beacon_req_msg {
>>>          struct wcn36xx_hal_msg_header header;
>>>
>>> +       /* length of the template + 6. Only qcom knows why */
>>> +       u32 beacon_length6;
>
> This field is new as you noticed....
>
>
>>> +
>>>          /* length of the template. */
>>>          u32 beacon_length;
>>>
>>> diff --git a/smd.c b/smd.c
>>> index 2a67816..bf7092a 100644
>>> --- a/smd.c
>>> +++ b/smd.c
>>> @@ -1298,19 +1298,17 @@ int wcn36xx_smd_send_beacon(struct wcn36xx *wcn,
>>> struct ieee80211_vif *vif,
>>>          mutex_lock(&wcn->hal_mutex);
>>>          INIT_HAL_MSG(msg_body, WCN36XX_HAL_SEND_BEACON_REQ);
>>>
>>> -       /* TODO need to find out why this is needed? */
>>> -       msg_body.beacon_length = skb_beacon->len + 6;
>>> +       msg_body.beacon_length = skb_beacon->len;
>>> +       /* TODO need to find out why + 6 is needed */
>>> +       msg_body.beacon_length6 = msg_body.beacon_length + 6;
>>>
>>> -       if (BEACON_TEMPLATE_SIZE > msg_body.beacon_length) {
>>> -               memcpy(&msg_body.beacon, &skb_beacon->len, sizeof(u32));
>>> -               memcpy(&(msg_body.beacon[4]), skb_beacon->data,
>>> -                      skb_beacon->len);
>
> But I remove this strange stuff to keep the balance. The first memcpy
> corresponds to the assignment of msg_body.beacon_length
>
> Cheers,
>
> Pontus



-- 
Best regards,
Eugene