Re: [PATCH v2 1/2] scan: refactor start_next_scan_request to not send duplicate requests

[Denis Kenzior <denkenz@gmail.com>]

[-- Attachment #1: Type: text/plain, Size: 1997 bytes --]

Hi Alvin,

> I traced the following events leading to a duplicate scan request being sent. 
> The scenario assumes that the scan request queue is empty in the beginning.
> 
> 
> station_enter_state: disconnected -> autoconnect_quick
>    -> station_quick_scan_trigger
>      -> station_scan_trigger: notify cb = station_quick_scan_results
>        -> scan_passive
>          -> scan_common
>            -> scan_request_send_trigger
>            -> l_queue_push_tail: scan request queue size = 1
> scan_notify: quick scan triggered
> network_connect: requested over dbus
>    -> network_connect_psk
>      -> station_connect_network
>        -> station_enter_state: autoconnect_quick -> connecting
>          -> periodic_scan_stop
>            -> scan_periodic_stop
>              -> scan_cancel
> scan_notify: quick scan aborted
>    -> scan_finished
>      -> l_queue_remove: scan request queue size = 0
>      -> station_quick_scan_results: quick scan's notify callback
>      . -> station_enter_state: connecting -> autoconnect_full

Aha, so this is the real culprit.  I fixed this as well in commit 
532f6b154e03b5fa222502098b5688a1784cc4e3.

>      .   -> scan_periodic_start
>      .     -> scan_periodic_queue
>      .       -> scan_{active_full,passive}: idk which; doesn't matter
>      .         -> scan_common
>      .           -> scan_request_send_trigger
>      .           -> l_queue_push_tail: scan request queue size = 1

Okay, so we start a periodic scan and in scan_finished call 
start_next_scan_request again.

>      -> start_next_scan_request: head of queue has already been sent
>        -> scan_request_send_trigger: send it again
> 
> 

And since we don't check that one is already going, we issue the trigger again. 
Makes sense to me now.  Much appreciated.

Regards,
-Denis