CC2340R5: No ble scan disable event callback

Hello,

Thanks for reaching out.

Hi David,
Thanks for you response.
However, as you can see on the last picture above, we start ble scan at 15:01:49 with timeout 60s, that is no stop event until 15:09:12, morn than 8min.

BR,

Hello,

Would you mind specifying the SDK version you are using and how you are implementing the stress testing? If you are using an example as a reference which one is it and what parameters are you configuring the BLEAppUtil_scanStart API call with.

Thanks

David.

Hi David,

We use SDK simplelink_lowpower_f3_sdk_8_40_02_01.

We inplement stress testing as below:
  - start ble scan by invoke BLEAppUtil_scanStart 

  - in BLEAPPUTIL_ADV_REPORT event , we will stored scan response to array, once array be full, will notify others thread to store array to flash, during this time , will ignore new scan response until store to flash done.
  - in BLEAPPUTIL_SCAN_DISABLED event, we will print some log.

  - do some sleep(10s) and then repeat above steps as stress test

And below paramters are for BLEAppUtil_scanStart  API call:

const BLEAppUtil_ScanStart_t observer_scan_start_params = {
    /*! Zero for continuously scanning */
    .scanPeriod = 0, /* Units of 1.28sec */

    /*! Scan Duration shall be greater than to scan interval,*/
    /*! Zero continuously scanning. */
    .scanDuration = 6000, /* Units of 10ms */

    /*! If non-zero, the list of advertising reports will be */
    /*! generated and come with @ref GAP_EVT_SCAN_DISABLED. */
    .maxNumReport = 0
};

const BLEAppUtil_ScanInit_t observer_scan_init_params = {
    /*! Opt SCAN_PRIM_PHY_1M | SCAN_PRIM_PHY_CODED */
    .primPhy = SCAN_PRIM_PHY_1M,

    /*! Opt SCAN_TYPE_ACTIVE | SCAN_TYPE_PASSIVE */
    .scanType = SCAN_TYPE_PASSIVE,

    /*! Scan interval shall be greater than or equal to scan window */
    .scanInterval = 800, /* Units of 625 us */

    /*! Scan window shall be less than or equal to scan interval */
    .scanWindow = 800, /* Units of 625 us */

    /*! Select which fields of an advertising report will be stored */
    /*! in the AdvRptList, For mor field see @ref Gap_scanner.h */
    .advReportFields = SCAN_ADVRPT_FLD_ADDRESS,
 
    /*! Opt SCAN_PRIM_PHY_1M | SCAN_PRIM_PHY_CODED */
    .scanPhys = SCAN_PRIM_PHY_1M,

    /*! Opt SCAN_FLT_POLICY_ALL | SCAN_FLT_POLICY_WL | SCAN_FLT_POLICY_ALL_RPA
    * | SCAN_FLT_POLICY_WL_RPA */
    .fltPolicy = SCAN_FLT_POLICY_ALL,

    /*! For more filter PDU @ref Gap_scanner.h */
    .fltPduType = (SCAN_FLT_PDU_NONCONNECTABLE_ONLY | \
                           SCAN_FLT_PDU_NONSCANNABLE_ONLY | \
                           SCAN_FLT_PDU_UNDIRECTED_ONLY | \
                           SCAN_FLT_PDU_ADV_ONLY | \
                           SCAN_FLT_PDU_LEGACY_ONLY),

    /*! Opt SCAN_FLT_RSSI_ALL | SCAN_FLT_RSSI_NONE */
    .fltMinRssi = SCAN_FLT_RSSI_ALL,

    /*! Opt SCAN_FLT_DISC_NONE | SCAN_FLT_DISC_GENERAL | SCAN_FLT_DISC_LIMITED
    * | SCAN_FLT_DISC_ALL | SCAN_FLT_DISC_DISABLE */
    .fltDiscMode = SCAN_FLT_DISC_DISABLE,

    /*! Opt SCAN_FLT_DUP_ENABLE | SCAN_FLT_DUP_DISABLE | SCAN_FLT_DUP_RESET */
    .fltDup = SCAN_FLT_DUP_DISABLE
}

BR,

Hello,

Thanks for sharing. I am trying to reproduce what you see but I am constantly getting a scan disable event. However I am not implementing the adv report to flash saving process. How are you notifying the other task (semaphore?) Would you mind not executing the 2 step to see if the issue does not reproduces, pointing that the issue might be related to the task that saves the array into flash? I would like to take it from there.

Thanks.

David.

Hi David,

Thanks for your response.
Yeah, in most of time we also get the scan disable event normally. 
When the array is full, we will set a flag and notify other task by freertos TaskNotify function xTaskNotify(), when the new ble scan BLEAPPUTIL_ADV_REPORT event coming it will be ignore until the flag was reset.
We will try to follow your suggesttion to test exclude step 2, but it maybe will need a long time to check if it will reproduce this issue again or not.

BR,

Hello,

Thanks, it would be quite useful to know that. I will keep testing without it as well.

BR,

David.