RE: HardFault on sd_ble_gap_adv_stop()

Andy — Wed, 12 Jun 2019 14:00:30 GMT

Nice, thanks again!

RE: HardFault on sd_ble_gap_adv_stop()

Vidar Berg — Wed, 12 Jun 2019 13:48:13 GMT

Sure:)

The last bits of the PSR register shows the current ISR number (ARM doc): So 0x01000016 corresponds to ISR number 0x16. Then I subtract this number by 16 as the non-programmable interrupts start at minus 15

0x16 -> 22 - 16 = 6.

Then you can look up ISR number 6 in the nrf52.h/nrf52xx.h header file:

/* =========================================================================================================================== */
/* ================                                Interrupt Number Definition                                ================ */
/* =========================================================================================================================== */

typedef enum {
/* =======================================  ARM Cortex-M4 Specific Interrupt Numbers  ======================================== */
  Reset_IRQn                = -15,              /*!< -15  Reset Vector, invoked on Power up and warm reset                     */
  NonMaskableInt_IRQn       = -14,              /*!< -14  Non maskable Interrupt, cannot be stopped or preempted               */
  HardFault_IRQn            = -13,              /*!< -13  Hard Fault, all classes of Fault                                     */
  MemoryManagement_IRQn     = -12,              /*!< -12  Memory Management, MPU mismatch, including Access Violation
                                                     and No Match                                                              */
  BusFault_IRQn             = -11,              /*!< -11  Bus Fault, Pre-Fetch-, Memory Access Fault, other address/memory
                                                     related Fault                                                             */
  UsageFault_IRQn           = -10,              /*!< -10  Usage Fault, i.e. Undef Instruction, Illegal State Transition        */
  SVCall_IRQn               =  -5,              /*!< -5 System Service Call via SVC instruction                                */
  DebugMonitor_IRQn         =  -4,              /*!< -4 Debug Monitor                                                          */
  PendSV_IRQn               =  -2,              /*!< -2 Pendable request for system service                                    */
  SysTick_IRQn              =  -1,              /*!< -1 System Tick Timer                                                      */
/* ===========================================  nrf52 Specific Interrupt Numbers  ============================================ */
  POWER_CLOCK_IRQn          =   0,              /*!< 0  POWER_CLOCK                                                            */
  RADIO_IRQn                =   1,              /*!< 1  RADIO                                                                  */
  UARTE0_UART0_IRQn         =   2,              /*!< 2  UARTE0_UART0                                                           */
  SPIM0_SPIS0_TWIM0_TWIS0_SPI0_TWI0_IRQn=   3,  /*!< 3  SPIM0_SPIS0_TWIM0_TWIS0_SPI0_TWI0                                      */
  SPIM1_SPIS1_TWIM1_TWIS1_SPI1_TWI1_IRQn=   4,  /*!< 4  SPIM1_SPIS1_TWIM1_TWIS1_SPI1_TWI1                                      */
  NFCT_IRQn                 =   5,              /*!< 5  NFCT                                                                   */
  GPIOTE_IRQn               =   6,              /*!< 6  GPIOTE                                                                 */
  SAADC_IRQn                =   7,              /*!< 7  SAADC                                                                  */
  TIMER0_IRQn               =   8,              /*!< 8  TIMER0                                                                 */
  TIMER1_IRQn               =   9,              /*!< 9  TIMER1                                                                 */

RE: HardFault on sd_ble_gap_adv_stop()

Andy — Wed, 12 Jun 2019 13:37:59 GMT

Yep, this was it! Thanks again for your help 🙂

Before I verify the answer, could you explain how were you able to tell that the call came from the GPIOTE IRQ handler? How can I use that information in the future?

RE: HardFault on sd_ble_gap_adv_stop()

Andy — Wed, 12 Jun 2019 13:15:29 GMT

You're right! I'll move the call out of the IRQ and will update in a bit. It sounds very likely that that's the issue.

Thanks for your speedy help! That's a speed record for me :D

RE: HardFault on sd_ble_gap_adv_stop()

Vidar Berg — Wed, 12 Jun 2019 13:07:33 GMT

Hi,

It appears to be called from your GPIOTE IRQ handler based on the PSR value. What's the interrupt priority of this IRQ? I'm wondering if it might be higher than the SV call?