BMI160 sample on custom nRF9160

RE: BMI160 sample on custom nRF9160

Hakon — Thu, 15 Apr 2021 13:33:05 GMT

[quote user="okwestern"]What would you think is the best route? To continue troubleshooting the driver I've enclosed above (which is a part of zephyr), or post the new driver and get help correcting the driver/sensor api link?[/quote]

I think you should keep debugging the original code and try figuring out why it triggers MPU fault. There are a couple of things you can try.

You should include the following lines in prj.conf

CONFIG_DEBUG=y
CONFIG_THREAD_NAME=y
CONFIG_EXTRA_EXCEPTION_INFO=y

for more detailed debug information.

You can also look up the faulting instruction address, by running the command below,

arm-none-eabi-addr2line 0x43204950 -e zephyr.elf

RE: BMI160 sample on custom nRF9160

okwestern — Tue, 13 Apr 2021 13:46:43 GMT

I completely re-wrote the SPI part of the bmi160 library to accept chip select based on the "lis2dh" driver, and now it completes the initiation and startup, and tries to read (fetch_sample), but it only gets static temp/acc/gyro data (doesn't change with movement). And same exact readings for three boards. I'm suspecting the link between my driver and the sensor API.

[00:00:00.402,191] [0m<inf> BMI160: Compiled: Apr  6 2021 05:40:23[0m
[00:00:00.402,191] [0m<inf> BMI160: master found: SPI_1[0m
[00:00:00.402,221] [0m<dbg> BMI160.bmi160_spi_init: SPI GPIO CS configured on GPIO_0:8[0m
[00:00:00.402,221] [1;31m<err> BMI160: gpio conf ok...[0m
[00:00:00.402,221] [0m<inf> BMI160: pins sat[0m
[00:00:00.502,349] [0m<dbg> BMI160.bmi160_init: reboot chip ok..[0m
[00:00:00.503,417] [0m<dbg> BMI160.bmi160_init: read from 0x7F ok..[0m
[00:00:00.503,570] [0m<dbg> BMI160.bmi160_init: read chip id ok..[0m
[00:00:00.503,570] [0m<dbg> BMI160.bmi160_init: chip id: d1[0m
[00:00:00.561,859] [0m<dbg> BMI160.bmi160_init: set power mode ok...[0m
[00:00:00.561,920] [0m<dbg> BMI160.bmi160_init: set default range for accelerometer ok...[0m
[00:00:00.561,981] [0m<dbg> BMI160.bmi160_init: set default range for gyroscope ok...[0m
[00:00:00.562,103] [0m<dbg> BMI160.bmi160_init: set accel's default ODR ok...[0m
[00:00:00.562,225] [0m<dbg> BMI160.bmi160_init: set gyro's default ODR ok...[0m
*** Booting Zephyr OS build zephyr-v2.5.0-352-gd96c517bab0d  ***

BMI160 driver/sample...

Gyro: Device found.

set accelerometer range ok.

set gyro range ok.

HW calibration ok.

Testing the polling mode.

Gyro (rad/s): X=2.596426, Y=0, Z=0

Acc (m/s^2): X=0, Y=0, Z=0

Temperature (Celsius): 33.038420

Gyro (rad/s): X=2.596426, Y=0, Z=0

Acc (m/s^2): X=0, Y=0, Z=0

Temperature (Celsius): 33.038420

Gyro (rad/s): X=2.596426, Y=0, Z=0

Acc (m/s^2): X=0, Y=0, Z=0

Temperature (Celsius): 33.038420

What would you think is the best route? To continue troubleshooting the driver I've enclosed above (which is a part of zephyr), or post the new driver and get help correcting the driver/sensor api link?

RE: BMI160 sample on custom nRF9160

okwestern — Tue, 13 Apr 2021 13:39:13 GMT

That didn't help...

[00:00:00.754,821] .[0m<inf> BMI160: SPI master controller found: SPI_1..[0m
[00:00:00.754,913] .[0m<inf> BMI160: reboot OK..[0m
[00:00:00.755,981] .[0m<inf> BMI160: Can read from 0x7F...[0m
[00:00:00.756,134] .[0m<inf> BMI160: Read chip OK..[0m
[00:00:00.856,262] .[0m<inf> BMI160: CHIP_ID OK (0xd1)!.[0m
[00:00:01.014,953] .[0m<inf> BMI160: set power mode ok..[0m
[00:00:01.015,014] .[0m<inf> BMI160: set default accelerometer range ok..[0m
[00:00:01.015,075] .[0m<inf> BMI160: set default gyroscope range ok..[0m
[00:00:01.015,228] .[0m<inf> BMI160: set accel's default ODR ok..[0m
[00:00:01.015,350] .[0m<inf> BMI160: set set gyro's default ODR ok..[0m
[00:00:01.015,441] .[1;31m<err> spi_nrfx_spim: Slave mode is not supported on SPI_1.[0m
[00:00:01.015,441] .[1;31m<err> os: ***** MPU FAULT *****.[0m
[00:00:01.015,441] .[1;31m<err> os:   Instruction Access Violation.[0m
[00:00:01.015,472] .[1;31m<err> os: r0/a1:  0x20000140  r1/a2:  0x20000518  r2/a3:  0x00000001.[0m
[00:00:01.015,472] .[1;31m<err> os: r3/a4:  0x43204950 r12/ip:  0x00000000 r14/lr:  0x00004ea9.[0m
[00:00:01.015,472] .[1;31m<err> os:  xpsr:  0x08000000.[0m
[00:00:01.015,472] .[1;31m<err> os: Faulting instruction address (r15/pc): 0x43204950.[0m
[00:00:01.015,472] .[1;31m<err> os: >>> ZEPHYR FATAL ERROR 0: CPU exception on CPU 0.[0m
[00:00:01.015,502] .[1;31m<err> os: Current thread: 0x200002b8 (unknown).[0m
[00:00:01.137,115] .[1;31m<err> os: Halting system.[0m

RE: BMI160 sample on custom nRF9160

Hakon — Tue, 13 Apr 2021 13:21:57 GMT

Hello,

can you try increasing the main stack size? You can set the stack size in prj.conf;

CONFIG_MAIN_STACK_SIZE=4096