Failure of the BMA4xx driver to compile

RE: Failure of the BMA4xx driver to compile

Kazi Afroza Sultana — Fri, 29 Aug 2025 12:30:45 GMT

Hello Christopher,

I understand it's frustrating. Sorry for the delayed reply.

BMA400 sensor has a register map and configuration that is significantly different from other BMA4xx family members such as the BMA422. The proper config for this driver shouild be CONFIG_BMA4XX. This config will not work for BMA422.

RE: Failure of the BMA4xx driver to compile

chrisstv — Thu, 28 Aug 2025 08:25:30 GMT

I think this case can be closed. Since the BMA4XX driver is not compatible with the BMA400 device.

Thanks

RE: Failure of the BMA4xx driver to compile

chrisstv — Fri, 22 Aug 2025 14:27:53 GMT

Hi Thanks, not exactly what I wanted to hear...

I agree with @Charlie-83, Please, Please add a Warning I have wasted many hours trying to get this driver to work with a BMA400!!!🙄☹

RE: Failure of the BMA4xx driver to compile

ivan iushkov — Fri, 22 Aug 2025 12:27:32 GMT

zephyr doesn't have BMA400 driver, see the issue:
https://github.com/zephyrproject-rtos/zephyr/issues/85462

RE: Failure of the BMA4xx driver to compile

Priyanka — Fri, 15 Aug 2025 10:57:11 GMT

Hi Chris,

Apologies for the delay and thank you for the update. I have escalated this internally and will surely get back to you as soon as possible.

-Priyanka

RE: Failure of the BMA4xx driver to compile

chrisstv — Thu, 14 Aug 2025 23:52:28 GMT

Further update.....

I have just installed the latest SDK and Toolchain (3.10) and the error still exists even though the BMA4xx driver has been rewritten. So I suspect the error may lie in the "sensor" code.

[00:00:00.252,288] <dbg> bma4xx: bma4xx_chip_init: chip_id is 0x90
[00:00:00.258,789] <wrn> bma4xx: Driver tested for BMA422. Check for unintended operation.
[00:00:00.283,508] <inf> bma4xx: FIFO ENABLED
[00:00:00.288,208] <dbg> bma4xx: bma4xx_configure: FIFO_CONFIG1 (0x49) 0x50
[00:00:00.297,607] <dbg> bma4xx: bma4xx_configure: FIFO_WTM_0( (0x46)) (WM Low) 0x0
[00:00:00.306,182] <dbg> bma4xx: bma4xx_configure: FIFO_WTM_1 (0x47) (WM High) 0x0
*** Booting nRF Connect SDK v3.1.0-6c6e5b32496e ***
*** Using Zephyr OS v4.1.99-1612683d4010 ***
Found bma400@14
[00:00:00.325,195] <err> os: ***** USAGE FAULT *****
[00:00:00.330,810] <err> os:   Illegal use of the EPSR
[00:00:00.336,608] <err> os: r0/a1:  0x0000ad3c  r1/a2:  0x00000040  r2/a3:  0x00000000
[00:00:00.345,245] <err> os: r3/a4:  0x00000000 r12/ip:  0x00002745 r14/lr:  0x000009d7
[00:00:00.353,851] <err> os:  xpsr:  0x00000000
[00:00:00.359,039] <err> os: Faulting instruction address (r15/pc): 0x00000000
[00:00:00.366,912] <err> os: >>> ZEPHYR FATAL ERROR 35: Unknown error on CPU 0
[00:00:00.374,755] <err> os: Current thread: 0x20000760 (unknown)
[00:00:00.381,469] <err> os: Halting system

The new driver has a bma4xx_def.h file which incorrectly assigns the wrong register addreses. Lines 79 and 80...

/* BMA4xx I2C register */
//#define BMA4XX_REG_I2C_SDO_HIGH (0x19)	//wrong
//#define BMA4XX_REG_I2C_SDO_LOW  (0x18)	//wrong
#define BMA4XX_REG_I2C_SDO_HIGH (0x15)	// Datasheet says this
#define BMA4XX_REG_I2C_SDO_LOW  (0x14)	// Datasheet says this

Chris

RE: Failure of the BMA4xx driver to compile

chrisstv — Thu, 14 Aug 2025 21:44:01 GMT

Hi Priyanka,

I opened a new ticket for the Temperature bug as requested.

Any news on the creashing error? This is really holding up our development at the moment!

After more digging I noticed a warning, which I cannot understand!! As far as I can tell everything is enabled. The device tree shows all present and enabled.

CONFIG_BMA4XX was assigned the value y, but got the value n. Missing dependencies: DT_HAS_BOSCH_BMA4XX_ENABLED

I have no idea if this is anything to do with the bug.

Chris

RE: Failure of the BMA4xx driver to compile

Priyanka — Thu, 31 Jul 2025 08:33:49 GMT

Hi Chris,

Kazi is away on vacation and I will be looking into the case for he time being. I have inquired internally and will get back to you as soon as I get a good response.

For the new bug that you have experienced, please begin a new ticket since it's different from the current one so that someone with related expertise can handle it.

This will also help keep Devzone clean for future seekers.

-Priyanka

RE: Failure of the BMA4xx driver to compile

chrisstv — Thu, 24 Jul 2025 19:11:41 GMT

Hi Kazi

Thanks,

Unfortunately I found another bug no related to the issue above!

When using "CONFIG_BMA4XX_TEMPERATURE=y" there is a compilation failure.

Line 621 of bma4xx.c:

		rc = bma4xx_get_shift(SENSOR_CHAN_DIE_TEMP, 0, &out->shift);

Unfortunately the first argument of the function tries to use "SENSOR_CHAN_DIE_TEMP" which is a "int16", but the first argument of the function is a "Two Word Struct" so results in a compilation failure.

Line 597 calls the function correctly:

		rc = bma4xx_get_shift((struct sensor_chan_spec){.chan_type = SENSOR_CHAN_ACCEL_XYZ,
								.chan_idx = 0},
				      header->accel_fs, &out->shift);

So I think line 621 function call needs to be changed to follow function call on Line 597 Eg:

		rc = bma4xx_get_shift((struct sensor_chan_spec){.chan_type = SENSOR_CHAN_DIE_TEMP,
								.chan_idx = 0},  0, &out->shift);

Regards

Chris

RE: Failure of the BMA4xx driver to compile

Kazi Afroza Sultana — Thu, 24 Jul 2025 14:25:27 GMT

Hello,

I have asked developer. I will get back to you later.

Thanks.

BR
Kazi

RE: Failure of the BMA4xx driver to compile

chrisstv — Fri, 18 Jul 2025 14:38:30 GMT

I've tried 2.8.0, 2.9.0, 3.0.0, and 3.0.2. All give the same result.

[00:00:00.249,420] <dbg> bma4xx: bma4xx_chip_init: chip_id is 0x90
[00:00:00.271,118] <wrn> bma4xx: Driver tested for BMA422. Check for unintended operation.
*** Booting nRF Connect SDK v3.0.2-89ba1294ac9b ***
*** Using Zephyr OS v4.0.99-f791c49f492c ***
Found bma400@14
[00:00:00.293,975] <err> os: ***** USAGE FAULT *****
[00:00:00.299,804] <err> os:   Illegal use of the EPSR
[00:00:00.305,816] <err> os: r0/a1:  0x00009b80  r1/a2:  0x0000003f  r2/a3:  0x00000000
[00:00:00.314,788] <err> os: r3/a4:  0x00000000 r12/ip:  0x000025e9 r14/lr:  0x00000693
[00:00:00.323,760] <err> os:  xpsr:  0x00000000
[00:00:00.329,162] <err> os: Faulting instruction address (r15/pc): 0x00000000
[00:00:00.337,310] <err> os: >>> ZEPHYR FATAL ERROR 35: Unknown error on CPU 0
[00:00:00.345,458] <err> os: Current thread: 0x200007c8 (unknown)
[00:00:00.352,447] <err> os: Halting system

Regards

Chris

RE: Failure of the BMA4xx driver to compile

Kazi Afroza Sultana — Thu, 17 Jul 2025 12:26:21 GMT

Hello,

Can you try with latest version of NCS (V3.0.2)?

RE: Failure of the BMA4xx driver to compile

chrisstv — Tue, 15 Jul 2025 18:42:00 GMT

Hi,

Update: I have now managed to compile with the BMA4xx driver, but have now run into issues using the sensor.

BOOT WITH BOTH SENSOR COMMANDS DISABLED

*** Booting nRF Connect SDK v2.7.0-5cb85570ca43 ***
*** Using Zephyr OS v3.6.99-100befc70c74 ***
Found bma400@14
Sensor Reporting!
AX: 256.000032; AY: 0.023725; AZ: -1.037648;[00:00:00.406,311] <dbg> bma4xx: bma4xx_chip_init: chip_id is 0x90
[00:00:00.413,146] <wrn> bma4xx: Driver tested for BMA422. Check for unintended operation.

BOOT WITH SENSOR_SAMPLE_FETCH ENABLED

*** Booting nRF Connect SDK v2.7.0-5cb85570ca43 ***
*** Using Zephyr OS v3.6.99-100befc70c74 ***
Found bma400@14
[00:00:00.435,974] <err> os: ***** USAGE FAULT *****
[00:00:00.441,802] <err> os:   Illegal use of the EPSR
[00:00:00.447,845] <err> os: r0/a1:  0x00009170  r1/a2:  0x0000003b  r2/a3:  0x00000000
[00:00:00.456,817] <err> os: r3/a4:  0x00000000 r12/ip:  0x00002335 r14/lr:  0x00000693
[00:00:00.465,820] <err> os:  xpsr:  0x00000000
[00:00:00.471,221] <err> os: Faulting instruction address (r15/pc): 0x00000000
[00:00:00.479,400] <err> os: >>> ZEPHYR FATAL ERROR 35: Unknown error on CPU 0
[00:00:00.487,548] <err> os: Current thread: 0x20000278 (unknown)
[00:00:00.494,567] <err> os: Halting system
[00:00:00.407,775] <dbg> bma4xx: bma4xx_chip_init: chip_id is 0x90
[00:00:00.414,611] <wrn> bma4xx: Driver tested for BMA422. Check for unintended operation.

BOOT WITH SENSOR_CHANNEL_GET ENABLED

*** Booting nRF Connect SDK v2.7.0-5cb85570ca43 ***
*** Using Zephyr OS v3.6.99-100befc70c74 ***
Found bma400@14
[00:00:00.437,408] <err> os: ***** USAGE FAULT *****
[00:00:00.443,237] <err> os:   Illegal use of the EPSR
[00:00:00.449,279] <err> os: r0/a1:  0x00009174  r1/a2:  0x00000003  r2/a3:  0x20001ea0
[00:00:00.458,282] <err> os: r3/a4:  0x00000000 r12/ip:  0x00002339 r14/lr:  0x00000695
[00:00:00.467,285] <err> os:  xpsr:  0x00000000
[00:00:00.472,686] <err> os: Faulting instruction address (r15/pc): 0x00000000
[00:00:00.480,865] <err> os: >>> ZEPHYR FATAL ERROR 35: Unknown error on CPU 0
[00:00:00.489,044] <err> os: Current thread: 0x20000278 (unknown)
[00:00:00.496,063] <err> os: Halting system
[00:00:00.410,308] <dbg> bma4xx: bma4xx_chip_init: chip_id is 0x90
[00:00:00.417,144] <wrn> bma4xx: Driver tested for BMA422. Check for unintended operation.

BOOT WITH BOTH SENSOR COMMANDS DISABLED

*** Booting nRF Connect SDK v2.7.0-5cb85570ca43 ***
*** Using Zephyr OS v3.6.99-100befc70c74 ***
Found bma400@14
Sensor Reporting!
AX: 256.000032; AY: 0.023725; AZ: -1.037648;

Above output is from 4 boots with different configurations of the sensor.... lines enabled. In all 4 cases the chip is identified ID:90, and Found bma400@14. which is correct.

So neither "sensor_sample_fetch" or "sensor_channel_get" work with the bma4xx driver, both get fatal errors.

I have created a sample test case, attached. I have removed the build directory to reduce size.

I was building on nRF52840dk-nRF52840 development board, nRF-connect 2.7.0 I have a BMA400 connected to the standard i2c0 pins, p0.26 and p0.27.

devzone.nordicsemi.com/.../5875.bma4xx_5F00_example.zip

RE: Failure of the BMA4xx driver to compile

chrisstv — Mon, 14 Jul 2025 17:10:52 GMT

Hi,

Yes: per my reply above I think it was the: CONFIG_SENSOR_ASYNC_API=y, line that was the main missing link.

After digging around in the driver files I found this in the: bma4xx/Kconfig file:

	depends on DT_HAS_BOSCH_BMA4XX_ENABLED
	depends on SENSOR_ASYNC_API

Thanks

Chris

RE: Failure of the BMA4xx driver to compile

Kazi Afroza Sultana — Mon, 14 Jul 2025 16:04:16 GMT

Hello,

This is common error '': undefined reference to `__device_dts_ord_105'' when driver is not enabled for the sensors in the prj.conf file.

Try adding CONFIG_BMA400=y and CONFIG_BMA400_I2C=y in the prj.conf file and run the sample again.

RE: Failure of the BMA4xx driver to compile

chrisstv — Mon, 14 Jul 2025 12:50:04 GMT

Hi Thanks,

I may have tracked down the issue after trawling through the build directory I discovered that there had been no attempt to compile the driver so the cause of the error was no "bma4xx.obj" file. The error only showed at link time...

It appear that two new CONFIG entries are required for this driver but not the previous Bosch drivers.

CONFIG_SENSOR_ASYNC_API=y
CONFIG_BMA4XX=y

As I'm at work I can only try this on the testcase I submitted which now compiles. I will try in the full code this evening.

This is not mentioned here, maybe it should be!

docs.nordicsemi.com/.../bosch_bma4xx-i2c.html

Hope this helps other users.

Chris