lis3mdl_init_interrupt triggers __ASSERT in lis3mdl_sample_fetch

[TilmannUnte]

Describe the bug
I'm using the LIS3MDL magnetometer connected to I2C with the optional trigger attached to a GPIO pin on my MCU.

With CONFIG_LIS3MDL_TRIGGER enabled, lis3mdl_init calls lis3mdl_init_interrupt.
This function contains a call to lis3mdl_sample_fetch:
if (lis3mdl_sample_fetch(dev, SENSOR_CHAN_ALL) < 0) {

However, lis3mdl_sample_fetch contains the following assertion:
__ASSERT_NO_MSG(chan == SENSOR_CHAN_ALL);

Therefore, the initialization always results in a fatal error, when CONFIG_LIS3MDL_TRIGGER and CONFIG_ASSERT are both set.

I consider this to be a bug, because everything works fine when CONFIG_ASSERT is not set!
Unless I have missed an important detail, I think it would be fine to remove the assertion to fix this bug.

To Reproduce

• suggested board: disco_l475_iot1
• clone my test program: https://github.com/TilmannUnte/zephyr_lis3mdl_test
• build, flash and start serial console
• press blue user button on board and wait a few seconds for fatal error to appear
• deactivate CONFIG_ASSERT in prj.conf and retry, the error should be gone

Expected behavior
The LIS3MDL driver should not trip a (pointless?) assertion when setting up the interrupt line.

Impact
With this bug, The CONFIG_LIS3MDL_TRIGGER and CONFIG_ASSERT settings are mutually exclusive.
This is very annoying in complex applications, because assertions have safety critical roles and I cannot just deactivate them.

Logs and console output
These are from my production code, not the test program I linked to, but the error is equivalent:

ASSERTION FAIL [chan == SENSOR_CHAN_ALL] @ WEST_TOPDIR/zephyr/drivers/sensor/st/lis3mdl/lis3mdl.c:69
[00:00:02.348,000] <err> os: r0/a1:  0x00000004  r1/a2:  0x00000045  r2/a3:  0x00000004
[00:00:02.348,000] <err> os: r3/a4:  0x00000004 r12/ip:  0x00005bbd r14/lr:  0x0800981f
[00:00:02.348,000] <err> os:  xpsr:  0x01000000
[00:00:02.348,000] <err> os: s[ 0]:  0x00000000  s[ 1]:  0x08013df5  s[ 2]:  0x00000000  s[ 3]:  0x08002c41
[00:00:02.348,000] <err> os: s[ 4]:  0x00000000  s[ 5]:  0x08002cc7  s[ 6]:  0x08002c41  s[ 7]:  0x00000000
[00:00:02.348,000] <err> os: s[ 8]:  0x00000000  s[ 9]:  0x08002c41  s[10]:  0x00000000  s[11]:  0x0801629b
[00:00:02.348,000] <err> os: s[12]:  0x00000000  s[13]:  0xdafb43d1  s[14]:  0x52ce17f1  s[15]:  0xf7af9383
[00:00:02.348,000] <err> os: fpscr:  0x148621a2
[00:00:02.348,000] <err> os: Faulting instruction address (r15/pc): 0x08016724
[00:00:02.348,000] <err> os: >>> ZEPHYR FATAL ERROR 4: Kernel panic on CPU 0
[00:00:02.348,000] <err> os: Current thread: 0x20002d80 (lis3mdl_tcb)
[00:00:02.373,000] <err> os: Halting system

(gdb) break lis3mdl.c:69
Breakpoint 1 at 0x80097c4: file /home/untetilm/zephyrproject/zephyr/drivers/sensor/st/lis3mdl/lis3mdl.c, line 69.
Note: automatically using hardware breakpoints for read-only addresses.
(gdb) continue
Continuing.

Breakpoint 1, lis3mdl_sample_fetch (dev=dev@entry=0x801bae0 <__device_dts_ord_107>, chan=chan@entry=SENSOR_CHAN_ALL) at /home/untetilm/zephyrproject/zephyr/drivers/sensor/st/lis3mdl/lis3mdl.c:69
69		__ASSERT_NO_MSG(chan == SENSOR_CHAN_ALL);
(gdb) info args
dev = 0x801bae0 <__device_dts_ord_107>
chan = SENSOR_CHAN_ALL
(gdb) bt
#0  lis3mdl_sample_fetch (dev=dev@entry=0x801bae0 <__device_dts_ord_107>, chan=chan@entry=SENSOR_CHAN_ALL) at /home/untetilm/zephyrproject/zephyr/drivers/sensor/st/lis3mdl/lis3mdl.c:69
#1  0x08009e7c in lis3mdl_init_interrupt (dev=dev@entry=0x801bae0 <__device_dts_ord_107>) at /home/untetilm/zephyrproject/zephyr/drivers/sensor/st/lis3mdl/lis3mdl_trigger.c:136
#2  0x080099ec in lis3mdl_init (dev=0x801bae0 <__device_dts_ord_107>) at /home/untetilm/zephyrproject/zephyr/drivers/sensor/st/lis3mdl/lis3mdl.c:155
[rest omitted]

Environment:

• OS: Ubuntu Linux 22.04
• Zephyr SDK 0.17
• Zephyr commit: 90d8710