drivers: sensors: bmi08x: add initial support for bmi08x

[XenuIsWatching]

This adds support for the bosch bmi085 and bmi088. This also includes support for data sync mode. The binary blob in bmi08x_config_file.h is under the BSD-3 clause which comes from the github repo https://github.com/BoschSensortec/BMI08x-Sensor-API/blob/master/LICENSE

Signed-off-by: Ryan McClelland ryanmcclelland@meta.com

[carlescufi]

I want to note that the binary blob for configuration is something we already allow in similar drivers:

• https://github.com/zephyrproject-rtos/zephyr/blob/main/drivers/sensor/bmi270/bmi270_config_file.h

My recommendation is to document that this is allowed.

[MaureenHelm]

The source copyright and license should be preserved (Bosch, BSD-3-Clause)

[XenuIsWatching]

fixed.... but it appears to of broken one of the CI checks

[Cherish-Gww]

I didn't find it in the datasheet of bmi088 and bmi085 BMI08X_REG_ACCEL_INT1_MAP: 0x56 and BMI08X_REG_ACCEL_INT2_MAP: 0x57. the datasheet said that 0x55-0x57 is reserved.

[XenuIsWatching]

it is there.... Just about every register that says "reserved" in that data actually does have a purpose (just not a very important one...) if you look in this Bosch public BMI08x bare metal driver.. you see those registers actually defined in there. I could change it to the more easily found register in the data sheet... but it's really the same thing
https://github.com/BoschSensortec/BMI08x-Sensor-API

[Cherish-Gww]

I really do not like "reserved" register, I can`t find those "reserved" register describe.
https://github.com/boschsensortec/BMI08x-Sensor-API/blob/631a966891885643f180e18e94870411b8f4c3fc/bmi08a.c#L2198

[XenuIsWatching]

it's here:
https://github.com/boschsensortec/BMI08x-Sensor-API/blob/631a966891885643f180e18e94870411b8f4c3fc/bmi08a.c#L2225

[Cherish-Gww]

I think this is not data ready interrupt in set_accel_data_ready_int
by writing BMI08X_REG_ACCEL_INT1_MAP and BMI08X_REG_ACCEL_INT2_MAP register.

I did a test, and the synchronization interrupt could not be generated data ready interrupt, When I change write BMI08X_REG_ACCEL_INT1_INT2_MAP_DATA register, The interrupt have active.

in datasheet:
To increase flexibility, both gyroscope and accelerometer can not only be operated individually, but tied
together for data synchronization purposes. The on-chip features comprise FIFOs for acceleration and
gyroscope data and interrupt controllers.

[XenuIsWatching]

I see, I assume you're not using data sync?... I'll replace it with this

#if BMI08X_ACCEL_ANY_INST_HAS_DATA_SYNC
	if (config->data_sync != 0) {
		/* set accel ints */
		if (bmi08x_accel_byte_write(dev, BMI08X_REG_ACCEL_INT1_MAP, cfg->int1_map) < 0) {
			LOG_ERR("Failed to map interrupts.");
			return -EIO;
		}
		if (bmi08x_accel_byte_write(dev, BMI08X_REG_ACCEL_INT2_MAP, cfg->int2_map) < 0) {
			LOG_ERR("Failed to map interrupts.");
			return -EIO;
		}
	} else
#endif
	{
		uint8_t map_data = ((cfg->int2_map << BMI08X_ACCEL_INT2_DRDY_POS) |
			(cfg->int1_map << BMI08X_ACCEL_INT1_DRDY_POS));
		if (bmi08x_accel_byte_write(dev, BMI08X_REG_ACCEL_INT1_INT2_MAP_DATA, map_data) < 0) {
			LOG_ERR("Failed to map interrupts.");
			return -EIO;
		}
	}

[Cherish-Gww]

I think it would be better to set the sensor synchronization function in the attribute. What do you think?
Macros can only be selected from one of two options, and using attribute functions can be set at any time.

[XenuIsWatching]

Sorry, ...but I can't tell exactly what you're saying here??

Data sync is done in the DTS so it is instanced. You can have multiple BMI08X connected and one can have data sync on while the other doesn't. With a KConfig, it's either ALL or NONE.

[Cherish-Gww]

BMI08X_ACCEL_ANY_INST_HAS_DATA_SYNC

bmi08x_attr_set

[XenuIsWatching]

I emailed a bosch FAE about this a while ago... data sync can't easily be set at runtime (unless if you write complex code to reset it and reinitialize it...) It also must be set at initialization

I see that the config file is acceptable under 3-clause BSD

[carlescufi]

@XenuIsWatching please rebase

[XenuIsWatching]

@XenuIsWatching please rebase

rebased, but it blew away the reviews

[bperseghetti]

@MaureenHelm @carlescufi sorry to bug you all again on this, but are you able to re-approve and merge now that it's rebased?

[bperseghetti]

@XenuIsWatching thanks for the fixes! Can't wait wait to get this merged in!

[carlescufi]

refreshing my +1

[bbilas]

One more ;)

[XenuIsWatching]

actually 3 more... I took a second look through it all

[carlescufi]

Refresh +1

[bperseghetti]

Thanks @bbilas! Are you able to merge this in now @carlescufi?

[carlescufi]

Thanks @bbilas! Are you able to merge this in now @carlescufi?

@MaureenHelm is back, so I delegate to her now.