>From: Jonathan Cameron <jic23@xxxxxxxxxx>
>Sent: Saturday, June 3, 2017 10:47
>To: Jean-Baptiste Maneyrol
>Cc: linux-iio
>Subject: Re: [PATCH v2] iio: imu: inv_mpu6050: test whoami against invalid values and do it first
>
>On Mon, 29 May 2017 13:17:14 +0000
>Jean-Baptiste Maneyrol <JManeyrol@xxxxxxxxxxxxxx> wrote:
>
>> SPI bus is never generating error during transfer, so to check if
>> a chip is correctly connected on a SPI bus we enforce whoami check
>> to be correct.
>> In this way we can assure SPI probe is failing if there is no chip
>> connected.
>>
>> Signed-off-by: Jean-Baptiste Maneyrol <jmaneyrol@xxxxxxxxxxxxxx>
>Hi Jean-Baptiste,
>
>Sorry, I have just read back the previous emails and realised
>I wasn't being very clear. What I was suggesting was that we should
>check against all valid ids that are known, not just the one registered.
>So many of invensense's parts are 'nearly' compatible that it is
>common to get different revisions of boards using different versions.
>As such we are sure to break boards if we insist the code is right.
>(for starters I know there are DTSs out there for the beaglebone blue
>claiming it has a 9150 when it actually always has a 9250)
>
>So what I was looking for was a check over all valid IDs. Perhaps
>stick a warning in the log if the wrong one is found and fix up
>which chip info we are using.
>
>I'm afraid I thought this was what you meant when I agreed in
>the v1 thread, whereas it's clear you were thinking to enforce people
>getting it 'right!' (sadly something we know they don't always do).
>
>Jonathan
Hi Jonathan,
you're right, I misunderstood what you were meaning. Sorry about that.
Now it's crystal clear, thanks for the explanation.
I will do a patch v3 checking against all known ids and if there is a
mismatch, emit a warning and print the real chip name.
Thanks.
Jean-Baptiste
> ---
> drivers/iio/imu/inv_mpu6050/inv_mpu_core.c | 21 +++++++++++----------
> 1 file changed, 11 insertions(+), 10 deletions(-)
>
> diff --git a/drivers/iio/imu/inv_mpu6050/inv_mpu_core.c b/drivers/iio/imu/inv_mpu6050/inv_mpu_core.c
> index 96dabbd..57b3442 100644
> --- a/drivers/iio/imu/inv_mpu6050/inv_mpu_core.c
> +++ b/drivers/iio/imu/inv_mpu6050/inv_mpu_core.c
> @@ -774,23 +774,24 @@ static int inv_check_and_setup_chip(struct inv_mpu6050_state *st)
> st->hw = &hw_info[st->chip_type];
> st->reg = hw_info[st->chip_type].reg;
>
> - /* reset to make sure previous state are not there */
> - result = regmap_write(st->map, st->reg->pwr_mgmt_1,
> - INV_MPU6050_BIT_H_RESET);
> - if (result)
> - return result;
> - msleep(INV_MPU6050_POWER_UP_TIME);
> -
> /* check chip self-identification */
> result = regmap_read(st->map, INV_MPU6050_REG_WHOAMI, ®val);
> if (result)
> return result;
> if (regval != st->hw->whoami) {
> - dev_warn(regmap_get_device(st->map),
> - "whoami mismatch got %#02x expected %#02hhx for %s\n",
> - regval, st->hw->whoami, st->hw->name);
> + dev_err(regmap_get_device(st->map),
> + "invalid whoami got %#02x expected %#02hhx for %s\n",
> + regval, st->hw->whoami, st->hw->name);
> + return -ENODEV;
> }
>
> + /* reset to make sure previous state are not there */
> + result = regmap_write(st->map, st->reg->pwr_mgmt_1,
> + INV_MPU6050_BIT_H_RESET);
> + if (result)
> + return result;
> + msleep(INV_MPU6050_POWER_UP_TIME);
> +
> /*
> * toggle power state. After reset, the sleep bit could be on
> * or off depending on the OTP settings. Toggling power would
--
To unsubscribe from this list: send the line "unsubscribe linux-iio" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
