On Thu, 16 Jul 2020 14:59:10 +0100
Lee Jones <lee.jo...@linaro.org> wrote:

> Kerneldoc expects attributes/parameters to be in '@*.: ' format and
> gets confused if the variable does not follow the type/attribute
> definitions.
> 
> Fixes the following W=1 kernel build warning(s):
> 
>  drivers/iio/gyro/adis16080.c:49: warning: Function parameter or member 
> 'lock' not described in 'adis16080_state'
>  drivers/iio/gyro/adis16080.c:49: warning: Function parameter or member 
> '____cacheline_aligned' not described in 'adis16080_state'
> 
> Cc: Michael Hennerich <michael.henner...@analog.com>
> Cc: Barry Song <21cn...@gmail.com>
> Signed-off-by: Lee Jones <lee.jo...@linaro.org>
Hmm. You are going to have an awful lot of those ____cacheline_aligned ones.

Might be worth thinking about whether we fix kernel-doc to cope with those.

A quick grep suggests the vast majority of users of this have it after
the element name.

@Jon Corbet :  What do you think?  Looks like there is special
handling already for ____cacheline_aligned_in_smp. Should we extend
that to this case?

Jonathan



> ---
>  drivers/iio/gyro/adis16080.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/iio/gyro/adis16080.c b/drivers/iio/gyro/adis16080.c
> index 1b84b8e112fe1..f38f9abcccbb5 100644
> --- a/drivers/iio/gyro/adis16080.c
> +++ b/drivers/iio/gyro/adis16080.c
> @@ -38,14 +38,14 @@ struct adis16080_chip_info {
>   * @us:                      actual spi_device to write data
>   * @info:            chip specific parameters
>   * @buf:             transmit or receive buffer
> - * @lock             lock to protect buffer during reads
> + * @lock:            lock to protect buffer during reads
>   **/
>  struct adis16080_state {
>       struct spi_device               *us;
>       const struct adis16080_chip_info *info;
>       struct mutex                    lock;
>  
> -     __be16 buf ____cacheline_aligned;
> +     __be16 ____cacheline_aligned buf;
>  };
>  
>  static int adis16080_read_sample(struct iio_dev *indio_dev,

Reply via email to