linux-kernel - Re: [PATCH] iio: magnetometer: ak8975: Fix 'Unexpected device' error

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20231002102745.0000540b@Huawei.com>
Date:   Mon, 2 Oct 2023 10:27:45 +0100
From:   Jonathan Cameron <Jonathan.Cameron@...wei.com>
To:     André Apitzsch <git@...tzsch.eu>
CC:     Jonathan Cameron <jic23@...nel.org>,
        Lars-Peter Clausen <lars@...afoo.de>,
        Andy Shevchenko <andriy.shevchenko@...ux.intel.com>,
        Biju Das <biju.das.jz@...renesas.com>,
        <linux-iio@...r.kernel.org>, <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] iio: magnetometer: ak8975: Fix 'Unexpected device'
 error

On Sun, 1 Oct 2023 18:09:56 +0200
André Apitzsch <git@...tzsch.eu> wrote:

> Explicity specify array indices to fix mapping between
> asahi_compass_chipset and ak_def_array.
> While at it, remove unneeded AKXXXX.
> 
> Fixes: 4f9ea93afde1 ("iio: magnetometer: ak8975: Convert enum->pointer for data in the match tables")
> Signed-off-by: André Apitzsch <git@...tzsch.eu>
> ---
>  drivers/iio/magnetometer/ak8975.c | 11 +++++------
>  1 file changed, 5 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/iio/magnetometer/ak8975.c b/drivers/iio/magnetometer/ak8975.c
> index 8cfceb007936..dd466c5fa621 100644
> --- a/drivers/iio/magnetometer/ak8975.c
> +++ b/drivers/iio/magnetometer/ak8975.c
> @@ -204,7 +204,6 @@ static long ak09912_raw_to_gauss(u16 data)
>  
>  /* Compatible Asahi Kasei Compass parts */
>  enum asahi_compass_chipset {
> -	AKXXXX		= 0,

When we see this 'spacer' it is normally there to avoid a
confusion between 'not defined' and set too the first element.

So look at device_get_match_data() implementation and how we tell
if it worked.  That will return 0 if we have an AK8975 which is
then detected as a failure to match. 

https://elixir.bootlin.com/linux/v6.6-rc3/source/drivers/iio/magnetometer/ak8975.c#L932

So we need the spacer until someone converts this driver to use
pointers instead for both of and ACPI tables. I also don't like that
the of path here is falling back to the i2c_device_id match.

The data should be set in ak8975_of_match[] as well.

Jonathan

>  	AK8975,
>  	AK8963,
>  	AK09911,
> @@ -248,7 +247,7 @@ struct ak_def {
>  };
>  
>  static const struct ak_def ak_def_array[] = {
> -	{
> +	[AK8975] = {
>  		.type = AK8975,
>  		.raw_to_gauss = ak8975_raw_to_gauss,
>  		.range = 4096,
> @@ -273,7 +272,7 @@ static const struct ak_def ak_def_array[] = {
>  			AK8975_REG_HYL,
>  			AK8975_REG_HZL},
>  	},
> -	{
> +	[AK8963] = {
>  		.type = AK8963,
>  		.raw_to_gauss = ak8963_09911_raw_to_gauss,
>  		.range = 8190,
> @@ -298,7 +297,7 @@ static const struct ak_def ak_def_array[] = {
>  			AK8975_REG_HYL,
>  			AK8975_REG_HZL},
>  	},
> -	{
> +	[AK09911] = {
>  		.type = AK09911,
>  		.raw_to_gauss = ak8963_09911_raw_to_gauss,
>  		.range = 8192,
> @@ -323,7 +322,7 @@ static const struct ak_def ak_def_array[] = {
>  			AK09912_REG_HYL,
>  			AK09912_REG_HZL},
>  	},
> -	{
> +	[AK09912] = {
>  		.type = AK09912,
>  		.raw_to_gauss = ak09912_raw_to_gauss,
>  		.range = 32752,
> @@ -348,7 +347,7 @@ static const struct ak_def ak_def_array[] = {
>  			AK09912_REG_HYL,
>  			AK09912_REG_HZL},
>  	},
> -	{
> +	[AK09916] = {
>  		.type = AK09916,
>  		.raw_to_gauss = ak09912_raw_to_gauss,
>  		.range = 32752,
> 
> ---
> base-commit: df964ce9ef9fea10cf131bf6bad8658fde7956f6
> change-id: 20231001-ak_magnetometer-b063098082dd
> 
> Best regards,