linux-kernel - Re: [PATCH 3/3] hwmon: scmi: Enhance error reporting with dev_err

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: <d48395d1-627e-48cb-8eae-391cab27a9d9@roeck-us.net>
Date: Tue, 5 Aug 2025 06:20:38 -0700
From: Guenter Roeck <linux@...ck-us.net>
To: "a.shimko" <artyom.shimko@...il.com>, linux-hwmon@...r.kernel.org
Cc: linux-kernel@...r.kernel.org, sudeep.holla@....com,
 cristian.marussi@....com, jdelvare@...e.com, guenter.roeck@...ux.com
Subject: Re: [PATCH 3/3] hwmon: scmi: Enhance error reporting with
 dev_err_probe

On 8/5/25 05:43, a.shimko wrote:
> From: Artem Shimko <artyom.shimko@...il.com>
> 
> Replace error returns with dev_err_probe() throughout driver:
> - Add descriptive error messages for all failure cases
> - Include relevant context (sensor IDs, types etc)
> - Standardize error reporting format
> 
> Improved messages include:
> - "No valid sensor info for index %d"
> - "Failed to allocate channel info array"
> - "SCMI protocol ops not initialized"
> 
> Signed-off-by: Artem Shimko <artyom.shimko@...il.com>
> ---
>   drivers/hwmon/scmi-hwmon.c | 35 ++++++++++++++++++++++++-----------
>   1 file changed, 24 insertions(+), 11 deletions(-)
> 
> diff --git a/drivers/hwmon/scmi-hwmon.c b/drivers/hwmon/scmi-hwmon.c
> index d03174922e65..081502418dfa 100644
> --- a/drivers/hwmon/scmi-hwmon.c
> +++ b/drivers/hwmon/scmi-hwmon.c
> @@ -240,26 +240,36 @@ static int scmi_hwmon_probe(struct scmi_device *sdev)
>   	struct scmi_protocol_handle *ph;
>   
>   	if (!handle)
> -		return -ENODEV;
> +		return dev_err_probe(dev, -ENODEV, "SCMI device handle is NULL\n");

-ENODEV is not supposed to log an error message.

>   
>   	sensor_ops = handle->devm_protocol_get(sdev, SCMI_PROTOCOL_SENSOR, &ph);
> -	if (IS_ERR(sensor_ops))
> -		return PTR_ERR(sensor_ops);
> +	if (IS_ERR_OR_NULL(sensor_ops)) {

This never returns NULL, and AFAIS no other driver checks for it.
Why add this unnecessary check ?

> +		if (IS_ERR(sensor_ops))
> +			return dev_err_probe(dev, PTR_ERR(sensor_ops),
> +					     "SCMI sensor protocol acquisition failed\n");
> +		return dev_err_probe(dev, -EPROTO,
> +				     "SCMI sensor protocol ops structure unexpectedly NULL\n");
> +	}
> +
> +	if (!sensor_ops->info_get || !sensor_ops->count_get)
> +		return dev_err_probe(dev, -ENOENT,
> +				     "SCMI sensor protocol operations are not initialized\n");

It is not this driver's responsibility to validate sensor_ops.

>   
>   	nr_sensors = sensor_ops->count_get(ph);
>   	if (!nr_sensors)
> -		return -EIO;
> +		return dev_err_probe(dev, -EIO, "No sensors found\n");
>   
>   	scmi_sensors = devm_kzalloc(dev, sizeof(*scmi_sensors), GFP_KERNEL);
>   	if (!scmi_sensors)
> -		return -ENOMEM;
> +		return dev_err_probe(dev, -ENOMEM, "Failed to allocate scmi_sensors structure\n");

-ENOMEM is not supposed to log an error message.

>   
>   	scmi_sensors->ph = ph;
>   
>   	for (i = 0; i < nr_sensors; i++) {
>   		sensor = sensor_ops->info_get(ph, i);
>   		if (!sensor)
> -			return -EINVAL;
> +			return dev_err_probe(dev, -EINVAL,
> +					     "Failed to get sensor info for sensor %d\n", i);
>   
>   		switch (sensor->type) {
>   		case TEMPERATURE_C:
> @@ -285,12 +295,12 @@ static int scmi_hwmon_probe(struct scmi_device *sdev)
>   	scmi_hwmon_chan = devm_kcalloc(dev, nr_types, sizeof(*scmi_hwmon_chan),
>   				       GFP_KERNEL);
>   	if (!scmi_hwmon_chan)
> -		return -ENOMEM;
> +		return dev_err_probe(dev, -ENOMEM, "Failed to allocate channel info array\n");

Same as above.

>   
>   	ptr_scmi_ci = devm_kcalloc(dev, nr_types + 1, sizeof(*ptr_scmi_ci),
>   				   GFP_KERNEL);
>   	if (!ptr_scmi_ci)
> -		return -ENOMEM;
> +		return dev_err_probe(dev, -ENOMEM, "Failed to allocate channel info pointers\n");

Same as above.

>   
>   	scmi_chip_info.info = ptr_scmi_ci;
>   	chip_info = &scmi_chip_info;
> @@ -307,7 +317,8 @@ static int scmi_hwmon_probe(struct scmi_device *sdev)
>   			devm_kcalloc(dev, nr_count[type],
>   				     sizeof(*scmi_sensors->info), GFP_KERNEL);
>   		if (!scmi_sensors->info[type])
> -			return -ENOMEM;
> +			return dev_err_probe(dev, -ENOMEM,
> +					     "Failed to allocate sensor info for type %d\n", type);

And again.

>   	}
>   
>   	for (i = nr_sensors - 1; i >= 0 ; i--) {
> @@ -336,7 +347,7 @@ static int scmi_hwmon_probe(struct scmi_device *sdev)
>   						     scmi_sensors, chip_info,
>   						     NULL);
>   	if (IS_ERR(hwdev))
> -		return PTR_ERR(hwdev);
> +		return dev_err_probe(dev, PTR_ERR(hwdev), "Failed to register hwmon device\n");
>   
>   	for (i = 0; i < nr_count_temp; i++) {
>   		int ret;
> @@ -352,7 +363,9 @@ static int scmi_hwmon_probe(struct scmi_device *sdev)
>   		ret = scmi_thermal_sensor_register(dev, ph, sensor);
>   		if (ret) {
>   			if (ret == -ENOMEM)
> -				return ret;
> +				return dev_err_probe(dev, ret,
> +						     "Failed to allocate memory for thermal zone\n");
> +
And again.

>   			dev_warn(dev,
>   				 "Thermal zone misconfigured for %s. err=%d\n",
>   				 sensor->name, ret);