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] [day] [month] [year] [list]
Message-ID: <30bad12b-13a3-46b1-b03f-877729fb01e5@roeck-us.net>
Date: Wed, 4 Feb 2026 14:39:41 -0800
From: Guenter Roeck <linux@...ck-us.net>
To: Kyle Hsieh <kylehsieh1995@...il.com>
Cc: linux-hwmon@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] hwmon: pmbus: tda38640: Add identification for
 PMBUS_VOUT_MODE

On 2/3/26 23:04, Kyle Hsieh wrote:
> Add support for detecting the PMBUS_VOUT_MODE for tda38640. Both
> linear and direct modes are now handled. Previously, only linear
> mode was assumed, which could trigger "Failed to identify chip
> capabilities" for chips in direct mode.
> 
> This change ensures proper identification of the chip regardless
> of its VOUT mode.
> 
> Signed-off-by: Kyle Hsieh <kylehsieh1995@...il.com>
> ---
> This patch adds proper detection of the PMBUS_VOUT_MODE for the
> tda38640 PMBus driver. Previously, the driver assumed linear mode
> only, which would trigger "Failed to identify chip capabilities"
> for chips operating in direct mode.
> 
> Changes:
> - Add support to identify both linear and direct VOUT modes.
> - Prevent misidentification and improve robustness of the driver.
> ---
>   drivers/hwmon/pmbus/tda38640.c | 26 +++++++++++++++++++++++++-
>   1 file changed, 25 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/hwmon/pmbus/tda38640.c b/drivers/hwmon/pmbus/tda38640.c
> index d902d39f49f4..410b7fc5aef5 100644
> --- a/drivers/hwmon/pmbus/tda38640.c
> +++ b/drivers/hwmon/pmbus/tda38640.c
> @@ -23,6 +23,30 @@ struct tda38640_data {
>   	u32 en_pin_lvl;
>   };
>   
> +static int tda38640_identify(struct i2c_client *client,
> +			     struct pmbus_driver_info *info)
> +{
> +	int vout_mode;
> +
> +	vout_mode = pmbus_read_byte_data(client, 0, PMBUS_VOUT_MODE);
> +	if (vout_mode < 0 || vout_mode == 0xff)

This is odd for a chip supposedly supporting the command. Why check
for 0xff ?

> +		return vout_mode < 0 ? vout_mode : -ENODEV;
> +	switch (vout_mode >> 5) {
> +	case 0: /* Linear */
> +		info->format[PSC_VOLTAGE_OUT] = linear;

This is already set, so there no need to update it.

> +		break;
> +	case 2: /* Direct */
> +		info->format[PSC_VOLTAGE_OUT] = direct;
> +		info->m[PSC_VOLTAGE_OUT] = 1;
> +		info->b[PSC_VOLTAGE_OUT] = 0;
> +		info->R[PSC_VOLTAGE_OUT] = 3;
> +		break;

The datasheet doesn't claim support for direct mode. Where is this documented,
and where do the coefficients come from ?

Also, I am concerned that this interferes with svid mode.
We'll have to ensure that this is not the case before proceeding.

Thanks,
Guenter

> +	default:
> +		return -ENODEV;
> +	}
> +	return 0;
> +}
> +
>   #define to_tda38640_data(x)  container_of(x, struct tda38640_data, info)
>   
>   /*
> @@ -142,7 +166,6 @@ static int svid_mode(struct i2c_client *client, struct tda38640_data *data)
>   static struct pmbus_driver_info tda38640_info = {
>   	.pages = 1,
>   	.format[PSC_VOLTAGE_IN] = linear,
> -	.format[PSC_VOLTAGE_OUT] = linear,
>   	.format[PSC_CURRENT_OUT] = linear,
>   	.format[PSC_CURRENT_IN] = linear,
>   	.format[PSC_POWER] = linear,
> @@ -153,6 +176,7 @@ static struct pmbus_driver_info tda38640_info = {
>   	    | PMBUS_HAVE_VOUT | PMBUS_HAVE_STATUS_VOUT
>   	    | PMBUS_HAVE_IOUT | PMBUS_HAVE_STATUS_IOUT
>   	    | PMBUS_HAVE_POUT | PMBUS_HAVE_PIN,
> +	.identify = tda38640_identify,
>   #if IS_ENABLED(CONFIG_SENSORS_TDA38640_REGULATOR)
>   	.num_regulators = 1,
>   	.reg_desc = tda38640_reg_desc,
> 
> ---
> base-commit: 5fd0a1df5d05ad066e5618ccdd3d0fa6cb686c27
> change-id: 20260204-tda38640_direct_mode-d6e47121a259
> 
> Best regards,


Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ