On Thu, Oct 28, 2010 at 11:52:39AM -0400, Jean Delvare wrote:
> The fan control feature of the w83795 driver is insufficiently
> reviewed and tested for public consumption at this time, so make it
> optional and disabled by default. We will change the default when
> review and testing is deemed sufficient. Ultimately the option will
> go away.
>
> Signed-off-by: Jean Delvare <khali@xxxxxxxxxxxx>
> Cc: Guenter Roeck <guenter.roeck@xxxxxxxxxxxx>
Minor comment below. Otherwise
Acked-by: Guenter Roeck <guenter.roeck@xxxxxxxxxxxx>
> ---
> With this, the only non-standard attribute showing by default is "chassis". I'll
> fix it now and send a patch immediately.
>
> drivers/hwmon/Kconfig | 17 +++++++++++++++++
> drivers/hwmon/w83795.c | 10 ++++++++++
> 2 files changed, 27 insertions(+)
>
> --- linux-2.6.37-rc0.orig/drivers/hwmon/Kconfig 2010-10-28 16:26:10.000000000 +0200
> +++ linux-2.6.37-rc0/drivers/hwmon/Kconfig 2010-10-28 17:24:25.000000000 +0200
> @@ -1041,6 +1041,23 @@ config SENSORS_W83795
> This driver can also be built as a module. If so, the module
> will be called w83795.
>
> +config SENSORS_W83795_FANCTRL
> + boolean "Include fan control support (DANGEROUS)"
> + depends on SENSORS_W83795 && EXPERIMENTAL
> + default n
> + help
> + If you say yes here, support for the both manual and automatic
> + fan control features will be included in the driver.
> +
> + This part of the code wasn't carefully reviewed and tested yet,
> + so enabling this option is strongly discouraged on production
> + servers. Only developers and testers should enable it for the
> + time being.
> +
> + Please also note that this option will create sysfs attribute
> + files which may change in the future, so you shouldn't rely
> + on them being stable.
> +
> config SENSORS_W83L785TS
> tristate "Winbond W83L785TS-S"
> depends on I2C && EXPERIMENTAL
> --- linux-2.6.37-rc0.orig/drivers/hwmon/w83795.c 2010-10-28 16:41:32.000000000 +0200
> +++ linux-2.6.37-rc0/drivers/hwmon/w83795.c 2010-10-28 17:39:12.000000000 +0200
> @@ -1455,6 +1455,7 @@ store_in(struct device *dev, struct devi
> }
>
>
> +#ifdef CONFIG_SENSORS_W83795_FANCTRL
> static ssize_t
> show_sf_setup(struct device *dev, struct device_attribute *attr, char *buf)
> {
> @@ -1506,6 +1507,7 @@ store_sf_setup(struct device *dev, struc
> mutex_unlock(&data->update_lock);
> return count;
> }
> +#endif
>
>
> #define NOT_USED -1
> @@ -1711,6 +1713,7 @@ static const struct sensor_device_attrib
> store_chassis_clear, ALARM_STATUS, 46),
> SENSOR_ATTR_2(beep_enable, S_IWUSR | S_IRUGO, show_alarm_beep,
> store_beep, BEEP_ENABLE, 47),
> +#ifdef CONFIG_SENSORS_W83795_FANCTRL
> SENSOR_ATTR_2(speed_cruise_tolerance, S_IWUSR | S_IRUGO, show_fanin,
> store_fanin, FANIN_TOL, NOT_USED),
> SENSOR_ATTR_2(pwm_default, S_IWUSR | S_IRUGO, show_sf_setup,
> @@ -1719,6 +1722,7 @@ static const struct sensor_device_attrib
> store_sf_setup, SETUP_PWM_UPTIME, NOT_USED),
> SENSOR_ATTR_2(pwm_downtime, S_IWUSR | S_IRUGO, show_sf_setup,
> store_sf_setup, SETUP_PWM_DOWNTIME, NOT_USED),
> +#endif
> };
>
> /*
> @@ -1872,6 +1876,7 @@ static int w83795_handle_files(struct de
> return err;
> }
>
> +#ifdef CONFIG_SENSORS_W83795_FANCTRL
> for (i = 0; i < data->has_pwm; i++) {
> for (j = 0; j < ARRAY_SIZE(w83795_pwm[0]); j++) {
> err = fn(dev, &w83795_pwm[i][j].dev_attr);
> @@ -1879,11 +1884,16 @@ static int w83795_handle_files(struct de
> return err;
> }
> }
> +#endif
>
> for (i = 0; i < ARRAY_SIZE(w83795_temp); i++) {
> if (!(data->has_temp & (1 << i)))
> continue;
> +#ifdef CONFIG_SENSORS_W83795_FANCTRL
> for (j = 0; j < ARRAY_SIZE(w83795_temp[0]); j++) {
> +#else
> + for (j = 0; j < 8; j++) {
> +#endif
Can you keep using ARRAY_SIZE here and if necessary make array elements conditional ?
> err = fn(dev, &w83795_temp[i][j].dev_attr);
> if (err)
> return err;
>
>
> --
> Jean Delvare
_______________________________________________
lm-sensors mailing list
lm-sensors@xxxxxxxxxxxxxx
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
