[PATCH] hwmon: (pmbus) Improve initialization of 'currpage' and 'currphase'

[PATCH] hwmon: (pmbus) Improve initialization of 'currpage' and 'currphase'

[Guenter Roeck]

The 'currpage' and 'currphase' variables in struct pmbus_data are used by
the PMBus core to determine if the phase or page value has changed. Both
are initialized with values which are never expected to be set in the code
to ensure that the first page/phase write operation is actually performed.

This is not well explained and occasionally causes confusion. Change the
type of both variables to s16 and initialize with -1 to ensure that the
initial value never matches a requested value, and clarify that this
value means "unknown/unset".

Cc: Alex Qiu <xqiu@google.com>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
---
 drivers/hwmon/pmbus/pmbus_core.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/hwmon/pmbus/pmbus_core.c b/drivers/hwmon/pmbus/pmbus_core.c
index 8d321bf7d15b..a420877ba533 100644
--- a/drivers/hwmon/pmbus/pmbus_core.c
+++ b/drivers/hwmon/pmbus/pmbus_core.c
@@ -109,8 +109,8 @@ struct pmbus_data {
 	bool has_status_word;		/* device uses STATUS_WORD register */
 	int (*read_status)(struct i2c_client *client, int page);
 
-	u8 currpage;
-	u8 currphase;	/* current phase, 0xff for all */
+	s16 currpage;	/* current page, -1 for unknown/unset */
+	s16 currphase;	/* current phase, 0xff for all, -1 for unknown/unset */
 };
 
 struct pmbus_debugfs_entry {
@@ -2529,8 +2529,8 @@ int pmbus_do_probe(struct i2c_client *client, const struct i2c_device_id *id,
 	if (pdata)
 		data->flags = pdata->flags;
 	data->info = info;
-	data->currpage = 0xff;
-	data->currphase = 0xfe;
+	data->currpage = -1;
+	data->currphase = -1;
 
 	ret = pmbus_init_common(client, data, info);
 	if (ret < 0)
-- 
2.17.1

Re: [PATCH] hwmon: (pmbus) Improve initialization of 'currpage' and 'currphase'

[Alex Qiu]

Hi Guenter,

LGTM; minor nits:

On Thu, May 7, 2020 at 10:44 AM Guenter Roeck <linux@roeck-us.net> wrote:
>
> The 'currpage' and 'currphase' variables in struct pmbus_data are used by
> the PMBus core to determine if the phase or page value has changed. Both
> are initialized with values which are never expected to be set in the code
> to ensure that the first page/phase write operation is actually performed.
>
> This is not well explained and occasionally causes confusion. Change the
> type of both variables to s16 and initialize with -1 to ensure that the
> initial value never matches a requested value, and clarify that this
> value means "unknown/unset".
>
> Cc: Alex Qiu <xqiu@google.com>
> Signed-off-by: Guenter Roeck <linux@roeck-us.net>
> ---
>  drivers/hwmon/pmbus/pmbus_core.c | 8 ++++----
>  1 file changed, 4 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/hwmon/pmbus/pmbus_core.c b/drivers/hwmon/pmbus/pmbus_core.c
> index 8d321bf7d15b..a420877ba533 100644
> --- a/drivers/hwmon/pmbus/pmbus_core.c
> +++ b/drivers/hwmon/pmbus/pmbus_core.c
> @@ -109,8 +109,8 @@ struct pmbus_data {
>         bool has_status_word;           /* device uses STATUS_WORD register */
>         int (*read_status)(struct i2c_client *client, int page);
>
> -       u8 currpage;
> -       u8 currphase;   /* current phase, 0xff for all */
> +       s16 currpage;   /* current page, -1 for unknown/unset */
> +       s16 currphase;  /* current phase, 0xff for all, -1 for unknown/unset */
I'm not sure if assigning macros for -1 and 0xff would be a good idea?
>  };
>
>  struct pmbus_debugfs_entry {
> @@ -2529,8 +2529,8 @@ int pmbus_do_probe(struct i2c_client *client, const struct i2c_device_id *id,
>         if (pdata)
>                 data->flags = pdata->flags;
>         data->info = info;
> -       data->currpage = 0xff;
> -       data->currphase = 0xfe;
> +       data->currpage = -1;
> +       data->currphase = -1;
>
>         ret = pmbus_init_common(client, data, info);
>         if (ret < 0)
> --
> 2.17.1
>

Thx!

- Alex Qiu

Re: [PATCH] hwmon: (pmbus) Improve initialization of 'currpage' and 'currphase'

[Guenter Roeck]

On Thu, May 07, 2020 at 01:19:58PM -0700, Alex Qiu wrote:
> Hi Guenter,
> 
> LGTM; minor nits:
> 
> On Thu, May 7, 2020 at 10:44 AM Guenter Roeck <linux@roeck-us.net> wrote:
> >
> > The 'currpage' and 'currphase' variables in struct pmbus_data are used by
> > the PMBus core to determine if the phase or page value has changed. Both
> > are initialized with values which are never expected to be set in the code
> > to ensure that the first page/phase write operation is actually performed.
> >
> > This is not well explained and occasionally causes confusion. Change the
> > type of both variables to s16 and initialize with -1 to ensure that the
> > initial value never matches a requested value, and clarify that this
> > value means "unknown/unset".
> >
> > Cc: Alex Qiu <xqiu@google.com>
> > Signed-off-by: Guenter Roeck <linux@roeck-us.net>
> > ---
> >  drivers/hwmon/pmbus/pmbus_core.c | 8 ++++----
> >  1 file changed, 4 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/hwmon/pmbus/pmbus_core.c b/drivers/hwmon/pmbus/pmbus_core.c
> > index 8d321bf7d15b..a420877ba533 100644
> > --- a/drivers/hwmon/pmbus/pmbus_core.c
> > +++ b/drivers/hwmon/pmbus/pmbus_core.c
> > @@ -109,8 +109,8 @@ struct pmbus_data {
> >         bool has_status_word;           /* device uses STATUS_WORD register */
> >         int (*read_status)(struct i2c_client *client, int page);
> >
> > -       u8 currpage;
> > -       u8 currphase;   /* current phase, 0xff for all */
> > +       s16 currpage;   /* current page, -1 for unknown/unset */
> > +       s16 currphase;  /* current phase, 0xff for all, -1 for unknown/unset */
> I'm not sure if assigning macros for -1 and 0xff would be a good idea?

The driver uses hardcoded 0xff in various places. We could replace it
with defines, but that should be a separate patch.

Thanks,
Guenter