2021-04-09 07:18:13

by Jiapeng Chong

Subject: [PATCH] HID: hid-sensor-custom: remove useless variable

Fix the following gcc warning:

drivers/hid/hid-sensor-custom.c:400:7: warning: variable ‘ret’ set but
not used [-Wunused-but-set-variable].

Reported-by: Abaci Robot <[email protected]>
Signed-off-by: Jiapeng Chong <[email protected]>
---
drivers/hid/hid-sensor-custom.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/hid/hid-sensor-custom.c b/drivers/hid/hid-sensor-custom.c
index 2628bc5..e430673 100644
--- a/drivers/hid/hid-sensor-custom.c
+++ b/drivers/hid/hid-sensor-custom.c
@@ -397,15 +397,14 @@ static ssize_t store_value(struct device *dev, struct device_attribute *attr,

if (!strncmp(name, "value", strlen("value"))) {
u32 report_id;
- int ret;

if (kstrtoint(buf, 0, &value) != 0)
return -EINVAL;

report_id = sensor_inst->fields[field_index].attribute.
report_id;
- ret = sensor_hub_set_feature(sensor_inst->hsdev, report_id,
- index, sizeof(value), &value);
+ sensor_hub_set_feature(sensor_inst->hsdev, report_id, index,
+ sizeof(value), &value);
} else
return -EINVAL;

--
1.8.3.1

2021-04-09 18:20:32

by srinivas pandruvada

Subject: Re: [PATCH] HID: hid-sensor-custom: remove useless variable

On Fri, 2021-04-09 at 15:15 +0800, Jiapeng Chong wrote:
> Fix the following gcc warning:
>
> drivers/hid/hid-sensor-custom.c:400:7: warning: variable ‘ret’ set
> but
> not used [-Wunused-but-set-variable].
>
> Reported-by: Abaci Robot <[email protected]>
> Signed-off-by: Jiapeng Chong <[email protected]>
Acked-by: Srinivas Pandruvada <[email protected]

> ---
> drivers/hid/hid-sensor-custom.c | 5 ++---
> 1 file changed, 2 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/hid/hid-sensor-custom.c b/drivers/hid/hid-
> sensor-custom.c
> index 2628bc5..e430673 100644
> --- a/drivers/hid/hid-sensor-custom.c
> +++ b/drivers/hid/hid-sensor-custom.c
> @@ -397,15 +397,14 @@ static ssize_t store_value(struct device *dev,
> struct device_attribute *attr,
>
> if (!strncmp(name, "value", strlen("value"))) {
> u32 report_id;
> - int ret;
>
> if (kstrtoint(buf, 0, &value) != 0)
> return -EINVAL;
>
> report_id = sensor_inst->fields[field_index].attribute.
> report_
> id;
> - ret = sensor_hub_set_feature(sensor_inst->hsdev,
> report_id,
> - index, sizeof(value),
> &value);
> + sensor_hub_set_feature(sensor_inst->hsdev, report_id,
> index,
> + sizeof(value), &value);
> } else
> return -EINVAL;
>

2021-04-11 17:45:05

by srinivas pandruvada

Subject: Re: [PATCH] HID: hid-sensor-custom: remove useless variable

On Sun, 2021-04-11 at 14:56 +0100, Jonathan Cameron wrote:
> On Fri, 09 Apr 2021 11:19:12 -0700
> Srinivas Pandruvada <[email protected]> wrote:
>
> > On Fri, 2021-04-09 at 15:15 +0800, Jiapeng Chong wrote:
> > > Fix the following gcc warning:
> > >
> > > drivers/hid/hid-sensor-custom.c:400:7: warning: variable ‘ret’
> > > set
> > > but
> > > not used [-Wunused-but-set-variable].
> > >
> > > Reported-by: Abaci Robot <[email protected]>
> > > Signed-off-by: Jiapeng Chong <[email protected]>
> > Acked-by: Srinivas Pandruvada <[email protected]
>
> Perhaps better to return ret if it is non zero?
> I can't immediately figure out if there is a reason we know that
> can't
> happen.
Only time it can fail when there is no report descriptor or the field
index is >= report->maxfield.
But since the attribute is registered from the report descriptor and
index, this can't happen.
But we can enhance sensor_hub_set_feature() to fail when
hid_set_field() fails. There is one case where field->logical_minimum
< 0 and value is out of range.

Thanks,
Srinivas


>
> Jonathan
>
> > > ---
> > > drivers/hid/hid-sensor-custom.c | 5 ++---
> > > 1 file changed, 2 insertions(+), 3 deletions(-)
> > >
> > > diff --git a/drivers/hid/hid-sensor-custom.c b/drivers/hid/hid-
> > > sensor-custom.c
> > > index 2628bc5..e430673 100644
> > > --- a/drivers/hid/hid-sensor-custom.c
> > > +++ b/drivers/hid/hid-sensor-custom.c
> > > @@ -397,15 +397,14 @@ static ssize_t store_value(struct device
> > > *dev,
> > > struct device_attribute *attr,
> > >
> > > if (!strncmp(name, "value", strlen("value"))) {
> > > u32 report_id;
> > > - int ret;
> > >
> > > if (kstrtoint(buf, 0, &value) != 0)
> > > return -EINVAL;
> > >
> > > report_id = sensor_inst->fields[field_index].attribute.
> > > report_
> > > id;
> > > - ret = sensor_hub_set_feature(sensor_inst->hsdev,
> > > report_id,
> > > - index, sizeof(value),
> > > &value);
> > > + sensor_hub_set_feature(sensor_inst->hsdev, report_id,
> > > index,
> > > + sizeof(value), &value);
> > > } else
> > > return -EINVAL;
> > >

2021-04-11 18:52:42

by Jonathan Cameron

Subject: Re: [PATCH] HID: hid-sensor-custom: remove useless variable

On Fri, 09 Apr 2021 11:19:12 -0700
Srinivas Pandruvada <[email protected]> wrote:

> On Fri, 2021-04-09 at 15:15 +0800, Jiapeng Chong wrote:
> > Fix the following gcc warning:
> >
> > drivers/hid/hid-sensor-custom.c:400:7: warning: variable ‘ret’ set
> > but
> > not used [-Wunused-but-set-variable].
> >
> > Reported-by: Abaci Robot <[email protected]>
> > Signed-off-by: Jiapeng Chong <[email protected]>
> Acked-by: Srinivas Pandruvada <[email protected]

Perhaps better to return ret if it is non zero?
I can't immediately figure out if there is a reason we know that can't
happen.

Jonathan

>
> > ---
> > drivers/hid/hid-sensor-custom.c | 5 ++---
> > 1 file changed, 2 insertions(+), 3 deletions(-)
> >
> > diff --git a/drivers/hid/hid-sensor-custom.c b/drivers/hid/hid-
> > sensor-custom.c
> > index 2628bc5..e430673 100644
> > --- a/drivers/hid/hid-sensor-custom.c
> > +++ b/drivers/hid/hid-sensor-custom.c
> > @@ -397,15 +397,14 @@ static ssize_t store_value(struct device *dev,
> > struct device_attribute *attr,
> >
> > if (!strncmp(name, "value", strlen("value"))) {
> > u32 report_id;
> > - int ret;
> >
> > if (kstrtoint(buf, 0, &value) != 0)
> > return -EINVAL;
> >
> > report_id = sensor_inst->fields[field_index].attribute.
> > report_
> > id;
> > - ret = sensor_hub_set_feature(sensor_inst->hsdev,
> > report_id,
> > - index, sizeof(value),
> > &value);
> > + sensor_hub_set_feature(sensor_inst->hsdev, report_id,
> > index,
> > + sizeof(value), &value);
> > } else
> > return -EINVAL;
> >
>

2021-04-12 23:28:32

by Jonathan Cameron

Subject: Re: [PATCH] HID: hid-sensor-custom: remove useless variable

On Sun, 11 Apr 2021 09:06:35 -0700
Srinivas Pandruvada <[email protected]> wrote:

> On Sun, 2021-04-11 at 14:56 +0100, Jonathan Cameron wrote:
> > On Fri, 09 Apr 2021 11:19:12 -0700
> > Srinivas Pandruvada <[email protected]> wrote:
> >
> > > On Fri, 2021-04-09 at 15:15 +0800, Jiapeng Chong wrote:
> > > > Fix the following gcc warning:
> > > >
> > > > drivers/hid/hid-sensor-custom.c:400:7: warning: variable ‘ret’
> > > > set
> > > > but
> > > > not used [-Wunused-but-set-variable].
> > > >
> > > > Reported-by: Abaci Robot <[email protected]>
> > > > Signed-off-by: Jiapeng Chong <[email protected]>
> > > Acked-by: Srinivas Pandruvada <[email protected]
> >
> > Perhaps better to return ret if it is non zero?
> > I can't immediately figure out if there is a reason we know that
> > can't
> > happen.
> Only time it can fail when there is no report descriptor or the field
> index is >= report->maxfield.
> But since the attribute is registered from the report descriptor and
> index, this can't happen.
> But we can enhance sensor_hub_set_feature() to fail when
> hid_set_field() fails. There is one case where field->logical_minimum
> < 0 and value is out of range.

I'll go with what you think. Apply as is, or handle the
return value because we might at some later date return an error that
can actually happen from here?

Jonathan

>
> Thanks,
> Srinivas
>
>
> >
> > Jonathan
> >
> > > > ---
> > > > drivers/hid/hid-sensor-custom.c | 5 ++---
> > > > 1 file changed, 2 insertions(+), 3 deletions(-)
> > > >
> > > > diff --git a/drivers/hid/hid-sensor-custom.c b/drivers/hid/hid-
> > > > sensor-custom.c
> > > > index 2628bc5..e430673 100644
> > > > --- a/drivers/hid/hid-sensor-custom.c
> > > > +++ b/drivers/hid/hid-sensor-custom.c
> > > > @@ -397,15 +397,14 @@ static ssize_t store_value(struct device
> > > > *dev,
> > > > struct device_attribute *attr,
> > > >
> > > > if (!strncmp(name, "value", strlen("value"))) {
> > > > u32 report_id;
> > > > - int ret;
> > > >
> > > > if (kstrtoint(buf, 0, &value) != 0)
> > > > return -EINVAL;
> > > >
> > > > report_id = sensor_inst->fields[field_index].attribute.
> > > > report_
> > > > id;
> > > > - ret = sensor_hub_set_feature(sensor_inst->hsdev,
> > > > report_id,
> > > > - index, sizeof(value),
> > > > &value);
> > > > + sensor_hub_set_feature(sensor_inst->hsdev, report_id,
> > > > index,
> > > > + sizeof(value), &value);
> > > > } else
> > > > return -EINVAL;
> > > >
>

2021-04-13 11:38:25

by srinivas pandruvada

Subject: Re: [PATCH] HID: hid-sensor-custom: remove useless variable

On Mon, 2021-04-12 at 11:20 +0100, Jonathan Cameron wrote:
> On Sun, 11 Apr 2021 09:06:35 -0700
> Srinivas Pandruvada <[email protected]> wrote:
>
> > On Sun, 2021-04-11 at 14:56 +0100, Jonathan Cameron wrote:
> > > On Fri, 09 Apr 2021 11:19:12 -0700
> > > Srinivas Pandruvada <[email protected]> wrote:
> > >
> > > > On Fri, 2021-04-09 at 15:15 +0800, Jiapeng Chong wrote:
> > > > > Fix the following gcc warning:
> > > > >
> > > > > drivers/hid/hid-sensor-custom.c:400:7: warning: variable
> > > > > ‘ret’
> > > > > set
> > > > > but
> > > > > not used [-Wunused-but-set-variable].
> > > > >
> > > > > Reported-by: Abaci Robot <[email protected]>
> > > > > Signed-off-by: Jiapeng Chong <[email protected]
> > > > > >
> > > > Acked-by: Srinivas Pandruvada <
> > > > [email protected]
> > >
> > > Perhaps better to return ret if it is non zero?
> > > I can't immediately figure out if there is a reason we know that
> > > can't
> > > happen.
> > Only time it can fail when there is no report descriptor or the
> > field
> > index is >= report->maxfield.
> > But since the attribute is registered from the report descriptor
> > and
> > index, this can't happen.
> > But we can enhance sensor_hub_set_feature() to fail when
> > hid_set_field() fails. There is one case where field-
> > >logical_minimum
> > < 0 and value is out of range.
>
> I'll go with what you think. Apply as is, or handle the
> return value because we might at some later date return an error that
> can actually happen from here?
>
I will send you a patch this week to handle error condition.

Thanks,
Srinivas

> Jonathan
>
> > Thanks,
> > Srinivas
> >
> >
> > > Jonathan
> > >
> > > > > ---
> > > > > drivers/hid/hid-sensor-custom.c | 5 ++---
> > > > > 1 file changed, 2 insertions(+), 3 deletions(-)
> > > > >
> > > > > diff --git a/drivers/hid/hid-sensor-custom.c
> > > > > b/drivers/hid/hid-
> > > > > sensor-custom.c
> > > > > index 2628bc5..e430673 100644
> > > > > --- a/drivers/hid/hid-sensor-custom.c
> > > > > +++ b/drivers/hid/hid-sensor-custom.c
> > > > > @@ -397,15 +397,14 @@ static ssize_t store_value(struct
> > > > > device
> > > > > *dev,
> > > > > struct device_attribute *attr,
> > > > >
> > > > > if (!strncmp(name, "value", strlen("value"))) {
> > > > > u32 report_id;
> > > > > - int ret;
> > > > >
> > > > > if (kstrtoint(buf, 0, &value) != 0)
> > > > > return -EINVAL;
> > > > >
> > > > > report_id = sensor_inst-
> > > > > >fields[field_index].attribute.
> > > > >
> > > > > report_
> > > > > id;
> > > > > - ret = sensor_hub_set_feature(sensor_inst-
> > > > > >hsdev,
> > > > > report_id,
> > > > > - index,
> > > > > sizeof(value),
> > > > > &value);
> > > > > + sensor_hub_set_feature(sensor_inst->hsdev,
> > > > > report_id,
> > > > > index,
> > > > > + sizeof(value), &value);
> > > > > } else
> > > > > return -EINVAL;
> > > > >