`strncpy` is deprecated for use on NUL-terminated destination strings [1].
We should prefer more robust and less ambiguous string interfaces.
`prop` is defined as this string literal with size 30 (including null):
| #define SX9324_PROXRAW_DEF "semtech,ph01-proxraw-strength"
| char prop[] = SX9324_PROXRAW_DEF;
Each of the strncpy->strscpy replacements involve string literals with a
size less than 30 which means there are no current problems with how
strncpy is used. However, let's move away from using strncpy entirely.
A suitable replacement is `strscpy` [2] due to the fact that it
guarantees NUL-termination on the destination buffer without
unnecessarily NUL-padding.
Moreover, let's opt for the more conventional `sizeof()` as opposed to
`ARRAY_SIZE` for these simple strings.
Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings [1]
Link: https://manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html [2]
Link: https://github.com/KSPP/linux/issues/90
Cc: [email protected]
Signed-off-by: Justin Stitt <[email protected]>
---
FWIW: It seems fragile to base future `prop` stores on the
size of it's default string.
Note: build-tested
---
drivers/iio/proximity/sx9324.c | 12 ++++--------
1 file changed, 4 insertions(+), 8 deletions(-)
diff --git a/drivers/iio/proximity/sx9324.c b/drivers/iio/proximity/sx9324.c
index 438f9c9aba6e..25ac2733bcef 100644
--- a/drivers/iio/proximity/sx9324.c
+++ b/drivers/iio/proximity/sx9324.c
@@ -937,11 +937,9 @@ sx9324_get_default_reg(struct device *dev, int idx,
case SX9324_REG_AFE_CTRL4:
case SX9324_REG_AFE_CTRL7:
if (reg_def->reg == SX9324_REG_AFE_CTRL4)
- strncpy(prop, "semtech,ph01-resolution",
- ARRAY_SIZE(prop));
+ strscpy(prop, "semtech,ph01-resolution", sizeof(prop));
else
- strncpy(prop, "semtech,ph23-resolution",
- ARRAY_SIZE(prop));
+ strscpy(prop, "semtech,ph23-resolution", sizeof(prop));
ret = device_property_read_u32(dev, prop, &raw);
if (ret)
@@ -1012,11 +1010,9 @@ sx9324_get_default_reg(struct device *dev, int idx,
case SX9324_REG_PROX_CTRL0:
case SX9324_REG_PROX_CTRL1:
if (reg_def->reg == SX9324_REG_PROX_CTRL0)
- strncpy(prop, "semtech,ph01-proxraw-strength",
- ARRAY_SIZE(prop));
+ strscpy(prop, "semtech,ph01-proxraw-strength", sizeof(prop));
else
- strncpy(prop, "semtech,ph23-proxraw-strength",
- ARRAY_SIZE(prop));
+ strscpy(prop, "semtech,ph23-proxraw-strength", sizeof(prop));
ret = device_property_read_u32(dev, prop, &raw);
if (ret)
break;
---
base-commit: 2cf0f715623872823a72e451243bbf555d10d032
change-id: 20230921-strncpy-drivers-iio-proximity-sx9324-c-8c3437676039
Best regards,
--
Justin Stitt <[email protected]>
On Thu, 21 Sep 2023 07:01:01 +0000
Justin Stitt <[email protected]> wrote:
> `strncpy` is deprecated for use on NUL-terminated destination strings [1].
>
> We should prefer more robust and less ambiguous string interfaces.
>
> `prop` is defined as this string literal with size 30 (including null):
> | #define SX9324_PROXRAW_DEF "semtech,ph01-proxraw-strength"
> | char prop[] = SX9324_PROXRAW_DEF;
>
> Each of the strncpy->strscpy replacements involve string literals with a
> size less than 30 which means there are no current problems with how
> strncpy is used. However, let's move away from using strncpy entirely.
>
> A suitable replacement is `strscpy` [2] due to the fact that it
> guarantees NUL-termination on the destination buffer without
> unnecessarily NUL-padding.
>
> Moreover, let's opt for the more conventional `sizeof()` as opposed to
> `ARRAY_SIZE` for these simple strings.
>
> Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings [1]
> Link: https://manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html [2]
> Link: https://github.com/KSPP/linux/issues/90
> Cc: [email protected]
> Signed-off-by: Justin Stitt <[email protected]>
> ---
> FWIW: It seems fragile to base future `prop` stores on the
> size of it's default string.
Agreed - can we just get rid of the copying? Just set a const char *
to point to appropriate string instead and use that?
The dance is just about reasonable for the case where there is
a string format being used to fill in the numbers, here I can't see any
advantage at all. (for the other case, a kasprintf() or similar is probably
more appropriate anyway) given this isn't a particular hot path.
Jonathan
>
> Note: build-tested
> ---
> drivers/iio/proximity/sx9324.c | 12 ++++--------
> 1 file changed, 4 insertions(+), 8 deletions(-)
>
> diff --git a/drivers/iio/proximity/sx9324.c b/drivers/iio/proximity/sx9324.c
> index 438f9c9aba6e..25ac2733bcef 100644
> --- a/drivers/iio/proximity/sx9324.c
> +++ b/drivers/iio/proximity/sx9324.c
> @@ -937,11 +937,9 @@ sx9324_get_default_reg(struct device *dev, int idx,
> case SX9324_REG_AFE_CTRL4:
> case SX9324_REG_AFE_CTRL7:
> if (reg_def->reg == SX9324_REG_AFE_CTRL4)
> - strncpy(prop, "semtech,ph01-resolution",
> - ARRAY_SIZE(prop));
> + strscpy(prop, "semtech,ph01-resolution", sizeof(prop));
> else
> - strncpy(prop, "semtech,ph23-resolution",
> - ARRAY_SIZE(prop));
> + strscpy(prop, "semtech,ph23-resolution", sizeof(prop));
>
> ret = device_property_read_u32(dev, prop, &raw);
> if (ret)
> @@ -1012,11 +1010,9 @@ sx9324_get_default_reg(struct device *dev, int idx,
> case SX9324_REG_PROX_CTRL0:
> case SX9324_REG_PROX_CTRL1:
> if (reg_def->reg == SX9324_REG_PROX_CTRL0)
> - strncpy(prop, "semtech,ph01-proxraw-strength",
> - ARRAY_SIZE(prop));
> + strscpy(prop, "semtech,ph01-proxraw-strength", sizeof(prop));
> else
> - strncpy(prop, "semtech,ph23-proxraw-strength",
> - ARRAY_SIZE(prop));
> + strscpy(prop, "semtech,ph23-proxraw-strength", sizeof(prop));
> ret = device_property_read_u32(dev, prop, &raw);
> if (ret)
> break;
>
> ---
> base-commit: 2cf0f715623872823a72e451243bbf555d10d032
> change-id: 20230921-strncpy-drivers-iio-proximity-sx9324-c-8c3437676039
>
> Best regards,
> --
> Justin Stitt <[email protected]>
>
Hi Jonathan,
On Sat, Sep 23, 2023 at 10:47 AM Jonathan Cameron <[email protected]> wrote:
>
> On Thu, 21 Sep 2023 07:01:01 +0000
> Justin Stitt <[email protected]> wrote:
>
> > `strncpy` is deprecated for use on NUL-terminated destination strings [1].
> >
> > We should prefer more robust and less ambiguous string interfaces.
> >
> > `prop` is defined as this string literal with size 30 (including null):
> > | #define SX9324_PROXRAW_DEF "semtech,ph01-proxraw-strength"
> > | char prop[] = SX9324_PROXRAW_DEF;
> >
> > Each of the strncpy->strscpy replacements involve string literals with a
> > size less than 30 which means there are no current problems with how
> > strncpy is used. However, let's move away from using strncpy entirely.
> >
> > A suitable replacement is `strscpy` [2] due to the fact that it
> > guarantees NUL-termination on the destination buffer without
> > unnecessarily NUL-padding.
> >
> > Moreover, let's opt for the more conventional `sizeof()` as opposed to
> > `ARRAY_SIZE` for these simple strings.
> >
> > Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings [1]
> > Link: https://manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html [2]
> > Link: https://github.com/KSPP/linux/issues/90
> > Cc: [email protected]
> > Signed-off-by: Justin Stitt <[email protected]>
> > ---
> > FWIW: It seems fragile to base future `prop` stores on the
> > size of it's default string.
>
> Agreed - can we just get rid of the copying? Just set a const char *
> to point to appropriate string instead and use that?
>
> The dance is just about reasonable for the case where there is
> a string format being used to fill in the numbers, here I can't see any
> advantage at all. (for the other case, a kasprintf() or similar is probably
> more appropriate anyway) given this isn't a particular hot path.
I sent a [v2]! Can you see if this matches your vision here?
>
> Jonathan
> >
> > Note: build-tested
> > ---
> > drivers/iio/proximity/sx9324.c | 12 ++++--------
> > 1 file changed, 4 insertions(+), 8 deletions(-)
> >
> > diff --git a/drivers/iio/proximity/sx9324.c b/drivers/iio/proximity/sx9324.c
> > index 438f9c9aba6e..25ac2733bcef 100644
> > --- a/drivers/iio/proximity/sx9324.c
> > +++ b/drivers/iio/proximity/sx9324.c
> > @@ -937,11 +937,9 @@ sx9324_get_default_reg(struct device *dev, int idx,
> > case SX9324_REG_AFE_CTRL4:
> > case SX9324_REG_AFE_CTRL7:
> > if (reg_def->reg == SX9324_REG_AFE_CTRL4)
> > - strncpy(prop, "semtech,ph01-resolution",
> > - ARRAY_SIZE(prop));
> > + strscpy(prop, "semtech,ph01-resolution", sizeof(prop));
> > else
> > - strncpy(prop, "semtech,ph23-resolution",
> > - ARRAY_SIZE(prop));
> > + strscpy(prop, "semtech,ph23-resolution", sizeof(prop));
> >
> > ret = device_property_read_u32(dev, prop, &raw);
> > if (ret)
> > @@ -1012,11 +1010,9 @@ sx9324_get_default_reg(struct device *dev, int idx,
> > case SX9324_REG_PROX_CTRL0:
> > case SX9324_REG_PROX_CTRL1:
> > if (reg_def->reg == SX9324_REG_PROX_CTRL0)
> > - strncpy(prop, "semtech,ph01-proxraw-strength",
> > - ARRAY_SIZE(prop));
> > + strscpy(prop, "semtech,ph01-proxraw-strength", sizeof(prop));
> > else
> > - strncpy(prop, "semtech,ph23-proxraw-strength",
> > - ARRAY_SIZE(prop));
> > + strscpy(prop, "semtech,ph23-proxraw-strength", sizeof(prop));
> > ret = device_property_read_u32(dev, prop, &raw);
> > if (ret)
> > break;
> >
> > ---
> > base-commit: 2cf0f715623872823a72e451243bbf555d10d032
> > change-id: 20230921-strncpy-drivers-iio-proximity-sx9324-c-8c3437676039
> >
> > Best regards,
> > --
> > Justin Stitt <[email protected]>
> >
>
[v2]: https://lore.kernel.org/r/20231026-strncpy-drivers-iio-proximity-sx9324-c-v2-1-cee6e5db700c@google.com
Thanks
Justin