2023-09-25 09:29:52

by Justin Stitt

Subject: [PATCH v4] hwmon: (acpi_power_meter) replace open-coded kmemdup_nul

`strncpy` is deprecated for use on NUL-terminated destination strings [1].

Let's refactor this kcalloc() + strncpy() into a kmemdup_nul() which has
more obvious behavior and is less error prone.

Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings [1]
Link: https://github.com/KSPP/linux/issues/90
Cc: [email protected]
Signed-off-by: Justin Stitt <[email protected]>
---
Changes in v4:
- drop +1 from length arg (thanks Kees)
- reword subject line (thanks Kees)
- rebase onto 6465e260f4879080
- Link to v3: https://lore.kernel.org/r/20230921-strncpy-drivers-hwmon-acpi_power_meter-c-v3-1-307552c6ec3f@google.com

Changes in v3:
- refactor to use kmemdup_nul() (thanks Thomas and Kees)
- change commit msg to reflect ^
- rebase onto 2cf0f71562387282
- Link to v2: https://lore.kernel.org/r/20230919-strncpy-drivers-hwmon-acpi_power_meter-c-v2-1-8348432d6442@google.com

Changes in v2:
- use memcpy over strscpy (thanks Kees)
- Link to v1: https://lore.kernel.org/r/20230914-strncpy-drivers-hwmon-acpi_power_meter-c-v1-1-905297479fe8@google.com
---
drivers/hwmon/acpi_power_meter.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/hwmon/acpi_power_meter.c b/drivers/hwmon/acpi_power_meter.c
index fa28d447f0df..c13b5c8a0433 100644
--- a/drivers/hwmon/acpi_power_meter.c
+++ b/drivers/hwmon/acpi_power_meter.c
@@ -796,14 +796,13 @@ static int read_capabilities(struct acpi_power_meter_resource *resource)
goto error;
}

- *str = kcalloc(element->string.length + 1, sizeof(u8),
- GFP_KERNEL);
+ *str = kmemdup_nul(element->string.pointer, element->string.length,
+ GFP_KERNEL);
if (!*str) {
res = -ENOMEM;
goto error;
}

- strncpy(*str, element->string.pointer, element->string.length);
str++;
}


---
base-commit: 6465e260f48790807eef06b583b38ca9789b6072
change-id: 20230914-strncpy-drivers-hwmon-acpi_power_meter-c-c9f2d8053bef

Best regards,
--
Justin Stitt <[email protected]>

2023-09-25 22:16:16

by Kees Cook

Subject: Re: [PATCH v4] hwmon: (acpi_power_meter) replace open-coded kmemdup_nul

On Mon, Sep 25, 2023 at 03:43:23AM +0000, Justin Stitt wrote:
> `strncpy` is deprecated for use on NUL-terminated destination strings [1].
>
> Let's refactor this kcalloc() + strncpy() into a kmemdup_nul() which has
> more obvious behavior and is less error prone.
>
> Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings [1]
> Link: https://github.com/KSPP/linux/issues/90
> Cc: [email protected]
> Signed-off-by: Justin Stitt <[email protected]>
> ---
> Changes in v4:
> - drop +1 from length arg (thanks Kees)
> - reword subject line (thanks Kees)
> - rebase onto 6465e260f4879080
> - Link to v3: https://lore.kernel.org/r/20230921-strncpy-drivers-hwmon-acpi_power_meter-c-v3-1-307552c6ec3f@google.com
>
> Changes in v3:
> - refactor to use kmemdup_nul() (thanks Thomas and Kees)
> - change commit msg to reflect ^
> - rebase onto 2cf0f71562387282
> - Link to v2: https://lore.kernel.org/r/20230919-strncpy-drivers-hwmon-acpi_power_meter-c-v2-1-8348432d6442@google.com
>
> Changes in v2:
> - use memcpy over strscpy (thanks Kees)
> - Link to v1: https://lore.kernel.org/r/20230914-strncpy-drivers-hwmon-acpi_power_meter-c-v1-1-905297479fe8@google.com
> ---
> drivers/hwmon/acpi_power_meter.c | 5 ++---
> 1 file changed, 2 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/hwmon/acpi_power_meter.c b/drivers/hwmon/acpi_power_meter.c
> index fa28d447f0df..c13b5c8a0433 100644
> --- a/drivers/hwmon/acpi_power_meter.c
> +++ b/drivers/hwmon/acpi_power_meter.c
> @@ -796,14 +796,13 @@ static int read_capabilities(struct acpi_power_meter_resource *resource)
> goto error;
> }
>
> - *str = kcalloc(element->string.length + 1, sizeof(u8),
> - GFP_KERNEL);
> + *str = kmemdup_nul(element->string.pointer, element->string.length,
> + GFP_KERNEL);

This whitespace looks weird -- I'd expect this to line up with
"element", like this:

> + *str = kmemdup_nul(element->string.pointer, element->string.length,
> + GFP_KERNEL);

Otherwise, yes, looks correct.

--
Kees Cook