[PATCH] image: Ensure image header name is null terminated

[PATCH] image: Ensure image header name is null terminated

[Joel Stanley]

When building with GCC 12:

../include/image.h:779:9: warning: ‘strncpy’ specified bound 32 equals destination size [-Wstringop-truncation]
  779 |         strncpy(image_get_name(hdr), name, IH_NMLEN);
      |         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Ensure the copied string is null terminated by always setting the final
byte to 0. Shorten the strncpy to IH_NMLEN-1 as we will always overwrite
the last byte.

We can't use strlcpy as this is code is built on the host as well as the
target.

Fixes: b97a2a0a21f2 ("[new uImage] Define a API for image handling operations")
Signed-off-by: Joel Stanley <joel@jms.id.au>
---
 include/image.h | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/include/image.h b/include/image.h
index e4c6a50b885f..665b2278b7fb 100644
--- a/include/image.h
+++ b/include/image.h
@@ -776,7 +776,10 @@ image_set_hdr_b(comp)		/* image_set_comp */
 
 static inline void image_set_name(image_header_t *hdr, const char *name)
 {
-	strncpy(image_get_name(hdr), name, IH_NMLEN);
+	char *hdr_name = image_get_name(hdr);
+
+	strncpy(hdr_name, name, IH_NMLEN - 1);
+	hdr_name[IH_NMLEN - 1] = '\0';
 }
 
 int image_check_hcrc(const image_header_t *hdr);
-- 
2.35.1

Re: [PATCH] image: Ensure image header name is null terminated

[Wolfgang Denk]

Dear Joel,

In message <20220823055907.416060-1-joel@jms.id.au> you wrote:
> When building with GCC 12:
>
> ../include/image.h:779:9: warning: ‘strncpy’ specified bound 32 equals destination size [-Wstringop-truncation]
>   779 |         strncpy(image_get_name(hdr), name, IH_NMLEN);
>       |         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
> Ensure the copied string is null terminated by always setting the final
> byte to 0. Shorten the strncpy to IH_NMLEN-1 as we will always overwrite
> the last byte.
>
> We can't use strlcpy as this is code is built on the host as well as the
> target.
>
> Fixes: b97a2a0a21f2 ("[new uImage] Define a API for image handling operations")
> Signed-off-by: Joel Stanley <joel@jms.id.au>
> ---
>  include/image.h | 5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
>
> diff --git a/include/image.h b/include/image.h
> index e4c6a50b885f..665b2278b7fb 100644
> --- a/include/image.h
> +++ b/include/image.h
> @@ -776,7 +776,10 @@ image_set_hdr_b(comp)		/* image_set_comp */
>  
>  static inline void image_set_name(image_header_t *hdr, const char *name)
>  {
> -	strncpy(image_get_name(hdr), name, IH_NMLEN);
> +	char *hdr_name = image_get_name(hdr);
> +
> +	strncpy(hdr_name, name, IH_NMLEN - 1);
> +	hdr_name[IH_NMLEN - 1] = '\0';
>  }

Why don't you use strlcpy() instead?  This covers exactly the
situation we have here.

Best regards,

Wolfgang Denk

-- 
DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich,  Office: Kirchenstr. 5, 82194 Groebenzell, Germany
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd@denx.de
panic: can't find /

Re: [PATCH] image: Ensure image header name is null terminated

[John Keeping]

On Tue, Aug 23, 2022 at 03:59:07PM +1000, Joel Stanley wrote:
> When building with GCC 12:
> 
> ../include/image.h:779:9: warning: ‘strncpy’ specified bound 32 equals destination size [-Wstringop-truncation]
>   779 |         strncpy(image_get_name(hdr), name, IH_NMLEN);
>       |         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> 
> Ensure the copied string is null terminated by always setting the final
> byte to 0. Shorten the strncpy to IH_NMLEN-1 as we will always overwrite
> the last byte.
> 
> We can't use strlcpy as this is code is built on the host as well as the
> target.

Since this is in the header, isn't the point that it doesn't need to be
null-terminated?

When printing we're careful to use:

	"%.*s", IH_NMLEN, ...

so I think the warning is wrong here - we want both of the strncpy()
behaviours that are normally considered strange:

- it's okay not to null terminate as this is an explicitly sized field

- we want to pad the whole field with zeroes if the string is short

> Fixes: b97a2a0a21f2 ("[new uImage] Define a API for image handling operations")
> Signed-off-by: Joel Stanley <joel@jms.id.au>
> ---
>  include/image.h | 5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
> 
> diff --git a/include/image.h b/include/image.h
> index e4c6a50b885f..665b2278b7fb 100644
> --- a/include/image.h
> +++ b/include/image.h
> @@ -776,7 +776,10 @@ image_set_hdr_b(comp)		/* image_set_comp */
>  
>  static inline void image_set_name(image_header_t *hdr, const char *name)
>  {
> -	strncpy(image_get_name(hdr), name, IH_NMLEN);
> +	char *hdr_name = image_get_name(hdr);
> +
> +	strncpy(hdr_name, name, IH_NMLEN - 1);
> +	hdr_name[IH_NMLEN - 1] = '\0';
>  }
>  
>  int image_check_hcrc(const image_header_t *hdr);
> -- 
> 2.35.1
> 

Re: [PATCH] image: Ensure image header name is null terminated

[Simon Glass]

Hi John,

On Tue, 23 Aug 2022 at 03:46, John Keeping <john@metanate.com> wrote:
>
> On Tue, Aug 23, 2022 at 03:59:07PM +1000, Joel Stanley wrote:
> > When building with GCC 12:
> >
> > ../include/image.h:779:9: warning: ‘strncpy’ specified bound 32 equals destination size [-Wstringop-truncation]
> >   779 |         strncpy(image_get_name(hdr), name, IH_NMLEN);
> >       |         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> >
> > Ensure the copied string is null terminated by always setting the final
> > byte to 0. Shorten the strncpy to IH_NMLEN-1 as we will always overwrite
> > the last byte.
> >
> > We can't use strlcpy as this is code is built on the host as well as the
> > target.
>
> Since this is in the header, isn't the point that it doesn't need to be
> null-terminated?
>
> When printing we're careful to use:
>
>         "%.*s", IH_NMLEN, ...
>
> so I think the warning is wrong here - we want both of the strncpy()
> behaviours that are normally considered strange:
>
> - it's okay not to null terminate as this is an explicitly sized field
>
> - we want to pad the whole field with zeroes if the string is short

That's my understanding too. We are careful to avoid expecting a
terminator. I am not sure what to do with the warning though

Regards,
Simon


>
> > Fixes: b97a2a0a21f2 ("[new uImage] Define a API for image handling operations")
> > Signed-off-by: Joel Stanley <joel@jms.id.au>
> > ---
> >  include/image.h | 5 ++++-
> >  1 file changed, 4 insertions(+), 1 deletion(-)
> >
> > diff --git a/include/image.h b/include/image.h
> > index e4c6a50b885f..665b2278b7fb 100644
> > --- a/include/image.h
> > +++ b/include/image.h
> > @@ -776,7 +776,10 @@ image_set_hdr_b(comp)            /* image_set_comp */
> >
> >  static inline void image_set_name(image_header_t *hdr, const char *name)
> >  {
> > -     strncpy(image_get_name(hdr), name, IH_NMLEN);
> > +     char *hdr_name = image_get_name(hdr);
> > +
> > +     strncpy(hdr_name, name, IH_NMLEN - 1);
> > +     hdr_name[IH_NMLEN - 1] = '\0';
> >  }
> >
> >  int image_check_hcrc(const image_header_t *hdr);
> > --
> > 2.35.1
> >

Re: [PATCH] image: Ensure image header name is null terminated

[Rasmus Villemoes]

On 23/08/2022 15.38, Simon Glass wrote:
> Hi John,
> 
> On Tue, 23 Aug 2022 at 03:46, John Keeping <john@metanate.com> wrote:
>>
>> On Tue, Aug 23, 2022 at 03:59:07PM +1000, Joel Stanley wrote:
>>> When building with GCC 12:
>>>
>>> ../include/image.h:779:9: warning: ‘strncpy’ specified bound 32 equals destination size [-Wstringop-truncation]
>>>   779 |         strncpy(image_get_name(hdr), name, IH_NMLEN);
>>>       |         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>>>
>>> Ensure the copied string is null terminated by always setting the final
>>> byte to 0. Shorten the strncpy to IH_NMLEN-1 as we will always overwrite
>>> the last byte.
>>>
>>> We can't use strlcpy as this is code is built on the host as well as the
>>> target.
>>
>> Since this is in the header, isn't the point that it doesn't need to be
>> null-terminated?
>>
>> When printing we're careful to use:
>>
>>         "%.*s", IH_NMLEN, ...
>>
>> so I think the warning is wrong here - we want both of the strncpy()
>> behaviours that are normally considered strange:
>>
>> - it's okay not to null terminate as this is an explicitly sized field
>>
>> - we want to pad the whole field with zeroes if the string is short
> 
> That's my understanding too. We are careful to avoid expecting a
> terminator. I am not sure what to do with the warning though

Maybe this could be some inspiration:

info gcc

'nonstring'
     The 'nonstring' variable attribute specifies that an object or
     member declaration with type array of 'char', 'signed char', or
     'unsigned char', or pointer to such a type is intended to store
     character arrays that do not necessarily contain a terminating
     'NUL'.  This is useful in detecting uses of such arrays or pointers
     with functions that expect 'NUL'-terminated strings, and to avoid
     warnings when such an array or pointer is used as an argument to a
     bounded string manipulation function such as 'strncpy'.  For
     example, without the attribute, GCC will issue a warning for the
     'strncpy' call below because it may truncate the copy without
     appending the terminating 'NUL' character.  Using the attribute
     makes it possible to suppress the warning.  However, when the array
     is declared with the attribute the call to 'strlen' is diagnosed
     because when the array doesn't contain a 'NUL'-terminated string
     the call is undefined.  To copy, compare, of search non-string
     character arrays use the 'memcpy', 'memcmp', 'memchr', and other
     functions that operate on arrays of bytes.  In addition, calling
     'strnlen' and 'strndup' with such arrays is safe provided a
     suitable bound is specified, and not diagnosed.

          struct Data
          {
            char name [32] __attribute__ ((nonstring));
          };

          int f (struct Data *pd, const char *s)
          {
            strncpy (pd->name, s, sizeof pd->name);
            ...
            return strlen (pd->name);   // unsafe, gets a warning
          }

[https://gcc.gnu.org/onlinedocs/gcc-12.2.0/gcc/Common-Variable-Attributes.html#Common-Variable-Attributes]

Re: [PATCH] image: Ensure image header name is null terminated

[Tom Rini]

[-- Attachment #1: Type: text/plain, Size: 1014 bytes --]

On Tue, Aug 23, 2022 at 03:59:07PM +1000, Joel Stanley wrote:

> When building with GCC 12:
> 
> ../include/image.h:779:9: warning: ‘strncpy’ specified bound 32 equals destination size [-Wstringop-truncation]
>   779 |         strncpy(image_get_name(hdr), name, IH_NMLEN);
>       |         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> 
> Ensure the copied string is null terminated by always setting the final
> byte to 0. Shorten the strncpy to IH_NMLEN-1 as we will always overwrite
> the last byte.
> 
> We can't use strlcpy as this is code is built on the host as well as the
> target.
> 
> Fixes: b97a2a0a21f2 ("[new uImage] Define a API for image handling operations")
> Signed-off-by: Joel Stanley <joel@jms.id.au>

So this breaks some tests:
https://source.denx.de/u-boot/u-boot/-/jobs/496773#L201
and it's not clear to me if the problem is the tests or the fix itself
(should we be doing a buffer of IH_NMLEN+1 and ensuring that's NULL
terminated? I don't know).

-- 
Tom

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 659 bytes --]

Re: [PATCH] image: Ensure image header name is null terminated

[Simon Glass]

Hi,

On Wed, 14 Sept 2022 at 16:11, Tom Rini <trini@konsulko.com> wrote:
>
> On Tue, Aug 23, 2022 at 03:59:07PM +1000, Joel Stanley wrote:
>
> > When building with GCC 12:
> >
> > ../include/image.h:779:9: warning: ‘strncpy’ specified bound 32 equals destination size [-Wstringop-truncation]
> >   779 |         strncpy(image_get_name(hdr), name, IH_NMLEN);
> >       |         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> >
> > Ensure the copied string is null terminated by always setting the final
> > byte to 0. Shorten the strncpy to IH_NMLEN-1 as we will always overwrite
> > the last byte.
> >
> > We can't use strlcpy as this is code is built on the host as well as the
> > target.
> >
> > Fixes: b97a2a0a21f2 ("[new uImage] Define a API for image handling operations")
> > Signed-off-by: Joel Stanley <joel@jms.id.au>
>
> So this breaks some tests:
> https://source.denx.de/u-boot/u-boot/-/jobs/496773#L201
> and it's not clear to me if the problem is the tests or the fix itself
> (should we be doing a buffer of IH_NMLEN+1 and ensuring that's NULL
> terminated? I don't know).

My reading of it is that the field is of length IH_NMLEN and there is
only a terminator if the string is shorter than that.

So I don't think this patch is correct / needed. Perhaps we can find a
way to silence the warning, e.g. using memcyp(xx,yy, min(IH_NMLEN,
strnlen(yy, ...) + 1)) ?

Regards,
Simon