[PATCH net-next] ipv4: ip_tunnel: Replace strcpy use with strscpy

[PATCH net-next] ipv4: ip_tunnel: Replace strcpy use with strscpy

[Ruben Wauters]

Use of strcpy is decpreated, replaces the use of strcpy with strscpy as
recommended.

I am aware there is an explicit bounds check above, however using
strscpy protects against buffer overflows in any future code, and there
is no good reason I can see to not use it.

Signed-off-by: Ruben Wauters <rubenru09@aol.com>
---
 net/ipv4/ip_tunnel.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/ipv4/ip_tunnel.c b/net/ipv4/ip_tunnel.c
index 3913ec89ad20..9724bbbd0e0a 100644
--- a/net/ipv4/ip_tunnel.c
+++ b/net/ipv4/ip_tunnel.c
@@ -247,7 +247,7 @@ static struct net_device *__ip_tunnel_create(struct net *net,
 	} else {
 		if (strlen(ops->kind) > (IFNAMSIZ - 3))
 			goto failed;
-		strcpy(name, ops->kind);
+		strscpy(name, ops->kind);
 		strcat(name, "%d");
 	}
 
-- 
2.48.1

Re: [PATCH net-next] ipv4: ip_tunnel: Replace strcpy use with strscpy

[Simon Horman]

On Thu, May 01, 2025 at 02:23:00AM +0100, Ruben Wauters wrote:
> Use of strcpy is decpreated, replaces the use of strcpy with strscpy as
> recommended.
> 
> I am aware there is an explicit bounds check above, however using
> strscpy protects against buffer overflows in any future code, and there
> is no good reason I can see to not use it.

Thanks, I agree. This patch doesn't buy us safety. But it doesn't lose
us anything. And allows the code to move towards best practice.

One thing I notices is that this change is is inconsistent with the call to
the 3-argument variant of strscpy a few lines above - it should also be hte
2-argument version. Maybe that could be changed too. Maybe in a
separate patch.

It is customary when making such changes to add a note that
strscpy() was chosen because the code expects a NUL-terminated string
without zero-padding. (Which is the case due to the call to strcat().)
Perhaps you could add some text to the commit message of v2 of this patch?

> Signed-off-by: Ruben Wauters <rubenru09@aol.com>

Reviewed-by: Simon Horman <horms@kernel.org>

> ---
>  net/ipv4/ip_tunnel.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/net/ipv4/ip_tunnel.c b/net/ipv4/ip_tunnel.c
> index 3913ec89ad20..9724bbbd0e0a 100644
> --- a/net/ipv4/ip_tunnel.c
> +++ b/net/ipv4/ip_tunnel.c
> @@ -247,7 +247,7 @@ static struct net_device *__ip_tunnel_create(struct net *net,
>  	} else {
>  		if (strlen(ops->kind) > (IFNAMSIZ - 3))
>  			goto failed;
> -		strcpy(name, ops->kind);
> +		strscpy(name, ops->kind);
>  		strcat(name, "%d");
>  	}
>  
> -- 
> 2.48.1
> 
> 

Re: [PATCH net-next] ipv4: ip_tunnel: Replace strcpy use with strscpy

[Ruben Wauters]

On Thu, 2025-05-01 at 16:39 +0100, Simon Horman wrote:
> On Thu, May 01, 2025 at 02:23:00AM +0100, Ruben Wauters wrote:
> > Use of strcpy is decpreated, replaces the use of strcpy with
> > strscpy as
> > recommended.
> > 
> > I am aware there is an explicit bounds check above, however using
> > strscpy protects against buffer overflows in any future code, and
> > there
> > is no good reason I can see to not use it.
> 
> Thanks, I agree. This patch doesn't buy us safety. But it doesn't
> lose
> us anything. And allows the code to move towards best practice.
> 
> One thing I notices is that this change is is inconsistent with the
> call to
> the 3-argument variant of strscpy a few lines above - it should also
> be hte
> 2-argument version. Maybe that could be changed too. Maybe in a
> separate patch.


I can remove the size parameter from the above strscpy to make it
consistent in v2.

> It is customary when making such changes to add a note that
> strscpy() was chosen because the code expects a NUL-terminated string
> without zero-padding. (Which is the case due to the call to
> strcat().)
> Perhaps you could add some text to the commit message of v2 of this
> patch?

Apologies, I wasn't aware of this, I can add the text to v2.

Just a point of clarification I wanted to ask, for v2 of the patch,
should I include the Reviewed-by tag below? or should I remove it as
there has been changes?


> > Signed-off-by: Ruben Wauters <rubenru09@aol.com>
> 
> Reviewed-by: Simon Horman <horms@kernel.org>
> 
> > ---
> >  net/ipv4/ip_tunnel.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/net/ipv4/ip_tunnel.c b/net/ipv4/ip_tunnel.c
> > index 3913ec89ad20..9724bbbd0e0a 100644
> > --- a/net/ipv4/ip_tunnel.c
> > +++ b/net/ipv4/ip_tunnel.c
> > @@ -247,7 +247,7 @@ static struct net_device
> > *__ip_tunnel_create(struct net *net,
> >  	} else {
> >  		if (strlen(ops->kind) > (IFNAMSIZ - 3))
> >  			goto failed;
> > -		strcpy(name, ops->kind);
> > +		strscpy(name, ops->kind);
> >  		strcat(name, "%d");
> >  	}
> >  
> > -- 
> > 2.48.1
> > 
> > 
Ruben Wauters

Re: [PATCH net-next] ipv4: ip_tunnel: Replace strcpy use with strscpy

[Simon Horman]

On Thu, May 01, 2025 at 05:51:08PM +0100, Ruben Wauters wrote:
> On Thu, 2025-05-01 at 16:39 +0100, Simon Horman wrote:
> > On Thu, May 01, 2025 at 02:23:00AM +0100, Ruben Wauters wrote:
> > > Use of strcpy is decpreated, replaces the use of strcpy with
> > > strscpy as
> > > recommended.
> > > 
> > > I am aware there is an explicit bounds check above, however using
> > > strscpy protects against buffer overflows in any future code, and
> > > there
> > > is no good reason I can see to not use it.
> > 
> > Thanks, I agree. This patch doesn't buy us safety. But it doesn't
> > lose
> > us anything. And allows the code to move towards best practice.
> > 
> > One thing I notices is that this change is is inconsistent with the
> > call to
> > the 3-argument variant of strscpy a few lines above - it should also
> > be hte
> > 2-argument version. Maybe that could be changed too. Maybe in a
> > separate patch.
> 
> 
> I can remove the size parameter from the above strscpy to make it
> consistent in v2.
> 
> > It is customary when making such changes to add a note that
> > strscpy() was chosen because the code expects a NUL-terminated string
> > without zero-padding. (Which is the case due to the call to
> > strcat().)
> > Perhaps you could add some text to the commit message of v2 of this
> > patch?
> 
> Apologies, I wasn't aware of this, I can add the text to v2.
> 
> Just a point of clarification I wanted to ask, for v2 of the patch,
> should I include the Reviewed-by tag below? or should I remove it as
> there has been changes?

I think you can include it unless the changes turn
out to be materially different to what has been discussed
in this thread.

But if in doubt, drop it.

> 
> 
> > > Signed-off-by: Ruben Wauters <rubenru09@aol.com>
> > 
> > Reviewed-by: Simon Horman <horms@kernel.org>

...