1. Patchew
  2. linux
  3. [PATCH net-next v2 0/5] netconsole: reuse netpoll_parse_ip_addr in configfs helpers
  4. View patch

[PATCH net-next v2 3/5] netconsole: add support for strings with new line in netpoll_parse_ip_addr

Breno Leitao posted 5 patches 2 months, 2 weeks ago
There is a newer version of this series
[PATCH net-next v2 3/5] netconsole: add support for strings with new line in netpoll_parse_ip_addr
Posted by Breno Leitao 2 months, 2 weeks ago 
The current IP address parsing logic fails when the input string
contains a trailing newline character. This can occur when IP
addresses are provided through configfs, which contains newlines in
a const buffer.

Teach netpoll_parse_ip_addr() how to ignore newlines at the end of the
IPs. Also, simplify the code by:

 * No need to check for separators. Try to parse ipv4, if it fails try
   ipv6 similarly to ceph_pton()
 * If ipv6 is not supported, don't call in6_pton() at all.

Signed-off-by: Breno Leitao <leitao@debian.org>
---
 drivers/net/netconsole.c | 23 ++++++++++++-----------
 1 file changed, 12 insertions(+), 11 deletions(-)

diff --git a/drivers/net/netconsole.c b/drivers/net/netconsole.c
index 8d1b93264e0fd..f2c2b8852c603 100644
--- a/drivers/net/netconsole.c
+++ b/drivers/net/netconsole.c
@@ -303,20 +303,21 @@ static void netconsole_print_banner(struct netpoll *np)
 static int netpoll_parse_ip_addr(const char *str, union inet_addr *addr)
 {
 	const char *end;
+	int len;
 
-	if (!strchr(str, ':') &&
-	    in4_pton(str, -1, (void *)addr, -1, &end) > 0) {
-		if (!*end)
-			return 0;
-	}
-	if (in6_pton(str, -1, addr->in6.s6_addr, -1, &end) > 0) {
-#if IS_ENABLED(CONFIG_IPV6)
-		if (!*end)
-			return 1;
-#else
+	len = strlen(str);
+	if (!len)
 		return -1;
+
+	if (str[len - 1] == '\n')
+		len -= 1;
+
+	if (in4_pton(str, len, (void *)addr, -1, &end) > 0)
+		return 0;
+#if IS_ENABLED(CONFIG_IPV6)
+	if (in6_pton(str, len, addr->in6.s6_addr, -1, &end) > 0)
+		return 1;
 #endif
-	}
 	return -1;
 }
 

-- 
2.47.1
Re: [PATCH net-next v2 3/5] netconsole: add support for strings with new line in netpoll_parse_ip_addr
Posted by Simon Horman 2 months, 2 weeks ago 
On Mon, Jul 21, 2025 at 06:02:03AM -0700, Breno Leitao wrote:
> The current IP address parsing logic fails when the input string
> contains a trailing newline character. This can occur when IP
> addresses are provided through configfs, which contains newlines in
> a const buffer.
> 
> Teach netpoll_parse_ip_addr() how to ignore newlines at the end of the
> IPs. Also, simplify the code by:
> 
>  * No need to check for separators. Try to parse ipv4, if it fails try
>    ipv6 similarly to ceph_pton()
>  * If ipv6 is not supported, don't call in6_pton() at all.
> 
> Signed-off-by: Breno Leitao <leitao@debian.org>

My suggestion below not withstanding, this looks good to me.

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

> ---
>  drivers/net/netconsole.c | 23 ++++++++++++-----------
>  1 file changed, 12 insertions(+), 11 deletions(-)
> 
> diff --git a/drivers/net/netconsole.c b/drivers/net/netconsole.c
> index 8d1b93264e0fd..f2c2b8852c603 100644
> --- a/drivers/net/netconsole.c
> +++ b/drivers/net/netconsole.c
> @@ -303,20 +303,21 @@ static void netconsole_print_banner(struct netpoll *np)
>  static int netpoll_parse_ip_addr(const char *str, union inet_addr *addr)
>  {
>  	const char *end;
> +	int len;
>  
> -	if (!strchr(str, ':') &&
> -	    in4_pton(str, -1, (void *)addr, -1, &end) > 0) {
> -		if (!*end)
> -			return 0;
> -	}
> -	if (in6_pton(str, -1, addr->in6.s6_addr, -1, &end) > 0) {
> -#if IS_ENABLED(CONFIG_IPV6)
> -		if (!*end)
> -			return 1;
> -#else
> +	len = strlen(str);
> +	if (!len)
>  		return -1;
> +
> +	if (str[len - 1] == '\n')
> +		len -= 1;
> +
> +	if (in4_pton(str, len, (void *)addr, -1, &end) > 0)
> +		return 0;
> +#if IS_ENABLED(CONFIG_IPV6)
> +	if (in6_pton(str, len, addr->in6.s6_addr, -1, &end) > 0)
> +		return 1;
>  #endif

I don't think it needs to block progress.
But FWIIW, I think it would be nice to increase
build coverage and express this as:

	if (IS_ENABLED(CONFIG_IPV6) &&
	    in6_pton(str, len, addr->in6.s6_addr, -1, &end) > 0)
		return 1;
Re: [PATCH net-next v2 3/5] netconsole: add support for strings with new line in netpoll_parse_ip_addr
Posted by Breno Leitao 2 months, 2 weeks ago 
On Wed, Jul 23, 2025 at 03:54:11PM +0100, Simon Horman wrote:
> On Mon, Jul 21, 2025 at 06:02:03AM -0700, Breno Leitao wrote:

> > --- a/drivers/net/netconsole.c
> > +++ b/drivers/net/netconsole.c
> > @@ -303,20 +303,21 @@ static void netconsole_print_banner(struct netpoll *np)
> >  static int netpoll_parse_ip_addr(const char *str, union inet_addr *addr)
> >  {
> >  	const char *end;
> > +	int len;
> >  
> > -	if (!strchr(str, ':') &&
> > -	    in4_pton(str, -1, (void *)addr, -1, &end) > 0) {
> > -		if (!*end)
> > -			return 0;
> > -	}
> > -	if (in6_pton(str, -1, addr->in6.s6_addr, -1, &end) > 0) {
> > -#if IS_ENABLED(CONFIG_IPV6)
> > -		if (!*end)
> > -			return 1;
> > -#else
> > +	len = strlen(str);
> > +	if (!len)
> >  		return -1;
> > +
> > +	if (str[len - 1] == '\n')
> > +		len -= 1;
> > +
> > +	if (in4_pton(str, len, (void *)addr, -1, &end) > 0)
> > +		return 0;
> > +#if IS_ENABLED(CONFIG_IPV6)
> > +	if (in6_pton(str, len, addr->in6.s6_addr, -1, &end) > 0)
> > +		return 1;
> >  #endif
> 
> I don't think it needs to block progress.
> But FWIIW, I think it would be nice to increase
> build coverage and express this as:

Agree. While testing with IPv6 disabled, the netcons selftest exploded,
so, this explose a bug in the selftest. This is now fixed in:

	https://lore.kernel.org/all/20250723-netcons_test_ipv6-v1-1-41c9092f93f9@debian.org/

Thanks for the review,
--breno
Re: [PATCH net-next v2 3/5] netconsole: add support for strings with new line in netpoll_parse_ip_addr
Posted by Simon Horman 2 months, 1 week ago 
On Wed, Jul 23, 2025 at 10:37:59AM -0700, Breno Leitao wrote:
> On Wed, Jul 23, 2025 at 03:54:11PM +0100, Simon Horman wrote:
> > On Mon, Jul 21, 2025 at 06:02:03AM -0700, Breno Leitao wrote:
> 
> > > --- a/drivers/net/netconsole.c
> > > +++ b/drivers/net/netconsole.c
> > > @@ -303,20 +303,21 @@ static void netconsole_print_banner(struct netpoll *np)
> > >  static int netpoll_parse_ip_addr(const char *str, union inet_addr *addr)
> > >  {
> > >  	const char *end;
> > > +	int len;
> > >  
> > > -	if (!strchr(str, ':') &&
> > > -	    in4_pton(str, -1, (void *)addr, -1, &end) > 0) {
> > > -		if (!*end)
> > > -			return 0;
> > > -	}
> > > -	if (in6_pton(str, -1, addr->in6.s6_addr, -1, &end) > 0) {
> > > -#if IS_ENABLED(CONFIG_IPV6)
> > > -		if (!*end)
> > > -			return 1;
> > > -#else
> > > +	len = strlen(str);
> > > +	if (!len)
> > >  		return -1;
> > > +
> > > +	if (str[len - 1] == '\n')
> > > +		len -= 1;
> > > +
> > > +	if (in4_pton(str, len, (void *)addr, -1, &end) > 0)
> > > +		return 0;
> > > +#if IS_ENABLED(CONFIG_IPV6)
> > > +	if (in6_pton(str, len, addr->in6.s6_addr, -1, &end) > 0)
> > > +		return 1;
> > >  #endif
> > 
> > I don't think it needs to block progress.
> > But FWIIW, I think it would be nice to increase
> > build coverage and express this as:
> 
> Agree. While testing with IPv6 disabled, the netcons selftest exploded,
> so, this explose a bug in the selftest. This is now fixed in:
> 
> 	https://lore.kernel.org/all/20250723-netcons_test_ipv6-v1-1-41c9092f93f9@debian.org/

Nice, good to find bugs.

Patchew 2.1-devel-b67e4c2

© 2016 - 2025 Red Hat, Inc.