[PATCH net-next 0/4] netconsole: validate a target's IP address configuration

[PATCH net-next 0/4] netconsole: validate a target's IP address configuration

[Gustavo Luiz Duarte]

This series adds two validations to the target configuration when the
user tries to enable it: first whether remote_ip was set, and second
whether local_ip and remote_ip address families match. Refuse to enable
the target if any of those validations fail.

These validations are already done for the target passed on the
command-line, so this aligns dynamic targets with the command-line
behavior.

The first two patches replace the per-target 'ipv6' flag with a
per-address 'family' field, which makes it easier to detect these error
conditions. Patches 3 and 4 implement the actual validations.

Signed-off-by: Gustavo Luiz Duarte <gustavold@gmail.com>
---
Gustavo Luiz Duarte (4):
      netconsole: add an address family to struct inet_addr
      netconsole: use the address family instead of the ipv6 flag
      netconsole: reject enabling a target with no remote IP address
      netconsole: reject a target mixing IPv4 and IPv6 addresses

 drivers/net/netconsole.c | 117 ++++++++++++++++++++---------------------------
 include/linux/netpoll.h  |  10 ++--
 2 files changed, 57 insertions(+), 70 deletions(-)
---
base-commit: a23b36233d4103def55dc8cf65698106d0bd1e62
change-id: 20260730-netcons_ipv6-565d55f55729

Best regards,
--  
Gustavo Luiz Duarte <gustavold@gmail.com>

[PATCH net-next 1/4] netconsole: add an address family to struct inet_addr

[Gustavo Luiz Duarte]

netconsole_target stores a single 'bool ipv6' to denote the target's
address family. This makes it hard to detect conditions like "no address
set" or ipv4/ipv6 mixup between local_ip and remote_ip.

Add a 'family' field to inet_addr so each address stores its own address
family: AF_UNSPEC while unset, else AF_INET or AF_INET6.

Nothing reads the new field yet. The next patch switches the users over
and removes the bool. No functional change.

Signed-off-by: Gustavo Luiz Duarte <gustavold@gmail.com>
---
 drivers/net/netconsole.c | 22 ++++++++++++++--------
 include/linux/netpoll.h  | 10 +++++++---
 2 files changed, 21 insertions(+), 11 deletions(-)

diff --git a/drivers/net/netconsole.c b/drivers/net/netconsole.c
index 03913302328c..d615a9256787 100644
--- a/drivers/net/netconsole.c
+++ b/drivers/net/netconsole.c
@@ -212,7 +212,7 @@ struct netconsole_target {
 	bool			extended;
 	bool			release;
 	struct netpoll		np;
-	union inet_addr		local_ip, remote_ip;
+	struct inet_addr	local_ip, remote_ip;
 	bool			ipv6;
 	u16			local_port, remote_port;
 	u8			remote_mac[ETH_ALEN];
@@ -415,6 +415,7 @@ static int netcons_take_ipv6(struct netconsole_target *nt,
 				continue;
 			/* Got the IP, let's return */
 			nt->local_ip.in6 = ifp->addr;
+			nt->local_ip.family = AF_INET6;
 			err = 0;
 			break;
 		}
@@ -456,6 +457,7 @@ static int netcons_take_ipv4(struct netconsole_target *nt,
 	}
 
 	nt->local_ip.ip = ifa->ifa_local;
+	nt->local_ip.family = AF_INET;
 	np_info(np, "local IP %pI4\n", &nt->local_ip.ip);
 
 	return 0;
@@ -731,7 +733,7 @@ static void netconsole_print_banner(struct netconsole_target *nt)
 /* Parse the string and populate the `inet_addr` union. Return 0 if IPv4 is
  * populated, 1 if IPv6 is populated, and -1 upon failure.
  */
-static int netpoll_parse_ip_addr(const char *str, union inet_addr *addr)
+static int netpoll_parse_ip_addr(const char *str, struct inet_addr *addr)
 {
 	const char *end = NULL;
 	int len;
@@ -743,14 +745,18 @@ static int netpoll_parse_ip_addr(const char *str, union inet_addr *addr)
 	if (str[len - 1] == '\n')
 		len -= 1;
 
-	if (in4_pton(str, len, (void *)addr, -1, &end) > 0 &&
-	    (!end || *end == 0 || *end == '\n'))
+	if (in4_pton(str, len, (void *)&addr->ip, -1, &end) > 0 &&
+	    (!end || *end == 0 || *end == '\n')) {
+		addr->family = AF_INET;
 		return 0;
+	}
 
 	if (IS_ENABLED(CONFIG_IPV6) &&
-	    in6_pton(str, len, (void *)addr, -1, &end) > 0 &&
-	    (!end || *end == 0 || *end == '\n'))
+	    in6_pton(str, len, (void *)&addr->in6, -1, &end) > 0 &&
+	    (!end || *end == 0 || *end == '\n')) {
+		addr->family = AF_INET6;
 		return 1;
+	}
 
 	return -1;
 }
@@ -858,7 +864,7 @@ static ssize_t local_ip_show(struct config_item *item, char *buf)
 	if (nt->ipv6)
 		return sysfs_emit(buf, "%pI6c\n", &nt->local_ip.in6);
 	else
-		return sysfs_emit(buf, "%pI4\n", &nt->local_ip);
+		return sysfs_emit(buf, "%pI4\n", &nt->local_ip.ip);
 }
 
 static ssize_t remote_ip_show(struct config_item *item, char *buf)
@@ -868,7 +874,7 @@ static ssize_t remote_ip_show(struct config_item *item, char *buf)
 	if (nt->ipv6)
 		return sysfs_emit(buf, "%pI6c\n", &nt->remote_ip.in6);
 	else
-		return sysfs_emit(buf, "%pI4\n", &nt->remote_ip);
+		return sysfs_emit(buf, "%pI4\n", &nt->remote_ip.ip);
 }
 
 static ssize_t local_mac_show(struct config_item *item, char *buf)
diff --git a/include/linux/netpoll.h b/include/linux/netpoll.h
index 1c6b1eec5efd..de97f001a0f9 100644
--- a/include/linux/netpoll.h
+++ b/include/linux/netpoll.h
@@ -16,9 +16,13 @@
 #include <linux/ip.h>
 #include <linux/udp.h>
 
-union inet_addr {
-	__be32		ip;
-	struct in6_addr	in6;
+struct inet_addr {
+	/* Address family: AF_UNSPEC when unset, else AF_INET or AF_INET6 */
+	u8			family;
+	union {
+		__be32		ip;
+		struct in6_addr	in6;
+	};
 };
 
 struct netpoll {

-- 
2.55.0

Re: [PATCH net-next 1/4] netconsole: add an address family to struct inet_addr

[Breno Leitao]

On Wed, Aug 05, 2026 at 10:33:01PM +0100, Gustavo Luiz Duarte wrote:
> @@ -731,7 +733,7 @@ static void netconsole_print_banner(struct netconsole_target *nt)
>  /* Parse the string and populate the `inet_addr` union. Return 0 if IPv4 is
						    ^-> there is no more union.

I think there are other references for union that needs to be updated as
well.

> +static int netpoll_parse_ip_addr(const char *str, struct inet_addr *addr)
...
>  	return -1;

Should the failure path set addr->family = AF_UNSPEC?

> +++ b/include/linux/netpoll.h
> +struct inet_addr {
> +	/* Address family: AF_UNSPEC when unset, else AF_INET or AF_INET6 */
> +	u8			family;
> +	union {
> +		__be32		ip;
> +		struct in6_addr	in6;
> +	};
>  };

Since a1116396476f6 ("netconsole: move local_ip/remote_ip/ipv6 to
netconsole_target") struct netpoll has no address member, and netconsole
is the only user left in the tree:

    drivers/net/netconsole.c:   union inet_addr         local_ip, remote_ip;
    drivers/net/netconsole.c:static int netpoll_parse_ip_addr(const char *str, union inet_addr *addr)

Since you touched it, can you move it to netconsole headers, please?

--breno

Re: [PATCH net-next 1/4] netconsole: add an address family to struct inet_addr

[Gustavo Luiz Duarte]

On Thu, Aug 6, 2026 at 4:49 PM Breno Leitao <leitao@debian.org> wrote:
>
> On Wed, Aug 05, 2026 at 10:33:01PM +0100, Gustavo Luiz Duarte wrote:
> > @@ -731,7 +733,7 @@ static void netconsole_print_banner(struct netconsole_target *nt)
> >  /* Parse the string and populate the `inet_addr` union. Return 0 if IPv4 is
>                                                     ^-> there is no more union.
>
> I think there are other references for union that needs to be updated as
> well.

I fixed this in the next patch, but let me do it here for consistency.
I didn't find any other reference to union.

>
> > +static int netpoll_parse_ip_addr(const char *str, struct inet_addr *addr)
> ...
> >       return -1;
>
> Should the failure path set addr->family = AF_UNSPEC?

The failure path leaves addr untouched, so if the user writes an invalid value
to local_ip or remote_ip the write fails and the original value is maintained.

>
> > +++ b/include/linux/netpoll.h
> > +struct inet_addr {
> > +     /* Address family: AF_UNSPEC when unset, else AF_INET or AF_INET6 */
> > +     u8                      family;
> > +     union {
> > +             __be32          ip;
> > +             struct in6_addr in6;
> > +     };
> >  };
>
> Since a1116396476f6 ("netconsole: move local_ip/remote_ip/ipv6 to
> netconsole_target") struct netpoll has no address member, and netconsole
> is the only user left in the tree:
>
>     drivers/net/netconsole.c:   union inet_addr         local_ip, remote_ip;
>     drivers/net/netconsole.c:static int netpoll_parse_ip_addr(const char *str, union inet_addr *addr)
>
> Since you touched it, can you move it to netconsole headers, please?

Sure, will do it in v2.

[PATCH net-next 2/4] netconsole: use the address family instead of the ipv6 flag

[Gustavo Luiz Duarte]

Now that we have the address family in inet_addr, use that and remove
nt->ipv6.

We no longer need netcons_local_ip_unset() to check that all bytes are
zeroes, as that is now denoted by (family == AF_UNSPEC).

Signed-off-by: Gustavo Luiz Duarte <gustavold@gmail.com>
---
 drivers/net/netconsole.c | 82 ++++++++++++++----------------------------------
 1 file changed, 23 insertions(+), 59 deletions(-)

diff --git a/drivers/net/netconsole.c b/drivers/net/netconsole.c
index d615a9256787..070bae7b4fd7 100644
--- a/drivers/net/netconsole.c
+++ b/drivers/net/netconsole.c
@@ -181,7 +181,6 @@ enum target_state {
  *		local_mac	(read-only)
  * @local_ip:	Source IP address of the target (read-write).
  * @remote_ip:	Destination IP address of the target (read-write).
- * @ipv6:	Whether the target addresses are IPv6 (read-write).
  * @local_port:	Source UDP port of the target (read-write).
  * @remote_port: Destination UDP port of the target (read-write).
  * @remote_mac:	Destination ethernet address of the target (read-write).
@@ -213,7 +212,6 @@ struct netconsole_target {
 	bool			release;
 	struct netpoll		np;
 	struct inet_addr	local_ip, remote_ip;
-	bool			ipv6;
 	u16			local_port, remote_port;
 	u8			remote_mac[ETH_ALEN];
 	/* protected by target_list_lock; +1 gives scnprintf() room for its
@@ -463,23 +461,6 @@ static int netcons_take_ipv4(struct netconsole_target *nt,
 	return 0;
 }
 
-/*
- * Test whether the caller left nt->local_ip unset, so that
- * netcons_netpoll_setup() should auto-populate it from the egress device.
- *
- * nt->local_ip is a union of __be32 (IPv4) and struct in6_addr (IPv6),
- * so an IPv6 address whose first 4 bytes are zero (e.g. ::1, ::2,
- * IPv4-mapped ::ffff:a.b.c.d) must not be tested via the IPv4 arm —
- * doing so would misclassify a caller-supplied address as unset and
- * silently overwrite it with whatever address the device exposes.
- */
-static bool netcons_local_ip_unset(const struct netconsole_target *nt)
-{
-	if (nt->ipv6)
-		return ipv6_addr_any(&nt->local_ip.in6);
-	return !nt->local_ip.ip;
-}
-
 static int netcons_netpoll_setup(struct netconsole_target *nt)
 {
 	struct net *net = current->nsproxy->net_ns;
@@ -525,16 +506,13 @@ static int netcons_netpoll_setup(struct netconsole_target *nt)
 		rtnl_lock();
 	}
 
-	if (netcons_local_ip_unset(nt)) {
-		if (!nt->ipv6) {
-			err = netcons_take_ipv4(nt, ndev);
-			if (err)
-				goto put;
-		} else {
+	if (nt->local_ip.family == AF_UNSPEC) {
+		if (nt->remote_ip.family == AF_INET6)
 			err = netcons_take_ipv6(nt, ndev);
-			if (err)
-				goto put;
-		}
+		else
+			err = netcons_take_ipv4(nt, ndev);
+		if (err)
+			goto put;
 		ip_overwritten = true;
 	}
 
@@ -716,22 +694,22 @@ static void netconsole_print_banner(struct netconsole_target *nt)
 	struct netpoll *np = &nt->np;
 
 	np_info(np, "local port %d\n", nt->local_port);
-	if (nt->ipv6)
+	if (nt->local_ip.family == AF_INET6)
 		np_info(np, "local IPv6 address %pI6c\n", &nt->local_ip.in6);
 	else
 		np_info(np, "local IPv4 address %pI4\n", &nt->local_ip.ip);
 	np_info(np, "interface name '%s'\n", np->dev_name);
 	np_info(np, "local ethernet address '%pM'\n", np->dev_mac);
 	np_info(np, "remote port %d\n", nt->remote_port);
-	if (nt->ipv6)
+	if (nt->remote_ip.family == AF_INET6)
 		np_info(np, "remote IPv6 address %pI6c\n", &nt->remote_ip.in6);
 	else
 		np_info(np, "remote IPv4 address %pI4\n", &nt->remote_ip.ip);
 	np_info(np, "remote ethernet address %pM\n", nt->remote_mac);
 }
 
-/* Parse the string and populate the `inet_addr` union. Return 0 if IPv4 is
- * populated, 1 if IPv6 is populated, and -1 upon failure.
+/* Parse the string and populate the `inet_addr` struct. Return 0 on success
+ * and -1 upon failure.
  */
 static int netpoll_parse_ip_addr(const char *str, struct inet_addr *addr)
 {
@@ -755,7 +733,7 @@ static int netpoll_parse_ip_addr(const char *str, struct inet_addr *addr)
 	    in6_pton(str, len, (void *)&addr->in6, -1, &end) > 0 &&
 	    (!end || *end == 0 || *end == '\n')) {
 		addr->family = AF_INET6;
-		return 1;
+		return 0;
 	}
 
 	return -1;
@@ -861,7 +839,7 @@ static ssize_t local_ip_show(struct config_item *item, char *buf)
 {
 	struct netconsole_target *nt = to_target(item);
 
-	if (nt->ipv6)
+	if (nt->local_ip.family == AF_INET6)
 		return sysfs_emit(buf, "%pI6c\n", &nt->local_ip.in6);
 	else
 		return sysfs_emit(buf, "%pI4\n", &nt->local_ip.ip);
@@ -871,7 +849,7 @@ static ssize_t remote_ip_show(struct config_item *item, char *buf)
 {
 	struct netconsole_target *nt = to_target(item);
 
-	if (nt->ipv6)
+	if (nt->remote_ip.family == AF_INET6)
 		return sysfs_emit(buf, "%pI6c\n", &nt->remote_ip.in6);
 	else
 		return sysfs_emit(buf, "%pI4\n", &nt->remote_ip.ip);
@@ -1219,7 +1197,6 @@ static ssize_t local_ip_store(struct config_item *item, const char *buf,
 {
 	struct netconsole_target *nt = to_target(item);
 	ssize_t ret = -EINVAL;
-	int ipv6;
 
 	dynamic_netconsole_mutex_lock();
 	if (nt->state == STATE_ENABLED) {
@@ -1228,10 +1205,8 @@ static ssize_t local_ip_store(struct config_item *item, const char *buf,
 		goto out_unlock;
 	}
 
-	ipv6 = netpoll_parse_ip_addr(buf, &nt->local_ip);
-	if (ipv6 == -1)
+	if (netpoll_parse_ip_addr(buf, &nt->local_ip) < 0)
 		goto out_unlock;
-	nt->ipv6 = !!ipv6;
 
 	ret = count;
 out_unlock:
@@ -1244,7 +1219,6 @@ static ssize_t remote_ip_store(struct config_item *item, const char *buf,
 {
 	struct netconsole_target *nt = to_target(item);
 	ssize_t ret = -EINVAL;
-	int ipv6;
 
 	dynamic_netconsole_mutex_lock();
 	if (nt->state == STATE_ENABLED) {
@@ -1253,10 +1227,8 @@ static ssize_t remote_ip_store(struct config_item *item, const char *buf,
 		goto out_unlock;
 	}
 
-	ipv6 = netpoll_parse_ip_addr(buf, &nt->remote_ip);
-	if (ipv6 == -1)
+	if (netpoll_parse_ip_addr(buf, &nt->remote_ip) < 0)
 		goto out_unlock;
-	nt->ipv6 = !!ipv6;
 
 	ret = count;
 out_unlock:
@@ -2067,7 +2039,7 @@ static void netpoll_udp_checksum(struct netconsole_target *nt,
 
 	/* check needs to be set, since it will be consumed in csum_partial */
 	udph->check = 0;
-	if (nt->ipv6)
+	if (nt->remote_ip.family == AF_INET6)
 		udph->check = csum_ipv6_magic(&nt->local_ip.in6,
 					      &nt->remote_ip.in6,
 					      udp_len, IPPROTO_UDP,
@@ -2108,7 +2080,7 @@ static void push_eth(struct netconsole_target *nt, struct sk_buff *skb)
 	skb_reset_mac_header(skb);
 	ether_addr_copy(eth->h_source, np->dev->dev_addr);
 	ether_addr_copy(eth->h_dest, nt->remote_mac);
-	if (nt->ipv6)
+	if (nt->remote_ip.family == AF_INET6)
 		eth->h_proto = htons(ETH_P_IPV6);
 	else
 		eth->h_proto = htons(ETH_P_IP);
@@ -2177,7 +2149,7 @@ static int netpoll_send_udp(struct netconsole_target *nt, const char *msg,
 		WARN_ON_ONCE(!irqs_disabled());
 
 	udp_len = len + sizeof(struct udphdr);
-	if (nt->ipv6)
+	if (nt->remote_ip.family == AF_INET6)
 		ip_len = udp_len + sizeof(struct ipv6hdr);
 	else
 		ip_len = udp_len + sizeof(struct iphdr);
@@ -2193,7 +2165,7 @@ static int netpoll_send_udp(struct netconsole_target *nt, const char *msg,
 	skb_put(skb, len);
 
 	push_udp(nt, skb, len);
-	if (nt->ipv6)
+	if (nt->remote_ip.family == AF_INET6)
 		push_ipv6(nt, skb, len);
 	else
 		push_ipv4(nt, skb, len);
@@ -2513,10 +2485,8 @@ __releases(&target_list_lock)
 static int netconsole_parser_cmdline(struct netconsole_target *nt, char *opt)
 {
 	struct netpoll *np = &nt->np;
-	bool ipversion_set = false;
 	char *cur = opt;
 	char *delim;
-	int ipv6;
 
 	if (*cur != '@') {
 		delim = strchr(cur, '@');
@@ -2530,16 +2500,12 @@ static int netconsole_parser_cmdline(struct netconsole_target *nt, char *opt)
 	cur++;
 
 	if (*cur != '/') {
-		ipversion_set = true;
 		delim = strchr(cur, '/');
 		if (!delim)
 			goto parse_failed;
 		*delim = 0;
-		ipv6 = netpoll_parse_ip_addr(cur, &nt->local_ip);
-		if (ipv6 < 0)
+		if (netpoll_parse_ip_addr(cur, &nt->local_ip) < 0)
 			goto parse_failed;
-		else
-			nt->ipv6 = (bool)ipv6;
 		cur = delim;
 	}
 	cur++;
@@ -2581,13 +2547,11 @@ static int netconsole_parser_cmdline(struct netconsole_target *nt, char *opt)
 	if (!delim)
 		goto parse_failed;
 	*delim = 0;
-	ipv6 = netpoll_parse_ip_addr(cur, &nt->remote_ip);
-	if (ipv6 < 0)
+	if (netpoll_parse_ip_addr(cur, &nt->remote_ip) < 0)
 		goto parse_failed;
-	else if (ipversion_set && nt->ipv6 != (bool)ipv6)
+	if (nt->local_ip.family != AF_UNSPEC &&
+	    nt->local_ip.family != nt->remote_ip.family)
 		goto parse_failed;
-	else
-		nt->ipv6 = (bool)ipv6;
 	cur = delim + 1;
 
 	if (*cur != 0) {

-- 
2.55.0

Re: [PATCH net-next 2/4] netconsole: use the address family instead of the ipv6 flag

[Breno Leitao]

On Wed, Aug 05, 2026 at 10:33:02PM +0100, Gustavo Luiz Duarte wrote:
> Now that we have the address family in inet_addr, use that and remove
> nt->ipv6.
> 
> We no longer need netcons_local_ip_unset() to check that all bytes are
> zeroes, as that is now denoted by (family == AF_UNSPEC).
> 
> Signed-off-by: Gustavo Luiz Duarte <gustavold@gmail.com>
> ---
>  drivers/net/netconsole.c | 82 ++++++++++++++----------------------------------
>  1 file changed, 23 insertions(+), 59 deletions(-)
> 
> diff --git a/drivers/net/netconsole.c b/drivers/net/netconsole.c
> index d615a9256787..070bae7b4fd7 100644
> --- a/drivers/net/netconsole.c
> +++ b/drivers/net/netconsole.c
> @@ -181,7 +181,6 @@ enum target_state {
>   *		local_mac	(read-only)
>   * @local_ip:	Source IP address of the target (read-write).
>   * @remote_ip:	Destination IP address of the target (read-write).
> - * @ipv6:	Whether the target addresses are IPv6 (read-write).
>   * @local_port:	Source UDP port of the target (read-write).
>   * @remote_port: Destination UDP port of the target (read-write).
>   * @remote_mac:	Destination ethernet address of the target (read-write).
> @@ -213,7 +212,6 @@ struct netconsole_target {
>  	bool			release;
>  	struct netpoll		np;
>  	struct inet_addr	local_ip, remote_ip;
> -	bool			ipv6;
>  	u16			local_port, remote_port;
>  	u8			remote_mac[ETH_ALEN];
>  	/* protected by target_list_lock; +1 gives scnprintf() room for its
> @@ -463,23 +461,6 @@ static int netcons_take_ipv4(struct netconsole_target *nt,
>  	return 0;
>  }
>  
> -/*
> - * Test whether the caller left nt->local_ip unset, so that
> - * netcons_netpoll_setup() should auto-populate it from the egress device.
> - *
> - * nt->local_ip is a union of __be32 (IPv4) and struct in6_addr (IPv6),
> - * so an IPv6 address whose first 4 bytes are zero (e.g. ::1, ::2,
> - * IPv4-mapped ::ffff:a.b.c.d) must not be tested via the IPv4 arm —
> - * doing so would misclassify a caller-supplied address as unset and
> - * silently overwrite it with whatever address the device exposes.
> - */
> -static bool netcons_local_ip_unset(const struct netconsole_target *nt)
> -{
> -	if (nt->ipv6)
> -		return ipv6_addr_any(&nt->local_ip.in6);
> -	return !nt->local_ip.ip;
> -}
> -
>  static int netcons_netpoll_setup(struct netconsole_target *nt)
>  {
>  	struct net *net = current->nsproxy->net_ns;
> @@ -525,16 +506,13 @@ static int netcons_netpoll_setup(struct netconsole_target *nt)
>  		rtnl_lock();
>  	}
>  
> -	if (netcons_local_ip_unset(nt)) {
> -		if (!nt->ipv6) {
> -			err = netcons_take_ipv4(nt, ndev);
> -			if (err)
> -				goto put;
> -		} else {
> +	if (nt->local_ip.family == AF_UNSPEC) {
> +		if (nt->remote_ip.family == AF_INET6)
>  			err = netcons_take_ipv6(nt, ndev);
> -			if (err)
> -				goto put;
> -		}
> +		else
> +			err = netcons_take_ipv4(nt, ndev);
> +		if (err)
> +			goto put;

nice, that is much clearer. 

>  		ip_overwritten = true;
>  	}
>  
> @@ -716,22 +694,22 @@ static void netconsole_print_banner(struct netconsole_target *nt)
>  	struct netpoll *np = &nt->np;
>  
>  	np_info(np, "local port %d\n", nt->local_port);
> -	if (nt->ipv6)
> +	if (nt->local_ip.family == AF_INET6)
>  		np_info(np, "local IPv6 address %pI6c\n", &nt->local_ip.in6);
>  	else
>  		np_info(np, "local IPv4 address %pI4\n", &nt->local_ip.ip);

Any chance it can be UNSPEC?

>  	np_info(np, "interface name '%s'\n", np->dev_name);
>  	np_info(np, "local ethernet address '%pM'\n", np->dev_mac);
>  	np_info(np, "remote port %d\n", nt->remote_port);
> -	if (nt->ipv6)
> +	if (nt->remote_ip.family == AF_INET6)
>  		np_info(np, "remote IPv6 address %pI6c\n", &nt->remote_ip.in6);
>  	else

Same question

> @@ -755,7 +733,7 @@ static int netpoll_parse_ip_addr(const char *str, struct inet_addr *addr)
>  	    in6_pton(str, len, (void *)&addr->in6, -1, &end) > 0 &&
>  	    (!end || *end == 0 || *end == '\n')) {
>  		addr->family = AF_INET6;

Why not setting family for for IPV4? In 

        if (in4_pton(str, len, (void *)addr, -1, &end) > 0 &&
            (!end || *end == 0 || *end == '\n'))
                return 0;

Re: [PATCH net-next 2/4] netconsole: use the address family instead of the ipv6 flag

[Gustavo Luiz Duarte]

On Thu, Aug 6, 2026 at 4:55 PM Breno Leitao <leitao@debian.org> wrote:
>
> On Wed, Aug 05, 2026 at 10:33:02PM +0100, Gustavo Luiz Duarte wrote:
> > Now that we have the address family in inet_addr, use that and remove
> > nt->ipv6.
> >
> > We no longer need netcons_local_ip_unset() to check that all bytes are
> > zeroes, as that is now denoted by (family == AF_UNSPEC).
> >
> > Signed-off-by: Gustavo Luiz Duarte <gustavold@gmail.com>
> > ---
> >  drivers/net/netconsole.c | 82 ++++++++++++++----------------------------------
> >  1 file changed, 23 insertions(+), 59 deletions(-)
> >
> > diff --git a/drivers/net/netconsole.c b/drivers/net/netconsole.c
> > index d615a9256787..070bae7b4fd7 100644
> > --- a/drivers/net/netconsole.c
> > +++ b/drivers/net/netconsole.c
> > @@ -181,7 +181,6 @@ enum target_state {
> >   *           local_mac       (read-only)
> >   * @local_ip:        Source IP address of the target (read-write).
> >   * @remote_ip:       Destination IP address of the target (read-write).
> > - * @ipv6:    Whether the target addresses are IPv6 (read-write).
> >   * @local_port:      Source UDP port of the target (read-write).
> >   * @remote_port: Destination UDP port of the target (read-write).
> >   * @remote_mac:      Destination ethernet address of the target (read-write).
> > @@ -213,7 +212,6 @@ struct netconsole_target {
> >       bool                    release;
> >       struct netpoll          np;
> >       struct inet_addr        local_ip, remote_ip;
> > -     bool                    ipv6;
> >       u16                     local_port, remote_port;
> >       u8                      remote_mac[ETH_ALEN];
> >       /* protected by target_list_lock; +1 gives scnprintf() room for its
> > @@ -463,23 +461,6 @@ static int netcons_take_ipv4(struct netconsole_target *nt,
> >       return 0;
> >  }
> >
> > -/*
> > - * Test whether the caller left nt->local_ip unset, so that
> > - * netcons_netpoll_setup() should auto-populate it from the egress device.
> > - *
> > - * nt->local_ip is a union of __be32 (IPv4) and struct in6_addr (IPv6),
> > - * so an IPv6 address whose first 4 bytes are zero (e.g. ::1, ::2,
> > - * IPv4-mapped ::ffff:a.b.c.d) must not be tested via the IPv4 arm —
> > - * doing so would misclassify a caller-supplied address as unset and
> > - * silently overwrite it with whatever address the device exposes.
> > - */
> > -static bool netcons_local_ip_unset(const struct netconsole_target *nt)
> > -{
> > -     if (nt->ipv6)
> > -             return ipv6_addr_any(&nt->local_ip.in6);
> > -     return !nt->local_ip.ip;
> > -}
> > -
> >  static int netcons_netpoll_setup(struct netconsole_target *nt)
> >  {
> >       struct net *net = current->nsproxy->net_ns;
> > @@ -525,16 +506,13 @@ static int netcons_netpoll_setup(struct netconsole_target *nt)
> >               rtnl_lock();
> >       }
> >
> > -     if (netcons_local_ip_unset(nt)) {
> > -             if (!nt->ipv6) {
> > -                     err = netcons_take_ipv4(nt, ndev);
> > -                     if (err)
> > -                             goto put;
> > -             } else {
> > +     if (nt->local_ip.family == AF_UNSPEC) {
> > +             if (nt->remote_ip.family == AF_INET6)
> >                       err = netcons_take_ipv6(nt, ndev);
> > -                     if (err)
> > -                             goto put;
> > -             }
> > +             else
> > +                     err = netcons_take_ipv4(nt, ndev);
> > +             if (err)
> > +                     goto put;
>
> nice, that is much clearer.
>
> >               ip_overwritten = true;
> >       }
> >
> > @@ -716,22 +694,22 @@ static void netconsole_print_banner(struct netconsole_target *nt)
> >       struct netpoll *np = &nt->np;
> >
> >       np_info(np, "local port %d\n", nt->local_port);
> > -     if (nt->ipv6)
> > +     if (nt->local_ip.family == AF_INET6)
> >               np_info(np, "local IPv6 address %pI6c\n", &nt->local_ip.in6);
> >       else
> >               np_info(np, "local IPv4 address %pI4\n", &nt->local_ip.ip);
>
> Any chance it can be UNSPEC?

Yes. Since this is a refactoring patch, I didn't want to change behavior here.
So I kept the original behavior which is print "0.0.0.0".

I was planning to send a separate patch set to change this:
a) Print an empty string if unset (AF_UNSPEC).
b) Allow the user to write an empty string to local_ip or remote_ip to
denote unset.

This can be useful in case the user is reconfiguring the target and
wants to make use of the local_ip automatic discovery logic from
netconsole.

>
> >       np_info(np, "interface name '%s'\n", np->dev_name);
> >       np_info(np, "local ethernet address '%pM'\n", np->dev_mac);
> >       np_info(np, "remote port %d\n", nt->remote_port);
> > -     if (nt->ipv6)
> > +     if (nt->remote_ip.family == AF_INET6)
> >               np_info(np, "remote IPv6 address %pI6c\n", &nt->remote_ip.in6);
> >       else
>
> Same question
>
> > @@ -755,7 +733,7 @@ static int netpoll_parse_ip_addr(const char *str, struct inet_addr *addr)
> >           in6_pton(str, len, (void *)&addr->in6, -1, &end) > 0 &&
> >           (!end || *end == 0 || *end == '\n')) {
> >               addr->family = AF_INET6;
>
> Why not setting family for for IPV4? In
>
>         if (in4_pton(str, len, (void *)addr, -1, &end) > 0 &&
>             (!end || *end == 0 || *end == '\n'))
>                 return 0;

I did that in the previous patch. Here we are just changing the return code
convention for netpoll_parse_ip_addr(). We no longer need to return the address
family because it is now baked into inet_address. So instead of returning 0 for
ipv4, 1 for ipv6, and -1 for failure, we just return -1 for failure and 0 for
success.

Re: [PATCH net-next 2/4] netconsole: use the address family instead of the ipv6 flag

[Breno Leitao]

On Fri, Aug 07, 2026 at 01:00:51PM +0100, Gustavo Luiz Duarte wrote:
> On Thu, Aug 6, 2026 at 4:55 PM Breno Leitao <leitao@debian.org> wrote:
> >
> > On Wed, Aug 05, 2026 at 10:33:02PM +0100, Gustavo Luiz Duarte wrote:
> > > Now that we have the address family in inet_addr, use that and remove
> > > nt->ipv6.
> > >
> > > We no longer need netcons_local_ip_unset() to check that all bytes are
> > > zeroes, as that is now denoted by (family == AF_UNSPEC).
> > >
> > > Signed-off-by: Gustavo Luiz Duarte <gustavold@gmail.com>
> > > ---
> > >  drivers/net/netconsole.c | 82 ++++++++++++++----------------------------------
> > >  1 file changed, 23 insertions(+), 59 deletions(-)
> > >
> > > diff --git a/drivers/net/netconsole.c b/drivers/net/netconsole.c
> > > index d615a9256787..070bae7b4fd7 100644
> > > --- a/drivers/net/netconsole.c
> > > +++ b/drivers/net/netconsole.c
> > > @@ -181,7 +181,6 @@ enum target_state {
> > >   *           local_mac       (read-only)
> > >   * @local_ip:        Source IP address of the target (read-write).
> > >   * @remote_ip:       Destination IP address of the target (read-write).
> > > - * @ipv6:    Whether the target addresses are IPv6 (read-write).
> > >   * @local_port:      Source UDP port of the target (read-write).
> > >   * @remote_port: Destination UDP port of the target (read-write).
> > >   * @remote_mac:      Destination ethernet address of the target (read-write).
> > > @@ -213,7 +212,6 @@ struct netconsole_target {
> > >       bool                    release;
> > >       struct netpoll          np;
> > >       struct inet_addr        local_ip, remote_ip;
> > > -     bool                    ipv6;
> > >       u16                     local_port, remote_port;
> > >       u8                      remote_mac[ETH_ALEN];
> > >       /* protected by target_list_lock; +1 gives scnprintf() room for its
> > > @@ -463,23 +461,6 @@ static int netcons_take_ipv4(struct netconsole_target *nt,
> > >       return 0;
> > >  }
> > >
> > > -/*
> > > - * Test whether the caller left nt->local_ip unset, so that
> > > - * netcons_netpoll_setup() should auto-populate it from the egress device.
> > > - *
> > > - * nt->local_ip is a union of __be32 (IPv4) and struct in6_addr (IPv6),
> > > - * so an IPv6 address whose first 4 bytes are zero (e.g. ::1, ::2,
> > > - * IPv4-mapped ::ffff:a.b.c.d) must not be tested via the IPv4 arm —
> > > - * doing so would misclassify a caller-supplied address as unset and
> > > - * silently overwrite it with whatever address the device exposes.
> > > - */
> > > -static bool netcons_local_ip_unset(const struct netconsole_target *nt)
> > > -{
> > > -     if (nt->ipv6)
> > > -             return ipv6_addr_any(&nt->local_ip.in6);
> > > -     return !nt->local_ip.ip;
> > > -}
> > > -
> > >  static int netcons_netpoll_setup(struct netconsole_target *nt)
> > >  {
> > >       struct net *net = current->nsproxy->net_ns;
> > > @@ -525,16 +506,13 @@ static int netcons_netpoll_setup(struct netconsole_target *nt)
> > >               rtnl_lock();
> > >       }
> > >
> > > -     if (netcons_local_ip_unset(nt)) {
> > > -             if (!nt->ipv6) {
> > > -                     err = netcons_take_ipv4(nt, ndev);
> > > -                     if (err)
> > > -                             goto put;
> > > -             } else {
> > > +     if (nt->local_ip.family == AF_UNSPEC) {
> > > +             if (nt->remote_ip.family == AF_INET6)
> > >                       err = netcons_take_ipv6(nt, ndev);
> > > -                     if (err)
> > > -                             goto put;
> > > -             }
> > > +             else
> > > +                     err = netcons_take_ipv4(nt, ndev);
> > > +             if (err)
> > > +                     goto put;
> >
> > nice, that is much clearer.
> >
> > >               ip_overwritten = true;
> > >       }
> > >
> > > @@ -716,22 +694,22 @@ static void netconsole_print_banner(struct netconsole_target *nt)
> > >       struct netpoll *np = &nt->np;
> > >
> > >       np_info(np, "local port %d\n", nt->local_port);
> > > -     if (nt->ipv6)
> > > +     if (nt->local_ip.family == AF_INET6)
> > >               np_info(np, "local IPv6 address %pI6c\n", &nt->local_ip.in6);
> > >       else
> > >               np_info(np, "local IPv4 address %pI4\n", &nt->local_ip.ip);
> >
> > Any chance it can be UNSPEC?
> 
> Yes. Since this is a refactoring patch, I didn't want to change behavior here.
> So I kept the original behavior which is print "0.0.0.0".
> 
> I was planning to send a separate patch set to change this:
> a) Print an empty string if unset (AF_UNSPEC).

Please add it to this patchset. 
> b) Allow the user to write an empty string to local_ip or remote_ip to
> denote unset.
> 
> This can be useful in case the user is reconfiguring the target and
> wants to make use of the local_ip automatic discovery logic from
> netconsole.

Cool, this could come later.

> > > @@ -755,7 +733,7 @@ static int netpoll_parse_ip_addr(const char *str, struct inet_addr *addr)
> > >           in6_pton(str, len, (void *)&addr->in6, -1, &end) > 0 &&
> > >           (!end || *end == 0 || *end == '\n')) {
> > >               addr->family = AF_INET6;
> >
> > Why not setting family for for IPV4? In
> >
> >         if (in4_pton(str, len, (void *)addr, -1, &end) > 0 &&
> >             (!end || *end == 0 || *end == '\n'))
> >                 return 0;
> 
> I did that in the previous patch. Here we are just changing the return code
> convention for netpoll_parse_ip_addr(). We no longer need to return the address
> family because it is now baked into inet_address. So instead of returning 0 for
> ipv4, 1 for ipv6, and -1 for failure, we just return -1 for failure and 0 for
> success.

Would it make sense to change it here, together and close to the
detection?

[PATCH net-next 3/4] netconsole: reject enabling a target with no remote IP address

[Gustavo Luiz Duarte]

The command-line path already requires a remote address, but if a user
creates a dynamic target and enables it without setting a remote
address, we currently try sending netconsole traffic to "0.0.0.0".

Refuse to enable a target if the remote address is unset.

Signed-off-by: Gustavo Luiz Duarte <gustavold@gmail.com>
---
 drivers/net/netconsole.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/net/netconsole.c b/drivers/net/netconsole.c
index 070bae7b4fd7..a41b0a5ce64f 100644
--- a/drivers/net/netconsole.c
+++ b/drivers/net/netconsole.c
@@ -471,6 +471,12 @@ static int netcons_netpoll_setup(struct netconsole_target *nt)
 	int err;
 
 	rtnl_lock();
+	if (nt->remote_ip.family == AF_UNSPEC) {
+		np_err(np, "remote IP address not configured, aborting\n");
+		err = -EDESTADDRREQ;
+		goto unlock;
+	}
+
 	if (np->dev_name[0])
 		ndev = __dev_get_by_name(net, np->dev_name);
 	else if (is_valid_ether_addr(np->dev_mac))

-- 
2.55.0

Re: [PATCH net-next 3/4] netconsole: reject enabling a target with no remote IP address

[Breno Leitao]

On Wed, Aug 05, 2026 at 10:33:03PM +0100, Gustavo Luiz Duarte wrote:
> The command-line path already requires a remote address, but if a user
> creates a dynamic target and enables it without setting a remote
> address, we currently try sending netconsole traffic to "0.0.0.0".
> 
> Refuse to enable a target if the remote address is unset.
> 
> Signed-off-by: Gustavo Luiz Duarte <gustavold@gmail.com>
> ---
>  drivers/net/netconsole.c | 6 ++++++
>  1 file changed, 6 insertions(+)
> 
> diff --git a/drivers/net/netconsole.c b/drivers/net/netconsole.c
> index 070bae7b4fd7..a41b0a5ce64f 100644
> --- a/drivers/net/netconsole.c
> +++ b/drivers/net/netconsole.c
> @@ -471,6 +471,12 @@ static int netcons_netpoll_setup(struct netconsole_target *nt)
>  	int err;
>  
>  	rtnl_lock();
> +	if (nt->remote_ip.family == AF_UNSPEC) {
> +		np_err(np, "remote IP address not configured, aborting\n");
> +		err = -EDESTADDRREQ;
> +		goto unlock;
> +	}

Why not doing it  before the rtnl lock?

Re: [PATCH net-next 3/4] netconsole: reject enabling a target with no remote IP address

[Gustavo Luiz Duarte]

On Thu, Aug 6, 2026 at 4:58 PM Breno Leitao <leitao@debian.org> wrote:
>
> On Wed, Aug 05, 2026 at 10:33:03PM +0100, Gustavo Luiz Duarte wrote:
> > The command-line path already requires a remote address, but if a user
> > creates a dynamic target and enables it without setting a remote
> > address, we currently try sending netconsole traffic to "0.0.0.0".
> >
> > Refuse to enable a target if the remote address is unset.
> >
> > Signed-off-by: Gustavo Luiz Duarte <gustavold@gmail.com>
> > ---
> >  drivers/net/netconsole.c | 6 ++++++
> >  1 file changed, 6 insertions(+)
> >
> > diff --git a/drivers/net/netconsole.c b/drivers/net/netconsole.c
> > index 070bae7b4fd7..a41b0a5ce64f 100644
> > --- a/drivers/net/netconsole.c
> > +++ b/drivers/net/netconsole.c
> > @@ -471,6 +471,12 @@ static int netcons_netpoll_setup(struct netconsole_target *nt)
> >       int err;
> >
> >       rtnl_lock();
> > +     if (nt->remote_ip.family == AF_UNSPEC) {
> > +             np_err(np, "remote IP address not configured, aborting\n");
> > +             err = -EDESTADDRREQ;
> > +             goto unlock;
> > +     }
>
> Why not doing it  before the rtnl lock?

Oh, you are right, there is no reason for holding the lock here.
Thanks for pointing it out.
I will fix it in v2.

[PATCH net-next 4/4] netconsole: reject a target mixing IPv4 and IPv6 addresses

[Gustavo Luiz Duarte]

The local_ip and remote_ip configfs attributes are written independently
and nothing stops a user from mixing ipv4 and ipv6. This leads to an
ipv4 address being zero-extended into an ipv6 header or an ipv6 address
being truncated to its first 4 bytes for an ipv4 header.

The command-line parser already refuses such a mismatch. This adds a
similar check when a dynamic target is being enabled.

Signed-off-by: Gustavo Luiz Duarte <gustavold@gmail.com>
---
 drivers/net/netconsole.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/drivers/net/netconsole.c b/drivers/net/netconsole.c
index a41b0a5ce64f..69621705d68d 100644
--- a/drivers/net/netconsole.c
+++ b/drivers/net/netconsole.c
@@ -477,6 +477,13 @@ static int netcons_netpoll_setup(struct netconsole_target *nt)
 		goto unlock;
 	}
 
+	if (nt->local_ip.family != AF_UNSPEC &&
+	    nt->local_ip.family != nt->remote_ip.family) {
+		np_err(np, "local and remote IP address families differ, aborting\n");
+		err = -EINVAL;
+		goto unlock;
+	}
+
 	if (np->dev_name[0])
 		ndev = __dev_get_by_name(net, np->dev_name);
 	else if (is_valid_ether_addr(np->dev_mac))

-- 
2.55.0