[PATCH net-next] ipv6: eliminate ndisc_ops_is_useropt()

[PATCH net-next] ipv6: eliminate ndisc_ops_is_useropt()

[Maciej Żenczykowski]

as it doesn't seem to offer anything of value.

There's only 1 trivial user:
  int lowpan_ndisc_is_useropt(u8 nd_opt_type) {
    return nd_opt_type == ND_OPT_6CO;
  }

but there's no harm to always treating that as
a useropt...

Cc: David Ahern <dsahern@kernel.org>
Cc: YOSHIFUJI Hideaki / 吉藤英明 <yoshfuji@linux-ipv6.org>
Cc: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Maciej Żenczykowski <maze@google.com>
---
 include/net/ndisc.h | 15 ---------------
 net/6lowpan/ndisc.c |  6 ------
 net/ipv6/ndisc.c    |  4 ++--
 3 files changed, 2 insertions(+), 23 deletions(-)

diff --git a/include/net/ndisc.h b/include/net/ndisc.h
index 7a533d5b1d59..3c88d5bc5eed 100644
--- a/include/net/ndisc.h
+++ b/include/net/ndisc.h
@@ -147,11 +147,6 @@ void __ndisc_fill_addr_option(struct sk_buff *skb, int type, const void *data,
  * The following hooks can be defined; unless noted otherwise, they are
  * optional and can be filled with a null pointer.
  *
- * int (*is_useropt)(u8 nd_opt_type):
- *     This function is called when IPv6 decide RA userspace options. if
- *     this function returns 1 then the option given by nd_opt_type will
- *     be handled as userspace option additional to the IPv6 options.
- *
  * int (*parse_options)(const struct net_device *dev,
  *			struct nd_opt_hdr *nd_opt,
  *			struct ndisc_options *ndopts):
@@ -200,7 +195,6 @@ void __ndisc_fill_addr_option(struct sk_buff *skb, int type, const void *data,
  *     addresses. E.g. 802.15.4 6LoWPAN.
  */
 struct ndisc_ops {
-	int	(*is_useropt)(u8 nd_opt_type);
 	int	(*parse_options)(const struct net_device *dev,
 				 struct nd_opt_hdr *nd_opt,
 				 struct ndisc_options *ndopts);
@@ -224,15 +218,6 @@ struct ndisc_ops {
 };
 
 #if IS_ENABLED(CONFIG_IPV6)
-static inline int ndisc_ops_is_useropt(const struct net_device *dev,
-				       u8 nd_opt_type)
-{
-	if (dev->ndisc_ops && dev->ndisc_ops->is_useropt)
-		return dev->ndisc_ops->is_useropt(nd_opt_type);
-	else
-		return 0;
-}
-
 static inline int ndisc_ops_parse_options(const struct net_device *dev,
 					  struct nd_opt_hdr *nd_opt,
 					  struct ndisc_options *ndopts)
diff --git a/net/6lowpan/ndisc.c b/net/6lowpan/ndisc.c
index 16be8f8b2f8c..c40b98f7743c 100644
--- a/net/6lowpan/ndisc.c
+++ b/net/6lowpan/ndisc.c
@@ -11,11 +11,6 @@
 
 #include "6lowpan_i.h"
 
-static int lowpan_ndisc_is_useropt(u8 nd_opt_type)
-{
-	return nd_opt_type == ND_OPT_6CO;
-}
-
 #if IS_ENABLED(CONFIG_IEEE802154_6LOWPAN)
 #define NDISC_802154_SHORT_ADDR_LENGTH	1
 static int lowpan_ndisc_parse_802154_options(const struct net_device *dev,
@@ -222,7 +217,6 @@ static void lowpan_ndisc_prefix_rcv_add_addr(struct net *net,
 #endif
 
 const struct ndisc_ops lowpan_ndisc_ops = {
-	.is_useropt		= lowpan_ndisc_is_useropt,
 #if IS_ENABLED(CONFIG_IEEE802154_6LOWPAN)
 	.parse_options		= lowpan_ndisc_parse_options,
 	.update			= lowpan_ndisc_update,
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index b8eec1b6cc2c..1e42e40fb379 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -200,9 +200,9 @@ static inline int ndisc_is_useropt(const struct net_device *dev,
 	return opt->nd_opt_type == ND_OPT_PREFIX_INFO ||
 		opt->nd_opt_type == ND_OPT_RDNSS ||
 		opt->nd_opt_type == ND_OPT_DNSSL ||
+		opt->nd_opt_type == ND_OPT_6CO ||
 		opt->nd_opt_type == ND_OPT_CAPTIVE_PORTAL ||
-		opt->nd_opt_type == ND_OPT_PREF64 ||
-		ndisc_ops_is_useropt(dev, opt->nd_opt_type);
+		opt->nd_opt_type == ND_OPT_PREF64;
 }
 
 static struct nd_opt_hdr *ndisc_next_useropt(const struct net_device *dev,
-- 
2.46.0.rc1.232.g9752f9e123-goog

Re: [PATCH net-next] ipv6: eliminate ndisc_ops_is_useropt()

[Paolo Abeni]

On 7/30/24 02:30, Maciej Żenczykowski wrote:
> as it doesn't seem to offer anything of value.
> 
> There's only 1 trivial user:
>    int lowpan_ndisc_is_useropt(u8 nd_opt_type) {
>      return nd_opt_type == ND_OPT_6CO;
>    }
> 
> but there's no harm to always treating that as
> a useropt...

AFAICS there will be an user-visible difference, as the user-space could 
start receiving "unexpected" notification for such opt even when the 
kernel is built with CONFIG_6LOWPAN=n.

The user-space should ignore unknown/unexpected notification,so it 
should be safe, but I'm a bit unsettled by the many 'should' ;)

Cheers,

Paolo

Re: [PATCH net-next] ipv6: eliminate ndisc_ops_is_useropt()

[Maciej Żenczykowski]

On Thu, Aug 1, 2024 at 6:54 AM Paolo Abeni <pabeni@redhat.com> wrote:
>
> On 7/30/24 02:30, Maciej Żenczykowski wrote:
> > as it doesn't seem to offer anything of value.
> >
> > There's only 1 trivial user:
> >    int lowpan_ndisc_is_useropt(u8 nd_opt_type) {
> >      return nd_opt_type == ND_OPT_6CO;
> >    }
> >
> > but there's no harm to always treating that as
> > a useropt...
>
> AFAICS there will be an user-visible difference, as the user-space could
> start receiving "unexpected" notification for such opt even when the
> kernel is built with CONFIG_6LOWPAN=n.
>
> The user-space should ignore unknown/unexpected notification,so it
> should be safe, but I'm a bit unsettled by the many 'should' ;)

We've added new options to this list in the past.  Even going so far
as shipping those changes via LTS... So IMHO this doesn't seem like a
real concern.  Also the option is very unlikely to be present in
normal non 6lowpan environments.

> Cheers,
>
> Paolo
>

Re: [PATCH net-next] ipv6: eliminate ndisc_ops_is_useropt()

[patchwork-bot+netdevbpf]

Hello:

This patch was applied to netdev/net-next.git (main)
by Jakub Kicinski <kuba@kernel.org>:

On Mon, 29 Jul 2024 17:30:10 -0700 you wrote:
> as it doesn't seem to offer anything of value.
> 
> There's only 1 trivial user:
>   int lowpan_ndisc_is_useropt(u8 nd_opt_type) {
>     return nd_opt_type == ND_OPT_6CO;
>   }
> 
> [...]

Here is the summary with links:
  - [net-next] ipv6: eliminate ndisc_ops_is_useropt()
    https://git.kernel.org/netdev/net-next/c/246ef40670b7

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html