LinuxLists.cc - [PATCH 0/1] Create multicast snooping sysctl option

2021-06-17 12:05:15

Subject: [PATCH 0/1] Create multicast snooping sysctl option

IGMP and MLD packets can be received on IP sockets but only if the
group has been explicitly joined. This makes snooping all multicast
packets in the ranges 224.0.0.0/8 and FF00::/8 difficult as each
group in these ranges would have to be added via socketopt calls
individually.

It is possible to get all IGMP and MLD packets by creating a packet
socket and using a BPF to grab only IGMP and MLD packets but this
removes some of the other useful options IP sockets have.

Define a new sysctl to allow one or more interfaces to be defined
as a IGMP and/or MLD Snooping device. This means all multicast
packets will be received on a socket bound to the interface without
being explicitly asked for.

Callum Sinclair (1):
net: Allow all multicast packets to be received on a interface.

Documentation/networking/ip-sysctl.rst | 8 ++++++++
include/linux/inetdevice.h | 1 +
include/linux/ipv6.h | 1 +
include/uapi/linux/ip.h | 1 +
include/uapi/linux/ipv6.h | 1 +
include/uapi/linux/netconf.h | 1 +
include/uapi/linux/sysctl.h | 1 +
net/ipv4/devinet.c | 7 +++++++
net/ipv4/igmp.c | 5 +++++
net/ipv6/addrconf.c | 14 ++++++++++++++
net/ipv6/mcast.c | 5 +++++
11 files changed, 45 insertions(+)

--
2.32.0

2021-06-17 12:07:27

by Callum Sinclair

[permalink] [raw]

Subject: [PATCH 1/1] net: Allow all multicast packets to be received on a interface.

To receive IGMP or MLD packets on a IP socket on any interface the
multicast group needs to be explicitly joined. This works well for when
the multicast group the user is interested in is known, but does not
provide an easy way to snoop all packets in the 224.0.0.0/8 or the
FF00::/8 range.

Define a new sysctl to allow a given interface to become a IGMP or MLD
snooper. When set the interface will allow any IGMP or MLD packet to be
received on sockets bound to these devices.
---
Documentation/networking/ip-sysctl.rst | 8 ++++++++
include/linux/inetdevice.h | 1 +
include/linux/ipv6.h | 1 +
include/uapi/linux/ip.h | 1 +
include/uapi/linux/ipv6.h | 1 +
include/uapi/linux/netconf.h | 1 +
include/uapi/linux/sysctl.h | 1 +
net/ipv4/devinet.c | 7 +++++++
net/ipv4/igmp.c | 5 +++++
net/ipv6/addrconf.c | 14 ++++++++++++++
net/ipv6/mcast.c | 5 +++++
11 files changed, 45 insertions(+)

diff --git a/Documentation/networking/ip-sysctl.rst b/Documentation/networking/ip-sysctl.rst
index a5c250044500..12f82da52684 100644
--- a/Documentation/networking/ip-sysctl.rst
+++ b/Documentation/networking/ip-sysctl.rst
@@ -1357,6 +1357,14 @@ mc_forwarding - BOOLEAN
conf/all/mc_forwarding must also be set to TRUE to enable multicast
routing for the interface

+mc_snooping - BOOLEAN
+ Enable multicast snooping on the interface. This allows any given
+ multicast group to be received without explicitly being joined.
+ The kernel needs to be compiled with CONFIG_MROUTE and/or
+ CONFIG_IPV6_MROUTE.
+ conf/all/mc_snooping must also be set to TRUE to enable multicast
+ snooping for the interface.
+
medium_id - INTEGER
Integer value used to differentiate the devices by the medium they
are attached to. Two devices can have different id values when
diff --git a/include/linux/inetdevice.h b/include/linux/inetdevice.h
index 53aa0343bf69..071edf7d4f9c 100644
--- a/include/linux/inetdevice.h
+++ b/include/linux/inetdevice.h
@@ -95,6 +95,7 @@ static inline void ipv4_devconf_setall(struct in_device *in_dev)

#define IN_DEV_FORWARD(in_dev) IN_DEV_CONF_GET((in_dev), FORWARDING)
#define IN_DEV_MFORWARD(in_dev) IN_DEV_ANDCONF((in_dev), MC_FORWARDING)
+#define IN_DEV_MSNOOPING(in_dev) IN_DEV_ANDCONF((in_dev), MC_SNOOPING)
#define IN_DEV_BFORWARD(in_dev) IN_DEV_ANDCONF((in_dev), BC_FORWARDING)
#define IN_DEV_RPFILTER(in_dev) IN_DEV_MAXCONF((in_dev), RP_FILTER)
#define IN_DEV_SRC_VMARK(in_dev) IN_DEV_ORCONF((in_dev), SRC_VMARK)
diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h
index 70b2ad3b9884..d88c34b1b3ae 100644
--- a/include/linux/ipv6.h
+++ b/include/linux/ipv6.h
@@ -52,6 +52,7 @@ struct ipv6_devconf {
#endif
#ifdef CONFIG_IPV6_MROUTE
__s32 mc_forwarding;
+ __s32 mc_snooping;
#endif
__s32 disable_ipv6;
__s32 drop_unicast_in_l2_multicast;
diff --git a/include/uapi/linux/ip.h b/include/uapi/linux/ip.h
index e42d13b55cf3..07956b4613d0 100644
--- a/include/uapi/linux/ip.h
+++ b/include/uapi/linux/ip.h
@@ -169,6 +169,7 @@ enum
IPV4_DEVCONF_DROP_UNICAST_IN_L2_MULTICAST,
IPV4_DEVCONF_DROP_GRATUITOUS_ARP,
IPV4_DEVCONF_BC_FORWARDING,
+ IPV4_DEVCONF_MC_SNOOPING,
__IPV4_DEVCONF_MAX
};

diff --git a/include/uapi/linux/ipv6.h b/include/uapi/linux/ipv6.h
index 70603775fe91..aa9389e1c1fd 100644
--- a/include/uapi/linux/ipv6.h
+++ b/include/uapi/linux/ipv6.h
@@ -190,6 +190,7 @@ enum {
DEVCONF_NDISC_TCLASS,
DEVCONF_RPL_SEG_ENABLED,
DEVCONF_RA_DEFRTR_METRIC,
+ DEVCONF_MC_SNOOPING,
DEVCONF_MAX
};

diff --git a/include/uapi/linux/netconf.h b/include/uapi/linux/netconf.h
index fac4edd55379..5259742a700b 100644
--- a/include/uapi/linux/netconf.h
+++ b/include/uapi/linux/netconf.h
@@ -19,6 +19,7 @@ enum {
NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN,
NETCONFA_INPUT,
NETCONFA_BC_FORWARDING,
+ NETCONFA_MC_SNOOPING,
__NETCONFA_MAX
};
#define NETCONFA_MAX (__NETCONFA_MAX - 1)
diff --git a/include/uapi/linux/sysctl.h b/include/uapi/linux/sysctl.h
index 1e05d3caa712..1b7be9dc78de 100644
--- a/include/uapi/linux/sysctl.h
+++ b/include/uapi/linux/sysctl.h
@@ -482,6 +482,7 @@ enum
NET_IPV4_CONF_PROMOTE_SECONDARIES=20,
NET_IPV4_CONF_ARP_ACCEPT=21,
NET_IPV4_CONF_ARP_NOTIFY=22,
+ NET_IPV4_CONF_MC_SNOOPING=23,
};

/* /proc/sys/net/ipv4/netfilter */
diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c
index 50deeff48c8b..3e4ac6aead9d 100644
--- a/net/ipv4/devinet.c
+++ b/net/ipv4/devinet.c
@@ -2014,6 +2014,8 @@ static int inet_netconf_msgsize_devconf(int type)
size += nla_total_size(4);
if (all || type == NETCONFA_MC_FORWARDING)
size += nla_total_size(4);
+ if (all || type == NETCONFA_MC_SNOOPING)
+ size += nla_total_size(4);
if (all || type == NETCONFA_BC_FORWARDING)
size += nla_total_size(4);
if (all || type == NETCONFA_PROXY_NEIGH)
@@ -2062,6 +2064,10 @@ static int inet_netconf_fill_devconf(struct sk_buff *skb, int ifindex,
nla_put_s32(skb, NETCONFA_MC_FORWARDING,
IPV4_DEVCONF(*devconf, MC_FORWARDING)) < 0)
goto nla_put_failure;
+ if ((all || type == NETCONFA_MC_SNOOPING) &&
+ nla_put_s32(skb, NETCONFA_MC_SNOOPING,
+ IPV4_DEVCONF(*devconf, NETCONFA_MC_SNOOPING)) < 0)
+ goto nla_put_failure;
if ((all || type == NETCONFA_BC_FORWARDING) &&
nla_put_s32(skb, NETCONFA_BC_FORWARDING,
IPV4_DEVCONF(*devconf, BC_FORWARDING)) < 0)
@@ -2506,6 +2512,7 @@ static struct devinet_sysctl_table {
DEVINET_SYSCTL_COMPLEX_ENTRY(FORWARDING, "forwarding",
devinet_sysctl_forward),
DEVINET_SYSCTL_RO_ENTRY(MC_FORWARDING, "mc_forwarding"),
+ DEVINET_SYSCTL_RW_ENTRY(MC_SNOOPING, "mc_snooping"),
DEVINET_SYSCTL_RW_ENTRY(BC_FORWARDING, "bc_forwarding"),

DEVINET_SYSCTL_RW_ENTRY(ACCEPT_REDIRECTS, "accept_redirects"),
diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c
index 7b272bbed2b4..cd5a837dfb0c 100644
--- a/net/ipv4/igmp.c
+++ b/net/ipv4/igmp.c
@@ -2692,6 +2692,11 @@ int ip_check_mc_rcu(struct in_device *in_dev, __be32 mc_addr, __be32 src_addr, u
struct ip_sf_list *psf;
int rv = 0;

+#ifdef CONFIG_IP_MROUTE
+ if (IN_DEV_MSNOOPING(in_dev))
+ return 1;
+#endif /* CONFIG_IP_MROUTE */
+
mc_hash = rcu_dereference(in_dev->mc_hash);
if (mc_hash) {
u32 hash = hash_32((__force u32)mc_addr, MC_HASH_SZ_LOG);
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 048570900fdf..b92ac4e8f37d 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -502,6 +502,8 @@ static int inet6_netconf_msgsize_devconf(int type)
#ifdef CONFIG_IPV6_MROUTE
if (all || type == NETCONFA_MC_FORWARDING)
size += nla_total_size(4);
+ if (all || type == NETCONFA_MC_SNOOPING)
+ size += nla_total_size(4);
#endif
if (all || type == NETCONFA_PROXY_NEIGH)
size += nla_total_size(4);
@@ -546,6 +548,10 @@ static int inet6_netconf_fill_devconf(struct sk_buff *skb, int ifindex,
nla_put_s32(skb, NETCONFA_MC_FORWARDING,
devconf->mc_forwarding) < 0)
goto nla_put_failure;
+ if ((all || type == NETCONFA_MC_SNOOPING) &&
+ nla_put_s32(skb, NETCONFA_MC_SNOOPING,
+ devconf->mc_snooping) < 0)
+ goto nla_put_failure;
#endif
if ((all || type == NETCONFA_PROXY_NEIGH) &&
nla_put_s32(skb, NETCONFA_PROXY_NEIGH, devconf->proxy_ndp) < 0)
@@ -5503,6 +5509,7 @@ static inline void ipv6_store_devconf(struct ipv6_devconf *cnf,
#endif
#ifdef CONFIG_IPV6_MROUTE
array[DEVCONF_MC_FORWARDING] = cnf->mc_forwarding;
+ array[DEVCONF_MC_SNOOPING] = cnf->mc_snooping;
#endif
array[DEVCONF_DISABLE_IPV6] = cnf->disable_ipv6;
array[DEVCONF_ACCEPT_DAD] = cnf->accept_dad;
@@ -6786,6 +6793,13 @@ static const struct ctl_table addrconf_sysctl[] = {
.mode = 0444,
.proc_handler = proc_dointvec,
},
+ {
+ .procname = "mc_snooping",
+ .data = &ipv6_devconf.mc_snooping,
+ .maxlen = sizeof(int),
+ .mode = 0644,
+ .proc_handler = proc_dointvec,
+ },
#endif
{
.procname = "disable_ipv6",
diff --git a/net/ipv6/mcast.c b/net/ipv6/mcast.c
index 54ec163fbafa..25046ee8276f 100644
--- a/net/ipv6/mcast.c
+++ b/net/ipv6/mcast.c
@@ -1013,6 +1013,11 @@ bool ipv6_chk_mcast_addr(struct net_device *dev, const struct in6_addr *group,
struct ifmcaddr6 *mc;
bool rv = false;

+#ifdef CONFIG_IPV6_MROUTE
+ if (dev_net(dev)->ipv6.devconf_all->mc_snooping)
+ return true;
+#endif
+
rcu_read_lock();
idev = __in6_dev_get(dev);
if (idev) {
--
2.32.0

2021-06-17 14:00:34

by Linus Lüssing

[permalink] [raw]

Subject: Re: [PATCH 1/1] net: Allow all multicast packets to be received on a interface.

Hi Callum,

On Thu, Jun 17, 2021 at 09:50:20PM +1200, Callum Sinclair wrote:
> +mc_snooping - BOOLEAN
> + Enable multicast snooping on the interface. This allows any given
> + multicast group to be received without explicitly being joined.
> + The kernel needs to be compiled with CONFIG_MROUTE and/or
> + CONFIG_IPV6_MROUTE.
> + conf/all/mc_snooping must also be set to TRUE to enable multicast
> + snooping for the interface.
> +

Generally this sounds like a useful feature. One note: When there
are snooping bridges/switches involved, you might run into issues
in receiving all multicast packets, as due to the missing IGMP/MLD
reports the snooping switches won't forward to you.

In that case, to conform to RFC4541 you would also need to become
the selected IGMP/MLD querier and send IGMP/MLD query messages. Or
better, you'd need to send Multicast Router Advertisements
(RFC4286). The latter is the recommended, more flexible way but
might not be supported by all multicast snooping switches yet.
The Linux bridge supports this.

There is a userspace tool called mrdisc you can use for MRD-Adv.
though: https://github.com/troglobit/mrdisc. So no need to
implement MRD Advertisements in the kernel with this patch (though
I could imagine that it might be a useful feature to have, having
MRD support out-of-the-box with this option). Just a note that some
IGMP/MLD Querier or MRD Adv. would be needed when IGMP/MLD snooping
switches are invoved might be nice to have in the mc_snooping
description for now, to avoid potential confusion later.

I'm also wondering if it could be useful to configure it via
setsockopt() per application instead of per device via sysctl. Either by
adding a new socket option. Or by allowing the any IP address
0.0.0.0 / :: with IP_ADD_MEMBERSHIP/IPV6_JOIN_GROUP. So that you
could use this for instance:

$ socat -u UDP6-RECV:1234,reuseaddr,ipv6-join-group="[::]:eth0" -
(currently :: fails with "Invalid argument")

I'm not sure however what the requirements for adding or extending
socket options are, if there are some POSIX standards that'd need
to be followed for compatibility with other OSes, for instance.

Hm, actually, I just noticed that there seem to be some multicast
related setsockopt()s already:

- PACKET_MR_PROMISC
- PACKET_MR_MULTICAST
- PACKET_MR_ALLMULTI

The last one seems to be what you are looking for, I think, the
manpage here says:

"PACKET_MR_ALLMULTI sets the socket up to receive all multicast
packets arriving at the interface"
https://www.man7.org/linux/man-pages/man7/packet.7.html

Or would you prefer to be able to use a layer 3 IP instead of
a layer 2 packet socket?

Regards, Linus

2021-06-17 18:02:38

by Andrew Lunn

[permalink] [raw]

Subject: Re: [PATCH 1/1] net: Allow all multicast packets to be received on a interface.

On Thu, Jun 17, 2021 at 09:50:20PM +1200, Callum Sinclair wrote:
> To receive IGMP or MLD packets on a IP socket on any interface the
> multicast group needs to be explicitly joined. This works well for when
> the multicast group the user is interested in is known, but does not
> provide an easy way to snoop all packets in the 224.0.0.0/8 or the
> FF00::/8 range.
>
> Define a new sysctl to allow a given interface to become a IGMP or MLD
> snooper. When set the interface will allow any IGMP or MLD packet to be
> received on sockets bound to these devices.

Hi Callum

What is the big picture here? Are you trying to move the snooping
algorithm into user space? User space will then add/remove Multicast
FIB entries to the bridge to control where mulitcast frames are sent?

In the past i have written a multicast routing daemon. It is a similar
problem. You need access to all the join/leaves. But the stack does
provide them, if you bind to the multicast routing socket. Why not use
that mechanism? Look in the mrouted sources for an example.

Andrew

2021-06-17 21:24:22

by kernel test robot

[permalink] [raw]

Subject: Re: [PATCH 1/1] net: Allow all multicast packets to be received on a interface.

Hi Callum,

Thank you for the patch! Yet something to improve:

[auto build test ERROR on linus/master]
[also build test ERROR on v5.13-rc6 next-20210617]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url: https://github.com/0day-ci/linux/commits/Callum-Sinclair/Create-multicast-snooping-sysctl-option/20210617-175212
base: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git 70585216fe7730d9fb5453d3e2804e149d0fe201
config: x86_64-randconfig-r024-20210617 (attached as .config)
compiler: clang version 13.0.0 (https://github.com/llvm/llvm-project 64720f57bea6a6bf033feef4a5751ab9c0c3b401)
reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
chmod +x ~/bin/make.cross
# install x86_64 cross compiling tool for clang build
# apt-get install binutils-x86-64-linux-gnu
# https://github.com/0day-ci/linux/commit/4220b6837f4315ff557bd44f7aada23b69e181b6
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Callum-Sinclair/Create-multicast-snooping-sysctl-option/20210617-175212
git checkout 4220b6837f4315ff557bd44f7aada23b69e181b6
# save the attached .config to linux build tree
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=x86_64

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <[email protected]>

All errors (new ones prefixed by >>):

>> net/ipv4/devinet.c:2069:4: error: use of undeclared identifier 'IPV4_DEVCONF_NETCONFA_MC_SNOOPING'; did you mean 'IPV4_DEVCONF_MC_SNOOPING'?
IPV4_DEVCONF(*devconf, NETCONFA_MC_SNOOPING)) < 0)
^
include/linux/inetdevice.h:53:45: note: expanded from macro 'IPV4_DEVCONF'
#define IPV4_DEVCONF(cnf, attr) ((cnf).data[IPV4_DEVCONF_ ## attr - 1])
^
<scratch space>:121:1: note: expanded from here
IPV4_DEVCONF_NETCONFA_MC_SNOOPING
^
include/uapi/linux/ip.h:172:2: note: 'IPV4_DEVCONF_MC_SNOOPING' declared here
IPV4_DEVCONF_MC_SNOOPING,
^
1 error generated.

vim +2069 net/ipv4/devinet.c

2028
2029 static int inet_netconf_fill_devconf(struct sk_buff *skb, int ifindex,
2030 struct ipv4_devconf *devconf, u32 portid,
2031 u32 seq, int event, unsigned int flags,
2032 int type)
2033 {
2034 struct nlmsghdr *nlh;
2035 struct netconfmsg *ncm;
2036 bool all = false;
2037
2038 nlh = nlmsg_put(skb, portid, seq, event, sizeof(struct netconfmsg),
2039 flags);
2040 if (!nlh)
2041 return -EMSGSIZE;
2042
2043 if (type == NETCONFA_ALL)
2044 all = true;
2045
2046 ncm = nlmsg_data(nlh);
2047 ncm->ncm_family = AF_INET;
2048
2049 if (nla_put_s32(skb, NETCONFA_IFINDEX, ifindex) < 0)
2050 goto nla_put_failure;
2051
2052 if (!devconf)
2053 goto out;
2054
2055 if ((all || type == NETCONFA_FORWARDING) &&
2056 nla_put_s32(skb, NETCONFA_FORWARDING,
2057 IPV4_DEVCONF(*devconf, FORWARDING)) < 0)
2058 goto nla_put_failure;
2059 if ((all || type == NETCONFA_RP_FILTER) &&
2060 nla_put_s32(skb, NETCONFA_RP_FILTER,
2061 IPV4_DEVCONF(*devconf, RP_FILTER)) < 0)
2062 goto nla_put_failure;
2063 if ((all || type == NETCONFA_MC_FORWARDING) &&
2064 nla_put_s32(skb, NETCONFA_MC_FORWARDING,
2065 IPV4_DEVCONF(*devconf, MC_FORWARDING)) < 0)
2066 goto nla_put_failure;
2067 if ((all || type == NETCONFA_MC_SNOOPING) &&
2068 nla_put_s32(skb, NETCONFA_MC_SNOOPING,
> 2069 IPV4_DEVCONF(*devconf, NETCONFA_MC_SNOOPING)) < 0)
2070 goto nla_put_failure;
2071 if ((all || type == NETCONFA_BC_FORWARDING) &&
2072 nla_put_s32(skb, NETCONFA_BC_FORWARDING,
2073 IPV4_DEVCONF(*devconf, BC_FORWARDING)) < 0)
2074 goto nla_put_failure;
2075 if ((all || type == NETCONFA_PROXY_NEIGH) &&
2076 nla_put_s32(skb, NETCONFA_PROXY_NEIGH,
2077 IPV4_DEVCONF(*devconf, PROXY_ARP)) < 0)
2078 goto nla_put_failure;
2079 if ((all || type == NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN) &&
2080 nla_put_s32(skb, NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN,
2081 IPV4_DEVCONF(*devconf, IGNORE_ROUTES_WITH_LINKDOWN)) < 0)
2082 goto nla_put_failure;
2083
2084 out:
2085 nlmsg_end(skb, nlh);
2086 return 0;
2087
2088 nla_put_failure:
2089 nlmsg_cancel(skb, nlh);
2090 return -EMSGSIZE;
2091 }
2092

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/[email protected]

Attachments:

(No filename) (4.81 kB)
.config.gz (38.59 kB)
Download all attachments

2021-06-17 21:36:17

by kernel test robot

[permalink] [raw]

Subject: Re: [PATCH 1/1] net: Allow all multicast packets to be received on a interface.

Attachments:

(No filename) (3.03 kB)
.config.gz (59.02 kB)
Download all attachments

2021-06-18 02:16:21

by kernel test robot

[permalink] [raw]

Subject: Re: [PATCH 1/1] net: Allow all multicast packets to be received on a interface.

Attachments:

(No filename) (2.85 kB)
.config.gz (40.91 kB)
Download all attachments

2021-06-18 03:06:00

by Callum Sinclair

[permalink] [raw]

Subject: Re: [PATCH 1/1] net: Allow all multicast packets to be received on a interface.

Hi Linus

> I'm also wondering if it could be useful to configure it via
> setsockopt() per application instead of per device via sysctl. Either by
> adding a new socket option. Or by allowing the any IP address
> 0.0.0.0 / :: with IP_ADD_MEMBERSHIP/IPV6_JOIN_GROUP. So that you
> could use this for instance:

Yes perhaps this would be a better way to get multicast snooping working with the existing
options. I can see that using a multicast routing IP socket I will receive all IGMP and MLD
data from that. I was just not creating the socket as a multicast routing socket.

> Or would you prefer to be able to use a layer 3 IP instead of
> a layer 2 packet socket?

Yes I was preferring to use a L3 IP socket instead of a L2 packet socket. This was to have
access to additional data from IP_PKTINFO.

Cheers
Callum
________________________________________
From: Linus L?ssing <[email protected]>
Sent: Friday, June 18, 2021 12:33 AM
To: Callum Sinclair
Cc: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH 1/1] net: Allow all multicast packets to be received on a interface.

Hi Callum,

On Thu, Jun 17, 2021 at 09:50:20PM +1200, Callum Sinclair wrote:
> +mc_snooping - BOOLEAN
> + Enable multicast snooping on the interface. This allows any given
> + multicast group to be received without explicitly being joined.
> + The kernel needs to be compiled with CONFIG_MROUTE and/or
> + CONFIG_IPV6_MROUTE.
> + conf/all/mc_snooping must also be set to TRUE to enable multicast
> + snooping for the interface.
> +

Generally this sounds like a useful feature. One note: When there
are snooping bridges/switches involved, you might run into issues
in receiving all multicast packets, as due to the missing IGMP/MLD
reports the snooping switches won't forward to you.

In that case, to conform to RFC4541 you would also need to become
the selected IGMP/MLD querier and send IGMP/MLD query messages. Or
better, you'd need to send Multicast Router Advertisements
(RFC4286). The latter is the recommended, more flexible way but
might not be supported by all multicast snooping switches yet.
The Linux bridge supports this.

There is a userspace tool called mrdisc you can use for MRD-Adv.
though: https://scanmail.trustwave.com/?c=20988&d=n8HL4MpWu6CIvz405pawlYFbPjGsj-TvRIl7ADnUOg&u=https%3a%2f%2fgithub%2ecom%2ftroglobit%2fmrdisc So no need to
implement MRD Advertisements in the kernel with this patch (though
I could imagine that it might be a useful feature to have, having
MRD support out-of-the-box with this option). Just a note that some
IGMP/MLD Querier or MRD Adv. would be needed when IGMP/MLD snooping
switches are invoved might be nice to have in the mc_snooping
description for now, to avoid potential confusion later.

I'm also wondering if it could be useful to configure it via
setsockopt() per application instead of per device via sysctl. Either by
adding a new socket option. Or by allowing the any IP address
0.0.0.0 / :: with IP_ADD_MEMBERSHIP/IPV6_JOIN_GROUP. So that you
could use this for instance:

$ socat -u UDP6-RECV:1234,reuseaddr,ipv6-join-group="[::]:eth0" -
(currently :: fails with "Invalid argument")

I'm not sure however what the requirements for adding or extending
socket options are, if there are some POSIX standards that'd need
to be followed for compatibility with other OSes, for instance.

Hm, actually, I just noticed that there seem to be some multicast
related setsockopt()s already:

- PACKET_MR_PROMISC
- PACKET_MR_MULTICAST
- PACKET_MR_ALLMULTI

The last one seems to be what you are looking for, I think, the
manpage here says:

"PACKET_MR_ALLMULTI sets the socket up to receive all multicast
packets arriving at the interface"
https://scanmail.trustwave.com/?c=20988&d=n8HL4MpWu6CIvz405pawlYFbPjGsj-TvRI4qVzmDbA&u=https%3a%2f%2fwww%2eman7%2eorg%2flinux%2fman-pages%2fman7%2fpacket%2e7%2ehtml

Or would you prefer to be able to use a layer 3 IP instead of
a layer 2 packet socket?

Regards, Linus

2021-06-18 04:24:49

by Callum Sinclair

[permalink] [raw]

Subject: Re: [PATCH 1/1] net: Allow all multicast packets to be received on a interface.

Hi Andrew

> What is the big picture here? Are you trying to move the snooping
> algorithm into user space? User space will then add/remove Multicast
> FIB entries to the bridge to control where mulitcast frames are sent?

Yes I want to run a IGMP, MLD and PIM implementation in userspace and
just use the kernel to send multicast frames to addresses that have been
installed into the multicast forwarding cache.

> In the past i have written a multicast routing daemon. It is a similar
> problem. You need access to all the join/leaves. But the stack does
> provide them, if you bind to the multicast routing socket. Why not use
> that mechanism? Look in the mrouted sources for an example.

Ah I can see that I get the IGMP and MLD packets now. I was just creating
the socket as a IP socket without multicast routing. Thanks for your help.

Cheers
Callum
________________________________________
From: Andrew Lunn <[email protected]>
Sent: Friday, June 18, 2021 2:18 AM
To: Callum Sinclair
Cc: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH 1/1] net: Allow all multicast packets to be received on a interface.

On Thu, Jun 17, 2021 at 09:50:20PM +1200, Callum Sinclair wrote:
> To receive IGMP or MLD packets on a IP socket on any interface the
> multicast group needs to be explicitly joined. This works well for when
> the multicast group the user is interested in is known, but does not
> provide an easy way to snoop all packets in the http://scanmail.trustwave.com/?c=20988&d=wNnL4EU-bOXOuxnfu9BLng8ncWxDIw3ACrur9S2N4w&u=http%3a%2f%2f224%2e0%2e0%2e0%2f8 or the
> FF00::/8 range.
>
> Define a new sysctl to allow a given interface to become a IGMP or MLD
> snooper. When set the interface will allow any IGMP or MLD packet to be
> received on sockets bound to these devices.

Hi Callum

What is the big picture here? Are you trying to move the snooping
algorithm into user space? User space will then add/remove Multicast
FIB entries to the bridge to control where mulitcast frames are sent?

In the past i have written a multicast routing daemon. It is a similar
problem. You need access to all the join/leaves. But the stack does
provide them, if you bind to the multicast routing socket. Why not use
that mechanism? Look in the mrouted sources for an example.

Andrew