| Message ID | 20221221-sockopt-port-range-v5-1-9fb2c00ad293@cloudflare.com (mailing list archive) |
|---|---|
| State | Handled Elsewhere |
| Delegated to: | Paul Moore |
| Headers | show |
| Series | Add IP_LOCAL_PORT_RANGE socket option | expand |
On Tue, Jan 24, 2023 at 11:05:26AM +0100, Jakub Sitnicki wrote: > Users who want to share a single public IP address for outgoing connections > between several hosts traditionally reach for SNAT. However, SNAT requires > state keeping on the node(s) performing the NAT. > > A stateless alternative exists, where a single IP address used for egress > can be shared between several hosts by partitioning the available ephemeral > port range. In such a setup: > > 1. Each host gets assigned a disjoint range of ephemeral ports. > 2. Applications open connections from the host-assigned port range. > 3. Return traffic gets routed to the host based on both, the destination IP > and the destination port. > > An application which wants to open an outgoing connection (connect) from a > given port range today can choose between two solutions: > > 1. Manually pick the source port by bind()'ing to it before connect()'ing > the socket. > > This approach has a couple of downsides: > > a) Search for a free port has to be implemented in the user-space. If > the chosen 4-tuple happens to be busy, the application needs to retry > from a different local port number. > > Detecting if 4-tuple is busy can be either easy (TCP) or hard > (UDP). In TCP case, the application simply has to check if connect() > returned an error (EADDRNOTAVAIL). That is assuming that the local > port sharing was enabled (REUSEADDR) by all the sockets. > > # Assume desired local port range is 60_000-60_511 > s = socket(AF_INET, SOCK_STREAM) > s.setsockopt(SOL_SOCKET, SO_REUSEADDR, 1) > s.bind(("192.0.2.1", 60_000)) > s.connect(("1.1.1.1", 53)) > # Fails only if 192.0.2.1:60000 -> 1.1.1.1:53 is busy > # Application must retry with another local port > > In case of UDP, the network stack allows binding more than one socket > to the same 4-tuple, when local port sharing is enabled > (REUSEADDR). Hence detecting the conflict is much harder and involves > querying sock_diag and toggling the REUSEADDR flag [1]. > > b) For TCP, bind()-ing to a port within the ephemeral port range means > that no connecting sockets, that is those which leave it to the > network stack to find a free local port at connect() time, can use > the this port. > > IOW, the bind hash bucket tb->fastreuse will be 0 or 1, and the port > will be skipped during the free port search at connect() time. > > 2. Isolate the app in a dedicated netns and use the use the per-netns > ip_local_port_range sysctl to adjust the ephemeral port range bounds. > > The per-netns setting affects all sockets, so this approach can be used > only if: > > - there is just one egress IP address, or > - the desired egress port range is the same for all egress IP addresses > used by the application. > > For TCP, this approach avoids the downsides of (1). Free port search and > 4-tuple conflict detection is done by the network stack: > > system("sysctl -w net.ipv4.ip_local_port_range='60000 60511'") > > s = socket(AF_INET, SOCK_STREAM) > s.setsockopt(SOL_IP, IP_BIND_ADDRESS_NO_PORT, 1) > s.bind(("192.0.2.1", 0)) > s.connect(("1.1.1.1", 53)) > # Fails if all 4-tuples 192.0.2.1:60000-60511 -> 1.1.1.1:53 are busy > > For UDP this approach has limited applicability. Setting the > IP_BIND_ADDRESS_NO_PORT socket option does not result in local source > port being shared with other connected UDP sockets. > > Hence relying on the network stack to find a free source port, limits the > number of outgoing UDP flows from a single IP address down to the number > of available ephemeral ports. > > To put it another way, partitioning the ephemeral port range between hosts > using the existing Linux networking API is cumbersome. > > To address this use case, add a new socket option at the SOL_IP level, > named IP_LOCAL_PORT_RANGE. The new option can be used to clamp down the > ephemeral port range for each socket individually. > > The option can be used only to narrow down the per-netns local port > range. If the per-socket range lies outside of the per-netns range, the > latter takes precedence. > > UAPI-wise, the low and high range bounds are passed to the kernel as a pair > of u16 values in host byte order packed into a u32. This avoids pointer > passing. > > PORT_LO = 40_000 > PORT_HI = 40_511 > > s = socket(AF_INET, SOCK_STREAM) > v = struct.pack("I", PORT_HI << 16 | PORT_LO) > s.setsockopt(SOL_IP, IP_LOCAL_PORT_RANGE, v) > s.bind(("127.0.0.1", 0)) > s.getsockname() > # Local address between ("127.0.0.1", 40_000) and ("127.0.0.1", 40_511), > # if there is a free port. EADDRINUSE otherwise. > > [1] https://github.com/cloudflare/cloudflare-blog/blob/232b432c1d57/2022-02-connectx/connectx.py#L116 > > v4 -> v5: > * Use the fact that netns port range starts at 1 when clamping. (Kuniyuki) > > v3 -> v4: > * Clarify that u16 values are in host byte order (Neal) > > v2 -> v3: > * Make SCTP bind()/bind_add() respect IP_LOCAL_PORT_RANGE option (Eric) > > v1 -> v2: > * Fix the corner case when the per-socket range doesn't overlap with the > per-netns range. Fallback correctly to the per-netns range. (Kuniyuki) You silently ignored my review comment. Let's repeat it again. Please put changelog after --- marker. Changelog doesn't belong to commit message. Thanks > > Reviewed-by: Marek Majkowski <marek@cloudflare.com> > Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com> > Signed-off-by: Jakub Sitnicki <jakub@cloudflare.com> > --- > include/net/inet_sock.h | 4 ++++ > include/net/ip.h | 3 ++- > include/uapi/linux/in.h | 1 + > net/ipv4/inet_connection_sock.c | 25 +++++++++++++++++++++++-- > net/ipv4/inet_hashtables.c | 2 +- > net/ipv4/ip_sockglue.c | 18 ++++++++++++++++++ > net/ipv4/udp.c | 2 +- > net/sctp/socket.c | 2 +- > 8 files changed, 51 insertions(+), 6 deletions(-)
On Tue, Jan 24, 2023 at 02:23 PM +02, Leon Romanovsky wrote: > On Tue, Jan 24, 2023 at 11:05:26AM +0100, Jakub Sitnicki wrote: >> Users who want to share a single public IP address for outgoing connections >> between several hosts traditionally reach for SNAT. However, SNAT requires >> state keeping on the node(s) performing the NAT. >> >> A stateless alternative exists, where a single IP address used for egress >> can be shared between several hosts by partitioning the available ephemeral >> port range. In such a setup: >> >> 1. Each host gets assigned a disjoint range of ephemeral ports. >> 2. Applications open connections from the host-assigned port range. >> 3. Return traffic gets routed to the host based on both, the destination IP >> and the destination port. >> >> An application which wants to open an outgoing connection (connect) from a >> given port range today can choose between two solutions: >> >> 1. Manually pick the source port by bind()'ing to it before connect()'ing >> the socket. >> >> This approach has a couple of downsides: >> >> a) Search for a free port has to be implemented in the user-space. If >> the chosen 4-tuple happens to be busy, the application needs to retry >> from a different local port number. >> >> Detecting if 4-tuple is busy can be either easy (TCP) or hard >> (UDP). In TCP case, the application simply has to check if connect() >> returned an error (EADDRNOTAVAIL). That is assuming that the local >> port sharing was enabled (REUSEADDR) by all the sockets. >> >> # Assume desired local port range is 60_000-60_511 >> s = socket(AF_INET, SOCK_STREAM) >> s.setsockopt(SOL_SOCKET, SO_REUSEADDR, 1) >> s.bind(("192.0.2.1", 60_000)) >> s.connect(("1.1.1.1", 53)) >> # Fails only if 192.0.2.1:60000 -> 1.1.1.1:53 is busy >> # Application must retry with another local port >> >> In case of UDP, the network stack allows binding more than one socket >> to the same 4-tuple, when local port sharing is enabled >> (REUSEADDR). Hence detecting the conflict is much harder and involves >> querying sock_diag and toggling the REUSEADDR flag [1]. >> >> b) For TCP, bind()-ing to a port within the ephemeral port range means >> that no connecting sockets, that is those which leave it to the >> network stack to find a free local port at connect() time, can use >> the this port. >> >> IOW, the bind hash bucket tb->fastreuse will be 0 or 1, and the port >> will be skipped during the free port search at connect() time. >> >> 2. Isolate the app in a dedicated netns and use the use the per-netns >> ip_local_port_range sysctl to adjust the ephemeral port range bounds. >> >> The per-netns setting affects all sockets, so this approach can be used >> only if: >> >> - there is just one egress IP address, or >> - the desired egress port range is the same for all egress IP addresses >> used by the application. >> >> For TCP, this approach avoids the downsides of (1). Free port search and >> 4-tuple conflict detection is done by the network stack: >> >> system("sysctl -w net.ipv4.ip_local_port_range='60000 60511'") >> >> s = socket(AF_INET, SOCK_STREAM) >> s.setsockopt(SOL_IP, IP_BIND_ADDRESS_NO_PORT, 1) >> s.bind(("192.0.2.1", 0)) >> s.connect(("1.1.1.1", 53)) >> # Fails if all 4-tuples 192.0.2.1:60000-60511 -> 1.1.1.1:53 are busy >> >> For UDP this approach has limited applicability. Setting the >> IP_BIND_ADDRESS_NO_PORT socket option does not result in local source >> port being shared with other connected UDP sockets. >> >> Hence relying on the network stack to find a free source port, limits the >> number of outgoing UDP flows from a single IP address down to the number >> of available ephemeral ports. >> >> To put it another way, partitioning the ephemeral port range between hosts >> using the existing Linux networking API is cumbersome. >> >> To address this use case, add a new socket option at the SOL_IP level, >> named IP_LOCAL_PORT_RANGE. The new option can be used to clamp down the >> ephemeral port range for each socket individually. >> >> The option can be used only to narrow down the per-netns local port >> range. If the per-socket range lies outside of the per-netns range, the >> latter takes precedence. >> >> UAPI-wise, the low and high range bounds are passed to the kernel as a pair >> of u16 values in host byte order packed into a u32. This avoids pointer >> passing. >> >> PORT_LO = 40_000 >> PORT_HI = 40_511 >> >> s = socket(AF_INET, SOCK_STREAM) >> v = struct.pack("I", PORT_HI << 16 | PORT_LO) >> s.setsockopt(SOL_IP, IP_LOCAL_PORT_RANGE, v) >> s.bind(("127.0.0.1", 0)) >> s.getsockname() >> # Local address between ("127.0.0.1", 40_000) and ("127.0.0.1", 40_511), >> # if there is a free port. EADDRINUSE otherwise. >> >> [1] https://github.com/cloudflare/cloudflare-blog/blob/232b432c1d57/2022-02-connectx/connectx.py#L116 >> >> v4 -> v5: >> * Use the fact that netns port range starts at 1 when clamping. (Kuniyuki) >> >> v3 -> v4: >> * Clarify that u16 values are in host byte order (Neal) >> >> v2 -> v3: >> * Make SCTP bind()/bind_add() respect IP_LOCAL_PORT_RANGE option (Eric) >> >> v1 -> v2: >> * Fix the corner case when the per-socket range doesn't overlap with the >> per-netns range. Fallback correctly to the per-netns range. (Kuniyuki) > > You silently ignored my review comment. > Let's repeat it again. Please put changelog after --- marker. Changelog > doesn't belong to commit message. I did not. I'm under the impression that you might have missed my follow up question if the changelog-above-trailer convention is still in place [1] and the clarification from Jakub K. [2]. I'm happy to adjust the changelog in whichever way that will make everyone content. However, ATM we don't have one, it seems. [1] https://lore.kernel.org/all/87sfg1vuqj.fsf@cloudflare.com/ [2] https://lore.kernel.org/all/20230123193526.065a9879@kernel.org/
On Tue, Jan 24, 2023 at 01:33:07PM +0100, Jakub Sitnicki wrote: > On Tue, Jan 24, 2023 at 02:23 PM +02, Leon Romanovsky wrote: > > On Tue, Jan 24, 2023 at 11:05:26AM +0100, Jakub Sitnicki wrote: > >> Users who want to share a single public IP address for outgoing connections > >> between several hosts traditionally reach for SNAT. However, SNAT requires > >> state keeping on the node(s) performing the NAT. > >> > >> A stateless alternative exists, where a single IP address used for egress > >> can be shared between several hosts by partitioning the available ephemeral > >> port range. In such a setup: > >> > >> 1. Each host gets assigned a disjoint range of ephemeral ports. > >> 2. Applications open connections from the host-assigned port range. > >> 3. Return traffic gets routed to the host based on both, the destination IP > >> and the destination port. > >> > >> An application which wants to open an outgoing connection (connect) from a > >> given port range today can choose between two solutions: > >> > >> 1. Manually pick the source port by bind()'ing to it before connect()'ing > >> the socket. > >> > >> This approach has a couple of downsides: > >> > >> a) Search for a free port has to be implemented in the user-space. If > >> the chosen 4-tuple happens to be busy, the application needs to retry > >> from a different local port number. > >> > >> Detecting if 4-tuple is busy can be either easy (TCP) or hard > >> (UDP). In TCP case, the application simply has to check if connect() > >> returned an error (EADDRNOTAVAIL). That is assuming that the local > >> port sharing was enabled (REUSEADDR) by all the sockets. > >> > >> # Assume desired local port range is 60_000-60_511 > >> s = socket(AF_INET, SOCK_STREAM) > >> s.setsockopt(SOL_SOCKET, SO_REUSEADDR, 1) > >> s.bind(("192.0.2.1", 60_000)) > >> s.connect(("1.1.1.1", 53)) > >> # Fails only if 192.0.2.1:60000 -> 1.1.1.1:53 is busy > >> # Application must retry with another local port > >> > >> In case of UDP, the network stack allows binding more than one socket > >> to the same 4-tuple, when local port sharing is enabled > >> (REUSEADDR). Hence detecting the conflict is much harder and involves > >> querying sock_diag and toggling the REUSEADDR flag [1]. > >> > >> b) For TCP, bind()-ing to a port within the ephemeral port range means > >> that no connecting sockets, that is those which leave it to the > >> network stack to find a free local port at connect() time, can use > >> the this port. > >> > >> IOW, the bind hash bucket tb->fastreuse will be 0 or 1, and the port > >> will be skipped during the free port search at connect() time. > >> > >> 2. Isolate the app in a dedicated netns and use the use the per-netns > >> ip_local_port_range sysctl to adjust the ephemeral port range bounds. > >> > >> The per-netns setting affects all sockets, so this approach can be used > >> only if: > >> > >> - there is just one egress IP address, or > >> - the desired egress port range is the same for all egress IP addresses > >> used by the application. > >> > >> For TCP, this approach avoids the downsides of (1). Free port search and > >> 4-tuple conflict detection is done by the network stack: > >> > >> system("sysctl -w net.ipv4.ip_local_port_range='60000 60511'") > >> > >> s = socket(AF_INET, SOCK_STREAM) > >> s.setsockopt(SOL_IP, IP_BIND_ADDRESS_NO_PORT, 1) > >> s.bind(("192.0.2.1", 0)) > >> s.connect(("1.1.1.1", 53)) > >> # Fails if all 4-tuples 192.0.2.1:60000-60511 -> 1.1.1.1:53 are busy > >> > >> For UDP this approach has limited applicability. Setting the > >> IP_BIND_ADDRESS_NO_PORT socket option does not result in local source > >> port being shared with other connected UDP sockets. > >> > >> Hence relying on the network stack to find a free source port, limits the > >> number of outgoing UDP flows from a single IP address down to the number > >> of available ephemeral ports. > >> > >> To put it another way, partitioning the ephemeral port range between hosts > >> using the existing Linux networking API is cumbersome. > >> > >> To address this use case, add a new socket option at the SOL_IP level, > >> named IP_LOCAL_PORT_RANGE. The new option can be used to clamp down the > >> ephemeral port range for each socket individually. > >> > >> The option can be used only to narrow down the per-netns local port > >> range. If the per-socket range lies outside of the per-netns range, the > >> latter takes precedence. > >> > >> UAPI-wise, the low and high range bounds are passed to the kernel as a pair > >> of u16 values in host byte order packed into a u32. This avoids pointer > >> passing. > >> > >> PORT_LO = 40_000 > >> PORT_HI = 40_511 > >> > >> s = socket(AF_INET, SOCK_STREAM) > >> v = struct.pack("I", PORT_HI << 16 | PORT_LO) > >> s.setsockopt(SOL_IP, IP_LOCAL_PORT_RANGE, v) > >> s.bind(("127.0.0.1", 0)) > >> s.getsockname() > >> # Local address between ("127.0.0.1", 40_000) and ("127.0.0.1", 40_511), > >> # if there is a free port. EADDRINUSE otherwise. > >> > >> [1] https://github.com/cloudflare/cloudflare-blog/blob/232b432c1d57/2022-02-connectx/connectx.py#L116 > >> > >> v4 -> v5: > >> * Use the fact that netns port range starts at 1 when clamping. (Kuniyuki) > >> > >> v3 -> v4: > >> * Clarify that u16 values are in host byte order (Neal) > >> > >> v2 -> v3: > >> * Make SCTP bind()/bind_add() respect IP_LOCAL_PORT_RANGE option (Eric) > >> > >> v1 -> v2: > >> * Fix the corner case when the per-socket range doesn't overlap with the > >> per-netns range. Fallback correctly to the per-netns range. (Kuniyuki) > > > > You silently ignored my review comment. > > Let's repeat it again. Please put changelog after --- marker. Changelog > > doesn't belong to commit message. > > I did not. I'm under the impression that you might have missed my follow > up question if the changelog-above-trailer convention is still in place > [1] and the clarification from Jakub K. [2]. Yes, I missed it. There is no value in seeing changelog in git log history as there is only one version of patches is applied at the end. Users don't care how many revisions you (or any developer) sent till the patches were accepted. > > I'm happy to adjust the changelog in whichever way that will make > everyone content. However, ATM we don't have one, it seems. We have, just some of us don't care. It doesn't mean they prefer to see changelog before ---. > > [1] https://lore.kernel.org/all/87sfg1vuqj.fsf@cloudflare.com/ > [2] https://lore.kernel.org/all/20230123193526.065a9879@kernel.org/
On Tue, Jan 24, 2023 at 03:04 PM +02, Leon Romanovsky wrote: > On Tue, Jan 24, 2023 at 01:33:07PM +0100, Jakub Sitnicki wrote: [...] >> I'm happy to adjust the changelog in whichever way that will make >> everyone content. However, ATM we don't have one, it seems. > > We have, just some of us don't care. It doesn't mean they prefer to see > changelog before ---. OK. Thanks for clarifying. Will sent v6 with changelog moved.
diff --git a/include/net/inet_sock.h b/include/net/inet_sock.h index bf5654ce711e..51857117ac09 100644 --- a/include/net/inet_sock.h +++ b/include/net/inet_sock.h @@ -249,6 +249,10 @@ struct inet_sock { __be32 mc_addr; struct ip_mc_socklist __rcu *mc_list; struct inet_cork_full cork; + struct { + __u16 lo; + __u16 hi; + } local_port_range; }; #define IPCORK_OPT 1 /* ip-options has been held in ipcork.opt */ diff --git a/include/net/ip.h b/include/net/ip.h index 144bdfbb25af..c3fffaa92d6e 100644 --- a/include/net/ip.h +++ b/include/net/ip.h @@ -340,7 +340,8 @@ static inline u64 snmp_fold_field64(void __percpu *mib, int offt, size_t syncp_o } \ } -void inet_get_local_port_range(struct net *net, int *low, int *high); +void inet_get_local_port_range(const struct net *net, int *low, int *high); +void inet_sk_get_local_port_range(const struct sock *sk, int *low, int *high); #ifdef CONFIG_SYSCTL static inline bool inet_is_local_reserved_port(struct net *net, unsigned short port) diff --git a/include/uapi/linux/in.h b/include/uapi/linux/in.h index 07a4cb149305..4b7f2df66b99 100644 --- a/include/uapi/linux/in.h +++ b/include/uapi/linux/in.h @@ -162,6 +162,7 @@ struct in_addr { #define MCAST_MSFILTER 48 #define IP_MULTICAST_ALL 49 #define IP_UNICAST_IF 50 +#define IP_LOCAL_PORT_RANGE 51 #define MCAST_EXCLUDE 0 #define MCAST_INCLUDE 1 diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index d1f837579398..6ed7e65de494 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -117,7 +117,7 @@ bool inet_rcv_saddr_any(const struct sock *sk) return !sk->sk_rcv_saddr; } -void inet_get_local_port_range(struct net *net, int *low, int *high) +void inet_get_local_port_range(const struct net *net, int *low, int *high) { unsigned int seq; @@ -130,6 +130,27 @@ void inet_get_local_port_range(struct net *net, int *low, int *high) } EXPORT_SYMBOL(inet_get_local_port_range); +void inet_sk_get_local_port_range(const struct sock *sk, int *low, int *high) +{ + const struct inet_sock *inet = inet_sk(sk); + const struct net *net = sock_net(sk); + int lo, hi, sk_lo, sk_hi; + + inet_get_local_port_range(net, &lo, &hi); + + sk_lo = inet->local_port_range.lo; + sk_hi = inet->local_port_range.hi; + + if (unlikely(lo <= sk_lo && sk_lo <= hi)) + lo = sk_lo; + if (unlikely(lo <= sk_hi && sk_hi <= hi)) + hi = sk_hi; + + *low = lo; + *high = hi; +} +EXPORT_SYMBOL(inet_sk_get_local_port_range); + static bool inet_use_bhash2_on_bind(const struct sock *sk) { #if IS_ENABLED(CONFIG_IPV6) @@ -316,7 +337,7 @@ inet_csk_find_open_port(const struct sock *sk, struct inet_bind_bucket **tb_ret, ports_exhausted: attempt_half = (sk->sk_reuse == SK_CAN_REUSE) ? 1 : 0; other_half_scan: - inet_get_local_port_range(net, &low, &high); + inet_sk_get_local_port_range(sk, &low, &high); high++; /* [32768, 60999] -> [32768, 61000[ */ if (high - low < 4) attempt_half = 0; diff --git a/net/ipv4/inet_hashtables.c b/net/ipv4/inet_hashtables.c index 7a13dd7f546b..e41fdc38ce19 100644 --- a/net/ipv4/inet_hashtables.c +++ b/net/ipv4/inet_hashtables.c @@ -1016,7 +1016,7 @@ int __inet_hash_connect(struct inet_timewait_death_row *death_row, l3mdev = inet_sk_bound_l3mdev(sk); - inet_get_local_port_range(net, &low, &high); + inet_sk_get_local_port_range(sk, &low, &high); high++; /* [32768, 60999] -> [32768, 61000[ */ remaining = high - low; if (likely(remaining > 1)) diff --git a/net/ipv4/ip_sockglue.c b/net/ipv4/ip_sockglue.c index 9f92ae35bb01..b511ff0adc0a 100644 --- a/net/ipv4/ip_sockglue.c +++ b/net/ipv4/ip_sockglue.c @@ -923,6 +923,7 @@ int do_ip_setsockopt(struct sock *sk, int level, int optname, case IP_CHECKSUM: case IP_RECVFRAGSIZE: case IP_RECVERR_RFC4884: + case IP_LOCAL_PORT_RANGE: if (optlen >= sizeof(int)) { if (copy_from_sockptr(&val, optval, sizeof(val))) return -EFAULT; @@ -1365,6 +1366,20 @@ int do_ip_setsockopt(struct sock *sk, int level, int optname, WRITE_ONCE(inet->min_ttl, val); break; + case IP_LOCAL_PORT_RANGE: + { + const __u16 lo = val; + const __u16 hi = val >> 16; + + if (optlen != sizeof(__u32)) + goto e_inval; + if (lo != 0 && hi != 0 && lo > hi) + goto e_inval; + + inet->local_port_range.lo = lo; + inet->local_port_range.hi = hi; + break; + } default: err = -ENOPROTOOPT; break; @@ -1743,6 +1758,9 @@ int do_ip_getsockopt(struct sock *sk, int level, int optname, case IP_MINTTL: val = inet->min_ttl; break; + case IP_LOCAL_PORT_RANGE: + val = inet->local_port_range.hi << 16 | inet->local_port_range.lo; + break; default: sockopt_release_sock(sk); return -ENOPROTOOPT; diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index 9592fe3e444a..c605d171eb2d 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -248,7 +248,7 @@ int udp_lib_get_port(struct sock *sk, unsigned short snum, int low, high, remaining; unsigned int rand; - inet_get_local_port_range(net, &low, &high); + inet_sk_get_local_port_range(sk, &low, &high); remaining = (high - low) + 1; rand = get_random_u32(); diff --git a/net/sctp/socket.c b/net/sctp/socket.c index a98511b676cd..b91616f819de 100644 --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@ -8322,7 +8322,7 @@ static int sctp_get_port_local(struct sock *sk, union sctp_addr *addr) int low, high, remaining, index; unsigned int rover; - inet_get_local_port_range(net, &low, &high); + inet_sk_get_local_port_range(sk, &low, &high); remaining = (high - low) + 1; rover = get_random_u32_below(remaining) + low;
Users who want to share a single public IP address for outgoing connections between several hosts traditionally reach for SNAT. However, SNAT requires state keeping on the node(s) performing the NAT. A stateless alternative exists, where a single IP address used for egress can be shared between several hosts by partitioning the available ephemeral port range. In such a setup: 1. Each host gets assigned a disjoint range of ephemeral ports. 2. Applications open connections from the host-assigned port range. 3. Return traffic gets routed to the host based on both, the destination IP and the destination port. An application which wants to open an outgoing connection (connect) from a given port range today can choose between two solutions: 1. Manually pick the source port by bind()'ing to it before connect()'ing the socket. This approach has a couple of downsides: a) Search for a free port has to be implemented in the user-space. If the chosen 4-tuple happens to be busy, the application needs to retry from a different local port number. Detecting if 4-tuple is busy can be either easy (TCP) or hard (UDP). In TCP case, the application simply has to check if connect() returned an error (EADDRNOTAVAIL). That is assuming that the local port sharing was enabled (REUSEADDR) by all the sockets. # Assume desired local port range is 60_000-60_511 s = socket(AF_INET, SOCK_STREAM) s.setsockopt(SOL_SOCKET, SO_REUSEADDR, 1) s.bind(("192.0.2.1", 60_000)) s.connect(("1.1.1.1", 53)) # Fails only if 192.0.2.1:60000 -> 1.1.1.1:53 is busy # Application must retry with another local port In case of UDP, the network stack allows binding more than one socket to the same 4-tuple, when local port sharing is enabled (REUSEADDR). Hence detecting the conflict is much harder and involves querying sock_diag and toggling the REUSEADDR flag [1]. b) For TCP, bind()-ing to a port within the ephemeral port range means that no connecting sockets, that is those which leave it to the network stack to find a free local port at connect() time, can use the this port. IOW, the bind hash bucket tb->fastreuse will be 0 or 1, and the port will be skipped during the free port search at connect() time. 2. Isolate the app in a dedicated netns and use the use the per-netns ip_local_port_range sysctl to adjust the ephemeral port range bounds. The per-netns setting affects all sockets, so this approach can be used only if: - there is just one egress IP address, or - the desired egress port range is the same for all egress IP addresses used by the application. For TCP, this approach avoids the downsides of (1). Free port search and 4-tuple conflict detection is done by the network stack: system("sysctl -w net.ipv4.ip_local_port_range='60000 60511'") s = socket(AF_INET, SOCK_STREAM) s.setsockopt(SOL_IP, IP_BIND_ADDRESS_NO_PORT, 1) s.bind(("192.0.2.1", 0)) s.connect(("1.1.1.1", 53)) # Fails if all 4-tuples 192.0.2.1:60000-60511 -> 1.1.1.1:53 are busy For UDP this approach has limited applicability. Setting the IP_BIND_ADDRESS_NO_PORT socket option does not result in local source port being shared with other connected UDP sockets. Hence relying on the network stack to find a free source port, limits the number of outgoing UDP flows from a single IP address down to the number of available ephemeral ports. To put it another way, partitioning the ephemeral port range between hosts using the existing Linux networking API is cumbersome. To address this use case, add a new socket option at the SOL_IP level, named IP_LOCAL_PORT_RANGE. The new option can be used to clamp down the ephemeral port range for each socket individually. The option can be used only to narrow down the per-netns local port range. If the per-socket range lies outside of the per-netns range, the latter takes precedence. UAPI-wise, the low and high range bounds are passed to the kernel as a pair of u16 values in host byte order packed into a u32. This avoids pointer passing. PORT_LO = 40_000 PORT_HI = 40_511 s = socket(AF_INET, SOCK_STREAM) v = struct.pack("I", PORT_HI << 16 | PORT_LO) s.setsockopt(SOL_IP, IP_LOCAL_PORT_RANGE, v) s.bind(("127.0.0.1", 0)) s.getsockname() # Local address between ("127.0.0.1", 40_000) and ("127.0.0.1", 40_511), # if there is a free port. EADDRINUSE otherwise. [1] https://github.com/cloudflare/cloudflare-blog/blob/232b432c1d57/2022-02-connectx/connectx.py#L116 v4 -> v5: * Use the fact that netns port range starts at 1 when clamping. (Kuniyuki) v3 -> v4: * Clarify that u16 values are in host byte order (Neal) v2 -> v3: * Make SCTP bind()/bind_add() respect IP_LOCAL_PORT_RANGE option (Eric) v1 -> v2: * Fix the corner case when the per-socket range doesn't overlap with the per-netns range. Fallback correctly to the per-netns range. (Kuniyuki) Reviewed-by: Marek Majkowski <marek@cloudflare.com> Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com> Signed-off-by: Jakub Sitnicki <jakub@cloudflare.com> --- include/net/inet_sock.h | 4 ++++ include/net/ip.h | 3 ++- include/uapi/linux/in.h | 1 + net/ipv4/inet_connection_sock.c | 25 +++++++++++++++++++++++-- net/ipv4/inet_hashtables.c | 2 +- net/ipv4/ip_sockglue.c | 18 ++++++++++++++++++ net/ipv4/udp.c | 2 +- net/sctp/socket.c | 2 +- 8 files changed, 51 insertions(+), 6 deletions(-)