* [PATCH v2 net] udp: Update reuse->has_conns under reuseport_lock.
@ 2022-10-12 18:52 Kuniyuki Iwashima
2022-10-12 18:59 ` Eric Dumazet
0 siblings, 1 reply; 7+ messages in thread
From: Kuniyuki Iwashima @ 2022-10-12 18:52 UTC (permalink / raw)
To: David S. Miller, Eric Dumazet, Jakub Kicinski, Paolo Abeni,
Hideaki YOSHIFUJI, David Ahern, Martin KaFai Lau
Cc: Craig Gallek, Willem de Bruijn, Kuniyuki Iwashima,
Kuniyuki Iwashima, netdev
When we call connect() for a UDP socket in a reuseport group, we have
to update sk->sk_reuseport_cb->has_conns to 1. Otherwise, the kernel
could select a unconnected socket wrongly for packets sent to the
connected socket.
However, the current way to set has_conns is illegal and possible to
trigger that problem. reuseport_has_conns() changes has_conns under
rcu_read_lock(), which upgrades the RCU reader to the updater. Then,
it must do the update under the updater's lock, reuseport_lock, but
it doesn't for now.
For this reason, there is a race below where we fail to set has_conns
resulting in the wrong socket selection. To avoid the race, let's split
the reader and updater with proper locking.
cpu1 cpu2
+----+ +----+
__ip[46]_datagram_connect() reuseport_grow()
. .
|- reuseport_has_conns(sk, true) |- more_reuse = __reuseport_alloc(more_socks_size)
| . |
| |- rcu_read_lock()
| |- reuse = rcu_dereference(sk->sk_reuseport_cb)
| |
| | | /* reuse->has_conns == 0 here */
| | |- more_reuse->has_conns = reuse->has_conns
| |- reuse->has_conns = 1 | /* more_reuse->has_conns SHOULD BE 1 HERE */
| | |
| | |- rcu_assign_pointer(reuse->socks[i]->sk_reuseport_cb,
| | | more_reuse)
| `- rcu_read_unlock() `- kfree_rcu(reuse, rcu)
|
|- sk->sk_state = TCP_ESTABLISHED
Fixes: acdcecc61285 ("udp: correct reuseport selection with connected sockets")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
---
v2:
* Fix build failure for CONFIG_IPV6=m
* Drop SO_INCOMING_CPU fix, which will be sent for net-next
after the v6.1 merge window
v1: https://lore.kernel.org/netdev/20221010174351.11024-1-kuniyu@amazon.com/
---
include/net/sock_reuseport.h | 11 +++++------
net/core/sock_reuseport.c | 15 +++++++++++++++
net/ipv4/datagram.c | 2 +-
net/ipv4/udp.c | 2 +-
net/ipv6/datagram.c | 2 +-
net/ipv6/udp.c | 2 +-
6 files changed, 24 insertions(+), 10 deletions(-)
diff --git a/include/net/sock_reuseport.h b/include/net/sock_reuseport.h
index 473b0b0fa4ab..efc9085c6892 100644
--- a/include/net/sock_reuseport.h
+++ b/include/net/sock_reuseport.h
@@ -43,21 +43,20 @@ struct sock *reuseport_migrate_sock(struct sock *sk,
extern int reuseport_attach_prog(struct sock *sk, struct bpf_prog *prog);
extern int reuseport_detach_prog(struct sock *sk);
-static inline bool reuseport_has_conns(struct sock *sk, bool set)
+static inline bool reuseport_has_conns(struct sock *sk)
{
struct sock_reuseport *reuse;
bool ret = false;
rcu_read_lock();
reuse = rcu_dereference(sk->sk_reuseport_cb);
- if (reuse) {
- if (set)
- reuse->has_conns = 1;
- ret = reuse->has_conns;
- }
+ if (reuse && reuse->has_conns)
+ ret = true;
rcu_read_unlock();
return ret;
}
+void reuseport_has_conns_set(struct sock *sk);
+
#endif /* _SOCK_REUSEPORT_H */
diff --git a/net/core/sock_reuseport.c b/net/core/sock_reuseport.c
index 5daa1fa54249..abb414ed4aa7 100644
--- a/net/core/sock_reuseport.c
+++ b/net/core/sock_reuseport.c
@@ -21,6 +21,21 @@ static DEFINE_IDA(reuseport_ida);
static int reuseport_resurrect(struct sock *sk, struct sock_reuseport *old_reuse,
struct sock_reuseport *reuse, bool bind_inany);
+void reuseport_has_conns_set(struct sock *sk)
+{
+ struct sock_reuseport *reuse;
+
+ if (!rcu_access_pointer(sk->sk_reuseport_cb))
+ return;
+
+ spin_lock(&reuseport_lock);
+ reuse = rcu_dereference_protected(sk->sk_reuseport_cb,
+ lockdep_is_held(&reuseport_lock));
+ reuse->has_conns = 1;
+ spin_unlock(&reuseport_lock);
+}
+EXPORT_SYMBOL(reuseport_has_conns_set);
+
static int reuseport_sock_index(struct sock *sk,
const struct sock_reuseport *reuse,
bool closed)
diff --git a/net/ipv4/datagram.c b/net/ipv4/datagram.c
index 405a8c2aea64..5e66add7befa 100644
--- a/net/ipv4/datagram.c
+++ b/net/ipv4/datagram.c
@@ -70,7 +70,7 @@ int __ip4_datagram_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len
}
inet->inet_daddr = fl4->daddr;
inet->inet_dport = usin->sin_port;
- reuseport_has_conns(sk, true);
+ reuseport_has_conns_set(sk);
sk->sk_state = TCP_ESTABLISHED;
sk_set_txhash(sk);
inet->inet_id = prandom_u32();
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index d63118ce5900..29228231b058 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -448,7 +448,7 @@ static struct sock *udp4_lib_lookup2(struct net *net,
result = lookup_reuseport(net, sk, skb,
saddr, sport, daddr, hnum);
/* Fall back to scoring if group has connections */
- if (result && !reuseport_has_conns(sk, false))
+ if (result && !reuseport_has_conns(sk))
return result;
result = result ? : sk;
diff --git a/net/ipv6/datagram.c b/net/ipv6/datagram.c
index df665d4e8f0f..5ecb56522f9d 100644
--- a/net/ipv6/datagram.c
+++ b/net/ipv6/datagram.c
@@ -256,7 +256,7 @@ int __ip6_datagram_connect(struct sock *sk, struct sockaddr *uaddr,
goto out;
}
- reuseport_has_conns(sk, true);
+ reuseport_has_conns_set(sk);
sk->sk_state = TCP_ESTABLISHED;
sk_set_txhash(sk);
out:
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
index 91e795bb9ade..56e4523a3004 100644
--- a/net/ipv6/udp.c
+++ b/net/ipv6/udp.c
@@ -182,7 +182,7 @@ static struct sock *udp6_lib_lookup2(struct net *net,
result = lookup_reuseport(net, sk, skb,
saddr, sport, daddr, hnum);
/* Fall back to scoring if group has connections */
- if (result && !reuseport_has_conns(sk, false))
+ if (result && !reuseport_has_conns(sk))
return result;
result = result ? : sk;
--
2.30.2
^ permalink raw reply related [flat|nested] 7+ messages in thread
* Re: [PATCH v2 net] udp: Update reuse->has_conns under reuseport_lock.
2022-10-12 18:52 [PATCH v2 net] udp: Update reuse->has_conns under reuseport_lock Kuniyuki Iwashima
@ 2022-10-12 18:59 ` Eric Dumazet
2022-10-12 19:27 ` Kuniyuki Iwashima
0 siblings, 1 reply; 7+ messages in thread
From: Eric Dumazet @ 2022-10-12 18:59 UTC (permalink / raw)
To: Kuniyuki Iwashima
Cc: David S. Miller, Jakub Kicinski, Paolo Abeni, Hideaki YOSHIFUJI,
David Ahern, Martin KaFai Lau, Craig Gallek, Willem de Bruijn,
Kuniyuki Iwashima, netdev
On Wed, Oct 12, 2022 at 11:53 AM Kuniyuki Iwashima <kuniyu@amazon.com> wrote:
>
> When we call connect() for a UDP socket in a reuseport group, we have
> to update sk->sk_reuseport_cb->has_conns to 1. Otherwise, the kernel
> could select a unconnected socket wrongly for packets sent to the
> connected socket.
>
> However, the current way to set has_conns is illegal and possible to
> trigger that problem. reuseport_has_conns() changes has_conns under
> rcu_read_lock(), which upgrades the RCU reader to the updater. Then,
> it must do the update under the updater's lock, reuseport_lock, but
> it doesn't for now.
>
> For this reason, there is a race below where we fail to set has_conns
> resulting in the wrong socket selection. To avoid the race, let's split
> the reader and updater with proper locking.
>
> cpu1 cpu2
> +----+ +----+
>
> __ip[46]_datagram_connect() reuseport_grow()
> . .
> |- reuseport_has_conns(sk, true) |- more_reuse = __reuseport_alloc(more_socks_size)
> | . |
> | |- rcu_read_lock()
> | |- reuse = rcu_dereference(sk->sk_reuseport_cb)
> | |
> | | | /* reuse->has_conns == 0 here */
> | | |- more_reuse->has_conns = reuse->has_conns
> | |- reuse->has_conns = 1 | /* more_reuse->has_conns SHOULD BE 1 HERE */
> | | |
> | | |- rcu_assign_pointer(reuse->socks[i]->sk_reuseport_cb,
> | | | more_reuse)
> | `- rcu_read_unlock() `- kfree_rcu(reuse, rcu)
> |
> |- sk->sk_state = TCP_ESTABLISHED
>
> Fixes: acdcecc61285 ("udp: correct reuseport selection with connected sockets")
> Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
> ---
> v2:
> * Fix build failure for CONFIG_IPV6=m
> * Drop SO_INCOMING_CPU fix, which will be sent for net-next
> after the v6.1 merge window
>
> v1: https://lore.kernel.org/netdev/20221010174351.11024-1-kuniyu@amazon.com/
> ---
> include/net/sock_reuseport.h | 11 +++++------
> net/core/sock_reuseport.c | 15 +++++++++++++++
> net/ipv4/datagram.c | 2 +-
> net/ipv4/udp.c | 2 +-
> net/ipv6/datagram.c | 2 +-
> net/ipv6/udp.c | 2 +-
> 6 files changed, 24 insertions(+), 10 deletions(-)
>
> diff --git a/include/net/sock_reuseport.h b/include/net/sock_reuseport.h
> index 473b0b0fa4ab..efc9085c6892 100644
> --- a/include/net/sock_reuseport.h
> +++ b/include/net/sock_reuseport.h
> @@ -43,21 +43,20 @@ struct sock *reuseport_migrate_sock(struct sock *sk,
> extern int reuseport_attach_prog(struct sock *sk, struct bpf_prog *prog);
> extern int reuseport_detach_prog(struct sock *sk);
>
> -static inline bool reuseport_has_conns(struct sock *sk, bool set)
> +static inline bool reuseport_has_conns(struct sock *sk)
> {
> struct sock_reuseport *reuse;
> bool ret = false;
>
> rcu_read_lock();
> reuse = rcu_dereference(sk->sk_reuseport_cb);
> - if (reuse) {
> - if (set)
> - reuse->has_conns = 1;
> - ret = reuse->has_conns;
> - }
> + if (reuse && reuse->has_conns)
> + ret = true;
> rcu_read_unlock();
>
> return ret;
> }
>
> +void reuseport_has_conns_set(struct sock *sk);
> +
> #endif /* _SOCK_REUSEPORT_H */
> diff --git a/net/core/sock_reuseport.c b/net/core/sock_reuseport.c
> index 5daa1fa54249..abb414ed4aa7 100644
> --- a/net/core/sock_reuseport.c
> +++ b/net/core/sock_reuseport.c
> @@ -21,6 +21,21 @@ static DEFINE_IDA(reuseport_ida);
> static int reuseport_resurrect(struct sock *sk, struct sock_reuseport *old_reuse,
> struct sock_reuseport *reuse, bool bind_inany);
>
> +void reuseport_has_conns_set(struct sock *sk)
> +{
> + struct sock_reuseport *reuse;
> +
> + if (!rcu_access_pointer(sk->sk_reuseport_cb))
> + return;
> +
> + spin_lock(&reuseport_lock);
> + reuse = rcu_dereference_protected(sk->sk_reuseport_cb,
> + lockdep_is_held(&reuseport_lock));
Could @reuse be NULL at this point ?
Previous test was performed without reuseport_lock being held.
> + reuse->has_conns = 1;
> + spin_unlock(&reuseport_lock);
> +}
> +EXPORT_SYMBOL(reuseport_has_conns_set);
> +
> static int reuseport_sock_index(struct sock *sk,
> const struct sock_reuseport *reuse,
> bool closed)
> diff --git a/net/ipv4/datagram.c b/net/ipv4/datagram.c
> index 405a8c2aea64..5e66add7befa 100644
> --- a/net/ipv4/datagram.c
> +++ b/net/ipv4/datagram.c
> @@ -70,7 +70,7 @@ int __ip4_datagram_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len
> }
> inet->inet_daddr = fl4->daddr;
> inet->inet_dport = usin->sin_port;
> - reuseport_has_conns(sk, true);
> + reuseport_has_conns_set(sk);
> sk->sk_state = TCP_ESTABLISHED;
> sk_set_txhash(sk);
> inet->inet_id = prandom_u32();
> diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
> index d63118ce5900..29228231b058 100644
> --- a/net/ipv4/udp.c
> +++ b/net/ipv4/udp.c
> @@ -448,7 +448,7 @@ static struct sock *udp4_lib_lookup2(struct net *net,
> result = lookup_reuseport(net, sk, skb,
> saddr, sport, daddr, hnum);
> /* Fall back to scoring if group has connections */
> - if (result && !reuseport_has_conns(sk, false))
> + if (result && !reuseport_has_conns(sk))
> return result;
>
> result = result ? : sk;
> diff --git a/net/ipv6/datagram.c b/net/ipv6/datagram.c
> index df665d4e8f0f..5ecb56522f9d 100644
> --- a/net/ipv6/datagram.c
> +++ b/net/ipv6/datagram.c
> @@ -256,7 +256,7 @@ int __ip6_datagram_connect(struct sock *sk, struct sockaddr *uaddr,
> goto out;
> }
>
> - reuseport_has_conns(sk, true);
> + reuseport_has_conns_set(sk);
> sk->sk_state = TCP_ESTABLISHED;
> sk_set_txhash(sk);
> out:
> diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
> index 91e795bb9ade..56e4523a3004 100644
> --- a/net/ipv6/udp.c
> +++ b/net/ipv6/udp.c
> @@ -182,7 +182,7 @@ static struct sock *udp6_lib_lookup2(struct net *net,
> result = lookup_reuseport(net, sk, skb,
> saddr, sport, daddr, hnum);
> /* Fall back to scoring if group has connections */
> - if (result && !reuseport_has_conns(sk, false))
> + if (result && !reuseport_has_conns(sk))
> return result;
>
> result = result ? : sk;
> --
> 2.30.2
>
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH v2 net] udp: Update reuse->has_conns under reuseport_lock.
2022-10-12 18:59 ` Eric Dumazet
@ 2022-10-12 19:27 ` Kuniyuki Iwashima
2022-10-13 16:09 ` Eric Dumazet
0 siblings, 1 reply; 7+ messages in thread
From: Kuniyuki Iwashima @ 2022-10-12 19:27 UTC (permalink / raw)
To: edumazet
Cc: davem, dsahern, kraig, kuba, kuni1840, kuniyu, martin.lau,
netdev, pabeni, willemb, yoshfuji
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 12 Oct 2022 11:59:43 -0700
> On Wed, Oct 12, 2022 at 11:53 AM Kuniyuki Iwashima <kuniyu@amazon.com> wrote:
> >
> > When we call connect() for a UDP socket in a reuseport group, we have
> > to update sk->sk_reuseport_cb->has_conns to 1. Otherwise, the kernel
> > could select a unconnected socket wrongly for packets sent to the
> > connected socket.
> >
> > However, the current way to set has_conns is illegal and possible to
> > trigger that problem. reuseport_has_conns() changes has_conns under
> > rcu_read_lock(), which upgrades the RCU reader to the updater. Then,
> > it must do the update under the updater's lock, reuseport_lock, but
> > it doesn't for now.
> >
> > For this reason, there is a race below where we fail to set has_conns
> > resulting in the wrong socket selection. To avoid the race, let's split
> > the reader and updater with proper locking.
> >
> > cpu1 cpu2
> > +----+ +----+
> >
> > __ip[46]_datagram_connect() reuseport_grow()
> > . .
> > |- reuseport_has_conns(sk, true) |- more_reuse = __reuseport_alloc(more_socks_size)
> > | . |
> > | |- rcu_read_lock()
> > | |- reuse = rcu_dereference(sk->sk_reuseport_cb)
> > | |
> > | | | /* reuse->has_conns == 0 here */
> > | | |- more_reuse->has_conns = reuse->has_conns
> > | |- reuse->has_conns = 1 | /* more_reuse->has_conns SHOULD BE 1 HERE */
> > | | |
> > | | |- rcu_assign_pointer(reuse->socks[i]->sk_reuseport_cb,
> > | | | more_reuse)
> > | `- rcu_read_unlock() `- kfree_rcu(reuse, rcu)
> > |
> > |- sk->sk_state = TCP_ESTABLISHED
> >
> > Fixes: acdcecc61285 ("udp: correct reuseport selection with connected sockets")
> > Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
> > ---
> > v2:
> > * Fix build failure for CONFIG_IPV6=m
> > * Drop SO_INCOMING_CPU fix, which will be sent for net-next
> > after the v6.1 merge window
> >
> > v1: https://lore.kernel.org/netdev/20221010174351.11024-1-kuniyu@amazon.com/
> > ---
> > include/net/sock_reuseport.h | 11 +++++------
> > net/core/sock_reuseport.c | 15 +++++++++++++++
> > net/ipv4/datagram.c | 2 +-
> > net/ipv4/udp.c | 2 +-
> > net/ipv6/datagram.c | 2 +-
> > net/ipv6/udp.c | 2 +-
> > 6 files changed, 24 insertions(+), 10 deletions(-)
> >
> > diff --git a/include/net/sock_reuseport.h b/include/net/sock_reuseport.h
> > index 473b0b0fa4ab..efc9085c6892 100644
> > --- a/include/net/sock_reuseport.h
> > +++ b/include/net/sock_reuseport.h
> > @@ -43,21 +43,20 @@ struct sock *reuseport_migrate_sock(struct sock *sk,
> > extern int reuseport_attach_prog(struct sock *sk, struct bpf_prog *prog);
> > extern int reuseport_detach_prog(struct sock *sk);
> >
> > -static inline bool reuseport_has_conns(struct sock *sk, bool set)
> > +static inline bool reuseport_has_conns(struct sock *sk)
> > {
> > struct sock_reuseport *reuse;
> > bool ret = false;
> >
> > rcu_read_lock();
> > reuse = rcu_dereference(sk->sk_reuseport_cb);
> > - if (reuse) {
> > - if (set)
> > - reuse->has_conns = 1;
> > - ret = reuse->has_conns;
> > - }
> > + if (reuse && reuse->has_conns)
> > + ret = true;
> > rcu_read_unlock();
> >
> > return ret;
> > }
> >
> > +void reuseport_has_conns_set(struct sock *sk);
> > +
> > #endif /* _SOCK_REUSEPORT_H */
> > diff --git a/net/core/sock_reuseport.c b/net/core/sock_reuseport.c
> > index 5daa1fa54249..abb414ed4aa7 100644
> > --- a/net/core/sock_reuseport.c
> > +++ b/net/core/sock_reuseport.c
> > @@ -21,6 +21,21 @@ static DEFINE_IDA(reuseport_ida);
> > static int reuseport_resurrect(struct sock *sk, struct sock_reuseport *old_reuse,
> > struct sock_reuseport *reuse, bool bind_inany);
> >
> > +void reuseport_has_conns_set(struct sock *sk)
> > +{
> > + struct sock_reuseport *reuse;
> > +
> > + if (!rcu_access_pointer(sk->sk_reuseport_cb))
> > + return;
> > +
> > + spin_lock(&reuseport_lock);
> > + reuse = rcu_dereference_protected(sk->sk_reuseport_cb,
> > + lockdep_is_held(&reuseport_lock));
>
> Could @reuse be NULL at this point ?
>
> Previous test was performed without reuseport_lock being held.
Usually, sk_reuseport_cb is changed under lock_sock().
The only exception is reuseport_grow() & TCP reqsk migration case.
1) shutdown() TCP listener, which is moved into the latter part of
reuse->socks[] to migrate reqsk.
2) New listen() overflows reuse->socks[] and call reuseport_grow().
3) reuse->max_socks overflows u16 with the new listener.
4) reuseport_grow() pops the old shutdown()ed listener from the array
and update its sk->sk_reuseport_cb as NULL without lock_sock().
shutdown()ed sk->sk_reuseport_cb can be changed without lock_sock().
But, reuseport_has_conns_set() is called only for UDP and under
lock_sock(), so @reuse never be NULL in this case.
> > + reuse->has_conns = 1;
> > + spin_unlock(&reuseport_lock);
> > +}
> > +EXPORT_SYMBOL(reuseport_has_conns_set);
> > +
> > static int reuseport_sock_index(struct sock *sk,
> > const struct sock_reuseport *reuse,
> > bool closed)
> > diff --git a/net/ipv4/datagram.c b/net/ipv4/datagram.c
> > index 405a8c2aea64..5e66add7befa 100644
> > --- a/net/ipv4/datagram.c
> > +++ b/net/ipv4/datagram.c
> > @@ -70,7 +70,7 @@ int __ip4_datagram_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len
> > }
> > inet->inet_daddr = fl4->daddr;
> > inet->inet_dport = usin->sin_port;
> > - reuseport_has_conns(sk, true);
> > + reuseport_has_conns_set(sk);
> > sk->sk_state = TCP_ESTABLISHED;
> > sk_set_txhash(sk);
> > inet->inet_id = prandom_u32();
> > diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
> > index d63118ce5900..29228231b058 100644
> > --- a/net/ipv4/udp.c
> > +++ b/net/ipv4/udp.c
> > @@ -448,7 +448,7 @@ static struct sock *udp4_lib_lookup2(struct net *net,
> > result = lookup_reuseport(net, sk, skb,
> > saddr, sport, daddr, hnum);
> > /* Fall back to scoring if group has connections */
> > - if (result && !reuseport_has_conns(sk, false))
> > + if (result && !reuseport_has_conns(sk))
> > return result;
> >
> > result = result ? : sk;
> > diff --git a/net/ipv6/datagram.c b/net/ipv6/datagram.c
> > index df665d4e8f0f..5ecb56522f9d 100644
> > --- a/net/ipv6/datagram.c
> > +++ b/net/ipv6/datagram.c
> > @@ -256,7 +256,7 @@ int __ip6_datagram_connect(struct sock *sk, struct sockaddr *uaddr,
> > goto out;
> > }
> >
> > - reuseport_has_conns(sk, true);
> > + reuseport_has_conns_set(sk);
> > sk->sk_state = TCP_ESTABLISHED;
> > sk_set_txhash(sk);
> > out:
> > diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
> > index 91e795bb9ade..56e4523a3004 100644
> > --- a/net/ipv6/udp.c
> > +++ b/net/ipv6/udp.c
> > @@ -182,7 +182,7 @@ static struct sock *udp6_lib_lookup2(struct net *net,
> > result = lookup_reuseport(net, sk, skb,
> > saddr, sport, daddr, hnum);
> > /* Fall back to scoring if group has connections */
> > - if (result && !reuseport_has_conns(sk, false))
> > + if (result && !reuseport_has_conns(sk))
> > return result;
> >
> > result = result ? : sk;
> > --
> > 2.30.2
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH v2 net] udp: Update reuse->has_conns under reuseport_lock.
2022-10-12 19:27 ` Kuniyuki Iwashima
@ 2022-10-13 16:09 ` Eric Dumazet
2022-10-13 16:51 ` Kuniyuki Iwashima
2022-10-13 17:41 ` Martin KaFai Lau
0 siblings, 2 replies; 7+ messages in thread
From: Eric Dumazet @ 2022-10-13 16:09 UTC (permalink / raw)
To: Kuniyuki Iwashima
Cc: davem, dsahern, kraig, kuba, kuni1840, martin.lau, netdev,
pabeni, willemb, yoshfuji
On Wed, Oct 12, 2022 at 12:28 PM Kuniyuki Iwashima <kuniyu@amazon.com> wrote:
>
> From: Eric Dumazet <edumazet@google.com>
> Date: Wed, 12 Oct 2022 11:59:43 -0700
> > On Wed, Oct 12, 2022 at 11:53 AM Kuniyuki Iwashima <kuniyu@amazon.com> wrote:
> > >
> > > When we call connect() for a UDP socket in a reuseport group, we have
> > > to update sk->sk_reuseport_cb->has_conns to 1. Otherwise, the kernel
> > > could select a unconnected socket wrongly for packets sent to the
> > > connected socket.
> > >
> > > However, the current way to set has_conns is illegal and possible to
> > > trigger that problem. reuseport_has_conns() changes has_conns under
> > > rcu_read_lock(), which upgrades the RCU reader to the updater. Then,
> > > it must do the update under the updater's lock, reuseport_lock, but
> > > it doesn't for now.
> > >
> > > For this reason, there is a race below where we fail to set has_conns
> > > resulting in the wrong socket selection. To avoid the race, let's split
> > > the reader and updater with proper locking.
> > >
> > > cpu1 cpu2
> > > +----+ +----+
> > >
> > > __ip[46]_datagram_connect() reuseport_grow()
> > > . .
> > > |- reuseport_has_conns(sk, true) |- more_reuse = __reuseport_alloc(more_socks_size)
> > > | . |
> > > | |- rcu_read_lock()
> > > | |- reuse = rcu_dereference(sk->sk_reuseport_cb)
> > > | |
> > > | | | /* reuse->has_conns == 0 here */
> > > | | |- more_reuse->has_conns = reuse->has_conns
> > > | |- reuse->has_conns = 1 | /* more_reuse->has_conns SHOULD BE 1 HERE */
> > > | | |
> > > | | |- rcu_assign_pointer(reuse->socks[i]->sk_reuseport_cb,
> > > | | | more_reuse)
> > > | `- rcu_read_unlock() `- kfree_rcu(reuse, rcu)
> > > |
> > > |- sk->sk_state = TCP_ESTABLISHED
> > >
> > > Fixes: acdcecc61285 ("udp: correct reuseport selection with connected sockets")
> > > Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
> > > ---
> > > v2:
> > > * Fix build failure for CONFIG_IPV6=m
> > > * Drop SO_INCOMING_CPU fix, which will be sent for net-next
> > > after the v6.1 merge window
> > >
> > > v1: https://lore.kernel.org/netdev/20221010174351.11024-1-kuniyu@amazon.com/
> > > ---
> > > include/net/sock_reuseport.h | 11 +++++------
> > > net/core/sock_reuseport.c | 15 +++++++++++++++
> > > net/ipv4/datagram.c | 2 +-
> > > net/ipv4/udp.c | 2 +-
> > > net/ipv6/datagram.c | 2 +-
> > > net/ipv6/udp.c | 2 +-
> > > 6 files changed, 24 insertions(+), 10 deletions(-)
> > >
> > > diff --git a/include/net/sock_reuseport.h b/include/net/sock_reuseport.h
> > > index 473b0b0fa4ab..efc9085c6892 100644
> > > --- a/include/net/sock_reuseport.h
> > > +++ b/include/net/sock_reuseport.h
> > > @@ -43,21 +43,20 @@ struct sock *reuseport_migrate_sock(struct sock *sk,
> > > extern int reuseport_attach_prog(struct sock *sk, struct bpf_prog *prog);
> > > extern int reuseport_detach_prog(struct sock *sk);
> > >
> > > -static inline bool reuseport_has_conns(struct sock *sk, bool set)
> > > +static inline bool reuseport_has_conns(struct sock *sk)
> > > {
> > > struct sock_reuseport *reuse;
> > > bool ret = false;
> > >
> > > rcu_read_lock();
> > > reuse = rcu_dereference(sk->sk_reuseport_cb);
> > > - if (reuse) {
> > > - if (set)
> > > - reuse->has_conns = 1;
> > > - ret = reuse->has_conns;
> > > - }
> > > + if (reuse && reuse->has_conns)
> > > + ret = true;
> > > rcu_read_unlock();
> > >
> > > return ret;
> > > }
> > >
> > > +void reuseport_has_conns_set(struct sock *sk);
> > > +
> > > #endif /* _SOCK_REUSEPORT_H */
> > > diff --git a/net/core/sock_reuseport.c b/net/core/sock_reuseport.c
> > > index 5daa1fa54249..abb414ed4aa7 100644
> > > --- a/net/core/sock_reuseport.c
> > > +++ b/net/core/sock_reuseport.c
> > > @@ -21,6 +21,21 @@ static DEFINE_IDA(reuseport_ida);
> > > static int reuseport_resurrect(struct sock *sk, struct sock_reuseport *old_reuse,
> > > struct sock_reuseport *reuse, bool bind_inany);
> > >
> > > +void reuseport_has_conns_set(struct sock *sk)
> > > +{
> > > + struct sock_reuseport *reuse;
> > > +
> > > + if (!rcu_access_pointer(sk->sk_reuseport_cb))
> > > + return;
> > > +
> > > + spin_lock(&reuseport_lock);
> > > + reuse = rcu_dereference_protected(sk->sk_reuseport_cb,
> > > + lockdep_is_held(&reuseport_lock));
> >
> > Could @reuse be NULL at this point ?
> >
> > Previous test was performed without reuseport_lock being held.
>
> Usually, sk_reuseport_cb is changed under lock_sock().
>
> The only exception is reuseport_grow() & TCP reqsk migration case.
>
> 1) shutdown() TCP listener, which is moved into the latter part of
> reuse->socks[] to migrate reqsk.
>
> 2) New listen() overflows reuse->socks[] and call reuseport_grow().
>
> 3) reuse->max_socks overflows u16 with the new listener.
>
> 4) reuseport_grow() pops the old shutdown()ed listener from the array
> and update its sk->sk_reuseport_cb as NULL without lock_sock().
>
> shutdown()ed sk->sk_reuseport_cb can be changed without lock_sock().
>
> But, reuseport_has_conns_set() is called only for UDP and under
> lock_sock(), so @reuse never be NULL in this case.
Given the complexity of this code and how much time is needed to
review all possibilities, please add an additional
if (reuse)
reuse->has_conns = 1;
I doubt anyone will object to such safety measures.
Thanks.
>
>
> > > + reuse->has_conns = 1;
> > > + spin_unlock(&reuseport_lock);
> > > +}
> > > +EXPORT_SYMBOL(reuseport_has_conns_set);
> > > +
> > > static int reuseport_sock_index(struct sock *sk,
> > > const struct sock_reuseport *reuse,
> > > bool closed)
> > > diff --git a/net/ipv4/datagram.c b/net/ipv4/datagram.c
> > > index 405a8c2aea64..5e66add7befa 100644
> > > --- a/net/ipv4/datagram.c
> > > +++ b/net/ipv4/datagram.c
> > > @@ -70,7 +70,7 @@ int __ip4_datagram_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len
> > > }
> > > inet->inet_daddr = fl4->daddr;
> > > inet->inet_dport = usin->sin_port;
> > > - reuseport_has_conns(sk, true);
> > > + reuseport_has_conns_set(sk);
> > > sk->sk_state = TCP_ESTABLISHED;
> > > sk_set_txhash(sk);
> > > inet->inet_id = prandom_u32();
> > > diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
> > > index d63118ce5900..29228231b058 100644
> > > --- a/net/ipv4/udp.c
> > > +++ b/net/ipv4/udp.c
> > > @@ -448,7 +448,7 @@ static struct sock *udp4_lib_lookup2(struct net *net,
> > > result = lookup_reuseport(net, sk, skb,
> > > saddr, sport, daddr, hnum);
> > > /* Fall back to scoring if group has connections */
> > > - if (result && !reuseport_has_conns(sk, false))
> > > + if (result && !reuseport_has_conns(sk))
> > > return result;
> > >
> > > result = result ? : sk;
> > > diff --git a/net/ipv6/datagram.c b/net/ipv6/datagram.c
> > > index df665d4e8f0f..5ecb56522f9d 100644
> > > --- a/net/ipv6/datagram.c
> > > +++ b/net/ipv6/datagram.c
> > > @@ -256,7 +256,7 @@ int __ip6_datagram_connect(struct sock *sk, struct sockaddr *uaddr,
> > > goto out;
> > > }
> > >
> > > - reuseport_has_conns(sk, true);
> > > + reuseport_has_conns_set(sk);
> > > sk->sk_state = TCP_ESTABLISHED;
> > > sk_set_txhash(sk);
> > > out:
> > > diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
> > > index 91e795bb9ade..56e4523a3004 100644
> > > --- a/net/ipv6/udp.c
> > > +++ b/net/ipv6/udp.c
> > > @@ -182,7 +182,7 @@ static struct sock *udp6_lib_lookup2(struct net *net,
> > > result = lookup_reuseport(net, sk, skb,
> > > saddr, sport, daddr, hnum);
> > > /* Fall back to scoring if group has connections */
> > > - if (result && !reuseport_has_conns(sk, false))
> > > + if (result && !reuseport_has_conns(sk))
> > > return result;
> > >
> > > result = result ? : sk;
> > > --
> > > 2.30.2
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH v2 net] udp: Update reuse->has_conns under reuseport_lock.
2022-10-13 16:09 ` Eric Dumazet
@ 2022-10-13 16:51 ` Kuniyuki Iwashima
2022-10-13 17:41 ` Martin KaFai Lau
1 sibling, 0 replies; 7+ messages in thread
From: Kuniyuki Iwashima @ 2022-10-13 16:51 UTC (permalink / raw)
To: edumazet
Cc: davem, dsahern, kraig, kuba, kuni1840, kuniyu, martin.lau,
netdev, pabeni, willemb, yoshfuji
From: Eric Dumazet <edumazet@google.com>
Date: Thu, 13 Oct 2022 09:09:31 -0700
> On Wed, Oct 12, 2022 at 12:28 PM Kuniyuki Iwashima <kuniyu@amazon.com> wrote:
> >
> > From: Eric Dumazet <edumazet@google.com>
> > Date: Wed, 12 Oct 2022 11:59:43 -0700
> > > On Wed, Oct 12, 2022 at 11:53 AM Kuniyuki Iwashima <kuniyu@amazon.com> wrote:
> > > >
> > > > When we call connect() for a UDP socket in a reuseport group, we have
> > > > to update sk->sk_reuseport_cb->has_conns to 1. Otherwise, the kernel
> > > > could select a unconnected socket wrongly for packets sent to the
> > > > connected socket.
> > > >
> > > > However, the current way to set has_conns is illegal and possible to
> > > > trigger that problem. reuseport_has_conns() changes has_conns under
> > > > rcu_read_lock(), which upgrades the RCU reader to the updater. Then,
> > > > it must do the update under the updater's lock, reuseport_lock, but
> > > > it doesn't for now.
> > > >
> > > > For this reason, there is a race below where we fail to set has_conns
> > > > resulting in the wrong socket selection. To avoid the race, let's split
> > > > the reader and updater with proper locking.
> > > >
> > > > cpu1 cpu2
> > > > +----+ +----+
> > > >
> > > > __ip[46]_datagram_connect() reuseport_grow()
> > > > . .
> > > > |- reuseport_has_conns(sk, true) |- more_reuse = __reuseport_alloc(more_socks_size)
> > > > | . |
> > > > | |- rcu_read_lock()
> > > > | |- reuse = rcu_dereference(sk->sk_reuseport_cb)
> > > > | |
> > > > | | | /* reuse->has_conns == 0 here */
> > > > | | |- more_reuse->has_conns = reuse->has_conns
> > > > | |- reuse->has_conns = 1 | /* more_reuse->has_conns SHOULD BE 1 HERE */
> > > > | | |
> > > > | | |- rcu_assign_pointer(reuse->socks[i]->sk_reuseport_cb,
> > > > | | | more_reuse)
> > > > | `- rcu_read_unlock() `- kfree_rcu(reuse, rcu)
> > > > |
> > > > |- sk->sk_state = TCP_ESTABLISHED
> > > >
> > > > Fixes: acdcecc61285 ("udp: correct reuseport selection with connected sockets")
> > > > Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
> > > > ---
> > > > v2:
> > > > * Fix build failure for CONFIG_IPV6=m
> > > > * Drop SO_INCOMING_CPU fix, which will be sent for net-next
> > > > after the v6.1 merge window
> > > >
> > > > v1: https://lore.kernel.org/netdev/20221010174351.11024-1-kuniyu@amazon.com/
> > > > ---
> > > > include/net/sock_reuseport.h | 11 +++++------
> > > > net/core/sock_reuseport.c | 15 +++++++++++++++
> > > > net/ipv4/datagram.c | 2 +-
> > > > net/ipv4/udp.c | 2 +-
> > > > net/ipv6/datagram.c | 2 +-
> > > > net/ipv6/udp.c | 2 +-
> > > > 6 files changed, 24 insertions(+), 10 deletions(-)
> > > >
> > > > diff --git a/include/net/sock_reuseport.h b/include/net/sock_reuseport.h
> > > > index 473b0b0fa4ab..efc9085c6892 100644
> > > > --- a/include/net/sock_reuseport.h
> > > > +++ b/include/net/sock_reuseport.h
> > > > @@ -43,21 +43,20 @@ struct sock *reuseport_migrate_sock(struct sock *sk,
> > > > extern int reuseport_attach_prog(struct sock *sk, struct bpf_prog *prog);
> > > > extern int reuseport_detach_prog(struct sock *sk);
> > > >
> > > > -static inline bool reuseport_has_conns(struct sock *sk, bool set)
> > > > +static inline bool reuseport_has_conns(struct sock *sk)
> > > > {
> > > > struct sock_reuseport *reuse;
> > > > bool ret = false;
> > > >
> > > > rcu_read_lock();
> > > > reuse = rcu_dereference(sk->sk_reuseport_cb);
> > > > - if (reuse) {
> > > > - if (set)
> > > > - reuse->has_conns = 1;
> > > > - ret = reuse->has_conns;
> > > > - }
> > > > + if (reuse && reuse->has_conns)
> > > > + ret = true;
> > > > rcu_read_unlock();
> > > >
> > > > return ret;
> > > > }
> > > >
> > > > +void reuseport_has_conns_set(struct sock *sk);
> > > > +
> > > > #endif /* _SOCK_REUSEPORT_H */
> > > > diff --git a/net/core/sock_reuseport.c b/net/core/sock_reuseport.c
> > > > index 5daa1fa54249..abb414ed4aa7 100644
> > > > --- a/net/core/sock_reuseport.c
> > > > +++ b/net/core/sock_reuseport.c
> > > > @@ -21,6 +21,21 @@ static DEFINE_IDA(reuseport_ida);
> > > > static int reuseport_resurrect(struct sock *sk, struct sock_reuseport *old_reuse,
> > > > struct sock_reuseport *reuse, bool bind_inany);
> > > >
> > > > +void reuseport_has_conns_set(struct sock *sk)
> > > > +{
> > > > + struct sock_reuseport *reuse;
> > > > +
> > > > + if (!rcu_access_pointer(sk->sk_reuseport_cb))
> > > > + return;
> > > > +
> > > > + spin_lock(&reuseport_lock);
> > > > + reuse = rcu_dereference_protected(sk->sk_reuseport_cb,
> > > > + lockdep_is_held(&reuseport_lock));
> > >
> > > Could @reuse be NULL at this point ?
> > >
> > > Previous test was performed without reuseport_lock being held.
> >
> > Usually, sk_reuseport_cb is changed under lock_sock().
> >
> > The only exception is reuseport_grow() & TCP reqsk migration case.
> >
> > 1) shutdown() TCP listener, which is moved into the latter part of
> > reuse->socks[] to migrate reqsk.
> >
> > 2) New listen() overflows reuse->socks[] and call reuseport_grow().
> >
> > 3) reuse->max_socks overflows u16 with the new listener.
> >
> > 4) reuseport_grow() pops the old shutdown()ed listener from the array
> > and update its sk->sk_reuseport_cb as NULL without lock_sock().
> >
> > shutdown()ed sk->sk_reuseport_cb can be changed without lock_sock().
> >
> > But, reuseport_has_conns_set() is called only for UDP and under
> > lock_sock(), so @reuse never be NULL in this case.
>
> Given the complexity of this code and how much time is needed to
> review all possibilities, please add an additional
>
> if (reuse)
> reuse->has_conns = 1;
>
> I doubt anyone will object to such safety measures.
I see.
If no one has any objections, I'll respin tomorrow with likely().
Thank you.
>
> Thanks.
>
> >
> >
> > > > + reuse->has_conns = 1;
> > > > + spin_unlock(&reuseport_lock);
> > > > +}
> > > > +EXPORT_SYMBOL(reuseport_has_conns_set);
> > > > +
> > > > static int reuseport_sock_index(struct sock *sk,
> > > > const struct sock_reuseport *reuse,
> > > > bool closed)
> > > > diff --git a/net/ipv4/datagram.c b/net/ipv4/datagram.c
> > > > index 405a8c2aea64..5e66add7befa 100644
> > > > --- a/net/ipv4/datagram.c
> > > > +++ b/net/ipv4/datagram.c
> > > > @@ -70,7 +70,7 @@ int __ip4_datagram_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len
> > > > }
> > > > inet->inet_daddr = fl4->daddr;
> > > > inet->inet_dport = usin->sin_port;
> > > > - reuseport_has_conns(sk, true);
> > > > + reuseport_has_conns_set(sk);
> > > > sk->sk_state = TCP_ESTABLISHED;
> > > > sk_set_txhash(sk);
> > > > inet->inet_id = prandom_u32();
> > > > diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
> > > > index d63118ce5900..29228231b058 100644
> > > > --- a/net/ipv4/udp.c
> > > > +++ b/net/ipv4/udp.c
> > > > @@ -448,7 +448,7 @@ static struct sock *udp4_lib_lookup2(struct net *net,
> > > > result = lookup_reuseport(net, sk, skb,
> > > > saddr, sport, daddr, hnum);
> > > > /* Fall back to scoring if group has connections */
> > > > - if (result && !reuseport_has_conns(sk, false))
> > > > + if (result && !reuseport_has_conns(sk))
> > > > return result;
> > > >
> > > > result = result ? : sk;
> > > > diff --git a/net/ipv6/datagram.c b/net/ipv6/datagram.c
> > > > index df665d4e8f0f..5ecb56522f9d 100644
> > > > --- a/net/ipv6/datagram.c
> > > > +++ b/net/ipv6/datagram.c
> > > > @@ -256,7 +256,7 @@ int __ip6_datagram_connect(struct sock *sk, struct sockaddr *uaddr,
> > > > goto out;
> > > > }
> > > >
> > > > - reuseport_has_conns(sk, true);
> > > > + reuseport_has_conns_set(sk);
> > > > sk->sk_state = TCP_ESTABLISHED;
> > > > sk_set_txhash(sk);
> > > > out:
> > > > diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
> > > > index 91e795bb9ade..56e4523a3004 100644
> > > > --- a/net/ipv6/udp.c
> > > > +++ b/net/ipv6/udp.c
> > > > @@ -182,7 +182,7 @@ static struct sock *udp6_lib_lookup2(struct net *net,
> > > > result = lookup_reuseport(net, sk, skb,
> > > > saddr, sport, daddr, hnum);
> > > > /* Fall back to scoring if group has connections */
> > > > - if (result && !reuseport_has_conns(sk, false))
> > > > + if (result && !reuseport_has_conns(sk))
> > > > return result;
> > > >
> > > > result = result ? : sk;
> > > > --
> > > > 2.30.2
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH v2 net] udp: Update reuse->has_conns under reuseport_lock.
2022-10-13 16:09 ` Eric Dumazet
2022-10-13 16:51 ` Kuniyuki Iwashima
@ 2022-10-13 17:41 ` Martin KaFai Lau
2022-10-13 18:46 ` Kuniyuki Iwashima
1 sibling, 1 reply; 7+ messages in thread
From: Martin KaFai Lau @ 2022-10-13 17:41 UTC (permalink / raw)
To: Kuniyuki Iwashima
Cc: davem, dsahern, Eric Dumazet, kraig, kuba, kuni1840, martin.lau,
netdev, pabeni, willemb, yoshfuji
On 10/13/22 9:09 AM, Eric Dumazet wrote:
>>>> diff --git a/net/core/sock_reuseport.c b/net/core/sock_reuseport.c
>>>> index 5daa1fa54249..abb414ed4aa7 100644
>>>> --- a/net/core/sock_reuseport.c
>>>> +++ b/net/core/sock_reuseport.c
>>>> @@ -21,6 +21,21 @@ static DEFINE_IDA(reuseport_ida);
>>>> static int reuseport_resurrect(struct sock *sk, struct sock_reuseport *old_reuse,
>>>> struct sock_reuseport *reuse, bool bind_inany);
>>>>
>>>> +void reuseport_has_conns_set(struct sock *sk)
>>>> +{
>>>> + struct sock_reuseport *reuse;
>>>> +
>>>> + if (!rcu_access_pointer(sk->sk_reuseport_cb))
>>>> + return;
>>>> +
>>>> + spin_lock(&reuseport_lock);
It seems other paths are still using the spin_lock_bh(). It will be useful to
have a few words here why _bh() is not needed.
>>>> + reuse = rcu_dereference_protected(sk->sk_reuseport_cb,
>>>> + lockdep_is_held(&reuseport_lock));
>>>
>>> Could @reuse be NULL at this point ?
>>>
>>> Previous test was performed without reuseport_lock being held.
>>
>> Usually, sk_reuseport_cb is changed under lock_sock().
>>
>> The only exception is reuseport_grow() & TCP reqsk migration case.
>>
>> 1) shutdown() TCP listener, which is moved into the latter part of
>> reuse->socks[] to migrate reqsk.
>>
>> 2) New listen() overflows reuse->socks[] and call reuseport_grow().
>>
>> 3) reuse->max_socks overflows u16 with the new listener.
>>
>> 4) reuseport_grow() pops the old shutdown()ed listener from the array
>> and update its sk->sk_reuseport_cb as NULL without lock_sock().
>>
>> shutdown()ed sk->sk_reuseport_cb can be changed without lock_sock().
>>
>> But, reuseport_has_conns_set() is called only for UDP and under
>> lock_sock(), so @reuse never be NULL in this case.
>
> Given the complexity of this code and how much time is needed to
> review all possibilities, please add an additional
>
> if (reuse)
> reuse->has_conns = 1;
+1
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH v2 net] udp: Update reuse->has_conns under reuseport_lock.
2022-10-13 17:41 ` Martin KaFai Lau
@ 2022-10-13 18:46 ` Kuniyuki Iwashima
0 siblings, 0 replies; 7+ messages in thread
From: Kuniyuki Iwashima @ 2022-10-13 18:46 UTC (permalink / raw)
To: martin.lau
Cc: davem, dsahern, edumazet, kraig, kuba, kuni1840, kuniyu,
martin.lau, netdev, pabeni, willemb
From: Martin KaFai Lau <martin.lau@linux.dev>
Date: Thu, 13 Oct 2022 10:41:53 -0700
> On 10/13/22 9:09 AM, Eric Dumazet wrote:
> >>>> diff --git a/net/core/sock_reuseport.c b/net/core/sock_reuseport.c
> >>>> index 5daa1fa54249..abb414ed4aa7 100644
> >>>> --- a/net/core/sock_reuseport.c
> >>>> +++ b/net/core/sock_reuseport.c
> >>>> @@ -21,6 +21,21 @@ static DEFINE_IDA(reuseport_ida);
> >>>> static int reuseport_resurrect(struct sock *sk, struct sock_reuseport *old_reuse,
> >>>> struct sock_reuseport *reuse, bool bind_inany);
> >>>>
> >>>> +void reuseport_has_conns_set(struct sock *sk)
> >>>> +{
> >>>> + struct sock_reuseport *reuse;
> >>>> +
> >>>> + if (!rcu_access_pointer(sk->sk_reuseport_cb))
> >>>> + return;
> >>>> +
> >>>> + spin_lock(&reuseport_lock);
>
> It seems other paths are still using the spin_lock_bh(). It will be useful to
> have a few words here why _bh() is not needed.
I think I forgot to add _bh(), but I'm now wondering what is the hlist
lock mentioned in reuseport_alloc()...
> >>>> + reuse = rcu_dereference_protected(sk->sk_reuseport_cb,
> >>>> + lockdep_is_held(&reuseport_lock));
> >>>
> >>> Could @reuse be NULL at this point ?
> >>>
> >>> Previous test was performed without reuseport_lock being held.
> >>
> >> Usually, sk_reuseport_cb is changed under lock_sock().
> >>
> >> The only exception is reuseport_grow() & TCP reqsk migration case.
> >>
> >> 1) shutdown() TCP listener, which is moved into the latter part of
> >> reuse->socks[] to migrate reqsk.
> >>
> >> 2) New listen() overflows reuse->socks[] and call reuseport_grow().
> >>
> >> 3) reuse->max_socks overflows u16 with the new listener.
> >>
> >> 4) reuseport_grow() pops the old shutdown()ed listener from the array
> >> and update its sk->sk_reuseport_cb as NULL without lock_sock().
> >>
> >> shutdown()ed sk->sk_reuseport_cb can be changed without lock_sock().
> >>
> >> But, reuseport_has_conns_set() is called only for UDP and under
> >> lock_sock(), so @reuse never be NULL in this case.
> >
> > Given the complexity of this code and how much time is needed to
> > review all possibilities, please add an additional
> >
> > if (reuse)
> > reuse->has_conns = 1;
>
> +1
Acked.
Thank you.
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2022-10-13 18:56 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-10-12 18:52 [PATCH v2 net] udp: Update reuse->has_conns under reuseport_lock Kuniyuki Iwashima
2022-10-12 18:59 ` Eric Dumazet
2022-10-12 19:27 ` Kuniyuki Iwashima
2022-10-13 16:09 ` Eric Dumazet
2022-10-13 16:51 ` Kuniyuki Iwashima
2022-10-13 17:41 ` Martin KaFai Lau
2022-10-13 18:46 ` Kuniyuki Iwashima
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).