[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20220727060921.2373314-1-kafai@fb.com>
Date: Tue, 26 Jul 2022 23:09:21 -0700
From: Martin KaFai Lau <kafai@...com>
To: <bpf@...r.kernel.org>, <netdev@...r.kernel.org>
CC: Alexei Starovoitov <ast@...nel.org>,
Andrii Nakryiko <andrii@...nel.org>,
Daniel Borkmann <daniel@...earbox.net>,
David Miller <davem@...emloft.net>,
Eric Dumazet <edumazet@...gle.com>,
Jakub Kicinski <kuba@...nel.org>, <kernel-team@...com>,
Paolo Abeni <pabeni@...hat.com>
Subject: [PATCH bpf-next 04/14] bpf: net: Avoid do_tcp_setsockopt() taking sk lock when called from bpf
Similar to the earlier patch that avoids sock_setsockopt() from
taking sk lock when called from bpf. This patch changes
do_tcp_setsockopt() to use the {lock,release}_sock_sockopt().
This patch also changes do_tcp_setsockopt() to check optval.is_bpf
when passing the cap_net_admin arg to
tcp_set_congestion_control(..., cap_net_admin). It is
the same as how bpf_setsockopt(TCP_CONGESTION) is calling
tcp_set_congestion_control() now.
Signed-off-by: Martin KaFai Lau <kafai@...com>
---
net/ipv4/tcp.c | 19 ++++++++++---------
1 file changed, 10 insertions(+), 9 deletions(-)
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index ba2bdc811374..7f8d81befa8e 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -3462,11 +3462,12 @@ static int do_tcp_setsockopt(struct sock *sk, int level, int optname,
return -EFAULT;
name[val] = 0;
- lock_sock(sk);
- err = tcp_set_congestion_control(sk, name, true,
- ns_capable(sock_net(sk)->user_ns,
- CAP_NET_ADMIN));
- release_sock(sk);
+ lock_sock_sockopt(sk, optval);
+ err = tcp_set_congestion_control(sk, name, !optval.is_bpf,
+ optval.is_bpf ?
+ true : ns_capable(sock_net(sk)->user_ns,
+ CAP_NET_ADMIN));
+ release_sock_sockopt(sk, optval);
return err;
}
case TCP_ULP: {
@@ -3482,9 +3483,9 @@ static int do_tcp_setsockopt(struct sock *sk, int level, int optname,
return -EFAULT;
name[val] = 0;
- lock_sock(sk);
+ lock_sock_sockopt(sk, optval);
err = tcp_set_ulp(sk, name);
- release_sock(sk);
+ release_sock_sockopt(sk, optval);
return err;
}
case TCP_FASTOPEN_KEY: {
@@ -3517,7 +3518,7 @@ static int do_tcp_setsockopt(struct sock *sk, int level, int optname,
if (copy_from_sockptr(&val, optval, sizeof(val)))
return -EFAULT;
- lock_sock(sk);
+ lock_sock_sockopt(sk, optval);
switch (optname) {
case TCP_MAXSEG:
@@ -3739,7 +3740,7 @@ static int do_tcp_setsockopt(struct sock *sk, int level, int optname,
break;
}
- release_sock(sk);
+ release_sock_sockopt(sk, optval);
return err;
}
--
2.30.2
Powered by blists - more mailing lists