On 8/22/24 1:45 PM, John Fastabend wrote:
Cong Wang wrote:
On Wed, Aug 21, 2024 at 03:55:33PM +0100, Simon Horman wrote:
On Tue, Aug 20, 2024 at 08:07:44PM -0700, Cong Wang wrote:
From: Cong Wang <cong.wang@xxxxxxxxxxxxx>
When we cork messages in psock->cork, the last message triggers the
flushing will result in sending a sk_msg larger than the current
message size. In this case, in tcp_bpf_send_verdict(), 'copied' becomes
negative at least in the following case:
468 case __SK_DROP:
469 default:
470 sk_msg_free_partial(sk, msg, tosend);
471 sk_msg_apply_bytes(psock, tosend);
472 *copied -= (tosend + delta); // <==== HERE
473 return -EACCES;
Therefore, it could lead to the following BUG with a proper value of
'copied' (thanks to syzbot). We should not use negative 'copied' as a
return value here.
------------[ cut here ]------------
kernel BUG at net/socket.c:733!
Internal error: Oops - BUG: 00000000f2000800 [#1] PREEMPT SMP
Modules linked in:
CPU: 0 UID: 0 PID: 3265 Comm: syz-executor510 Not tainted 6.11.0-rc3-syzkaller-00060-gd07b43284ab3 #0
Hardware name: linux,dummy-virt (DT)
pstate: 61400009 (nZCv daif +PAN -UAO -TCO +DIT -SSBS BTYPE=--)
pc : sock_sendmsg_nosec net/socket.c:733 [inline]
pc : sock_sendmsg_nosec net/socket.c:728 [inline]
pc : __sock_sendmsg+0x5c/0x60 net/socket.c:745
lr : sock_sendmsg_nosec net/socket.c:730 [inline]
lr : __sock_sendmsg+0x54/0x60 net/socket.c:745
sp : ffff800088ea3b30
x29: ffff800088ea3b30 x28: fbf00000062bc900 x27: 0000000000000000
x26: ffff800088ea3bc0 x25: ffff800088ea3bc0 x24: 0000000000000000
x23: f9f00000048dc000 x22: 0000000000000000 x21: ffff800088ea3d90
x20: f9f00000048dc000 x19: ffff800088ea3d90 x18: 0000000000000001
x17: 0000000000000000 x16: 0000000000000000 x15: 000000002002ffaf
x14: 0000000000000000 x13: 0000000000000000 x12: 0000000000000000
x11: 0000000000000000 x10: ffff8000815849c0 x9 : ffff8000815b49c0
x8 : 0000000000000000 x7 : 000000000000003f x6 : 0000000000000000
x5 : 00000000000007e0 x4 : fff07ffffd239000 x3 : fbf00000062bc900
x2 : 0000000000000000 x1 : 0000000000000000 x0 : 00000000fffffdef
Call trace:
sock_sendmsg_nosec net/socket.c:733 [inline]
__sock_sendmsg+0x5c/0x60 net/socket.c:745
____sys_sendmsg+0x274/0x2ac net/socket.c:2597
___sys_sendmsg+0xac/0x100 net/socket.c:2651
__sys_sendmsg+0x84/0xe0 net/socket.c:2680
__do_sys_sendmsg net/socket.c:2689 [inline]
__se_sys_sendmsg net/socket.c:2687 [inline]
__arm64_sys_sendmsg+0x24/0x30 net/socket.c:2687
__invoke_syscall arch/arm64/kernel/syscall.c:35 [inline]
invoke_syscall+0x48/0x110 arch/arm64/kernel/syscall.c:49
el0_svc_common.constprop.0+0x40/0xe0 arch/arm64/kernel/syscall.c:132
do_el0_svc+0x1c/0x28 arch/arm64/kernel/syscall.c:151
el0_svc+0x34/0xec arch/arm64/kernel/entry-common.c:712
el0t_64_sync_handler+0x100/0x12c arch/arm64/kernel/entry-common.c:730
el0t_64_sync+0x19c/0x1a0 arch/arm64/kernel/entry.S:598
Code: f9404463 d63f0060 3108441f 54fffe81 (d4210000)
---[ end trace 0000000000000000 ]---
Fixes: 4f738adba30a ("bpf: create tcp_bpf_ulp allowing BPF to monitor socket TX/RX data")
Reported-by: syzbot+58c03971700330ce14d8@xxxxxxxxxxxxxxxxxxxxxxxxx
Cc: John Fastabend <john.fastabend@xxxxxxxxx>
Cc: Jakub Sitnicki <jakub@xxxxxxxxxxxxxx>
Signed-off-by: Cong Wang <cong.wang@xxxxxxxxxxxxx>
---
net/ipv4/tcp_bpf.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/ipv4/tcp_bpf.c b/net/ipv4/tcp_bpf.c
index 53b0d62fd2c2..fe6178715ba0 100644
--- a/net/ipv4/tcp_bpf.c
+++ b/net/ipv4/tcp_bpf.c
@@ -577,7 +577,7 @@ static int tcp_bpf_sendmsg(struct sock *sk, struct msghdr *msg, size_t size)
err = sk_stream_error(sk, msg->msg_flags, err);
release_sock(sk);
sk_psock_put(sk, psock);
- return copied ? copied : err;
+ return copied > 0 ? copied : err;
Does it make more sense to make the condition err:
is err 0 iif everything is ok? (completely untested!)
Mind to elaborate?
From my point of view, 'copied' is to handle partial transmission, for
example:
0. User wants to send 2 * 1K bytes with sendmsg()
1. Kernel already sent the first 1K successfully
2. Kernel got some error when sending the 2nd 1K
In this scenario, we should return 1K instead of the error to the caller to
indicate this partial transmission situation, otherwise we could not
distinguish it with a compete failure (that is, 0 byte sent).
Yep, if we don't return the positive value on partial send we will confuse
apps and they will probably resent data.
From my side this looks good.
Reviewed-by: John Fastabend <john.fastabend@xxxxxxxxx>
Acked-by: Martin KaFai Lau <martin.lau@xxxxxxxxxx>
Jakub, can you directly land it to the net tree?
