android/android_kernel_samsung_sm8650
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Revert "bpf, sockmap: TCP data stall on recv before accept"

Browse Source 
This reverts commit ab90b68f65.

It breaks the Android KABI and will be brought back at a later time when
it is safe to do so.

Bug: 161946584
Change-Id: If61ba37db2a3ea1642c9e5c9295e2d7d4a94207b
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
This commit is contained in:
Greg Kroah-Hartman 2023-06-14 16:16:00 +00:00
parent c7e4973846
commit 3ce63059c1
1 changed files with 0 additions and 20 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
net/ipv4/tcp_bpf.c
View File

@ -212,26 +212,6 @@ static int tcp_bpf_recvmsg_parser(struct sock *sk,
return tcp_recvmsg(sk, msg, len, flags, addr_len);
lock_sock(sk);
/* We may have received data on the sk_receive_queue pre-accept and
* then we can not use read_skb in this context because we haven't
* assigned a sk_socket yet so have no link to the ops. The work-around
* is to check the sk_receive_queue and in these cases read skbs off
* queue again. The read_skb hook is not running at this point because
* of lock_sock so we avoid having multiple runners in read_skb.
*/
if (unlikely(!skb_queue_empty(&sk->sk_receive_queue))) {
tcp_data_ready(sk);
/* This handles the ENOMEM errors if we both receive data
* pre accept and are already under memory pressure. At least
* let user know to retry.
*/
if (unlikely(!skb_queue_empty(&sk->sk_receive_queue))) {
copied = -EAGAIN;
goto out;
}
}
msg_bytes_ready:
copied = sk_msg_recvmsg(sk, psock, msg, len, flags);
/* The typical case for EFAULT is the socket was gracefully