tcp: fix incorrect ca_state in tail loss probe
Subsystems (3)
PATCH 1/1 tcp: fix incorrect ca_state in tail loss probe
On receiving an ACK that covers the loss probe sequence, TLP immediately sets the congestion state to Open, even though some packets are not recovered and retransmisssion are on the way. The later ACks may trigger a WARN_ON check in step D of tcp_fastretrans_alert(), e.g.,
https://bugzilla.redhat.com/show_bug.cgi?id=989251
The fix is to follow the similar procedure in recovery by calling tcp_try_keep_open(). The sender switches to Open state if no packets are retransmissted. Otherwise it goes to Disorder and let subsequent ACKs move the state to Recovery or Open.
--- net/ipv4/tcp_input.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
net/ipv4/tcp_input.c +1 -1
@@ -3291,7 +3291,7 @@ static void tcp_process_tlp_ack(struct sock *sk, u32 ack, int flag)tcp_init_cwnd_reduction (sk, true);tcp_set_ca_state (sk,TCP_CA_CWR );tcp_end_cwnd_reduction (sk); -tcp_set_ca_state (sk, TCP_CA_Open); +tcp_try_keep_open (sk);NET_INC_STATS_BH (sock_net (sk),LINUX_MIB_TCPLOSSPROBERECOVERY ); }
Signature · 1 line
1.8.4
Quoted 6 lines · Yuchung Cheng
On receiving an ACK that covers the loss probe sequence, TLP immediately sets the congestion state to Open, even though some packets are not recovered and retransmisssion are on the way. The later ACks may trigger a
WARN_ONcheck in step D oftcp_fastretrans_alert(), e.g.,
https://bugzilla.redhat.com/show_bug.cgi?id=989251
Quoted 8 lines
The fix is to follow the similar procedure in recovery by calling
tcp_try_keep_open(). The sender switches to Open state if no packets are retransmissted. Otherwise it goes to Disorder and let subsequent ACKs move the state to Recovery or Open.Reported-By Michael Sterrett <[email protected]>Tested-by Dormando <[email protected]>Signed-off-by Yuchung Cheng <[email protected]>
Nice catch!
neal
Nandita Dukkipati
Acked-byQuoted 13 lines · Neal Cardwell
Quoted 11 lines · Yuchung Cheng
On receiving an ACK that covers the loss probe sequence, TLP immediately sets the congestion state to Open, even though some packets are not recovered and retransmisssion are on the way. The later ACks may trigger a
WARN_ONcheck in step D oftcp_fastretrans_alert(), e.g.,
https://bugzilla.redhat.com/show_bug.cgi?id=989251The fix is to follow the similar procedure in recovery by calling
tcp_try_keep_open(). The sender switches to Open state if no packets are retransmissted. Otherwise it goes to Disorder and let subsequent ACKs move the state to Recovery or Open.
Quoted 5 lines
Quoted 3 lines
Reported-By Michael Sterrett <[email protected]>Tested-by Dormando <[email protected]>Signed-off-by Yuchung Cheng <[email protected]>Acked-by Neal Cardwell <[email protected]>