Skip to contentSkip to conversation

dco: send SIGUSR1 upon ping timeout

PATCH In review Antonio Quartulli 3 participants 3 messages Jan 12, 2023

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 dco: send SIGUSR1 upon ping timeout

When a peer is removed with reason "ping expire", we should kill the instance with SIGUSR1 and not SIGTERM

--

Arne, I am not 100% sure why but it seems for ping-restart we always use SIGUSR1, right? but the DCO handling code was apparently using SIGTERM.

What do you think?

---
 src/openvpn/multi.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
src/openvpn/multi.c +3 -1
index 99123c39..10efffec 100644
@@ -3234,10 +3234,12 @@ process_incoming_del_peer(struct multi_context *m, struct multi_instance *mi,
                           dco_context_t *dco)
 {
     const char *reason = "ovpn-dco: unknown reason";
+    int signal = SIGTERM;
     switch (dco->dco_del_peer_reason)
     {
         case OVPN_DEL_PEER_REASON_EXPIRED:
             reason = "ovpn-dco: ping expired";
+            signal = SIGUSR1;
             break;
 
         case OVPN_DEL_PEER_REASON_TRANSPORT_ERROR:
@@ -3270,7 +3272,7 @@ process_incoming_del_peer(struct multi_context *m, struct multi_instance *mi,
     mi->context.sig->signal_text = reason;
     mi->context.c2.dco_read_bytes = dco->dco_read_bytes;
     mi->context.c2.dco_write_bytes = dco->dco_write_bytes;
-    multi_signal_instance(m, mi, SIGTERM);
+    multi_signal_instance(m, mi, signal);
 }
 
 bool
Signature · 1 line
2.38.2
#1 lore raw

Am 12.01.23 um 01:04 schrieb Antonio Quartulli:

Quoted 11 lines

When a peer is removed with reason "ping expire", we should kill the instance with SIGUSR1 and not SIGTERM

--

Arne, I am not 100% sure why but it seems for ping-restart we always use SIGUSR1, right? but the DCO handling code was apparently using SIGTERM.

What do you think?

This makes no difference in p2mp mode.

Arne

#2 lore raw

Hi,

Quoted 7 lines · Arne Schwabe

Am 12.01.23 um 01:04 schrieb Antonio Quartulli:

Quoted 4 lines

Arne, I am not 100% sure why but it seems for ping-restart we always use SIGUSR1, right? but the DCO handling code was apparently using SIGTERM.

What do you think?

This makes no difference in p2mp mode.

This is how I remember the code, from the "incoming explicit-exit-notify vs. p2p/p2mp server" testing. Logging is different for SIGUSR1 vs. SIGTERM instances ("restarting / terminating client instance" or something like that) but *then* the same thing happens.

You can't "restart" an incoming session ;-)

-> putting this to "Rejected" in patchwork.

gert

Signature · 6 lines
"If was one thing all people took for granted, was conviction that if you 
 feed honest figures into a computer, honest figures come out. Never doubted 
 it myself till I met a computer with a sense of humor."
                             Robert A. Heinlein, The Moon is a Harsh Mistress

Gert Doering - Munich, Germany                             [email protected]...