Hangs on unloading

[schwabe]

Sometimes on stopping openvpn, the interface deletion just "hangs":

2023-01-04 13:53:04 us=622820 2003:d0:270b:a600:215:5dff:fe02:811a [gremlin404326] Peer Connection Initiated with [AF_INET6]2003:d0:270b:a600:215:5dff:fe02:811a:52901
2023-01-04 13:53:04 us=622827 2003:d0:270b:a600:215:5dff:fe02:811a dco_update_keys: peer_id=-1
2023-01-04 13:53:04 us=623416 gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a MULTI_sva: pool returned IPv4=10.33.14.251, IPv6=fd00:f00f::1efa
2023-01-04 13:53:04 us=623456 gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a MULTI: Learn: 10.33.14.251 -> gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a
2023-01-04 13:53:04 us=623460 gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a MULTI: primary virtual IP for gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a: 10.33.14.251
2023-01-04 13:53:04 us=623468 gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a MULTI: Learn: fd00:f00f::1efa -> gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a
2023-01-04 13:53:04 us=623472 gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a MULTI: primary virtual IPv6 for gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a: fd00:f00f::1efa
2023-01-04 13:53:04 us=623480 gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a dco_new_peer: peer-id 828, fd 5, remote addr: [AF_INET6]2003:d0:270b:a600:215:5dff:fe02:811a:52901
2023-01-04 13:53:04 us=623516 gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a dco_new_peer: netlink out of memory error: No buffer space available (errno=105)
2023-01-04 13:53:04 us=623521 gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a Exiting due to fatal error
2023-01-04 13:53:04 us=623533 gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a Closing DCO interface
2023-01-04 13:53:04 us=623537 gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a net_addr_v4_del: 10.33.0.1 dev tun0
2023-01-04 13:53:04 us=624239 gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a sitnl_send: checking for received messages
2023-01-04 13:53:04 us=624272 gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a sitnl_send: rtnl: received 36 bytes
2023-01-04 13:53:04 us=624291 gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a net_addr_v6_del: fd00:f00f::2/64 dev tun0
2023-01-04 13:53:04 us=625271 gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a sitnl_send: checking for received messages
2023-01-04 13:53:04 us=625301 gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a sitnl_send: rtnl: received 36 bytes
2023-01-04 13:53:04 us=625319 gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a close_tun_dco
2023-01-04 13:53:04 us=625341 gremlin404326/2003:d0:270b:a600:215:5dff:fe02:811a net_iface_del: delete tun0

[ordex]

what does dmesg say after the hang has started? I am guessing it is the same as #18

[schwabe]

There is nothing in dmesg this time. Just the process hanging in D state

[ordex]

@schwabe is this still reproducible after the latest set of fixes?