tests/ip hangs

[milahu]

tests/ip hangs with 100% cpu load
in libmill-git-2dd13ae at make check

$ ip addr show
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
2: en: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen 1000
    link/ether xx:xx:xx:xx:xx:xx brd ff:ff:ff:ff:ff:ff
3: wl: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc htb state UP group default qlen 1000
    link/ether xx:xx:xx:xx:xx:xx brd ff:ff:ff:ff:ff:ff
    inet 192.168.2.102/24 brd 192.168.2.255 scope global noprefixroute wlp3s0
       valid_lft forever preferred_lft forever

$ sysctl -a -r 'net.ipv6.conf.*.disable_ipv6'
net.ipv6.conf.all.disable_ipv6 = 1
net.ipv6.conf.default.disable_ipv6 = 1
net.ipv6.conf.en.disable_ipv6 = 1
net.ipv6.conf.lo.disable_ipv6 = 1
net.ipv6.conf.wl.disable_ipv6 = 1

tc.bash stop did not help [stop traffic shaping]

$ diff ...
-3: wl: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc htb state UP group default qlen 1000
+3: wl: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000

me clueless

but at least the tests should be wrapped in timeout -k $sec $sec $cmd

[sustrik]

The tests have infinite timeouts. If DNS is not accessible, the test will hang.

[milahu]

If DNS is not accessible, the test will hang.

dns is working in all of my other apps

The tests have infinite timeouts.

which is not ideal with while(1) { ... } in ip.c

i replaced it with for(_i=0;_i<4;_i++) { ... ; sleep 1 }
and added some printfs ...

debug-tests-ip.txt

ideas?

[sustrik]

First things first: Can you resolve www.example.org from the machine in question?

dig +short www.example.org

[milahu]

Can you resolve www.example.org from the machine in question?

yes

already tried with the resolveip tool
now confirmed with dig

$ dig www.example.com | grep -e time -e SERVER -e HEADER -e example
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 61831
www.example.com.	38919	IN	A	93.184.216.34
;; Query time: 13 msec
;; SERVER: 192.168.2.1#53(192.168.2.1)

$ dig @8.8.8.8 www.example.com | grep -e time -e SERVER -e HEADER -e example
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 5835
www.example.com.	15320	IN	A	93.184.216.34
;; Query time: 11 msec
;; SERVER: 8.8.8.8#53(8.8.8.8)

$ time tor-resolve www.example.com
real average 300 ms

$ ping -c10 192.168.2.1
average 3.5 ms

$ ping -c10 8.8.8.8
average 11 ms

$ traceroute 8.8.8.8
[...]
 7  google-public-dns-a.google.com (8.8.8.8)  11.938 ms  11.619 ms  11.610 ms

my router is just slow at DNS, but it does work

i also tried ... as suggested by toybox ping

$ echo 0 9999999 | sudo tee /proc/sys/net/ipv4/ping_group_range

and

$ sudo ./tests/ip

without success

[sustrik]

Ok, so looking at your debug output it looks like the test exits al right:

ipremote. call dns_ai_close
tests/ip. resolve 6 done, errno = 99
tests/ip. return

What am I missing?

[milahu]

it looks like the test exits al right

only after i replaced while true with for(_i=0;_i<4;_i++)

now the loop runs 4 times for ipv4 and 4 times for ipv6, even with sleep 1 after every iteration

the lookup should be done in 13 ms but it takes forever with 100% cpu load, with the original code

also i dont see the ip address anywhere, it is always NULL

im gonna have a closer look at your dns lookup functions later

ps, libdill seems to work [note to self, verify]

[vieks]

I have the exact same problem.
see: #186