dhcpd crashing when configured to update DNS

[Philip Prindeville]

Filed a bug:

https://gitlab.isc.org/isc-projects/dhcp/-/issues/156

If I build isc-dhcp-4.4.1/bind/Makefile.in patched with “ac_cv_func_catgets=no” then this goes away.



> On Dec 10, 2020, at 4:16 PM, Philip Prindeville <philipp_subx at redfish-solutions.com> wrote:
> 
> I’m seeing the following when configuring the DHCP server to update DNS.  If this isn’t set, I don’t see it:
> 
> (gdb) info stack
> #0  0x00007f66d350aa4b in catgets (catd=0xffffffffffffffff, set_id=2, 
>    msg_id=1, s=0x5bd0cc "success") at src/locale/catgets.c:19
> #1  0x0000000000593aa9 in isc_result_tomany_helper.isra ()
> #2  0x000000000056a6b5 in req_response ()
> #3  0x0000000000599c64 in isc.taskmgr_dispatch ()
> #4  0x000000000059c78e in evloop ()
> #5  0x000000000059cb1c in isc.app_ctxrun ()
> #6  0x00000000004451eb in dispatch ()
> #7  0x000000000040529a in main ()
> (gdb) 
> 
> I’m using 4.4.1.
> 
> I was trying to follow this article:
> 
> https://www.talk-about-it.ca/setup-bind9-with-isc-dhcp-server-dynamic-host-registration/
> 
> My config looks like:
> 
> authoritative;
> ddns-domainname "redfish-solutions.com.";
> ddns-rev-domainname "in-addr.arpa.";
> ddns-update-style interim;
> ignore client-updates;
> update-static-leases on;
> use-host-decl-names on;
> option domain-name "redfish-solutions.com.";
> include "/tmp/run/dhcpd-rndc.key";
> update-optimization off;
> update-conflict-detection off;
> # include "/etc/bind/rndc.conf";
> 
> zone redfish-solutions.com. {
> 	primary 127.0.0.1;
> 	key rndc-key;
> }
> 
> zone 168.192.in-addr.arpa. {
> 	primary 127.0.0.1;
> 	key rndc-key;
> }
> 
> log-facility daemon;
> default-lease-time 3600;
> max-lease-time 86400;
> option domain-name "redfish-solutions.com";
> 
> # additional codes
> option classless-ipv4-route code 121 = array of { unsigned integer 8 };
> 
> subnet 192.168.1.0 netmask 255.255.255.0 {
> range 192.168.1.128 192.168.1.160;
> option subnet-mask 255.255.255.0;
> option broadcast-address 192.168.1.255;
> default-lease-time 43200;
> max-lease-time 43200;
> option routers 192.168.1.252;
> option domain-name-servers 192.168.1.252;
> option domain-search "redfish-solutions.com", "redfish-consulting.com";
> option ntp-servers 192.168.1.40, 192.168.1.252;
> }
> subnet 192.168.2.0 netmask 255.255.255.0 {
> range 192.168.2.16 192.168.2.254;
> option subnet-mask 255.255.255.0;
> option broadcast-address 192.168.2.255;
> default-lease-time 3600;
> max-lease-time 3600;
> option routers 192.168.2.1;
> option domain-name-servers 192.168.2.1;
> option domain-search "redfish-solutions.com", "redfish-consulting.com";
> option ntp-servers 192.168.1.40, 192.168.2.1;
> }
> host switch1 {
> hardware ethernet b4:75:0e:7f:e1:3e;
> fixed-address 192.168.1.2;
> option host-name "switch1”;
> }
> 
> ...
> 
> Yes, it’s a machine generated config coming from UCI files on OpenWRT master.
> 
> The architecture is x86_64, using MUSL as the C run-time.
> 
> Not sure why the stack trace is showing -1 (~0) as the pointer value for the message catalog.
> 
> Is there a simple way to disable NLS at configure time?
> 
> Thanks,
> 
> -Philip