Arch devs,
Not sure if this is worth a bug report to Arch? Upstream?, but for
the first time in 24 years named core dumped on me (bind 9.20.18-1). The
journal entry is:
Feb 12 02:54:22 valkyrie named[808]: timed out resolving
'240.78.147.in-addr.arpa/NS/IN': 66.76.175.100#53
Feb 12 02:54:23 valkyrie named[808]: timed out resolving
'240.78.147.in-addr.arpa/NS/IN': 208.180.42.100#53
Feb 12 02:54:24 valkyrie named[808]: timed out resolving
'240.78.147.in-addr.arpa/NS/IN': 8.8.4.4#53
Feb 12 02:54:24 valkyrie named[808]: network unreachable resolving
'240.78.147.in-addr.arpa/NS/IN': 2001:500:f0::63#53
Feb 12 02:54:25 valkyrie named[808]: host unreachable resolving
'119.240.78.147.in-addr.arpa/PTR/IN': 147.78.247.25#53
Feb 12 02:54:25 valkyrie named[808]: named: src/unix/udp.c:292:
uv__udp_recvmsg: Assertion `handle->recv_cb != NULL' failed.
Feb 12 02:54:25 valkyrie systemd-coredump[1815]: [🡕] Process 808
(named) of user 40 dumped core.
Stack trace of thread 849:
#0 0x00007fbcbc53da2c
n/a (libc.so.6 + 0x98a2c)
#1 0x00007fbcbc4e31a0
raise (libc.so.6 + 0x3e1a0)
#2 0x00007fbcbc4ca5fe
abort (libc.so.6 + 0x255fe)
#3 0x00007fbcbc4cb697
n/a (libc.so.6 + 0x26697)
#4 0x00007fbcbc4db0b2
__assert_fail (libc.so.6 + 0x360b2)
#5 0x00007fbcbcdc7246
n/a (libuv.so.1 + 0xa246)
#6 0x00007fbcbcdd1985
n/a (libuv.so.1 + 0x14985)
#7 0x00007fbcbcde7ad4
n/a (libuv.so.1 + 0x2aad4)
#8 0x00007fbcbcdd2b12
uv_run (libuv.so.1 + 0x15b12)
#9 0x00007fbcbd278700
n/a (libisc-9.20.18.so + 0x4b700)
#10 0x00007fbcbd28aef0
n/a (libisc-9.20.18.so + 0x5def0)
#11 0x00007fbcbc53b97a
n/a (libc.so.6 + 0x9697a)
#12 0x00007fbcbc5bf2bc
n/a (libc.so.6 + 0x11a2bc)
...
I've run bind since 2002, and on Arch since 2009 and never in all
that time had it crash.
The whois on 147.78.247.25 is also rather interesting, a
rent-a-server shop with US and Japanese corporate registration info.
Given the uniqueness of this crash, there probably should be a
report, but I'm not sure if it should go to Arch or ISC (or both with
the gitlab bug just tracking upstream)
What say the Arch folks? Bug needed? If so, start with Arch, ISC or
both? (or maybe freedesktop? but that didn't seem like it was involved
other than catching the crash)
--
David C. Rankin, J.D.,P.E.
