Comment 13 for bug 1432786

First observation is that the stack trace above *is* bogus. qemu-system-arm
for some reason behaves differently under gdb, triggering some kind
of (gdb?) bug. Please ignore the stack traces in comments 3, 5 & 6.
And my apologies for blaming gnutls.

The REAL stack trace points to ceph:

#0 0xb284846c in gettimeofday () from /lib/libc.so.6
#1 0xb405b868 in Cycles::init () at common/Cycles.cc:75
#2 0xb6b28500 in call_init.part () from /lib/ld-linux-armhf.so.3
#3 0xb6b28664 in _dl_init () from /lib/ld-linux-armhf.so.3
#4 0xb6b17b84 in _dl_start_user () from /lib/ld-linux-armhf.so.3