Comment 18 for bug 1168526

It does seem like this must be a kernel bug in epoll+signalfd (or a hard to spot misuse thereof in lxc).

When I instrument the signal_handler which is executed when epoll_wait returns a signalfd event, I do get a sigchld for the very first task which is spawned (a test to see if kernel supports pidns reboot), but do not get one for the container init - which becomes <defunct> and therefore clearly did in fact send a sigchld.

I'll send my tiny patch upstream, and test on other kernels for comparison.