Synchronous method in asynchronous signal callback

[Havoc Pennington]

On Tue, May 26, 2015 at 3:49 PM, Lennart Poettering <mzqohf at 0pointer.de> wrote:
> BTW, we'd probably have used glib in systemd if it wasn't for the fact
> that it's not OOM-safe, which makes it a non-option for us...
>

This is one of the main reasons dbus-daemon didn't use it, too, though
there were also "political" reasons (keeping the codebase disentangled
from gnome).

I became convinced in the process, though, that any
non-system-critical or remotely complex app that thinks it should or
could handle OOM is deeply misguided.
http://blog.ometer.com/2008/02/04/out-of-memory-handling-d-bus-experience/

GLib's "do not handle OOM" policy is correct for the 99% case. Many
who believe they should be handling OOM are probably wrong. And of
those who should be, many are probably making a token effort that's
broken in practice.

+1 to Simon's rant in general.

Havoc