I installed the patch below into the `emacs-24' branch.
This patch is causing a problem with emacsclient, at least on Cygwin. I've
tested both the GTK build and the w32 build. To reproduce:
$ emacs --daemon -Q
Starting Emacs daemon.
$ emacsclient.exe -c
Waiting for Emacs...
Now do `C-x 5 0' to close the client window; emacsclient is still running
and has to be killed from the shell.
Hmm... I can't imagine how this patch could have such an effect.
Can you confirm that the problem appears really with this precise patch
rather than some other one applied recently?