On Wed, Mar 09, 2016 at 06:49:43PM +0100, Paolo Bonzini wrote: > > > On 09/03/2016 18:28, Daniel P. Berrange wrote: > > @@ -3131,7 +3118,9 @@ static bool qemu_chr_open_socket_fd(CharDriverState > > *chr, Error **errp) > > if (qio_channel_socket_listen_sync(sioc, s->addr, errp) < 0) { > > goto fail; > > } > > - qemu_chr_finish_socket_connection(chr, sioc); > > + s->listen_ioc = sioc; > > + s->listen_tag = qio_channel_add_watch( > > + QIO_CHANNEL(s->listen_ioc), G_IO_IN, tcp_chr_accept, chr, > > NULL); > > } else if (s->reconnect_time) { > > qio_channel_socket_connect_async(sioc, s->addr, > > qemu_chr_socket_connected, > > Aha, yes, this could be it. If you move WSAEventSelect to > qio_channel_set_blocking, the previous patch will probably become > unnecessary.
Yes, I've tested that and you're correct, so I've dropped the previous patch. Regards, Daniel -- |: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|