On Sat, Nov 08, 2025 at 04:59:29PM -0600, Eric Blake wrote:
> An upcoming patch needs to pass more than just sioc as the opaque
> pointer to an AioContext; but since our AioContext code in general
> (and its QIO Channel wrapper code) lacks a notify callback present
> with GSource, we do not have the trivial option of just g_malloc'ing a
> small struct to hold all that data coupled with a notify of g_free.
> Instead, the data pointer must outlive the registered handler; in
> fact, having the data pointer have the same lifetime as QIONetListener
> is adequate.
>
> But the cleanest way to stick such a helper struct in QIONetListener
> will be to rearrange internal struct members. And that in turn means
> that all existing code that currently directly accesses
> listener->nsioc and listener->sioc[] should instead go through
> accessor functions, to be immune to the upcoming struct layout
> changes. So this patch adds accessor methods qio_net_listener_nsioc()
> and qio_net_listener_sioc(), and puts them to use.
>
> Signed-off-by: Eric Blake <[email protected]>
>
> ---
> v2: new patch
> ---
> include/io/net-listener.h | 24 ++++++++++++++++++++++++
> chardev/char-socket.c | 3 ++-
> io/net-listener.c | 12 ++++++++++++
> migration/socket.c | 5 +++--
> ui/vnc.c | 36 +++++++++++++++++++++++-------------
> 5 files changed, 64 insertions(+), 16 deletions(-)
>
> diff --git a/include/io/net-listener.h b/include/io/net-listener.h
> index 42fbfab5467..2605d6aae1e 100644
> --- a/include/io/net-listener.h
> +++ b/include/io/net-listener.h
> @@ -184,4 +184,28 @@ void qio_net_listener_disconnect(QIONetListener
> *listener);
> */
> bool qio_net_listener_is_connected(QIONetListener *listener);
>
> +
> +/**
> + * qio_net_listener_nsioc:
> + * @listener: the network listener object
> + *
> + * Determine the number of listener channels currently owned by the
> + * given listener.
> + *
> + * Returns: number of channels, or 0 if not listening
> + */
> +size_t qio_net_listener_nsioc(QIONetListener *listener);
> +
> +
> +/**
> + * qio_net_listener_sioc:
> + * @listener: the network listener object
> + * @n: index of the sioc to grab
> + *
> + * Accessor for the nth sioc owned by the listener.
> + *
> + * Returns: the requested listener, or #NULL if not in bounds
> + */
> +QIOChannelSocket *qio_net_listener_sioc(QIONetListener *listener, size_t n);
> +
> #endif /* QIO_NET_LISTENER_H */
> diff --git a/chardev/char-socket.c b/chardev/char-socket.c
> index 62852e3caf5..022ae47d726 100644
> --- a/chardev/char-socket.c
> +++ b/chardev/char-socket.c
> @@ -1255,7 +1255,8 @@ static int qmp_chardev_open_socket_server(Chardev *chr,
> }
>
> qapi_free_SocketAddress(s->addr);
> - s->addr = socket_local_address(s->listener->sioc[0]->fd, errp);
> + s->addr = socket_local_address(qio_net_listener_sioc(s->listener, 0)->fd,
> + errp);
Oh pre-existing bug / undesirable code pattern. This should be calling
qio_channel_socket_get_local_address, which avoids the needs to re-call
getsockname() as QIOChanelSocket has cached the local address.
Once we do that, then we would have 4 cases in this patch doing
qio_channel_socket_get_local_address(
qio_net_listener_sioc(listener, i), errp);
which would suggest adding a helper we could call as
qio_net_listener_get_local_address(listener, i, errp)
With regards,
Daniel
--
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org -o- https://fstop138.berrange.com :|
|: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
