Commit 95751324 authored by Marc-André Lureau's avatar Marc-André Lureau Committed by Jasper St. Pierre
Browse files

xrandr: use "hotplug_mode_update" property

Use the "hotplug_mode_update" connector property indicating that the
screen settings should be updated: get a new preferred mode on hotplug
events to handle dynamic guest resizing (where you resize the host
window and the guest resizes with it).

https://bugzilla.gnome.org/show_bug.cgi?id=711216
parent 17fd25e2
......@@ -817,6 +817,22 @@ meta_monitor_config_match_current (MetaMonitorConfig *self,
return ok;
}
gboolean
meta_monitor_manager_has_hotplug_mode_update (MetaMonitorManager *manager)
{
MetaOutput *outputs;
unsigned n_outputs;
unsigned int i;
outputs = meta_monitor_manager_get_outputs (manager, &n_outputs);
for (i = 0; i < n_outputs; i++)
if (outputs[i].hotplug_mode_update)
return TRUE;
return FALSE;
}
static MetaConfiguration *
meta_monitor_config_get_stored (MetaMonitorConfig *self,
MetaOutput *outputs,
......
......@@ -119,6 +119,9 @@ struct _MetaOutput
gpointer driver_private;
GDestroyNotify driver_notify;
/* get a new preferred mode on hotplug events, to handle dynamic guest resizing */
gboolean hotplug_mode_update;
};
struct _MetaCRTC
......@@ -407,6 +410,7 @@ void meta_monitor_manager_free_output_array (MetaOutput *old_outpu
int n_old_outputs);
void meta_monitor_manager_free_mode_array (MetaMonitorMode *old_modes,
int n_old_modes);
gboolean meta_monitor_manager_has_hotplug_mode_update (MetaMonitorManager *manager);
/* Returns true if transform causes width and height to be inverted
This is true for the odd transforms in the enum */
......
......@@ -311,6 +311,29 @@ read_output_edid (MetaMonitorManagerXrandr *manager_xrandr,
return NULL;
}
static gboolean
output_get_hotplug_mode_update (MetaMonitorManagerXrandr *manager_xrandr,
XID output_id)
{
MetaDisplay *display = meta_get_display ();
XRRPropertyInfo *info;
gboolean result = FALSE;
meta_error_trap_push (display);
info = XRRQueryOutputProperty (manager_xrandr->xdisplay, output_id,
display->atom_hotplug_mode_update);
meta_error_trap_pop (display);
if (info)
{
result = TRUE;
XFree (info);
}
return result;
}
static void
meta_monitor_manager_xrandr_read_current (MetaMonitorManager *manager)
{
......@@ -484,6 +507,8 @@ meta_monitor_manager_xrandr_read_current (MetaMonitorManager *manager)
meta_output->width_mm = output->mm_width;
meta_output->height_mm = output->mm_height;
meta_output->subpixel_order = COGL_SUBPIXEL_ORDER_UNKNOWN;
meta_output->hotplug_mode_update =
output_get_hotplug_mode_update (manager_xrandr, meta_output->output_id);
meta_output->n_modes = output->nmode;
meta_output->modes = g_new0 (MetaMonitorMode *, meta_output->n_modes);
......@@ -969,6 +994,16 @@ meta_monitor_manager_xrandr_set_crtc_gamma (MetaMonitorManager *manager,
XRRFreeGamma (gamma);
}
static void
meta_monitor_manager_xrandr_rebuild_derived (MetaMonitorManager *manager)
{
/* This will be a no-op if the change was from our side, as
we already called it in the DBus method handler */
meta_monitor_config_update_current (manager->config, manager);
meta_monitor_manager_rebuild_derived (manager);
}
static gboolean
meta_monitor_manager_xrandr_handle_xevent (MetaMonitorManager *manager,
XEvent *event)
......@@ -978,6 +1013,7 @@ meta_monitor_manager_xrandr_handle_xevent (MetaMonitorManager *manager,
MetaCRTC *old_crtcs;
MetaMonitorMode *old_modes;
unsigned int n_old_outputs, n_old_modes;
gboolean new_config;
if ((event->type - manager_xrandr->rr_event_base) != RRScreenChangeNotify)
return FALSE;
......@@ -994,31 +1030,36 @@ meta_monitor_manager_xrandr_handle_xevent (MetaMonitorManager *manager,
manager->serial++;
meta_monitor_manager_xrandr_read_current (manager);
/* Check if the current intended configuration has the same outputs
as the new real one, or if the event is a result of an XRandR call.
If so, we can go straight to rebuild the logical config and tell
the outside world.
Otherwise, this event was caused by hotplug, so give a chance to
MetaMonitorConfig.
Note that we need to check both the timestamps and the list of
outputs, because the X server might emit spurious events with
new configTimestamps (bug 702804), and the driver may have
changed the EDID for some other reason (old broken qxl and vbox
drivers...).
*/
if (manager_xrandr->resources->timestamp >= manager_xrandr->resources->configTimestamp ||
meta_monitor_config_match_current (manager->config, manager))
{
/* This will be a no-op if the change was from our side, as
we already called it in the DBus method handler */
meta_monitor_config_update_current (manager->config, manager);
new_config = manager_xrandr->resources->timestamp >=
manager_xrandr->resources->configTimestamp;
if (meta_monitor_manager_has_hotplug_mode_update (manager))
meta_monitor_manager_rebuild_derived (manager);
{
/* Check if the current intended configuration is a result of an
XRandR call. Otherwise, hotplug_mode_update tells us to get
a new preferred mode on hotplug events to handle dynamic
guest resizing. */
if (new_config)
meta_monitor_manager_xrandr_rebuild_derived (manager);
else
meta_monitor_config_make_default (manager->config, manager);
}
else
{
if (!meta_monitor_config_apply_stored (manager->config, manager))
/* Check if the current intended configuration has the same outputs
as the new real one, or if the event is a result of an XRandR call.
If so, we can go straight to rebuild the logical config and tell
the outside world.
Otherwise, this event was caused by hotplug, so give a chance to
MetaMonitorConfig.
Note that we need to check both the timestamps and the list of
outputs, because the X server might emit spurious events with new
configTimestamps (bug 702804), and the driver may have changed
the EDID for some other reason (old qxl and vbox drivers). */
if (new_config || meta_monitor_config_match_current (manager->config, manager))
meta_monitor_manager_xrandr_rebuild_derived (manager);
else if (!meta_monitor_config_apply_stored (manager->config, manager))
meta_monitor_config_make_default (manager->config, manager);
}
......
......@@ -82,6 +82,7 @@ item(VERSION)
item(ATOM_PAIR)
item(BACKLIGHT)
item(_XKB_RULES_NAMES)
item(hotplug_mode_update)
/* Oddities: These are used, and we need atoms for them,
* but when we need all _NET_WM hints (i.e. when we're making
......
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment