brl/mutter
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
bafa448666
mutter/clutter/clutter-master-clock.c

478 lines
14 KiB
C
Raw Normal View History

Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
/*
* Clutter.
*
* An OpenGL based 'interactive canvas' library.
*
* Authored By: Emmanuele Bassi <ebassi@linux.intel.com>
*
* Copyright (C) 2009 Intel Corporation.
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2 of the License, or (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with this library. If not, see <http://www.gnu.org/licenses/>.
*/
/*
* SECTION:clutter-master-clock
* @short_description: The master clock for all animations
*
* The #ClutterMasterClock class is responsible for advancing all
* #ClutterTimelines when a stage is being redrawn. The master clock
* makes sure that the scenegraph is always integrally updated before
* painting it.
*/
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include "clutter-master-clock.h"
#include "clutter-debug.h"
#include "clutter-private.h"
#define CLUTTER_MASTER_CLOCK_CLASS(klass) (G_TYPE_CHECK_CLASS_CAST ((klass), CLUTTER_TYPE_MASTER_CLOCK, ClutterMasterClockClass))
#define CLUTTER_IS_MASTER_CLOCK_CLASS(klass) (G_TYPE_CHECK_CLASS_TYPE ((klass), CLUTTER_TYPE_MASTER_CLOCK))
#define CLUTTER_MASTER_CLASS_GET_CLASS(obj) (G_TYPE_INSTANCE_GET_CLASS ((obj), CLUTTER_TYPE_MASTER_CLOCK, ClutterMasterClockClass))
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
typedef struct _ClutterClockSource ClutterClockSource;
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
typedef struct _ClutterMasterClockClass ClutterMasterClockClass;
struct _ClutterMasterClock
{
GObject parent_instance;
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
/* the list of timelines handled by the clock */
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
GSList *timelines;
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
/* the previous state of the clock, used to compute
* the delta
*/
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
GTimeVal prev_tick;
gulong msecs_delta;
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
/* an idle source, used by the Master Clock to queue
* a redraw on the stage and drive the animations
*/
GSource *source;
/* a guard, so that we dispatch the last redraw even
* after the last timeline has been completed
*/
guint last_advance : 1;
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
};
struct _ClutterMasterClockClass
{
GObjectClass parent_class;
};
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
struct _ClutterClockSource
{
GSource source;
ClutterMasterClock *master_clock;
};
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
static void on_timeline_started (ClutterTimeline *timeline,
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
ClutterMasterClock *master_clock);
static void on_timeline_completed (ClutterTimeline *timeline,
ClutterMasterClock *master_clock);
static void on_timeline_paused (ClutterTimeline *timeline,
ClutterMasterClock *master_clock);
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
static gboolean clutter_clock_prepare (GSource *source,
gint *timeout);
static gboolean clutter_clock_check (GSource *source);
static gboolean clutter_clock_dispatch (GSource *source,
GSourceFunc callback,
gpointer user_data);
static ClutterMasterClock *default_clock = NULL;
static GSourceFuncs clock_funcs = {
clutter_clock_prepare,
clutter_clock_check,
clutter_clock_dispatch,
NULL
};
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
G_DEFINE_TYPE (ClutterMasterClock, clutter_master_clock, G_TYPE_OBJECT);
[master clock] Handle Timeline::started signal correctly The "started" signal is sent first after the timeline has been set to the 'running' state. For this reason, checking if the clock has any running timelines running will always return true in the "started" signal handler: the timeline that sent the signal is running. What needs to be checked in the signal handler is if there are any timelines running other than the one that emitted the ::started signal, which we know is running anyway. This prevents frames from being lost at the beginning of an animation when a timeline is started after a quiescent period. Fixes bug: http://bugzilla.openedhand.com/show_bug.cgi?id=1617 Signed-off-by: Jonas Bonn <jonas@southpole.se> Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-05-29 09:25:19 -04:00
/*
* has_running_timeline:
* @master_clock: a #ClutterMasterClock
* @filter: a #ClutterTimeline or %NULL
*
* Checks if @master_clock has any running timeline; if @filter
* is not %NULL then the timeline will be filtered from the
* list of timelines held by @master_clock
*
* Return value: %TRUE if the #ClutterMasterClock has at least
* one running timeline
*/
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
static gboolean
[master clock] Handle Timeline::started signal correctly The "started" signal is sent first after the timeline has been set to the 'running' state. For this reason, checking if the clock has any running timelines running will always return true in the "started" signal handler: the timeline that sent the signal is running. What needs to be checked in the signal handler is if there are any timelines running other than the one that emitted the ::started signal, which we know is running anyway. This prevents frames from being lost at the beginning of an animation when a timeline is started after a quiescent period. Fixes bug: http://bugzilla.openedhand.com/show_bug.cgi?id=1617 Signed-off-by: Jonas Bonn <jonas@southpole.se> Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-05-29 09:25:19 -04:00
has_running_timeline (ClutterMasterClock *master_clock,
ClutterTimeline *filter)
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
{
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
GSList *l;
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
if (master_clock->last_advance)
return TRUE;
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
if (master_clock->timelines == NULL)
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
return FALSE;
for (l = master_clock->timelines; l != NULL; l = l->next)
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
{
[master clock] Handle Timeline::started signal correctly The "started" signal is sent first after the timeline has been set to the 'running' state. For this reason, checking if the clock has any running timelines running will always return true in the "started" signal handler: the timeline that sent the signal is running. What needs to be checked in the signal handler is if there are any timelines running other than the one that emitted the ::started signal, which we know is running anyway. This prevents frames from being lost at the beginning of an animation when a timeline is started after a quiescent period. Fixes bug: http://bugzilla.openedhand.com/show_bug.cgi?id=1617 Signed-off-by: Jonas Bonn <jonas@southpole.se> Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-05-29 09:25:19 -04:00
/* if we get a timeline then we should filter it
* from the list of timelines we want to check
*/
if (filter != NULL && filter == l->data)
continue;
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
if (clutter_timeline_is_playing (l->data))
return TRUE;
}
return FALSE;
}
[master clock] Handle Timeline::started signal correctly The "started" signal is sent first after the timeline has been set to the 'running' state. For this reason, checking if the clock has any running timelines running will always return true in the "started" signal handler: the timeline that sent the signal is running. What needs to be checked in the signal handler is if there are any timelines running other than the one that emitted the ::started signal, which we know is running anyway. This prevents frames from being lost at the beginning of an animation when a timeline is started after a quiescent period. Fixes bug: http://bugzilla.openedhand.com/show_bug.cgi?id=1617 Signed-off-by: Jonas Bonn <jonas@southpole.se> Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-05-29 09:25:19 -04:00
/*
* clutter_clock_source_new:
* @master_clock: a #ClutterMasterClock for the source
*
* The #ClutterClockSource is an idle GSource that will queue a redraw
* if @master_clock has at least a running #ClutterTimeline. The redraw
* will cause @master_clock to advance all timelines, thus advancing all
* animations as well.
*
* Return value: the newly created #GSource
*/
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
static GSource *
clutter_clock_source_new (ClutterMasterClock *master_clock)
{
GSource *source = g_source_new (&clock_funcs, sizeof (ClutterClockSource));
ClutterClockSource *clock_source = (ClutterClockSource *) source;
clock_source->master_clock = master_clock;
return source;
}
static gboolean
clutter_clock_prepare (GSource *source,
gint *timeout)
{
ClutterClockSource *clock_source = (ClutterClockSource *) source;
ClutterMasterClock *master_clock = clock_source->master_clock;
gboolean retval;
[master clock] Handle Timeline::started signal correctly The "started" signal is sent first after the timeline has been set to the 'running' state. For this reason, checking if the clock has any running timelines running will always return true in the "started" signal handler: the timeline that sent the signal is running. What needs to be checked in the signal handler is if there are any timelines running other than the one that emitted the ::started signal, which we know is running anyway. This prevents frames from being lost at the beginning of an animation when a timeline is started after a quiescent period. Fixes bug: http://bugzilla.openedhand.com/show_bug.cgi?id=1617 Signed-off-by: Jonas Bonn <jonas@southpole.se> Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-05-29 09:25:19 -04:00
/* just like an idle source, we are ready if nothing else is */
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
*timeout = -1;
[master clock] Handle Timeline::started signal correctly The "started" signal is sent first after the timeline has been set to the 'running' state. For this reason, checking if the clock has any running timelines running will always return true in the "started" signal handler: the timeline that sent the signal is running. What needs to be checked in the signal handler is if there are any timelines running other than the one that emitted the ::started signal, which we know is running anyway. This prevents frames from being lost at the beginning of an animation when a timeline is started after a quiescent period. Fixes bug: http://bugzilla.openedhand.com/show_bug.cgi?id=1617 Signed-off-by: Jonas Bonn <jonas@southpole.se> Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-05-29 09:25:19 -04:00
retval = has_running_timeline (master_clock, NULL);
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
return retval;
}
static gboolean
clutter_clock_check (GSource *source)
{
ClutterClockSource *clock_source = (ClutterClockSource *) source;
ClutterMasterClock *master_clock = clock_source->master_clock;
gboolean retval;
[master clock] Handle Timeline::started signal correctly The "started" signal is sent first after the timeline has been set to the 'running' state. For this reason, checking if the clock has any running timelines running will always return true in the "started" signal handler: the timeline that sent the signal is running. What needs to be checked in the signal handler is if there are any timelines running other than the one that emitted the ::started signal, which we know is running anyway. This prevents frames from being lost at the beginning of an animation when a timeline is started after a quiescent period. Fixes bug: http://bugzilla.openedhand.com/show_bug.cgi?id=1617 Signed-off-by: Jonas Bonn <jonas@southpole.se> Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-05-29 09:25:19 -04:00
retval = has_running_timeline (master_clock, NULL);
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
return retval;
}
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
static gboolean
clutter_clock_dispatch (GSource *source,
GSourceFunc callback,
gpointer user_data)
{
ClutterClockSource *clock_source = (ClutterClockSource *) source;
ClutterMasterClock *master_clock = clock_source->master_clock;
ClutterStageManager *stage_manager = clutter_stage_manager_get_default ();
[master clock] Use StageManager::peek_stages() Use the new StageManager::peek_stages() method to avoid a copy of the stages list.
2009-05-29 09:52:49 -04:00
const GSList *stages, *l;
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
CLUTTER_NOTE (SCHEDULER, "Master clock [tick]");
[master clock] Use StageManager::peek_stages() Use the new StageManager::peek_stages() method to avoid a copy of the stages list.
2009-05-29 09:52:49 -04:00
stages = clutter_stage_manager_peek_stages (stage_manager);
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
[master clock] Handle Timeline::started signal correctly The "started" signal is sent first after the timeline has been set to the 'running' state. For this reason, checking if the clock has any running timelines running will always return true in the "started" signal handler: the timeline that sent the signal is running. What needs to be checked in the signal handler is if there are any timelines running other than the one that emitted the ::started signal, which we know is running anyway. This prevents frames from being lost at the beginning of an animation when a timeline is started after a quiescent period. Fixes bug: http://bugzilla.openedhand.com/show_bug.cgi?id=1617 Signed-off-by: Jonas Bonn <jonas@southpole.se> Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-05-29 09:25:19 -04:00
/* queue a redraw for each stage; this will advance each timeline
* held by the master clock of the amount of milliseconds elapsed
* since the last redraw
*/
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
for (l = stages; l != NULL; l = l->next)
clutter_actor_queue_redraw (l->data);
[master clock] Handle Timeline::started signal correctly The "started" signal is sent first after the timeline has been set to the 'running' state. For this reason, checking if the clock has any running timelines running will always return true in the "started" signal handler: the timeline that sent the signal is running. What needs to be checked in the signal handler is if there are any timelines running other than the one that emitted the ::started signal, which we know is running anyway. This prevents frames from being lost at the beginning of an animation when a timeline is started after a quiescent period. Fixes bug: http://bugzilla.openedhand.com/show_bug.cgi?id=1617 Signed-off-by: Jonas Bonn <jonas@southpole.se> Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-05-29 09:25:19 -04:00
/* if this is the remainder of an advancement, needed for the last
* timeline to finish its run, then we need to reset the prev_tick
*/
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
if (master_clock->last_advance)
{
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
master_clock->prev_tick.tv_sec = 0;
[master clock] Handle Timeline::started signal correctly The "started" signal is sent first after the timeline has been set to the 'running' state. For this reason, checking if the clock has any running timelines running will always return true in the "started" signal handler: the timeline that sent the signal is running. What needs to be checked in the signal handler is if there are any timelines running other than the one that emitted the ::started signal, which we know is running anyway. This prevents frames from being lost at the beginning of an animation when a timeline is started after a quiescent period. Fixes bug: http://bugzilla.openedhand.com/show_bug.cgi?id=1617 Signed-off-by: Jonas Bonn <jonas@southpole.se> Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-05-29 09:25:19 -04:00
master_clock->last_advance = FALSE;
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
}
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
return TRUE;
}
static void
timeline_weak_ref (gpointer data,
GObject *object_pointer)
{
ClutterMasterClock *master_clock = data;
master_clock->timelines =
g_slist_remove (master_clock->timelines, object_pointer);
if (master_clock->timelines == NULL)
master_clock->prev_tick.tv_sec = 0;
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
}
static void
clutter_master_clock_finalize (GObject *gobject)
{
ClutterMasterClock *master_clock = CLUTTER_MASTER_CLOCK (gobject);
GSList *l;
for (l = master_clock->timelines; l != NULL; l = l->next)
{
ClutterTimeline *timeline = l->data;
g_object_weak_unref (G_OBJECT (timeline),
timeline_weak_ref,
master_clock);
g_signal_handlers_disconnect_by_func (timeline,
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
G_CALLBACK (on_timeline_started),
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
master_clock);
g_signal_handlers_disconnect_by_func (timeline,
G_CALLBACK (on_timeline_completed),
master_clock);
g_signal_handlers_disconnect_by_func (timeline,
G_CALLBACK (on_timeline_paused),
master_clock);
}
g_slist_free (master_clock->timelines);
G_OBJECT_CLASS (clutter_master_clock_parent_class)->finalize (gobject);
}
static void
clutter_master_clock_class_init (ClutterMasterClockClass *klass)
{
GObjectClass *gobject_class = G_OBJECT_CLASS (klass);
gobject_class->finalize = clutter_master_clock_finalize;
}
static void
clutter_master_clock_init (ClutterMasterClock *self)
{
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
GSource *source;
source = clutter_clock_source_new (self);
self->source = source;
g_source_set_priority (source, CLUTTER_PRIORITY_REDRAW);
g_source_set_can_recurse (source, FALSE);
g_source_attach (source, NULL);
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
}
/*
* _clutter_master_clock_get_default:
*
* Retrieves the default master clock. If this function has never
* been called before, the default master clock is created.
*
* Return value: the default master clock. The returned object is
* owned by Clutter and should not be modified or freed
*/
ClutterMasterClock *
_clutter_master_clock_get_default (void)
{
if (G_LIKELY (default_clock != NULL))
return default_clock;
default_clock = g_object_new (CLUTTER_TYPE_MASTER_CLOCK, NULL);
return default_clock;
}
static void
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
on_timeline_started (ClutterTimeline *timeline,
ClutterMasterClock *master_clock)
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
{
[master clock] Handle Timeline::started signal correctly The "started" signal is sent first after the timeline has been set to the 'running' state. For this reason, checking if the clock has any running timelines running will always return true in the "started" signal handler: the timeline that sent the signal is running. What needs to be checked in the signal handler is if there are any timelines running other than the one that emitted the ::started signal, which we know is running anyway. This prevents frames from being lost at the beginning of an animation when a timeline is started after a quiescent period. Fixes bug: http://bugzilla.openedhand.com/show_bug.cgi?id=1617 Signed-off-by: Jonas Bonn <jonas@southpole.se> Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-05-29 09:25:19 -04:00
/* we want to reset the prev_tick if this is the first
* timeline; since timeline is playing we need to filter
* it out, otherwise has_running_timeline() will return
* TRUE and prev_tick will not be unset
*/
if (!has_running_timeline (master_clock, timeline))
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
master_clock->prev_tick.tv_sec = 0;
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
}
static void
on_timeline_completed (ClutterTimeline *timeline,
ClutterMasterClock *master_clock)
{
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
/* if this is the last timeline we need to turn :last-advance
* on in order to queue the redraw of the scene for the last
* frame; otherwise the ClockSource will fail the prepare and
* check phases and the last frame will not be painted
*/
[master clock] Handle Timeline::started signal correctly The "started" signal is sent first after the timeline has been set to the 'running' state. For this reason, checking if the clock has any running timelines running will always return true in the "started" signal handler: the timeline that sent the signal is running. What needs to be checked in the signal handler is if there are any timelines running other than the one that emitted the ::started signal, which we know is running anyway. This prevents frames from being lost at the beginning of an animation when a timeline is started after a quiescent period. Fixes bug: http://bugzilla.openedhand.com/show_bug.cgi?id=1617 Signed-off-by: Jonas Bonn <jonas@southpole.se> Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-05-29 09:25:19 -04:00
if (!has_running_timeline (master_clock, NULL))
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
master_clock->last_advance = TRUE;
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
}
static void
on_timeline_paused (ClutterTimeline *timeline,
ClutterMasterClock *master_clock)
{
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
/* see the comment in on_timeline_completed */
[master clock] Handle Timeline::started signal correctly The "started" signal is sent first after the timeline has been set to the 'running' state. For this reason, checking if the clock has any running timelines running will always return true in the "started" signal handler: the timeline that sent the signal is running. What needs to be checked in the signal handler is if there are any timelines running other than the one that emitted the ::started signal, which we know is running anyway. This prevents frames from being lost at the beginning of an animation when a timeline is started after a quiescent period. Fixes bug: http://bugzilla.openedhand.com/show_bug.cgi?id=1617 Signed-off-by: Jonas Bonn <jonas@southpole.se> Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-05-29 09:25:19 -04:00
if (!has_running_timeline (master_clock, NULL))
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
master_clock->last_advance = TRUE;
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
}
/*
* _clutter_master_clock_add_timeline:
* @master_clock: a #ClutterMasterClock
* @timeline: a #ClutterTimeline
*
* Adds @timeline to the list of timelines held by the master
* clock. This function should be called during the instance
* creation phase of the timeline.
*/
void
_clutter_master_clock_add_timeline (ClutterMasterClock *master_clock,
ClutterTimeline *timeline)
{
gboolean is_first = FALSE;
if (g_slist_find (master_clock->timelines, timeline))
return;
is_first = (master_clock->timelines == NULL) ? TRUE : FALSE;
master_clock->timelines = g_slist_prepend (master_clock->timelines,
timeline);
g_object_weak_ref (G_OBJECT (timeline),
timeline_weak_ref,
master_clock);
g_signal_connect (timeline, "started",
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
G_CALLBACK (on_timeline_started),
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
master_clock);
g_signal_connect (timeline, "completed",
G_CALLBACK (on_timeline_completed),
master_clock);
g_signal_connect (timeline, "paused",
G_CALLBACK (on_timeline_paused),
master_clock);
}
/*
* _clutter_master_clock_remove_timeline:
* @master_clock: a #ClutterMasterClock
* @timeline: a #ClutterTimeline
*
* Removes @timeline from the list of timelines held by the
* master clock. This function should be called during the
* #ClutterTimeline finalization.
*/
void
_clutter_master_clock_remove_timeline (ClutterMasterClock *master_clock,
ClutterTimeline *timeline)
{
if (!g_slist_find (master_clock->timelines, timeline))
return;
master_clock->timelines = g_slist_remove (master_clock->timelines,
timeline);
g_object_weak_unref (G_OBJECT (timeline),
timeline_weak_ref,
master_clock);
g_signal_handlers_disconnect_by_func (timeline,
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
G_CALLBACK (on_timeline_started),
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
master_clock);
g_signal_handlers_disconnect_by_func (timeline,
G_CALLBACK (on_timeline_completed),
master_clock);
g_signal_handlers_disconnect_by_func (timeline,
G_CALLBACK (on_timeline_paused),
master_clock);
[master clock] Handle Timeline::started signal correctly The "started" signal is sent first after the timeline has been set to the 'running' state. For this reason, checking if the clock has any running timelines running will always return true in the "started" signal handler: the timeline that sent the signal is running. What needs to be checked in the signal handler is if there are any timelines running other than the one that emitted the ::started signal, which we know is running anyway. This prevents frames from being lost at the beginning of an animation when a timeline is started after a quiescent period. Fixes bug: http://bugzilla.openedhand.com/show_bug.cgi?id=1617 Signed-off-by: Jonas Bonn <jonas@southpole.se> Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-05-29 09:25:19 -04:00
/* last timeline: unset the prev_tick so that we can start
* from scratch when we add a new timeline
*/
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
if (master_clock->timelines == NULL)
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
master_clock->prev_tick.tv_sec = 0;
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
}
/*
* _clutter_master_clock_advance:
* @master_clock: a #ClutterMasterClock
*
* Advances all the timelines held by the master clock. This function
* should be called before calling clutter_redraw() to make sure that
* all the timelines are advanced and the scene is updated.
*/
void
_clutter_master_clock_advance (ClutterMasterClock *master_clock)
{
GTimeVal cur_tick = { 0, };
gulong msecs;
GSList *l;
g_return_if_fail (CLUTTER_IS_MASTER_CLOCK (master_clock));
[clock] Rework the master clock The master clock is currently advanced using a frame source driven by the default frame rate. This breaks the sync to vblank because the vblanking rate could be different than 60 Hz -- or it might be completely disabled (e.g. with CLUTTER_VBLANK=none). We should be using the main loop to check if we have timelines playing, and if so queue a redraw on the stages we own. We should also prepare the subsequent frame at the end of the redraw process, so if there are new redraw we will have the scene already in place. This makes Clutter redraw at the maximum frame rate, which is limited by the vblanking frequency.
2009-05-07 14:07:21 -04:00
if (master_clock->timelines == NULL)
return;
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
g_get_current_time (&cur_tick);
if (master_clock->prev_tick.tv_sec == 0)
master_clock->prev_tick = cur_tick;
msecs = (cur_tick.tv_sec - master_clock->prev_tick.tv_sec) * 1000
+ (cur_tick.tv_usec - master_clock->prev_tick.tv_usec) / 1000;
[master clock] Handle Timeline::started signal correctly The "started" signal is sent first after the timeline has been set to the 'running' state. For this reason, checking if the clock has any running timelines running will always return true in the "started" signal handler: the timeline that sent the signal is running. What needs to be checked in the signal handler is if there are any timelines running other than the one that emitted the ::started signal, which we know is running anyway. This prevents frames from being lost at the beginning of an animation when a timeline is started after a quiescent period. Fixes bug: http://bugzilla.openedhand.com/show_bug.cgi?id=1617 Signed-off-by: Jonas Bonn <jonas@southpole.se> Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-05-29 09:25:19 -04:00
if (msecs == 0)
return;
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
CLUTTER_NOTE (SCHEDULER, "Advancing %d timelines by %lu milliseconds",
g_slist_length (master_clock->timelines),
msecs);
for (l = master_clock->timelines; l != NULL; l = l->next)
{
ClutterTimeline *timeline = l->data;
if (clutter_timeline_is_playing (timeline))
[timeline] Expose the msec advancement The method of ClutterTimeline that advances the timeline by a delta (in millisecond) is going to be useful for testing the timeline's behaviour -- and unbreak the timeline test suite that was broken by the MasterClock merge.
2009-05-01 10:05:51 -04:00
clutter_timeline_advance_delta (timeline, msecs);
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
}
[master clock] Handle Timeline::started signal correctly The "started" signal is sent first after the timeline has been set to the 'running' state. For this reason, checking if the clock has any running timelines running will always return true in the "started" signal handler: the timeline that sent the signal is running. What needs to be checked in the signal handler is if there are any timelines running other than the one that emitted the ::started signal, which we know is running anyway. This prevents frames from being lost at the beginning of an animation when a timeline is started after a quiescent period. Fixes bug: http://bugzilla.openedhand.com/show_bug.cgi?id=1617 Signed-off-by: Jonas Bonn <jonas@southpole.se> Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-05-29 09:25:19 -04:00
/* store the previous state so that we can use
* it for the next advancement
*/
Use a single master "clock" to drive timelines Currently, all timelines install a timeout inside the TimeoutPool they share. Every time the main loop spins, all the timeouts are updated. This, in turn, will usually lead to redraws being queued on the stages. This behaviour leads to the potential starvation of timelines and to excessive redraws. One lesson learned from the games developers is that the scenegraph should be prepared in its entirety before the GL paint sequence is initiated. This means making sure that every ::new-frame signal handler is called before clutter_redraw() is invoked. In order to do so a TimeoutPool is not enough: we need a master clock. The clock will be responsible for advancing all the active timelines created inside a scene, but only when the stage is being redrawn. The sequence is: + queue_redraw() is invoked on an actor and bubbles up to the stage + if no redraw() has already been scheduled, install an idle handler with a known priority + inside the idle handler: - advance the master clock, which will in turn advance every playing timeline by the amount of milliseconds elapsed since the last redraw; this will make every playing timeline emit the ::new-frame signal - queue a relayout - call the redraw() method of the backend This way we trade multiple timeouts with a single frame source that only runs if a timeline is playing and queues redraws on the various stages.
2009-03-21 16:39:32 -04:00
master_clock->msecs_delta = msecs;
master_clock->prev_tick = cur_tick;
}
Reference in New Issue Copy Permalink