1
0
Fork 0
mutter-performance-source/src/compositor/mutter-window.c

1807 lines
49 KiB
C
Raw Normal View History

/* -*- mode: C; c-file-style: "gnu"; indent-tabs-mode: nil; -*- */
#define _ISOC99_SOURCE /* for roundf */
#include <math.h>
#include <X11/extensions/shape.h>
#include <X11/extensions/Xcomposite.h>
#include <X11/extensions/Xdamage.h>
#include <X11/extensions/Xrender.h>
#include <clutter/x11/clutter-x11.h>
#include "display.h"
#include "errors.h"
#include "frame.h"
#include "window.h"
#include "xprops.h"
#include "compositor-private.h"
#include "mutter-shaped-texture.h"
#include "mutter-window-private.h"
#include "shadow.h"
#include "tidy/tidy-texture-frame.h"
struct _MutterWindowPrivate
{
XWindowAttributes attrs;
MetaWindow *window;
Window xwindow;
MetaScreen *screen;
ClutterActor *actor;
ClutterActor *shadow;
Pixmap back_pixmap;
MetaCompWindowType type;
Damage damage;
guint8 opacity;
gchar * desc;
/* If the window is shaped, a region that matches the shape */
MetaRegion *shape_region;
/* A rectangular region with the unshaped extends of the window
* texture */
MetaRegion *bounding_region;
gint freeze_count;
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
/*
* These need to be counters rather than flags, since more plugins
* can implement same effect; the practicality of stacking effects
* might be dubious, but we have to at least handle it correctly.
*/
gint minimize_in_progress;
gint maximize_in_progress;
gint unmaximize_in_progress;
gint map_in_progress;
gint destroy_in_progress;
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
guint visible : 1;
guint mapped : 1;
guint shaped : 1;
guint argb32 : 1;
guint disposed : 1;
guint redecorating : 1;
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
guint needs_damage_all : 1;
guint received_damage : 1;
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
guint needs_pixmap : 1;
guint needs_reshape : 1;
guint size_changed : 1;
guint needs_destroy : 1;
guint no_shadow : 1;
guint no_more_x_calls : 1;
};
enum
{
PROP_MCW_META_WINDOW = 1,
PROP_MCW_META_SCREEN,
PROP_MCW_X_WINDOW,
PROP_MCW_X_WINDOW_ATTRIBUTES,
PROP_MCW_NO_SHADOW,
};
static void mutter_window_dispose (GObject *object);
static void mutter_window_finalize (GObject *object);
static void mutter_window_constructed (GObject *object);
static void mutter_window_set_property (GObject *object,
guint prop_id,
const GValue *value,
GParamSpec *pspec);
static void mutter_window_get_property (GObject *object,
guint prop_id,
GValue *value,
GParamSpec *pspec);
static void mutter_window_detach (MutterWindow *self);
static gboolean mutter_window_has_shadow (MutterWindow *self);
static void mutter_window_clear_shape_region (MutterWindow *self);
static void mutter_window_clear_bounding_region (MutterWindow *self);
static gboolean is_shaped (MetaDisplay *display,
Window xwindow);
/*
* Register GType wrapper for XWindowAttributes, so we do not have to
* query window attributes in the MutterWindow constructor but can pass
* them as a property to the constructor (so we can gracefully handle the case
* where no attributes can be retrieved).
*
* NB -- we only need a subset of the attributes; at some point we might want
* to just store the relevant values rather than the whole struct.
*/
#define META_TYPE_XATTRS (meta_xattrs_get_type ())
static GType meta_xattrs_get_type (void) G_GNUC_CONST;
static XWindowAttributes *
meta_xattrs_copy (const XWindowAttributes *attrs)
{
XWindowAttributes *result;
g_return_val_if_fail (attrs != NULL, NULL);
result = (XWindowAttributes*) g_malloc (sizeof (XWindowAttributes));
*result = *attrs;
return result;
}
static void
meta_xattrs_free (XWindowAttributes *attrs)
{
g_return_if_fail (attrs != NULL);
g_free (attrs);
}
static GType
meta_xattrs_get_type (void)
{
static GType our_type = 0;
if (!our_type)
our_type = g_boxed_type_register_static ("XWindowAttributes",
(GBoxedCopyFunc) meta_xattrs_copy,
(GBoxedFreeFunc) meta_xattrs_free);
return our_type;
}
G_DEFINE_TYPE (MutterWindow, mutter_window, CLUTTER_TYPE_GROUP);
static void
mutter_window_class_init (MutterWindowClass *klass)
{
GObjectClass *object_class = G_OBJECT_CLASS (klass);
GParamSpec *pspec;
g_type_class_add_private (klass, sizeof (MutterWindowPrivate));
object_class->dispose = mutter_window_dispose;
object_class->finalize = mutter_window_finalize;
object_class->set_property = mutter_window_set_property;
object_class->get_property = mutter_window_get_property;
object_class->constructed = mutter_window_constructed;
pspec = g_param_spec_object ("meta-window",
"MetaWindow",
"The displayed MetaWindow",
META_TYPE_WINDOW,
G_PARAM_READWRITE | G_PARAM_CONSTRUCT);
g_object_class_install_property (object_class,
PROP_MCW_META_WINDOW,
pspec);
pspec = g_param_spec_pointer ("meta-screen",
"MetaScreen",
"MetaScreen",
G_PARAM_READWRITE | G_PARAM_CONSTRUCT);
g_object_class_install_property (object_class,
PROP_MCW_META_SCREEN,
pspec);
pspec = g_param_spec_ulong ("x-window",
"Window",
"Window",
0,
G_MAXULONG,
0,
G_PARAM_READWRITE | G_PARAM_CONSTRUCT);
g_object_class_install_property (object_class,
PROP_MCW_X_WINDOW,
pspec);
pspec = g_param_spec_boxed ("x-window-attributes",
"XWindowAttributes",
"XWindowAttributes",
META_TYPE_XATTRS,
G_PARAM_READWRITE | G_PARAM_CONSTRUCT);
g_object_class_install_property (object_class,
PROP_MCW_X_WINDOW_ATTRIBUTES,
pspec);
pspec = g_param_spec_boolean ("no-shadow",
"No shadow",
"Do not add shaddow to this window",
FALSE,
G_PARAM_READWRITE | G_PARAM_CONSTRUCT);
g_object_class_install_property (object_class,
PROP_MCW_NO_SHADOW,
pspec);
}
static void
mutter_window_init (MutterWindow *self)
{
MutterWindowPrivate *priv;
priv = self->priv = G_TYPE_INSTANCE_GET_PRIVATE (self,
MUTTER_TYPE_COMP_WINDOW,
MutterWindowPrivate);
priv->opacity = 0xff;
}
static void
mutter_meta_window_decorated_notify (MetaWindow *mw,
GParamSpec *arg1,
gpointer data)
{
MutterWindow *self = MUTTER_WINDOW (data);
MutterWindowPrivate *priv = self->priv;
MetaFrame *frame = meta_window_get_frame (mw);
MetaScreen *screen = priv->screen;
MetaDisplay *display = meta_screen_get_display (screen);
Display *xdisplay = meta_display_get_xdisplay (display);
Window new_xwindow;
MetaCompScreen *info;
XWindowAttributes attrs;
/*
* Basically, we have to reconstruct the the internals of this object
* from scratch, as everything has changed.
*/
priv->redecorating = TRUE;
if (frame)
new_xwindow = meta_frame_get_xwindow (frame);
else
new_xwindow = meta_window_get_xwindow (mw);
mutter_window_detach (self);
info = meta_screen_get_compositor_data (screen);
/*
* First of all, clean up any resources we are currently using and will
* be replacing.
*/
if (priv->damage != None)
{
meta_error_trap_push (display);
XDamageDestroy (xdisplay, priv->damage);
meta_error_trap_pop (display, FALSE);
priv->damage = None;
}
g_free (priv->desc);
priv->desc = NULL;
priv->xwindow = new_xwindow;
if (!XGetWindowAttributes (xdisplay, new_xwindow, &attrs))
{
g_warning ("Could not obtain attributes for window 0x%x after "
"decoration change",
(guint) new_xwindow);
return;
}
g_object_set (self, "x-window-attributes", &attrs, NULL);
if (priv->shadow)
{
ClutterActor *p = clutter_actor_get_parent (priv->shadow);
if (CLUTTER_IS_CONTAINER (p))
clutter_container_remove_actor (CLUTTER_CONTAINER (p), priv->shadow);
else
clutter_actor_unparent (priv->shadow);
priv->shadow = NULL;
}
/*
* Recreate the contents.
*/
mutter_window_constructed (G_OBJECT (self));
}
static void
mutter_window_constructed (GObject *object)
{
MutterWindow *self = MUTTER_WINDOW (object);
MutterWindowPrivate *priv = self->priv;
MetaScreen *screen = priv->screen;
MetaDisplay *display = meta_screen_get_display (screen);
Window xwindow = priv->xwindow;
Display *xdisplay = meta_display_get_xdisplay (display);
XRenderPictFormat *format;
MetaCompositor *compositor;
compositor = meta_display_get_compositor (display);
mutter_window_update_window_type (self);
#ifdef HAVE_SHAPE
/* Listen for ShapeNotify events on the window */
if (meta_display_has_shape (display))
XShapeSelectInput (xdisplay, xwindow, ShapeNotifyMask);
#endif
priv->shaped = is_shaped (display, xwindow);
if (priv->attrs.class == InputOnly)
priv->damage = None;
else
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
priv->damage = XDamageCreate (xdisplay, xwindow,
XDamageReportBoundingBox);
format = XRenderFindVisualFormat (xdisplay, priv->attrs.visual);
if (format && format->type == PictTypeDirect && format->direct.alphaMask)
priv->argb32 = TRUE;
mutter_window_update_opacity (self);
if (mutter_window_has_shadow (self))
{
priv->shadow = mutter_create_shadow_frame (compositor);
clutter_container_add_actor (CLUTTER_CONTAINER (self), priv->shadow);
}
if (!priv->actor)
{
priv->actor = mutter_shaped_texture_new ();
clutter_container_add_actor (CLUTTER_CONTAINER (self), priv->actor);
/*
* Since we are holding a pointer to this actor independently of the
* ClutterContainer internals, and provide a public API to access it,
* add a reference here, so that if someone is messing about with us
* via the container interface, we do not end up with a dangling pointer.
* We will release it in dispose().
*/
g_object_ref (priv->actor);
g_signal_connect (priv->window, "notify::decorated",
G_CALLBACK (mutter_meta_window_decorated_notify), self);
}
else
{
/*
* This is the case where existing window is gaining/loosing frame.
* Just ensure the actor is top most (i.e., above shadow).
*/
clutter_actor_raise_top (priv->actor);
}
mutter_window_update_shape (self, priv->shaped);
}
static void
mutter_window_dispose (GObject *object)
{
MutterWindow *self = MUTTER_WINDOW (object);
MutterWindowPrivate *priv = self->priv;
MetaScreen *screen;
MetaDisplay *display;
Display *xdisplay;
MetaCompScreen *info;
if (priv->disposed)
return;
priv->disposed = TRUE;
screen = priv->screen;
display = meta_screen_get_display (screen);
xdisplay = meta_display_get_xdisplay (display);
info = meta_screen_get_compositor_data (screen);
mutter_window_detach (self);
mutter_window_clear_shape_region (self);
mutter_window_clear_bounding_region (self);
if (priv->damage != None)
{
meta_error_trap_push (display);
XDamageDestroy (xdisplay, priv->damage);
meta_error_trap_pop (display, FALSE);
priv->damage = None;
}
info->windows = g_list_remove (info->windows, (gconstpointer) self);
/*
* Release the extra reference we took on the actor.
*/
g_object_unref (priv->actor);
priv->actor = NULL;
G_OBJECT_CLASS (mutter_window_parent_class)->dispose (object);
}
static void
mutter_window_finalize (GObject *object)
{
MutterWindow *self = MUTTER_WINDOW (object);
MutterWindowPrivate *priv = self->priv;
g_free (priv->desc);
G_OBJECT_CLASS (mutter_window_parent_class)->finalize (object);
}
static void
mutter_window_set_property (GObject *object,
guint prop_id,
const GValue *value,
GParamSpec *pspec)
{
MutterWindow *self = MUTTER_WINDOW (object);
MutterWindowPrivate *priv = self->priv;
switch (prop_id)
{
case PROP_MCW_META_WINDOW:
priv->window = g_value_get_object (value);
break;
case PROP_MCW_META_SCREEN:
priv->screen = g_value_get_pointer (value);
break;
case PROP_MCW_X_WINDOW:
priv->xwindow = g_value_get_ulong (value);
break;
case PROP_MCW_X_WINDOW_ATTRIBUTES:
priv->attrs = *((XWindowAttributes*)g_value_get_boxed (value));
break;
case PROP_MCW_NO_SHADOW:
{
gboolean oldv = priv->no_shadow ? TRUE : FALSE;
gboolean newv = g_value_get_boolean (value);
if (oldv == newv)
return;
priv->no_shadow = newv;
if (newv && priv->shadow)
{
clutter_container_remove_actor (CLUTTER_CONTAINER (object),
priv->shadow);
priv->shadow = NULL;
}
else if (!newv && !priv->shadow && mutter_window_has_shadow (self))
{
gfloat w, h;
MetaDisplay *display = meta_screen_get_display (priv->screen);
MetaCompositor *compositor;
compositor = meta_display_get_compositor (display);
clutter_actor_get_size (CLUTTER_ACTOR (self), &w, &h);
priv->shadow = mutter_create_shadow_frame (compositor);
clutter_actor_set_size (priv->shadow, w, h);
clutter_container_add_actor (CLUTTER_CONTAINER (self), priv->shadow);
}
}
break;
default:
G_OBJECT_WARN_INVALID_PROPERTY_ID (object, prop_id, pspec);
break;
}
}
static void
mutter_window_get_property (GObject *object,
guint prop_id,
GValue *value,
GParamSpec *pspec)
{
MutterWindowPrivate *priv = MUTTER_WINDOW (object)->priv;
switch (prop_id)
{
case PROP_MCW_META_WINDOW:
g_value_set_object (value, priv->window);
break;
case PROP_MCW_META_SCREEN:
g_value_set_pointer (value, priv->screen);
break;
case PROP_MCW_X_WINDOW:
g_value_set_ulong (value, priv->xwindow);
break;
case PROP_MCW_X_WINDOW_ATTRIBUTES:
g_value_set_boxed (value, &priv->attrs);
break;
case PROP_MCW_NO_SHADOW:
g_value_set_boolean (value, priv->no_shadow);
break;
default:
G_OBJECT_WARN_INVALID_PROPERTY_ID (object, prop_id, pspec);
break;
}
}
void
mutter_window_update_window_type (MutterWindow *self)
{
MutterWindowPrivate *priv = self->priv;
priv->type = (MetaCompWindowType) meta_window_get_window_type (priv->window);
}
static gboolean
is_shaped (MetaDisplay *display, Window xwindow)
{
Display *xdisplay = meta_display_get_xdisplay (display);
gint xws, yws, xbs, ybs;
guint wws, hws, wbs, hbs;
gint bounding_shaped, clip_shaped;
if (meta_display_has_shape (display))
{
XShapeQueryExtents (xdisplay, xwindow, &bounding_shaped,
&xws, &yws, &wws, &hws, &clip_shaped,
&xbs, &ybs, &wbs, &hbs);
return (bounding_shaped != 0);
}
return FALSE;
}
static gboolean
mutter_window_has_shadow (MutterWindow *self)
{
MutterWindowPrivate * priv = self->priv;
if (priv->no_shadow)
return FALSE;
/*
* Always put a shadow around windows with a frame - This should override
* the restriction about not putting a shadow around shaped windows
* as the frame might be the reason the window is shaped
*/
if (priv->window)
{
if (meta_window_get_frame (priv->window))
{
meta_verbose ("Window 0x%x has shadow because it has a frame\n",
(guint)priv->xwindow);
return TRUE;
}
}
/*
* Do not add shadows to ARGB windows (since they are probably transparent)
*/
if (priv->argb32 || priv->opacity != 0xff)
{
meta_verbose ("Window 0x%x has no shadow as it is ARGB\n",
(guint)priv->xwindow);
return FALSE;
}
/*
* Never put a shadow around shaped windows
*/
if (priv->shaped)
{
meta_verbose ("Window 0x%x has no shadow as it is shaped\n",
(guint)priv->xwindow);
return FALSE;
}
/*
* Add shadows to override redirect windows (e.g., Gtk menus).
* This must have lower priority than window shape test.
*/
if (priv->attrs.override_redirect)
{
meta_verbose ("Window 0x%x has shadow because it is override redirect.\n",
(guint)priv->xwindow);
return TRUE;
}
/*
* Don't put shadow around DND icon windows
*/
if (priv->type == META_COMP_WINDOW_DND ||
priv->type == META_COMP_WINDOW_DESKTOP)
{
meta_verbose ("Window 0x%x has no shadow as it is DND or Desktop\n",
(guint)priv->xwindow);
return FALSE;
}
if (priv->type == META_COMP_WINDOW_MENU
#if 0
|| priv->type == META_COMP_WINDOW_DROPDOWN_MENU
#endif
)
{
meta_verbose ("Window 0x%x has shadow as it is a menu\n",
(guint)priv->xwindow);
return TRUE;
}
#if 0
if (priv->type == META_COMP_WINDOW_TOOLTIP)
{
meta_verbose ("Window 0x%x has shadow as it is a tooltip\n",
(guint)priv->xwindow);
return TRUE;
}
#endif
meta_verbose ("Window 0x%x has no shadow as it fell through\n",
(guint)priv->xwindow);
return FALSE;
}
Window
mutter_window_get_x_window (MutterWindow *self)
{
if (!self)
return None;
return self->priv->xwindow;
}
/**
* mutter_window_get_meta_window:
*
* Gets the MetaWindow object that the the MutterWindow is displaying
*
* Return value: (transfer none): the displayed MetaWindow
*/
MetaWindow *
mutter_window_get_meta_window (MutterWindow *self)
{
return self->priv->window;
}
/**
* mutter_window_get_texture:
*
* Gets the ClutterActor that is used to display the contents of the window
*
* Return value: (transfer none): the ClutterActor for the contents
*/
ClutterActor *
mutter_window_get_texture (MutterWindow *self)
{
return self->priv->actor;
}
MetaCompWindowType
mutter_window_get_window_type (MutterWindow *self)
{
if (!self)
return 0;
return self->priv->type;
}
gboolean
mutter_window_is_override_redirect (MutterWindow *self)
{
return meta_window_is_override_redirect (self->priv->window);
}
const char *mutter_window_get_description (MutterWindow *self)
{
/*
* For windows managed by the WM, we just defer to the WM for the window
* description. For override-redirect windows, we create the description
* ourselves, but only on demand.
*/
if (self->priv->window)
return meta_window_get_description (self->priv->window);
if (G_UNLIKELY (self->priv->desc == NULL))
{
self->priv->desc = g_strdup_printf ("Override Redirect (0x%x)",
(guint) self->priv->xwindow);
}
return self->priv->desc;
}
/**
* mutter_window_get_workspace:
* @self: #MutterWindow
*
* Returns the index of workspace on which this window is located; if the
* window is sticky, or is not currently located on any workspace, returns -1.
* This function is deprecated and should not be used in newly written code;
* meta_window_get_workspace() instead.
*
* Return value: (transfer none): index of workspace on which this window is
* located.
*/
gint
mutter_window_get_workspace (MutterWindow *self)
{
MutterWindowPrivate *priv;
MetaWorkspace *workspace;
if (!self)
return -1;
priv = self->priv;
if (!priv->window || meta_window_is_on_all_workspaces (priv->window))
return -1;
workspace = meta_window_get_workspace (priv->window);
if (!workspace)
return -1;
return meta_workspace_index (workspace);
}
gboolean
mutter_window_showing_on_its_workspace (MutterWindow *self)
{
if (!self)
return FALSE;
/* If override redirect: */
if (!self->priv->window)
return TRUE;
return meta_window_showing_on_its_workspace (self->priv->window);
}
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
static void
mutter_window_freeze (MutterWindow *self)
{
self->priv->freeze_count++;
}
static void
mutter_window_damage_all (MutterWindow *self)
{
MutterWindowPrivate *priv = self->priv;
ClutterX11TexturePixmap *texture_x11 = CLUTTER_X11_TEXTURE_PIXMAP (priv->actor);
guint pixmap_width = 0;
guint pixmap_height = 0;
if (!priv->needs_damage_all)
return;
g_object_get (texture_x11,
"pixmap-width", &pixmap_width,
"pixmap-height", &pixmap_height,
NULL);
clutter_x11_texture_pixmap_update_area (texture_x11,
0,
0,
pixmap_width,
pixmap_height);
priv->needs_damage_all = FALSE;
}
static void
mutter_window_thaw (MutterWindow *self)
{
self->priv->freeze_count--;
if (G_UNLIKELY (self->priv->freeze_count < 0))
{
g_warning ("Error in freeze/thaw accounting.");
self->priv->freeze_count = 0;
return;
}
if (self->priv->freeze_count)
return;
/* Since we ignore damage events while a window is frozen for certain effects
* we may need to issue an update_area() covering the whole pixmap if we
* don't know what real damage has happened. */
if (self->priv->needs_damage_all)
mutter_window_damage_all (self);
}
gboolean
mutter_window_effect_in_progress (MutterWindow *self)
{
return (self->priv->minimize_in_progress ||
self->priv->maximize_in_progress ||
self->priv->unmaximize_in_progress ||
self->priv->map_in_progress ||
self->priv->destroy_in_progress);
}
static void
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
mutter_window_queue_create_pixmap (MutterWindow *self)
{
MutterWindowPrivate *priv = self->priv;
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
priv->needs_pixmap = TRUE;
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
if (!priv->mapped)
return;
/* This will cause the compositor paint function to be run
* if the actor is visible or a clone of the actor is visible.
* if the actor isn't visible in any way, then we don't
* need to repair the window anyways, and can wait until
* the stage is redrawn for some other reason
*
* The compositor paint function repairs all windows.
*/
clutter_actor_queue_redraw (priv->actor);
}
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
static gboolean
is_freeze_thaw_effect (gulong event)
{
switch (event)
{
case MUTTER_PLUGIN_DESTROY:
case MUTTER_PLUGIN_MAXIMIZE:
case MUTTER_PLUGIN_UNMAXIMIZE:
return TRUE;
break;
default:
return FALSE;
}
}
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
static gboolean
start_simple_effect (MutterWindow *self,
gulong event)
{
MutterWindowPrivate *priv = self->priv;
MetaCompScreen *info = meta_screen_get_compositor_data (priv->screen);
gint *counter = NULL;
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
gboolean use_freeze_thaw = FALSE;
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
if (!info->plugin_mgr)
return FALSE;
switch (event)
{
case MUTTER_PLUGIN_MINIMIZE:
counter = &priv->minimize_in_progress;
break;
case MUTTER_PLUGIN_MAP:
counter = &priv->map_in_progress;
break;
case MUTTER_PLUGIN_DESTROY:
counter = &priv->destroy_in_progress;
break;
case MUTTER_PLUGIN_UNMAXIMIZE:
case MUTTER_PLUGIN_MAXIMIZE:
case MUTTER_PLUGIN_SWITCH_WORKSPACE:
g_assert_not_reached ();
break;
}
g_assert (counter);
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
use_freeze_thaw = is_freeze_thaw_effect (event);
if (use_freeze_thaw)
mutter_window_freeze (self);
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
(*counter)++;
if (!mutter_plugin_manager_event_simple (info->plugin_mgr,
self,
event))
{
(*counter)--;
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
if (use_freeze_thaw)
mutter_window_thaw (self);
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
return FALSE;
}
return TRUE;
}
static void
mutter_window_after_effects (MutterWindow *self)
{
MutterWindowPrivate *priv = self->priv;
if (priv->needs_destroy)
{
clutter_actor_destroy (CLUTTER_ACTOR (self));
return;
}
mutter_window_sync_visibility (self);
mutter_window_sync_actor_position (self);
if (!meta_window_is_mapped (priv->window))
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
mutter_window_detach (self);
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
if (priv->needs_pixmap)
clutter_actor_queue_redraw (priv->actor);
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
}
void
mutter_window_effect_completed (MutterWindow *self,
gulong event)
{
MutterWindowPrivate *priv = self->priv;
/* NB: Keep in mind that when effects get completed it possible
* that the corresponding MetaWindow may have be been destroyed.
* In this case priv->window will == NULL */
switch (event)
{
case MUTTER_PLUGIN_MINIMIZE:
{
priv->minimize_in_progress--;
if (priv->minimize_in_progress < 0)
{
g_warning ("Error in minimize accounting.");
priv->minimize_in_progress = 0;
}
}
break;
case MUTTER_PLUGIN_MAP:
/*
* Make sure that the actor is at the correct place in case
* the plugin fscked.
*/
priv->map_in_progress--;
if (priv->map_in_progress < 0)
{
g_warning ("Error in map accounting.");
priv->map_in_progress = 0;
}
break;
case MUTTER_PLUGIN_DESTROY:
priv->destroy_in_progress--;
if (priv->destroy_in_progress < 0)
{
g_warning ("Error in destroy accounting.");
priv->destroy_in_progress = 0;
}
break;
case MUTTER_PLUGIN_UNMAXIMIZE:
priv->unmaximize_in_progress--;
if (priv->unmaximize_in_progress < 0)
{
g_warning ("Error in unmaximize accounting.");
priv->unmaximize_in_progress = 0;
}
break;
case MUTTER_PLUGIN_MAXIMIZE:
priv->maximize_in_progress--;
if (priv->maximize_in_progress < 0)
{
g_warning ("Error in maximize accounting.");
priv->maximize_in_progress = 0;
}
break;
case MUTTER_PLUGIN_SWITCH_WORKSPACE:
g_assert_not_reached ();
break;
}
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
if (is_freeze_thaw_effect (event))
mutter_window_thaw (self);
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
if (!mutter_window_effect_in_progress (self))
mutter_window_after_effects (self);
}
/* Called to drop our reference to a window backing pixmap that we
* previously obtained with XCompositeNameWindowPixmap. We do this
* when the window is unmapped or when we want to update to a new
* pixmap for a new size.
*/
static void
mutter_window_detach (MutterWindow *self)
{
MutterWindowPrivate *priv = self->priv;
MetaScreen *screen = priv->screen;
MetaDisplay *display = meta_screen_get_display (screen);
Display *xdisplay = meta_display_get_xdisplay (display);
if (!priv->back_pixmap)
return;
/* Get rid of all references to the pixmap before freeing it; it's unclear whether
* you are supposed to be able to free a GLXPixmap after freeing the underlying
* pixmap, but it certainly doesn't work with current DRI/Mesa
*/
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
clutter_x11_texture_pixmap_set_pixmap (CLUTTER_X11_TEXTURE_PIXMAP (priv->actor),
None);
mutter_shaped_texture_clear (MUTTER_SHAPED_TEXTURE (priv->actor));
cogl_flush();
XFreePixmap (xdisplay, priv->back_pixmap);
priv->back_pixmap = None;
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
mutter_window_queue_create_pixmap (self);
}
void
mutter_window_destroy (MutterWindow *self)
{
MetaWindow *window;
MetaCompScreen *info;
MutterWindowPrivate *priv;
priv = self->priv;
window = priv->window;
meta_window_set_compositor_private (window, NULL);
/*
* We remove the window from internal lookup hashes and thus any other
* unmap events etc fail
*/
info = meta_screen_get_compositor_data (priv->screen);
info->windows = g_list_remove (info->windows, (gconstpointer) self);
if (priv->type == META_COMP_WINDOW_DROPDOWN_MENU ||
priv->type == META_COMP_WINDOW_POPUP_MENU ||
priv->type == META_COMP_WINDOW_TOOLTIP ||
priv->type == META_COMP_WINDOW_NOTIFICATION ||
priv->type == META_COMP_WINDOW_COMBO ||
priv->type == META_COMP_WINDOW_DND ||
priv->type == META_COMP_WINDOW_OVERRIDE_OTHER)
{
/*
* No effects, just kill it.
*/
clutter_actor_destroy (CLUTTER_ACTOR (self));
return;
}
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
priv->needs_destroy = TRUE;
/*
* Once the window destruction is initiated we can no longer perform any
* furter X-based operations. For example, if we have a Map effect running,
* we cannot query the window geometry once the effect completes. So, flag
* this.
*/
priv->no_more_x_calls = TRUE;
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
if (!mutter_window_effect_in_progress (self))
clutter_actor_destroy (CLUTTER_ACTOR (self));
}
void
mutter_window_sync_actor_position (MutterWindow *self)
{
MutterWindowPrivate *priv = self->priv;
MetaRectangle window_rect;
meta_window_get_outer_rect (priv->window, &window_rect);
if (priv->attrs.width != window_rect.width ||
priv->attrs.height != window_rect.height)
{
priv->size_changed = TRUE;
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
mutter_window_queue_create_pixmap (self);
}
/* XXX deprecated: please use meta_window_get_outer_rect instead */
priv->attrs.width = window_rect.width;
priv->attrs.height = window_rect.height;
priv->attrs.x = window_rect.x;
priv->attrs.y = window_rect.y;
if (mutter_window_effect_in_progress (self))
return;
clutter_actor_set_position (CLUTTER_ACTOR (self),
window_rect.x, window_rect.y);
clutter_actor_set_size (CLUTTER_ACTOR (self),
window_rect.width, window_rect.height);
}
void
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
mutter_window_show (MutterWindow *self,
MetaCompEffect effect)
{
MutterWindowPrivate *priv;
MetaCompScreen *info;
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
gulong event;
priv = self->priv;
info = meta_screen_get_compositor_data (priv->screen);
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
g_return_if_fail (!priv->visible);
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
self->priv->visible = TRUE;
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
event = 0;
switch (effect)
{
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
case META_COMP_EFFECT_CREATE:
event = MUTTER_PLUGIN_MAP;
break;
case META_COMP_EFFECT_UNMINIMIZE:
/* FIXME: should have MUTTER_PLUGIN_UNMINIMIZE */
event = MUTTER_PLUGIN_MAP;
break;
case META_COMP_EFFECT_NONE:
break;
case META_COMP_EFFECT_DESTROY:
case META_COMP_EFFECT_MINIMIZE:
g_assert_not_reached();
}
if (priv->redecorating ||
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
info->switch_workspace_in_progress ||
event == 0 ||
!start_simple_effect (self, event))
{
clutter_actor_show_all (CLUTTER_ACTOR (self));
priv->redecorating = FALSE;
}
}
void
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
mutter_window_hide (MutterWindow *self,
MetaCompEffect effect)
{
MutterWindowPrivate *priv;
MetaCompScreen *info;
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
gulong event;
priv = self->priv;
info = meta_screen_get_compositor_data (priv->screen);
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
g_return_if_fail (priv->visible);
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
priv->visible = FALSE;
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
/* If a plugin is animating a workspace transition, we have to
* hold off on hiding the window, and do it after the workspace
* switch completes
*/
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
if (info->switch_workspace_in_progress)
return;
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
event = 0;
switch (effect)
{
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
case META_COMP_EFFECT_DESTROY:
event = MUTTER_PLUGIN_DESTROY;
break;
case META_COMP_EFFECT_MINIMIZE:
event = MUTTER_PLUGIN_MINIMIZE;
break;
case META_COMP_EFFECT_NONE:
break;
case META_COMP_EFFECT_UNMINIMIZE:
case META_COMP_EFFECT_CREATE:
g_assert_not_reached();
}
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
if (event == 0 ||
!start_simple_effect (self, event))
clutter_actor_hide (CLUTTER_ACTOR (self));
}
void
mutter_window_maximize (MutterWindow *self,
MetaRectangle *old_rect,
MetaRectangle *new_rect)
{
MetaCompScreen *info = meta_screen_get_compositor_data (self->priv->screen);
/* The window has already been resized (in order to compute new_rect),
* which by side effect caused the actor to be resized. Restore it to the
* old size and position */
clutter_actor_set_position (CLUTTER_ACTOR (self), old_rect->x, old_rect->y);
clutter_actor_set_size (CLUTTER_ACTOR (self), old_rect->width, old_rect->height);
self->priv->maximize_in_progress++;
mutter_window_freeze (self);
if (!info->plugin_mgr ||
!mutter_plugin_manager_event_maximize (info->plugin_mgr,
self,
MUTTER_PLUGIN_MAXIMIZE,
new_rect->x, new_rect->y,
new_rect->width, new_rect->height))
{
self->priv->maximize_in_progress--;
mutter_window_thaw (self);
}
}
void
mutter_window_unmaximize (MutterWindow *self,
MetaRectangle *old_rect,
MetaRectangle *new_rect)
{
MetaCompScreen *info = meta_screen_get_compositor_data (self->priv->screen);
/* The window has already been resized (in order to compute new_rect),
* which by side effect caused the actor to be resized. Restore it to the
* old size and position */
clutter_actor_set_position (CLUTTER_ACTOR (self), old_rect->x, old_rect->y);
clutter_actor_set_size (CLUTTER_ACTOR (self), old_rect->width, old_rect->height);
self->priv->unmaximize_in_progress++;
mutter_window_freeze (self);
if (!info->plugin_mgr ||
!mutter_plugin_manager_event_maximize (info->plugin_mgr,
self,
MUTTER_PLUGIN_UNMAXIMIZE,
new_rect->x, new_rect->y,
new_rect->width, new_rect->height))
{
self->priv->unmaximize_in_progress--;
mutter_window_thaw (self);
}
}
MutterWindow *
mutter_window_new (MetaWindow *window)
{
MetaScreen *screen = meta_window_get_screen (window);
MetaDisplay *display = meta_screen_get_display (screen);
MetaCompScreen *info = meta_screen_get_compositor_data (screen);
MutterWindow *self;
MutterWindowPrivate *priv;
MetaFrame *frame;
Window top_window;
XWindowAttributes attrs;
frame = meta_window_get_frame (window);
if (frame)
top_window = meta_frame_get_xwindow (frame);
else
top_window = meta_window_get_xwindow (window);
meta_verbose ("add window: Meta %p, xwin 0x%x\n", window, (guint)top_window);
/* FIXME: Remove the redundant data we store in self->priv->attrs, and
* simply query metacity core for the data. */
if (!XGetWindowAttributes (meta_display_get_xdisplay (display), top_window, &attrs))
return NULL;
self = g_object_new (MUTTER_TYPE_COMP_WINDOW,
"meta-window", window,
"x-window", top_window,
"meta-screen", screen,
"x-window-attributes", &attrs,
NULL);
priv = self->priv;
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
priv->mapped = meta_window_toplevel_is_mapped (priv->window);
if (priv->mapped)
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
mutter_window_queue_create_pixmap (self);
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
mutter_window_sync_actor_position (self);
/* Hang our compositor window state off the MetaWindow for fast retrieval */
meta_window_set_compositor_private (window, G_OBJECT (self));
clutter_container_add_actor (CLUTTER_CONTAINER (info->window_group),
CLUTTER_ACTOR (self));
clutter_actor_hide (CLUTTER_ACTOR (self));
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
/* Initial position in the stack is arbitrary; stacking will be synced
* before we first paint.
*/
info->windows = g_list_append (info->windows, self);
return self;
}
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
void
mutter_window_mapped (MutterWindow *self)
{
MutterWindowPrivate *priv = self->priv;
g_return_if_fail (!priv->mapped);
priv->mapped = TRUE;
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
mutter_window_queue_create_pixmap (self);
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
}
void
mutter_window_unmapped (MutterWindow *self)
{
MutterWindowPrivate *priv = self->priv;
g_return_if_fail (priv->mapped);
priv->mapped = FALSE;
if (mutter_window_effect_in_progress (self))
return;
mutter_window_detach (self);
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
priv->needs_pixmap = FALSE;
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
}
static void
mutter_window_clear_shape_region (MutterWindow *self)
{
MutterWindowPrivate *priv = self->priv;
if (priv->shape_region)
{
meta_region_destroy (priv->shape_region);
priv->shape_region = NULL;
}
}
static void
mutter_window_clear_bounding_region (MutterWindow *self)
{
MutterWindowPrivate *priv = self->priv;
if (priv->bounding_region)
{
meta_region_destroy (priv->bounding_region);
priv->bounding_region = NULL;
}
}
static void
mutter_window_update_bounding_region (MutterWindow *self,
int width,
int height)
{
MutterWindowPrivate *priv = self->priv;
GdkRectangle bounding_rectangle = { 0, 0, width, height };
mutter_window_clear_bounding_region (self);
priv->bounding_region = meta_region_new_from_rectangle (&bounding_rectangle);
}
static void
mutter_window_update_shape_region (MutterWindow *self,
int n_rects,
XRectangle *rects)
{
MutterWindowPrivate *priv = self->priv;
int i;
mutter_window_clear_shape_region (self);
priv->shape_region = meta_region_new ();
for (i = 0; i < n_rects; i++)
{
GdkRectangle rect = { rects[i].x, rects[i].y, rects[i].width, rects[i].height };
meta_region_union_rectangle (priv->shape_region, &rect);
}
}
/**
* mutter_window_get_obscured_region:
* @self: a #MutterWindow
*
* Gets the region that is completely obscured by the window. Coordinates
* are relative to the upper-left of the window.
*
* Return value: (transfer none): the area obscured by the window,
* %NULL is the same as an empty region.
*/
MetaRegion *
mutter_window_get_obscured_region (MutterWindow *self)
{
MutterWindowPrivate *priv = self->priv;
if (!priv->argb32 && priv->back_pixmap)
{
if (priv->shaped)
return priv->shape_region;
else
return priv->bounding_region;
}
else
return NULL;
}
#if 0
/* Print out a region; useful for debugging */
static void
dump_region (MetaRegion *region)
{
GdkRectangle *rects;
int n_rects;
int i;
meta_region_get_rectangles (region, &rects, &n_rects);
g_print ("[");
for (i = 0; i < n_rects; i++)
{
g_print ("+%d+%dx%dx%d ",
rects[i].x, rects[i].y, rects[i].width, rects[i].height);
}
g_print ("]\n");
g_free (rects);
}
#endif
/**
* mutter_window_set_visible_region:
* @self: a #MutterWindow
* @visible_region: the region of the screen that isn't completely
* obscured.
*
* Provides a hint as to what areas of the window need to be
* drawn. Regions not in @visible_region are completely obscured.
* This will be set before painting then unset afterwards.
*/
void
mutter_window_set_visible_region (MutterWindow *self,
MetaRegion *visible_region)
{
MutterWindowPrivate *priv = self->priv;
MetaRegion *texture_clip_region = NULL;
/* Get the area of the window texture that would be drawn if
* we weren't obscured at all
*/
if (priv->shaped)
{
if (priv->shape_region)
texture_clip_region = meta_region_copy (priv->shape_region);
}
else
{
if (priv->bounding_region)
texture_clip_region = meta_region_copy (priv->bounding_region);
}
if (!texture_clip_region)
texture_clip_region = meta_region_new ();
/* Then intersect that with the visible region to get the region
* that we actually need to redraw.
*/
meta_region_intersect (texture_clip_region, visible_region);
/* Assumes ownership */
mutter_shaped_texture_set_clip_region (MUTTER_SHAPED_TEXTURE (priv->actor),
texture_clip_region);
}
/**
* mutter_window_set_visible_region_beneath:
* @self: a #MutterWindow
* @visible_region: the region of the screen that isn't completely
* obscured beneath the main window texture.
*
* Provides a hint as to what areas need to be drawn *beneath*
* the main window texture. This is the relevant visible region
* when drawing the shadow, properly accounting for areas of the
* shadow hid by the window itself. This will be set before painting
* then unset afterwards.
*/
void
mutter_window_set_visible_region_beneath (MutterWindow *self,
MetaRegion *beneath_region)
{
MutterWindowPrivate *priv = self->priv;
if (priv->shadow)
{
GdkRectangle shadow_rect;
ClutterActorBox box;
MetaOverlapType overlap;
/* We could compute an full clip region as we do for the window
* texture, but the shadow is relatively cheap to draw, and
* a little more complex to clip, so we just catch the case where
* the shadow is completely obscured and doesn't need to be drawn
* at all.
*/
clutter_actor_get_allocation_box (priv->shadow, &box);
shadow_rect.x = roundf (box.x1);
shadow_rect.y = roundf (box.y1);
shadow_rect.width = roundf (box.x2 - box.x1);
shadow_rect.height = roundf (box.y2 - box.y1);
overlap = meta_region_contains_rectangle (beneath_region, &shadow_rect);
tidy_texture_frame_set_needs_paint (TIDY_TEXTURE_FRAME (priv->shadow),
overlap != META_REGION_OVERLAP_OUT);
}
}
/**
* mutter_window_reset_visible_regions:
* @self: a #MutterWindow
*
* Unsets the regions set by mutter_window_reset_visible_region() and
*mutter_window_reset_visible_region_beneath()
*/
void
mutter_window_reset_visible_regions (MutterWindow *self)
{
MutterWindowPrivate *priv = self->priv;
mutter_shaped_texture_set_clip_region (MUTTER_SHAPED_TEXTURE (priv->actor),
NULL);
if (priv->shadow)
tidy_texture_frame_set_needs_paint (TIDY_TEXTURE_FRAME (priv->shadow), TRUE);
}
static void
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
check_needs_pixmap (MutterWindow *self)
{
MutterWindowPrivate *priv = self->priv;
MetaScreen *screen = priv->screen;
MetaDisplay *display = meta_screen_get_display (screen);
Display *xdisplay = meta_display_get_xdisplay (display);
MetaCompScreen *info = meta_screen_get_compositor_data (screen);
MetaCompositor *compositor;
Window xwindow = priv->xwindow;
gboolean full = FALSE;
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
if (!priv->needs_pixmap)
return;
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
if (!priv->mapped)
return;
if (xwindow == meta_screen_get_xroot (screen) ||
xwindow == clutter_x11_get_stage_window (CLUTTER_STAGE (info->stage)))
return;
compositor = meta_display_get_compositor (display);
if (priv->size_changed)
{
mutter_window_detach (self);
priv->size_changed = FALSE;
}
meta_error_trap_push (display);
if (priv->back_pixmap == None)
{
gint pxm_width, pxm_height;
meta_error_trap_push (display);
priv->back_pixmap = XCompositeNameWindowPixmap (xdisplay, xwindow);
if (meta_error_trap_pop_with_return (display, FALSE) != Success)
{
/* Probably a BadMatch if the window isn't viewable; we could
* GrabServer/GetWindowAttributes/NameWindowPixmap/UngrabServer/Sync
* to avoid this, but there's no reason to take two round trips
* when one will do. (We need that Sync if we want to handle failures
* for any reason other than !viewable. That's unlikely, but maybe
* we'll BadAlloc or something.)
*/
priv->back_pixmap = None;
}
if (priv->back_pixmap == None)
{
meta_verbose ("Unable to get named pixmap for %p\n", self);
mutter_window_update_bounding_region (self, 0, 0);
return;
}
if (compositor->no_mipmaps)
mutter_shaped_texture_set_create_mipmaps (MUTTER_SHAPED_TEXTURE (priv->actor),
FALSE);
clutter_x11_texture_pixmap_set_pixmap
(CLUTTER_X11_TEXTURE_PIXMAP (priv->actor),
priv->back_pixmap);
/*
* This only works *after* actually setting the pixmap, so we have to
* do it here.
* See: http://bugzilla.clutter-project.org/show_bug.cgi?id=2236
*/
if (!clutter_glx_texture_pixmap_using_extension (
CLUTTER_GLX_TEXTURE_PIXMAP (priv->actor)))
g_warning ("NOTE: Not using GLX TFP!\n");
g_object_get (priv->actor,
"pixmap-width", &pxm_width,
"pixmap-height", &pxm_height,
NULL);
if (priv->shadow)
clutter_actor_set_size (priv->shadow, pxm_width, pxm_height);
mutter_window_update_bounding_region (self, pxm_width, pxm_height);
full = TRUE;
}
meta_error_trap_pop (display, FALSE);
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
priv->needs_pixmap = FALSE;
}
static gboolean
is_frozen (MutterWindow *self)
{
return self->priv->freeze_count ? TRUE : FALSE;
}
void
mutter_window_process_damage (MutterWindow *self,
XDamageNotifyEvent *event)
{
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
MutterWindowPrivate *priv = self->priv;
ClutterX11TexturePixmap *texture_x11 = CLUTTER_X11_TEXTURE_PIXMAP (priv->actor);
priv->received_damage = TRUE;
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
if (is_frozen (self))
{
/* The window is frozen due to an effect in progress: we ignore damage
* here on the off chance that this will stop the corresponding
* texture_from_pixmap from being update.
*
* needs_damage_all tracks that some unknown damage happened while the
* window was frozen so that when the window becomes unfrozen we can
* issue a full window update to cover any lost damage.
*
* It should be noted that this is an unreliable mechanism since it's
* quite likely that drivers will aim to provide a zero-copy
* implementation of the texture_from_pixmap extension and in those cases
* any drawing done to the window is always immediately reflected in the
* texture regardless of damage event handling.
*/
priv->needs_damage_all = TRUE;
return;
}
clutter_x11_texture_pixmap_update_area (texture_x11,
event->area.x,
event->area.y,
event->area.width,
event->area.height);
}
void
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
mutter_window_sync_visibility (MutterWindow *self)
{
MutterWindowPrivate *priv = self->priv;
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
if (CLUTTER_ACTOR_IS_VISIBLE (self) != priv->visible)
{
Simplify relationship between mapping and visibility Previously, changes to the visibility of a window could be indicated by meta_compositor_map_window(), meta_compositor_unminimize_window(), meta_compositor_set_window_hidden(), etc, with the exact behavior depending on the 'live_hidden_windows' preference. Simplify this so that visibility is controlled by: meta_compositor_show_window() meta_compositor_hide_window() With an 'effect' parameter provided to indicate the appropriate effect (CREATE/UNMINIMIZE/MINIMIZE/DESTROY/NONE.) The map state of the window is signalled separately by: meta_compositor_map_window() meta_compositor_unmap_window() And is used only to control resource handling. Other changes: * The desired effect on show/hide is explicitly stored in MetaWindow, avoiding the need for the was_minimized flag. At idle, once we calculate the window state, we pass the effect to the compositor if it matches the new window state, and then clear the effect to start over for future map state changes. * meta_compositor_switch_workspace() is called before any windows are hidden or shown, allowing the compositor to avoid hiding or showing an effect for windows involved in the switch. http://bugzilla.gnome.org/show_bug.cgi?id=582341 * Handling of post-effect cleanups for MutterWindow are simplified - instead of trying to do different things based on the individual needs of different effects, we just wait until all effects complete and sync the window state to what it should be. * On unmap, once we destroy the pixmap, we tell ClutterX11Pixmap that we've done so, so it can clean up and unbind. (The unbinding doesn't seem to be working properly because of ClutterGLXPixmap or video driver issues.) http://bugzilla.gnome.org/show_bug.cgi?id=587251
2009-06-28 21:10:40 +00:00
if (priv->visible)
clutter_actor_show (CLUTTER_ACTOR (self));
else
clutter_actor_hide (CLUTTER_ACTOR (self));
}
}
static void
check_needs_reshape (MutterWindow *self)
{
MutterWindowPrivate *priv = self->priv;
if (!priv->needs_reshape)
return;
mutter_shaped_texture_clear_rectangles (MUTTER_SHAPED_TEXTURE (priv->actor));
mutter_window_clear_shape_region (self);
#ifdef HAVE_SHAPE
if (priv->shaped)
{
Display *xdisplay = meta_display_get_xdisplay (meta_window_get_display (priv->window));
XRectangle *rects;
int n_rects, ordering;
rects = XShapeGetRectangles (xdisplay,
priv->xwindow,
ShapeBounding,
&n_rects,
&ordering);
if (rects)
{
mutter_shaped_texture_add_rectangles (MUTTER_SHAPED_TEXTURE (priv->actor),
n_rects, rects);
mutter_window_update_shape_region (self, n_rects, rects);
XFree (rects);
}
}
#endif
priv->needs_reshape = FALSE;
}
void
mutter_window_update_shape (MutterWindow *self,
gboolean shaped)
{
MutterWindowPrivate *priv = self->priv;
priv->shaped = shaped;
priv->needs_reshape = TRUE;
clutter_actor_queue_redraw (priv->actor);
}
void
mutter_window_pre_paint (MutterWindow *self)
{
MutterWindowPrivate *priv = self->priv;
MetaScreen *screen = priv->screen;
MetaDisplay *display = meta_screen_get_display (screen);
Display *xdisplay = meta_display_get_xdisplay (display);
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
if (is_frozen (self))
{
/* The window is frozen due to a pending animation: we'll wait until
* the animation finishes to reshape and repair the window */
return;
}
if (priv->received_damage)
{
meta_error_trap_push (display);
XDamageSubtract (xdisplay, priv->damage, None, None);
meta_error_trap_pop (display, FALSE);
priv->received_damage = FALSE;
}
check_needs_reshape (self);
mutter-window: stream raw updates to ClutterX11TexturePixmap This changes the way we handle Damage events so instead of getting an event when the damage region of a pixmap becomes non-empty we now get sent all damage rectangles and stream those all though to ClutterX11TexturePixmap using clutter_x11_texture_pixmap_update_area() For Clutter 1.2, ClutterGLXTexturePixmap was updated so that calls to clutter_x11_texture_pixmap_update_area are now cheap (glXBindTexImageEXT calls are now deferred until just before painting) and since ClutterGLXTexturePixmap is now capable of queueing clipped redraws that can result in only updating a sub-region of the stage during a repaint cycle (and using glXCopySubBufferMESA to present the sub-region redraw to the front buffer) this should improve performance and reduced power consumption for a range of use cases. (For example viewing a website that has animated adverts doesn't force the whole screen to be redrawn for each frame of the advert) Besides being able to take advantage of glXCopySubBuffer to only update a small region of the stage the fact that this patch makes Mutter now request RawRectangles from the X server means we no longer do a synchronous X request for a complete Damage Region for every window damaged each frame. This should also improve performance. CLUTTER_PAINT=redraws can be used to visualize what parts of the stage are redrawn and with this patch applied I can open a terminal and as I type I see that only the damaged areas of the terminal are being redrawn.
2010-03-02 18:02:28 +00:00
check_needs_pixmap (self);
}
void
mutter_window_update_opacity (MutterWindow *self)
{
MutterWindowPrivate *priv = self->priv;
MetaDisplay *display = meta_screen_get_display (priv->screen);
MetaCompositor *compositor = meta_display_get_compositor (display);
Window xwin = meta_window_get_xwindow (priv->window);
gulong value;
guint8 opacity;
if (meta_prop_get_cardinal (display, xwin,
compositor->atom_net_wm_window_opacity,
&value))
{
opacity = (guint8)((gfloat)value * 255.0 / ((gfloat)0xffffffff));
}
else
opacity = 255;
self->priv->opacity = opacity;
clutter_actor_set_opacity (CLUTTER_ACTOR (self), opacity);
}