Previously the journal was always flushed at the end of
_cogl_rectangles_with_multitexture_coords, (i.e. the end of any
cogl_rectangle* calls) but now we have broadened the potential for batching
geometry. In ideal circumstances we will only flush once per scene.
In summary the journal works like this:
When you use any of the cogl_rectangle* APIs then nothing is emitted to the
GPU at this point, we just log one or more quads into the journal. A
journal entry consists of the quad coordinates, an associated material
reference, and a modelview matrix. Ideally the journal only gets flushed
once at the end of a scene, but in fact there are things to consider that
may cause unwanted flushing, including:
- modifying materials mid-scene
This is because each quad in the journal has an associated material
reference (i.e. not copy), so if you try and modify a material that is
already referenced in the journal we force a flush first)
NOTE: For now this means you should avoid using cogl_set_source_color()
since that currently uses a single shared material. Later we
should change it to use a pool of materials that is recycled
when the journal is flushed.
- modifying any state that isn't currently logged, such as depth, fog and
backface culling enables.
The first thing that happens when flushing, is to upload all the vertex data
associated with the journal into a single VBO.
We then go through a process of splitting up the journal into batches that
have compatible state so they can be emitted to the GPU together. This is
currently broken up into 3 levels so we can stagger the state changes:
1) we break the journal up according to changes in the number of material layers
associated with logged quads. The number of layers in a material determines
the stride of the associated vertices, so we have to update our vertex
array offsets at this level. (i.e. calling gl{Vertex,Color},Pointer etc)
2) we further split batches up according to material compatability. (e.g.
materials with different textures) We flush material state at this level.
3) Finally we split batches up according to modelview changes. At this level
we update the modelview matrix and actually emit the actual draw command.
This commit is largely about putting the initial design in-place; this will be
followed by other changes that take advantage of the extended batching.
The clutter_context_get_default() function is private, but shared
across Clutter. For this reason, it should be prefixed by '_' so
that the symbol is hidden from the shared object.
* clutter/clutter-actor.c:
(clutter_actor_set_min_width),
(clutter_actor_set_min_height),
(clutter_actor_set_natural_width),
(clutter_actor_set_natural_height): Ignore any override of the
minimum and natural size of the stage on backends that only
support static stages.
* clutter/clutter-stage.c (clutter_stage_allocate): Use the
preferred size of the ClutterStage implementation instead of
the display size.
* clutter/clutter-backend.[ch]: Remove get_display_size() and
clutter_backend_get_display_size().
* clutter/eglnative/clutter-backend-egl.c:
* clutter/fruity/clutter-backend-fruity.c:
* clutter/osx/clutter-backend-osx.c:
* clutter/sdl/clutter-backend-sdl.c:
* clutter/win32/clutter-backend-win32.c:
* clutter/x11/clutter-backend-x11.c: Remove get_display_size()
implementations.
signedness of old_xpos and old_ypos to get rid of compiler
warnings.
* clutter/win32/clutter-backend-win32.c
(clutter_backend_win32_get_features): Cast the result of
glGetString to a signed char pointer to avoid compiler warnings.
* clutter/win32/clutter-backend-win32.c: Reflect changes to the
GLX/X11 backend in revisions 2708-2709 and 2713-2715 which
simplify the backend a little.
* clutter/win32/clutter-stage-win32.c
(clutter_stage_win32_request_coords): Don't resize foreign
windows.
(clutter_stage_win32_unrealize): Don't destroy foreign windows.
(clutter_stage_win32_init): Added initialiser for is_foreign_win.
(clutter_win32_get_stage_from_window): Resort to looking in the
stage list if the window isn't the right window class so that it
can still find stages with foreign windows.
(clutter_win32_set_stage_foreign): New public function to set a
foreign window for a stage.
* clutter/win32/clutter-event-win32.c
(clutter_win32_disable_event_retrieval): New public function to
disable event retrieval.
(message_translate): Don't handle WM_SIZE or WM_MOVE for foreign
windows.
* clutter/win32/clutter-backend-win32.h (struct
_ClutterBackendWin32): Added a flag to disable event retrieval
* clutter/win32/clutter-backend-win32.c
(clutter_backend_win32_ensure_context): Update debug note to
include whether the stage is foreign or not.
subclassing code.
* clutter/win32/clutter-stage-win32.h:
* clutter/win32/clutter-stage-win32.c: Now inherits from
ClutterGroup and implements ClutterStageWindow instead of
inheriting directly from ClutterStage.
* clutter/win32/clutter-event-win32.c (message_translate): Now
takes an extra parameter to return whether DefWindowProc should be
called. This is needed to prevent the default WM_CLOSE handler
from destroying the window.
* clutter/win32/clutter-backend-win32.c
(clutter_backend_win32_dispose): Destroy all of the stages using
g_slist_foreach as per bug #871. Now also destroys the GL context.
(clutter_backend_win32_get_features): Added assertions to ensure
there is a valid GL context.
(clutter_backend_win32_ensure_context): Accepts NULL stage. Gets
implementation pointer from the stage.
* clutter/win32/clutter-backend-win32.c
(clutter_backend_win32_init): Added a call to
timeBeginPeriod. Without this the frame rates are terrible because
the glib timeouts are not accurate enough. However this requires
Glib >= 2.16.0 to take any effect because of a change in the way
g_poll is implemented. See revision 6597 of glib.
(clutter_backend_win32_finalize): Added a call to timeEndPeriod.
* configure.ac: Added -lwinmm to the library dependencies for the
Win32 backend.
* clutter/win32/clutter-win32.h:
* clutter/win32/clutter-stage-win32.h:
* clutter/win32/clutter-stage-win32.c:
* clutter/win32/clutter-event-win32.c:
* clutter/win32/clutter-backend-win32.h:
* clutter/win32/clutter-backend-win32.c:
Upgraded for multi-stage support.
* clutter/win32/clutter-stage-win32.c
(clutter_stage_win32_request_coords): Fixed so that it doesn't set
the position or size if it hasn't changed. This was causing
problems when the window was resized using the top left corner. In
that case the window receives resize and move messages separately
which caused the window to flash at a different size or position
while one message was handled before the other.
(clutter_stage_win32_realize): Added PFD_GENERIC_ACCELERATED to
the list of pixel format flags to force it to use hardware
acceleration.
2008-03-30 Neil Roberts <neil@o-hand.com>
* clutter-sections.txt: Added clutter_win32_get_stage_from_window
