The material cache will now only set the special combine mode if the
texture only has an alpha component. The atlased textures will have
all four components so it will leave the combine functions at the
default. This increases the chances of batching between glyphs and
images.
When using the global atlas, the glyph from cairo is now rendered into
an ARGB surface rather than an alpha-only surface.
Instead of creating just two materials (one for texturing and one for
solid primitives) the pango renderer now maintains a cache of
pipelines. The display list can request a pipeline for a texture from
the cache. The same pipeline cache is used by all display lists so
that the pipelines can be shared. This avoids changing the texture on
the material during a paint run.
It now avoids trying to reserve space for zero-sized glyphs. That
happens for example when the layout contains a space. This was causing
the regular glyph cache to be used because the global atlas does not
support zero-sized images. That would then break up the
batching. Instead it now still reserves an entry in the cache but
leaves the texture as COGL_INVALID_HANDLE.
When rendering a glyph from a texture, instead of adding the glyph's
texture handle to the display list it now retrieves the base texture
using _cogl_texture_foreach_subtexture_in_region and adds that
instead. That way the display can recognise that glyphs in the global
atlas are sharing the same texture and combine them into one VBO.
Whenever the glyph cache puts a glyph in the global atlas it will now
register for notifications of reorganisation of the global
atlases. When this happens it will forward this on as a notification
of reorganisation of the glyph cache.
This adds cogl_atlas_texture_* functions to register a callback that
will get invoked whenever any of the CoglAtlas's the textures use get
reorganized. The callback is global and is not tied to any particular
atlas texture.
If mipmapping is disabled, it will now try to create a standalone
atlas texture for a glyph rather than putting it in the atlas.
If the atlas texture can't be created then it will fallback to the
glyph cache.
This adds a new function called _cogl_atlas_texture_new_with_size. The
old new_from_bitmap function now just calls this and updates the
texture with the data.
If the texture can't be hardware repeated (ie, if it is sliced or it
has waste) then Cogl will reject the layer when rendering with a
VBO. In this case we should always fall back to rendering with
cogl_rectangle.
This commit is only needed temporarily because Cogl will end up
putting atlas textures in the display list. A later commit in the
series will make it so that the display list always has primitive
textures in it so this commit can be reverted.
This reverts the changes in 54d8aadf which combined the two glyph
caches into one. We want to start using separate caches again so that
we can non-mipmapped textures into the global atlas.
This extends cogl_onscreen_x11_set_foreign_xid to take a callback to a
function that details the event mask the Cogl requires the application
to select on foreign windows. This is required because Cogl, for
example, needs to track size changes of a window and may also in the
future want other notifications such as map/unmap.
Most applications wont need to use the foreign xwindow apis, but those
that do are required to pass a valid callback and update the event mask
of their window according to Cogl's requirements.
This adds Cogl API to show and hide onscreen framebuffers. We don't want
to go too far down the road of abstracting window system APIs with Cogl
since that would be out of its scope but the previous idea that we would
automatically map framebuffers on allocation except for those made from
foreign windows wasn't good enough. The problem is that we don't want to
make Clutter always create stages from foreign windows but with the
automatic map semantics then Clutter doesn't get an opportunity to
select for all the events it requires before mapping. This meant that we
wouldn't be delivered a mouse enter event for windows mapped underneath
the cursor which would break Clutters handling of button press events.
When building on windows for example we need to ensure we pass
-no-undefined to the linker. Although we were substituting a
COGL_EXTRA_LDFLAGS variable from our configure.ac we forgot to
reference that when linking cogl-pango.
For compatibility with the way we build Cogl as part of Clutter we now
substitute an empty MAINTAINER_CFLAGS variable. When building Cogl
standalone all our extra CFLAGS go through COGL_EXTRA_CFLAGS so the
separate MAINTAINER_CFLAGS aren't used, but automake will get confused
if a substitution isn't made.
This fixes the gdk-pixbuf check to not mistakenly check for the "xi"
package instead of gdk-pixbuf and remove a spurious listing "gl" in
COGL_PKG_REQUIRES which should only be there when we are using using
opengl not if we are using gles.
When building on windows for example we need to ensure we pass
-no-undefined to the linker. Although we were substituting a
COGL_EXTRA_LDFLAGS variable from our configure.ac we forgot to
reference that when linking cogl.
Until Cogl gains native win32/OSX support this remove the osx and win32
winsys files and instead we'll just rely on the stub-winsys.c to handle
these platforms. Since the only thing the platform specific files were
providing anyway was a get_proc_address function; it was trivial to
simply update the clutter backend code to handle this directly for now.
This is a workaround for a bug on OSX for some radeon hardware that
we can't verify and the referenced bug link is no longer valid.
If this is really still a problem then a new bug should be opened and we
can look at putting the fix in some more appropriate place than
cogl-gl.c
We want to be able to split Cogl out as a standalone project but there
are still some window systems that aren't natively supported by Cogl.
This allows Clutter to support those window systems directly but still
work with a standalone Cogl library.
This also ensures we set the SUPPORT_STUB conditional in clutter's
configure.ac when building for win32/osx and wayland.
For now we are going for the semantics that when a CoglOnscreen is first
allocated then it will automatically be mapped. This is for convenience
and if you don't want that behaviour then it is possible to instead
create an Onscreen from a foreign X window and in that case it wont be
mapped automatically.
This approach means that Cogl doesn't need onscreen_map/unmap functions
but it's possible we'll decide later that we can't avoid adding such
functions and we'll have to change these semantics.
Instead of using AC_DEFINE for the various COGL_HAS_PLATFORM defines
this now adds them to the COGL_DEFINES_SYMBOLS variable which gets
substituted into the public cogl-defines.h header.
This adds a simple standalone Cogl application that can be used to
smoke test a standalone build of Cogl without Clutter.
This also adds an x11-foreign app that shows how a toolkit can ask Cogl
to draw to an X Window that it owns instead of Cogl being responsible
for automatically creating and mapping an X Window for CoglOnscreen.
This allows more detailed control over the driver and winsys features
that Cogl should have. Cogl is designed so it can support multiple
window systems simultaneously so we have enable/disable options for
the drivers (gl vs gles1 vs gles2) and options for the individual window
systems; currently glx and egl. Egl is broken down into an option
for each platform.
The GDL API is used for example on intel ce4100 (aka Sodaville) based
systems as a way to allocate memory that can be composited using the
platforms overlay hardware. This updates the Cogl EGL winsys and the
support in Clutter so we can continue to support these platforms.
So that we can dynamically select what winsys backend to use at runtime
we need to have some indirection to how code accesses the winsys instead
of simply calling _cogl_winsys* functions that would collide if we
wanted to compile more than one backend into Cogl.
This moves the GLX specific code from cogl-texture-pixmap-x11.c into
cogl-winsys-glx.c. If we want the winsys components to by dynamically
loadable then we can't have GLX code scattered outside of
cogl-winsys-glx.c. This also sets us up for supporting the
EGL_texture_from_pixmap extension which is almost identical to the
GLX_texture_from_pixmap extension.
As was recently done for the GLX window system code, this commit moves
the EGL window system code down from the Clutter backend code into a
Cogl winsys.
Note: currently the cogl/configure.ac is hard coded to only build the GLX
winsys so currently this is only available when building Cogl as part
of Clutter.
The "DRM_SURFACELESS" EGL platform was invented when we were adding the
wayland backend to Clutter but in the end we added a dedicated backend
instead of extending the EGL backend so actually the platform name isn't
used.
Commit b061f737 moved _cogl_winsys_has_feature to the common winsys
code so there's no need to define it in the stub winsys any more. This
was breaking builds for backends using the stub winsys.
The comparison for finding onscreen framebuffers in
find_onscreen_for_xid had a small thinko so that it would ignore
framebuffers when the negation of the type is onscreen. This ends up
doing the right thing anyway because the onscreen type has the value 0
and the offscreen type has the value 1 but presumably it would fail if
we ever added any other framebuffer types.
The code for _cogl_winsys_has_feature will be identical in all of the
winsys backends for the time being, so it seems to make sense to have
it in the common cogl-winsys.c file.
Previously the mask of available winsys features was stored in a
CoglBitmask. That isn't the ideal type to use for this because it is
intended for a growable array of bits so it can allocate extra memory
if there are more than 31 flags set. For the winsys feature flags the
highest used bit is known at compile time so it makes sense to
allocate a fixed array instead. This is conceptually similar to the
CoglDebugFlags which are stored in an array of integers with macros to
test a bit in the array. This moves the macros used for CoglDebugFlags
to cogl-flags.h and makes them more generic so they can be shared with
CoglContext.
Instead of having cogl_renderer_xlib_add_filter and friends there is
now cogl_renderer_add_native_filter which can be used regardless of
the backend. The callback function for the filter now just takes a
void pointer instead of an XEvent pointer which should be interpreted
differently depending on the backend. For example, on Xlib it would
still be an XEvent but on Windows it could be a MSG. This simplifies
the code somewhat because the _cogl_xlib_add_filter no longer needs to
have its own filter list when a stub renderer is used because there is
always a renderer available.
cogl_renderer_xlib_handle_event has also been renamed to
cogl_renderer_handle_native_event. This just forwards the event on to
all of the listeners. The backend renderer is expected to register its
own event filter if it wants to process the events in some way.
Older drivers for PowerVR SGX hardware have the vendor-specific
GL_IMG_TEXTURE_NPOT extension instead of the
functionally-equivalent GL_OES_TEXTURE_NPOT extension.
We need to guard the usage of symbols related to the
GLX_INTEL_swap_event extension, to avoid breaking on platforms and/or
versions of Mesa that do not expose that extension.
It's generally useful to be able to query the width and height of a
framebuffer and we expect to need this in Clutter when we move the
eglnative backend code into Cogl since Clutter will need to read back
the fixed size of the framebuffer when realizing the stage.
This backend hasn't been used for years now and so because it is
untested code and almost certainly doesn't work any more it would be a
burdon to continue trying to maintain it. Considering that we are now
looking at moving OpenGL window system integration code down from
Clutter backends into Cogl that will be easier if we don't have to
consider this backend.
This adds an autogen.sh, configure.ac and build/autotool files etc under
clutter/cogl and makes some corresponding Makefile.am changes that make
it possible to build and install Cogl as a standalone library.
Some notable things about this are:
A standalone installation of Cogl installs 3 pkg-config files;
cogl-1.0.pc, cogl-gl-1.0.pc and cogl-2.0.pc. The second is only for
compatibility with what clutter installed though I'm not sure that
anything uses it so maybe we could remove it. cogl-1.0.pc is what
Clutter would use if it were updated to build against a standalone cogl
library. cogl-2.0.pc is what you would use if you were writing a
standalone Cogl application.
A standalone installation results in two libraries currently, libcogl.so
and libcogl-pango.so. Notably we don't include a major number in the
sonames because libcogl supports two major API versions; 1.x as used by
Clutter and the experimental 2.x API for standalone applications.
Parallel installation of later versions e.g. 3.x and beyond will be
supportable either with new sonames or if we can maintain ABI then we'll
continue to share libcogl.so.
The headers are similarly not installed into a directory with a major
version number since the same headers are shared to export the 1.x and
2.x APIs (The only difference is that cogl-2.0.pc ensures that
-DCOGL_ENABLE_EXPERIMENTAL_2_0_API is used). Parallel installation of
later versions is not precluded though since we can either continue
sharing or later add a major version suffix.
This migrates all the GLX window system code down from the Clutter
backend code into a Cogl winsys. Moving OpenGL window system binding
code down from Clutter into Cogl is the biggest blocker to having Cogl
become a standalone 3D graphics library, so this is an important step in
that direction.
As part of the process of splitting Cogl out as a standalone graphics
API we need to introduce some API concepts that will allow us to
initialize a new CoglContext when Clutter isn't there to handle that for
us...
The new objects roughly in the order that they are (optionally) involved
in constructing a context are: CoglRenderer, CoglOnscreenTemplate,
CoglSwapChain and CoglDisplay.
Conceptually a CoglRenderer represents a means for rendering. Cogl
supports rendering via OpenGL or OpenGL ES 1/2.0 and those APIs are
accessed through a number of different windowing APIs such as GLX, EGL,
SDL or WGL and more. Potentially in the future Cogl could render using
D3D or even by using libdrm and directly banging the hardware. All these
choices are wrapped up in the configuration of a CoglRenderer.
Conceptually a CoglDisplay represents a display pipeline for a renderer.
Although Cogl doesn't aim to provide a detailed abstraction of display
hardware, on some platforms we can give control over multiple display
planes (On TV platforms for instance video content may be on one plane
and 3D would be on another so a CoglDisplay lets you select the plane
up-front.)
Another aspect of CoglDisplay is that it lets us negotiate a display
pipeline that best supports the type of CoglOnscreen framebuffers we are
planning to create. For instance if you want transparent CoglOnscreen
framebuffers then we have to be sure the display pipeline wont discard
the alpha component of your framebuffers. Or if you want to use
double/tripple buffering that requires support from the display
pipeline.
CoglOnscreenTemplate and CoglSwapChain are how we describe our default
CoglOnscreen framebuffer configuration which can affect the
configuration of the display pipeline.
The default/simple way we expect most CoglContexts to be constructed
will be via something like:
if (!cogl_context_new (NULL, &error))
g_error ("Failed to construct a CoglContext: %s", error->message);
Where that NULL is for an optional "display" parameter and NULL says to
Cogl "please just try to do something sensible".
If you want some more control though you can manually construct a
CoglDisplay something like:
display = cogl_display_new (NULL, NULL);
cogl_gdl_display_set_plane (display, plane);
if (!cogl_display_setup (display, &error))
g_error ("Failed to setup a CoglDisplay: %s", error->message);
And in a similar fashion to cogl_context_new() you can optionally pass
a NULL "renderer" and/or a NULL "onscreen template" so Cogl will try to
just do something sensible.
If you need to change the CoglOnscreen defaults you can provide a
template something like:
chain = cogl_swap_chain_new ();
cogl_swap_chain_set_has_alpha (chain, TRUE);
cogl_swap_chain_set_length (chain, 3);
onscreen_template = cogl_onscreen_template_new (chain);
cogl_onscreen_template_set_pixel_format (onscreen_template,
COGL_PIXEL_FORMAT_RGB565);
display = cogl_display_new (NULL, onscreen_template);
if (!cogl_display_setup (display, &error))
g_error ("Failed to setup a CoglDisplay: %s", error->message);
This tries to make the naming style of files under cogl/winsys/
consistent with other cogl source files. In particular private header
files didn't have a '-private' infix.
This gives us a way to clearly track the internal Cogl API that Clutter
depends on. The aim is to split Cogl out from Clutter into a standalone
3D graphics API and eventually we want to get rid of any private
interfaces for Clutter so its useful to have a handle on that task.
Actually it's not as bad as I was expecting though.
This renames the two internal functions _cogl_get_draw/read_buffer
as cogl_get_draw_framebuffer and _cogl_get_read_framebuffer. The
former is now also exposed as experimental API.
The long term goal with the Cogl API is that we will get rid of the
default global context. As a step towards this, this patch tracks a
reference back to the context in each CoglFramebuffer so in a lot of
cases we can avoid using the _COGL_GET_CONTEXT macro.
There is no corresponding implementation of _cogl_features_init any more
so it was simply an oversight that the prototype wasn't removed when the
implementation was removed.
Recently _cogl_swap_buffers_notify was added (in 142b229c5c) so that
Cogl would be notified when Clutter performs a swap buffers request for
a given onscreen framebuffer. It was expected this would be required for
the recent cogl_read_pixel optimization that was implemented (ref
1bdb0e6e98) but in the end it wasn't used.
Since it wasn't used in the end this patch removes the API.
This moves the functionality of _cogl_create_context_driver from
driver/{gl,gles}/cogl-context-driver-{gl,gles}.c into
driver/{gl,gles}/cogl-{gl,gles}.c as a static function called
initialize_context_driver.
cogl-context-driver-{gl,gles}.[ch] have now been removed.
This adds a new experimental function (you need to define
COGL_ENABLE_EXPERIMENTAL_API to access it) which takes us towards being
able to have a standalone Cogl API. This is really a minor aesthetic
change for now since all the GL context creation code still lives in
Clutter but it's a step forward none the less.
Since our current designs introduce a CoglDisplay object as something
that would be passed to the context constructor this provides a stub
cogl-display.h with CoglDisplay typedef.
_cogl_context_get_default() which Clutter uses to access the Cogl
context has been modified to use cogl_context_new() to initialize
the default context.
There is one rather nasty hack used in this patch which is that the
implementation of cogl_context_new() has to forcibly make the allocated
context become the default context because currently all the code in
Cogl assumes it can access the context using _COGL_GET_CONTEXT including
code used to initialize the context.
In 047227fb cogl_atlas_new was changed so that it can take a flags
parameter to specify whether to clear the new atlases and whether to
copy images to the new atlas after reorganisation. This was done so
that the atlas code could be shared with the glyph cache. At some
point during the development of this patch the flag was just a single
boolean instead and this is accidentally how it is used from the glyph
cache. The glyph cache therefore passes 'TRUE' as the set of flags
which means it will only get the 'clear' flag and not the
'disable-migration' flag. When the glyph cache gets full it will
therefore try to copy the texture to the new atlas as well as
redrawing them with cairo. This causes problems because the glyph
cache needs to work in situations where there is no FBO support.
In _cogl_pipeline_prune_empty_layer_difference if the layer's parent
has no owner then it just takes ownership of it. However this could
theoretically end up taking ownership of the root layer because
according to the comment above in the same function that should never
have an owner. This patch just adds an extra check to ensure that the
unowned layer has a parent.
http://bugzilla.clutter-project.org/show_bug.cgi?id=2588
In _cogl_pipeline_prune_empty_layer_difference if we are reverting to
the immediate parent of an empty/redundant layer then it is not enough
to simply add a reference to the pipeline's ->layer_differences list
without also updating parent_layer->owner to point back to its new
owner.
This oversight was leading us to break the invariable that all layers
referenced in layer_differences have an owner and was also causing us to
break another invariable whereby after calling
_cogl_pipeline_layer_pre_change_notify the returned layer must always be
owned by the given 'required_owner'.
http://bugzilla.clutter-project.org/show_bug.cgi?id=2588
glib already has a data type to manage a list of callbacks called a
GHookList so we might as well use it instead of maintaining Cogl's own
type. The glib version may be slightly more efficient because it
avoids using a GList and instead encodes the prev and next pointers
directly in the GHook structure. It also has more features than
CoglCallbackList.
Previously we were applying the culling optimization to any actor
painted without considering that we may be painting to an offscreen
framebuffer where the stage clip isn't applicable.
For now we simply expose a getter for the current draw framebuffer
and we can assume that a return value of NULL corresponds to the
stage.
Note: This will need to be updated as stages start to be backed by real
CoglFramebuffer objects and so we won't get NULL in those cases.
Drawing and clipping to paths is generally quite expensive because the
geometry has to be tessellated into triangles in a single VBO which
breaks up the journal batching. If we can detect when the path
contains just a single rectangle then we can instead divert to calling
cogl_rectangle which will take advantage of the journal, or by pushing
a rectangle clip which usually ends up just using the scissor.
This patch adds a boolean to each path to mark when it is a
rectangle. It gets cleared whenever a node is added or gets set to
TRUE whenever cogl2_path_rectangle is called. This doesn't try to
catch cases where a rectangle is composed by cogl_path_line_to and
cogl_path_move_to commands.
In 9ff04e8a99 the builtin uniforms were moved to the common shader
boilerplate. However the common boilerplate is positioned before the
default precision specifier on GLES2 so it would fail to compile
because the uniforms end up with no precision in the fragment
shader. This patch just moves the precision specifier to above the
common boilerplate.
Instead of unconditionally combining the modelview and projection
matrices and then iterating each of the vertices to call
cogl_matrix_transform_point for each one in turn we now only combine the
matrices if there are more than 4 vertices (with less than 4 vertices
its less work to transform them separately) and we use the new
cogl_vertex_{transform,project}_points APIs which can hopefully
vectorize the transformations.
Finally the perspective divide and viewport scale is done in a separate
loop at the end and we don't do the spurious perspective divide and
viewport scale for the z component.
This adds two new experimental functions to cogl-matrix.c:
cogl_matrix_view_2d_in_perspective and cogl_matrix_view_2d_in_frustum
which can be used to setup a view transform that maps a 2D coordinate
system (0,0) top left and (width,height) bottom right to the current
viewport.
Toolkits such as Clutter that want to mix 2D and 3D drawing can use
these APIs to position a 2D coordinate system at an arbitrary depth
inside a 3D perspective projected viewing frustum.
OpenGL < 4.0 only supports integer based viewports and internally we
have a mixture of code using floats and integers for viewports. This
patch switches all viewports throughout clutter and cogl to be
represented using floats considering that in the future we may want to
take advantage of floating point viewports with modern hardware/drivers.
This makes a change to the original point_in_poly algorithm from:
http://www.ecse.rpi.edu/Homepages/wrf/Research/Short_Notes/pnpoly.html
The aim was to tune the test so that tests against screen aligned
rectangles are more resilient to some in-precision in how we transformed
that rectangle into screen coordinates. In particular gnome-shell was
finding that for some stage sizes then row 0 of the stage would become a
dead zone when going through the software picking fast-path and this was
because the y position of screen aligned rectangles could end up as
something like 0.00024 and the way the algorithm works it doesn't have
any epsilon/fuz factor to consider that in-precision.
We've avoided introducing an epsilon factor to the comparisons since we
feel there's a risk of changing some semantics in ways that might not be
desirable. One of those is that if you transform two polygons which
share an edge and test a point close to that edge then this algorithm
will currently give a positive result for only one polygon.
Another concern is the way this algorithm resolves the corner case where
the horizontal ray being cast to count edge crossings may cross directly
through a vertex. The solution is based on the "idea of Simulation of
Simplicity" and "pretends to shift the ray infinitesimally down so that
it either clearly intersects, or clearly doesn't touch". I'm not
familiar with the idea myself so I expect a misplaced epsilon is likely
to break that aspect of the algorithm.
The simple solution this patch applies is to pixel align the polygon
vertices which should eradicate most noise due to in-precision.
https://bugzilla.gnome.org/show_bug.cgi?id=641197
When using a pipeline and the journal to blit images between
framebuffers, it should disable blending. Otherwise it will end up
blending the source texture with uninitialised garbage in the
destination texture.
If an atlas texture's last reference is held by the journal or by the
last flushed pipeline then if an atlas migration is started it can
cause a crash. This is because the atlas migration will cause a
journal flush and can sometimes change the current pipeline which
means that the texture would be destroyed during migration.
This patch adds an extra 'post_reorganize' callback to the existing
'reorganize' callback (which is now renamed to 'pre_reorganize'). The
pre_reorganize callback is now called before the atlas grabs a list of
the current textures instead of after so that it doesn't matter if the
journal flush destroys some of those textures. The pre_reorganize
callback for CoglAtlasTexture grabs a reference to all of the textures
so that they can not be destroyed when the migration changes the
pipeline. In the post_reorganize callback the reference is removed
again.
http://bugzilla.clutter-project.org/show_bug.cgi?id=2538
When Cogl debugging is disabled then the 'waste' variable is not used
so it throws a compiler warning. This patch removes the variable and
the value is calculated directly as the parameter to COGL_NOTE.
Some code was doing pointer arithmetic on the return value from
cogl_buffer_map which is void* pointer. This is a GCC extension so we
should try to avoid it. This patch adds casts to guint8* where
appropriate.
Based on a patch by Fan, Chun-wei.
http://bugzilla.clutter-project.org/show_bug.cgi?id=2561
Instead of directly banging GL to migrate textures the atlas now uses
the CoglFramebuffer API. It will use one of four approaches; it can
set up two FBOs and use _cogl_blit_framebuffer to copy between them;
it can use a single target fbo and then render the source texture to
the FBO using a Cogl draw call; it can use a single FBO and call
glCopyTexSubImage2D; or it can fallback to reading all of the texture
data back to system memory and uploading it again with a sub texture
update.
Previously GL calls were used directly because Cogl wasn't able to
create a framebuffer without a stencil and depth buffer. However there
is now an internal version of cogl_offscreen_new_to_texture which
takes a set of flags to disable the two buffers.
The code for blitting has now been moved into a separate file called
cogl-blit.c because it has become quite long and it may be useful
outside of the atlas at some point.
The 4 different methods have a fixed order of preference which is:
* Texture render between two FBOs
* glBlitFramebuffer
* glCopyTexSubImage2D
* glGetTexImage + glTexSubImage2D
Once a method is succesfully used it is tried first for all subsequent
blits. The default default can be overridden by setting the
environment variable COGL_ATLAS_DEFAULT_BLIT_MODE to one of the
following values:
* texture-render
* framebuffer
* copy-tex-sub-image
* get-tex-data
This adds a declaration for _cogl_is_texture_2d to the private header
so that it can be used in cogl-blit.c to determine if the target
texture is a simple 2D texture.
This adds a function called _cogl_texture_2d_copy_from_framebuffer
which is a simple wrapper around glCopyTexSubImage2D. It is currently
specific to the texture 2D backend.
This adds the _cogl_blit_framebuffer internal function which is a
wrapper around glBlitFramebuffer. The API is changed from the GL
version of the function to reflect the limitations provided by the
GL_ANGLE_framebuffer_blit extension (eg, no scaling or mirroring).
This extension is the GLES equivalent of the GL_EXT_framebuffer_blit
extension except that it has some extra restrictions. We need to check
for some extension that provides glBlitFramebuffer so that we can
unconditionally use ctx->drv.pf_glBlitFramebuffer in both GL and GLES
code. Even with the restrictions, the extension provides enough
features for what Cogl needs.
Previously when _cogl_atlas_texture_migrate_out_of_atlas is called it
would unreference the atlas texture's sub-texture before calling
_cogl_atlas_copy_rectangle. This would leave the atlas texture in an
inconsistent state during the copy. This doesn't normally matter but
if the copy ends up doing a render then the atlas texture may end up
being referenced. In particular it would cause problems if the texture
is left in a texture unit because then Cogl may try to call
get_gl_texture even though the texture isn't actually being used for
rendering. To fix this the sub texture is now unrefed after the copy
call instead.
The current framebuffer is now internally separated so that there can
be a different draw and read buffer. This is required to use the
GL_EXT_framebuffer_blit extension. The current draw and read buffers
are stored as a pair in a single stack so that pushing the draw and
read buffer is done simultaneously with the new
_cogl_push_framebuffers internal function. Calling
cogl_pop_framebuffer will restore both the draw and read buffer to the
previous state. The public cogl_push_framebuffer function is layered
on top of the new function so that it just pushes the same buffer for
both drawing and reading.
When flushing the framebuffer state, the cogl_framebuffer_flush_state
function now tackes a pointer to both the draw and the read
buffer. Anywhere that was just flushing the state for the current
framebuffer with _cogl_get_framebuffer now needs to call both
_cogl_get_draw_buffer and _cogl_get_read_buffer.
When pushing a framebuffer it would previously push
COGL_INVALID_HANDLE to the top of the framebuffer stack so that when
it later calls cogl_set_framebuffer it will recognise that the
framebuffer is different and replace the top with the new
pointer. This isn't ideal because it breaks the code to flush the
journal because _cogl_framebuffer_flush_journal is called with the
value of the old pointer which is NULL. That function was checking for
a NULL pointer so it wouldn't actually flush. It also would mean that
if you pushed the same framebuffer twice we would end up dirtying
state unnecessarily. To fix this cogl_push_framebuffer now pushes a
reference to the current framebuffer instead.
After a dependent framebuffer is added to a framebuffer it was never
getting removed. Once the journal for a framebuffer is flushed we no
longer depend on any framebuffers so the list should be cleared. This
was causing leaks of offscreens and textures.
This adds a note to clarify that cogl_matrix_multiply allows you to
multiply the @a matrix in-place, so @a can equal @result but @b can't
equal @result.
When uploading the layer matrix to GL it wasn't first calling
glActiveTextureMatrix to set the right texture unit for the
layer. This would end up setting the texture matrix on whatever layer
happened to be previously active. This happened to work for
test-cogl-multitexture presumably because it was coincidentally
setting the layer matrix on the last used layer.
The pipeline private data is accessed both from the private data set
on a CoglPipeline and the destroy notify function of a weak material
that the vertex buffer creates when it needs to override the wrap
mode. However when a CoglPipeline is destroyed, the CoglObject code
first removes all of the private data set on the object and then the
CoglPipeline code gets invoked to destroy all of the weak children. At
this point the vertex buffer's weak override destroy notify function
will get invoked and try to use the private data which has already
been freed causing a crash.
This patch instead adds a reference count to the pipeline private data
stuct so that we can avoid freeing it until both the private data on
the pipeline has been destroyed and all of the weak materials are
destroyed.
http://bugzilla.clutter-project.org/show_bug.cgi?id=2544
In cogl_pipeline_set_layer_combine_constant it was comparing whether
the new color is the same as the old color using a memcmp on the
constant_color parameter. However the combine constant is stored in
the layer data as an array of four floats but the passed in color is a
CoglColor (which is currently an array of four guint8s). This was
causing valgrind errors and presumably also the check for setting the
same color twice would always fail.
This patch makes it do the conversion to a float array upfront before
the comparison.
cogl_matrix_project_points and cogl_matrix_transform_points had an
optimization for the common case where the stride parameters exactly
match the size of the corresponding structures. The code for both when
generated by gcc with -O2 on x86-64 use two registers to hold the
addresses of the input and output arrays. In the strided version these
pointers are incremented by adding the value of a register and in the
packed version they are incremented by adding an immediate value. I
think the difference in cost here would be negligible and it may even
be faster to add a register.
Also GCC appears to retain the loop counter in a register for the
strided version but in the packed version it can optimize it out and
directly use the input pointer as the counter. I think it would be
possible to reorder the code a bit to explicitly use the input pointer
as the counter if this were a problem.
Getting rid of the packed versions tidies up the code a bit and it
could potentially be faster if the code differences are small and we
get to avoid an extra conditional in cogl_matrix_transform_points.
When copying COMBINE state in
_cogl_pipeline_layer_init_multi_property_sparse_state we would read some
state from the destination layer (invalid data potentially), then
redundantly set the value back on the destination. This was picked up by
valgrind, and the code is now more careful about how it references the
src layer vs the destination layer.
There is currently a problem with per-framebuffer journals in that it's
possible to create a framebuffer from a texture which then gets rendered
too but the framebuffer (and corresponding journal) can be freed before
the texture gets used to draw with.
Conceptually we want to make sure when freeing a framebuffer that - if
it is associated with a texture - we flush the journal as the last thing
before really freeing the framebuffer's meta data. Technically though
this is awkward to implement since the obvious mechanism for us to be
notified about the framebuffer's destruction (by setting some user data
internally with a callback) notifies when the framebuffer has a
ref-count of 0. This means we'd have to be careful what we do with the
framebuffer to consider e.g. recursive destruction; anything that would
set more user data on the framebuffer while it is being destroyed and
ensuring nothing else gets notified of the framebuffer's destruction
before the journal has been flushed.
For simplicity, for now, this patch provides another solution which is
to flush framebuffer journals whenever we switch away from a given
framebuffer via cogl_set_framebuffer or cogl_push/pop_framebuffer. The
disadvantage of this approach is that we can't batch all the geometry of
a scene that involves intermediate renders to offscreen framebufers.
Clutter is doing this more and more with applications that use the
ClutterEffect APIs so this is a shame. Hopefully this will only be a
stop-gap solution while we consider how to reliably support journal
logging across framebuffer changes.
When flushing a clip stack that contains more than one rectangle which
needs to use the stencil buffer the code takes a different path so
that it can combine the new rectangle with the existing contents of
the stencil buffer. However it was not correctly flushing the
modelview and projection matrices so that rectangle would be in the
wrong place.
This adds a COGL_DEBUG=clipping option that reports how the clip is
being flushed. This is needed to determine whether the scissor,
stencil clip planes or software clipping is being used.
The CoglDebugFlags are now stored in an array of unsigned ints rather
than a single variable. The flags are accessed using macros instead of
directly peeking at the cogl_debug_flags variable. The index values
are stored in the enum rather than the actual mask values so that the
enum doesn't need to be more than 32 bits wide. The hope is that the
code to determine the index into the array can be optimized out by the
compiler so it should have exactly the same performance as the old
code.
