Bug: 27407637
Texture layers will set a GL_NONE rendertarget
until they have managed to get the real rendertarget
from SurfaceTexture. If, however, copyLayer is called
prior to any calls to updateTexImage() (for example, if
no frames have yet been produced), this will
result in trying to use a Texture with a GL_NONE rendertarget
which is bad.
Change-Id: Ib7285d799cb94ebd44f07a89f2ef2075d79ef71f
Full GLES error checking layer via -include
trickery. Change DEBUG_OPENGL to a level system.
HIGH = every GL call is error checked
MODERATE = checkpointing at interesting spots
LOW = only asserts there are no errors at the end of a frame
or when the FBO changes
NONE = AIN'T GOT NO TIME FOR ERRORS GOTTA GO FAST!
Change-Id: Ibe81aae93d942059c4ddf1cbb11c828b7ce4c10b
Several places were setting GL_UNPACK_ALIGNMENT
unneccessarily, whereas other places were assuming an
unpack alignment of 1. Since we never actually
do explicit row-alignment, set GL_UNPACK_ALIGNMENT
to 1 at context creation time and never change it
Bug: 26584230
Also turns on aggressive glGetError checking to
better catch potential problem zones
Change-Id: I190c8f0f0494a7f046d5ed769405c75d363be59a
bug:24670525
Removes silly 'do nothing if rects do not intersect' behavior, and
changes the name to clarify the difference (contrasting against
SkRect::intersect())
bug:24670525
Change-Id: Id2ca1cfea1a9d720d4dc70b251f426d9916f8b53
Adds remaining missing overrides and nullptr usages, missed due to
an extreme failure in tool usage.
Change-Id: I56abd72975a3999ad13330003c348db40f59aebf
bug:17702227
Add details useful to developers (such as layer size/View name), and
switch away from logging implementation names/details, since they
are generally not relevant to developers.
Change-Id: Iee605d182f241450f7e75a6d0c283d51fa1312f5
This moves the interface closer to android::Canvas. The only use of
return values was in the OpenGLRenderer subclass; that is replaced
with an internal dirty flag: returned from finish(), checked by
CanvasContext.
This is part of a series of CLs to refactor the Graphics JNI bindings.
BUG:15672762
R=djsollen@google.com,ccraik@google.com
Change-Id: Ifd533eb8839a254b0d3a5d04fc5a2905afdfc89e
Bug: 17208461
* Switch Layer to be VirtualLightRefBase instead of
Caches' side-channel ref-counting
* Include active layers in gfxinfo dump
* Run gfxinfo dump on the correct thread
* Dump gfxinfo on Layer creation failure
Change-Id: I28d195699e2334518e215ab28c7a17355aee9678
bug:17208461
They are destroyed via finalizer-enqueued destroy method, so it's not
valid to check that they've been destroyed at gl context destruction
time.
Change-Id: I670f69825547facd5f31d44acb406418881fee00
Merge management of ortho projection matrix with the viewport size,
since they should always be changed together.
Change-Id: Iccb8f30828f4fb7848999ac54852e7ed2d6f2eb1
bug:11748993
TextureView should always be drawn with linear filtering if drawing a
buffer sized differently from the layer.
This fixes a bug where TextureViews that were sized differently from
their contents wouldn't be drawn with texture filtering, causing
visible scaling artifacts.
Change-Id: I8a5d27452fe7269ec53896992f37cff51e3ce15a
The eventual goal is for the StatefulBaseRenderer to serve as the
common base class between the DisplayListRenderer and OpenGLRenderer.
This will separate DisplayList recording, Snapshot stack management,
and the GL in OpenGLRenderer.
Additionally, avoid sp<> parameters, and use const parameters in
several places, with the intent of greatly reducing the surface area
where renderer subclasses can modify snapshot stack.
Next steps:
-move bulk of clipping logic into StatefulBaseRenderer
-disable direct snapshot access
Change-Id: Ibc3c6747134ec7daf8ea535866239fa73b874390
An array of indices local to a layer would only be necessary if
we changed the way we resolve T-junctions. Since we only ever
draw quads, let's just use the indices we use everywhere else.
This change also uses the global indices array to render list
of colored rectangles to save on the number of vertices generated
CPU-side.
Change-Id: Ia6d1970b0e9247805af5a114ca2a84b5d0b7c282
This optimization saves up to 0.3ms per frame on the Play Store's
front page, on a Nexus 4 device.
Change-Id: Iaa4ef33c6e3b37e175efd5b9eea9ef59b43f14f3
Bug #9316260
The GL specification indicates that deleting a bound texture has
the side effect of binding the default texture (name=0). This change
replaces all calls to glDeleteTextures() by Caches::deleteTexture()
to properly keep track of texture bindings.
Change-Id: Ifbc60ef433e0f9776a668dd5bd5f0adbc65a77a0
This change uses a new OpenGL ES 3.0 feature to upload less data when
the font cache needs to be update. This can result in significant
performance improvements on device with large textures or with locales
that use a lot of glyphs (CJK for instance.)
This change also fixes various unpack alignment issues. The unpack
alignment, as well as the unpack row length, is not texture specific
but a global state that affect all glTex/SubImage2D calls. Some of
them were missing the appropriate glPixelStorei() call. This could
result in corrupted textures.
Change-Id: Iefb429d4d0d0b4e0faeadf27daafee6d30a21d85
Bug #8437401
A misplaced ref count decrement was causing a crash when attempting to
resize a layer to dimensions larger than the max texture size supported
by the GPU.
This change fixes the crash and clarifies the warnings to make it more
obvious what's happening.
Change-Id: I632dc1b90aaa2605969e10523491a81c4922d3dc
Bug #8378964
This change defers drawing into layers until after the renderer for FBO0
is ready to draw. At that point, all the precaching is done which means
all glyphs can be uploaded at once in the font caches.
Change-Id: Ie1f7a7ff30f76f06fb3dbc72c7d05e66207d1ecb
This change extracts the scale parameters of the current transform
to pass then to the font renderer. Rotation and perspective are
applied to the generated mesh inside the vertex shader. This limits
the number of glyphs we have to create in the font cache and thus
reduces memory churn.
Change-Id: Ic5b3bae2b2b0e0250a8ee723b071a1709725c749
bug:4128442
Currently we fix T-junctions for hardware layers, and for
Canvas.saveLayer if the transform isn't rectAsRect. Ideally, hardware
layers should lazily re-generate T-junction free geometry if the
transform becomes non-rectAsRect
Depends on frameworks/native change:
https://googleplex-android-review.googlesource.com/#/c/277668/
which adds Region::createTJunctionFreeRegion()
Change-Id: Ice75b4a1a4459ff835059ea81d70f7d1563fd96d
Bug #7146141
This change is needed to add a render buffer cache to avoid
creating and destroying stencil buffers on every frame.
This change also allows the renderer to use a 1 bit or 4 bit
stencil buffer whenever possible.
Finally this change fixes a bug introduced by a previous CL
which causes the stencil buffer to not be updated in certain
conditions. The fix relies on a new optional parameter in
drawColorRects() that can be used to avoid performing a
quickReject on rectangles generated by the clip region.
Change-Id: I2f55a8e807009887b276a83cde9f53fd5c01199f
Bug #7146141
The GL_QCOM_tiled_rendering extension requires careful use of
start/endTiling when attaching a renderbuffer dynamically.
Change-Id: I20036683ed3909ffaf40cc3d57a25257e35b6fa2
Bug #7146141
This change moves the resizeLayer() from LayerCache (where it should
never have been anyway) to Layer. This makes a little more sense.
Change-Id: I8b2f9c19c558e738405a58b9e71ec5799fc6be88
Bug #7146141
When non-rectangular clipping occurs in a layer the render buffer
used as the stencil buffer is not cached. If this happens on a
View's hardware layer the render buffer will live for as long
as the layer is bound to the view. When a stencil buffer is
required because of a call to Canvas.saveLayer() it will be allocated
on every frame. A future change will address this problem.
If "show GPU overdraw" is enabled, non-rectangular clips are not
supported anymore and we fall back to rectangular clips instead.
This is a limitation imposed by OpenGL ES that cannot be worked
around at this time.
This change also improves the Matrix4 implementation to easily
detect when a rect remains a rect after transform.
Change-Id: I0e69fb901792d38bc0c4ca1bf9fdb02d7db415b9
