This allows apps to find OpenGL ES 3.0 functions using
eglGetProcAddress() instead of dlopen/dlsym.
Bug: 9681677
Change-Id: I7ce6e1636bc47d6b0bf20a4e46bd67235714d129
in the common case this saves one instructions per jump
(which will help with the i-cache).
this change also gets rid of the "use slow tls" option,
which was useless. So at least now architectures that don't have
assembly bindings will perform much better.
Change-Id: I31be6c06ad2136b50ef3a1ac14682d7812ad40d2
destroyed but current-to-a-thread resources are only destroyed
when they're made not-current; however, we were not destroying
those when the thread itself terminated, causing these resources
to be leaked forever.
we now install a tls-key destructor that takes care of this
by calling eglReleaseThread upon thread termination.
Bug: 9209776
Change-Id: I88298a34e3a27488eb81eab76717715569c7d57c
for single library:
/vendor/lib/egl/GLES.so
for multiple libraries:
/vendor/lib/egl/EGL.so
/vendor/lib/egl/GLESv1_CM.so
/vendor/lib/egl/GLESv2.so
EGL_BOARD_CONFIG as well as egl.cfg are not needed anymore.
To facilitate the transition, the loader will also look
for the older naming scheme.
Bug: 8631636
Change-Id: Id4b113468df29bae72b323f76598229914e1c7a1
the code intended to filter out the software-renderer, but the
test as written was essentially a no-op.
the problem didn't happen most of the time because we had
updated egl.cfg to not even list the software renderer.
the test as written didn't generate a compile-time error
because String8 has a const char* cast operator; but the
end result was to compare pointers instead of the string
itself.
http://code.google.com/p/android/issues/detail?id=54606
Change-Id: I739dd1c838fbc24c5643a631fae19713a8ef1717
eglCreateSyncKHR requires a call to glFlush which we were not doing.
fixed by moving the code above eglSwapBuffers(), which both
fixes the problem and gives us a slightly better idea of when
the GPU is done drawing.
Change-Id: Ic826ef1fe25a6247742c3d49d0cb69f4031e3593
We now detect at runtime which sync features to use, which
allows us to remove a lot of the compile-time configuration
options. There is still one option though, to disable
KHR_fence_sync on some devices (which are more efficient
without it).
- added a backdoor to get the vendor's EGL strings
the new logic is:
- use always ANDROID_native_fence_sync if available
- fallback to KHR_fence_sync if available and not disabled
by the compile-time option
- use KHR_wait_sync if available and either of the above is
enabled
Change-Id: I9c4b49d9ff1151faf902cc93bd53ea5f205aaabf
- move all the code related to EGL extensions in one place
- add missing extension strings:
EGL_KHR_lock_surface
EGL_KHR_reusable_sync
- add public extensions strings and entry-points
EGL_ANDROID_wait_sync
EGL_ANDROID_presentation_time
- add missing entry-points for EGL_KHR_reusable_sync
Change-Id: Ifd98966b549e8efd8ef5385eba0efde8c4cbc77b
- added a ctor that updates and dumps the stack immediately
- added a "logtag" parameter to dump()
Change-Id: Ie51c256071d282591752243bdb4f68cf9ff8829d
Added EGL extension to set a timestamp on a surface.
Also, fix JNI encoding of "long" in glgen.
Bug 8191230
Change-Id: I38b7334bade3f8ff02bffe600bb74469ef22c164
* changes:
Symlink /system/lib/libGLESv3.so -> libGLESv2.so
Add ES3 support to libGLESv2 and tracing tools
Import OpenGL ES 3.0 headers from Khronos SVN
The platform has a unified GLES2/GLES3 wrapper in libGLESv2.so, which
is what bundled binaries should link against. The NDK will have
separate GLES2 and GLES3 libraries so that applications will get a
link error if they accidentally use GLES3 entry points in a GLES2
application. This symlink allows GLES3 NDK apps to load using the
unified wrapper.
Change-Id: Ic344ef63c334047fccebb55b7cb2fb23a4c2d61b
Since ES3 is backwards compatible with ES2, a new wrapper isn't
necessary, and the Khronos implementation guidelines recommend
supporting both versions with the same library.
Change-Id: If9bb02be60ce01cc5fe25d1f40c4e7f37244ebf6
This change has been made in Khronos SVN for GLES2 already, however it
is mildly controversial and discussion is on-going.
This change is necessary for Android because otherwise applications
cannot include both GLES2/gl2.h and GLES3/gl3.h as the glShaderSource()
prototype mismatches. Native programs may include framework headers
(which now include GLES3 headers only) but already have included GLES2
headers, creating an incompatibility.
Change-Id: I70a4075e079ee063e76cef5a227abb0ae70aebb4
Avoids irrelevant output differences based on the user's locale. The
current output files were generated with a C or POSIX locale, and sort
differently under en_US.UTF-8.
Change-Id: I0dd85475e9ee2d58e7fa9fe2a4f86ad5c8142e00
# By luliuhui
# Via Android Git Automerger (1) and others
* commit '873ab60e94811488caea1906eb98133329a3e28f':
Gfx-3d: fix GetProcAddress in egl extention
This patch can fix issue when Proc address return NULL.
Glbench10, egl-ext and conformance test need this patch to run pass.
Change-Id: I275c7cb6f77cb334c3ee7fa23cd696bba1c5a458
Author: Liuhui Lu <liuhui.lu@intel.com>
Signed-off-by: Shuo Gao <shuo.gao@intel.com>
Signed-off-by: Bruce Beare <bruce.j.beare@intel.com>
Signed-off-by: Jack Ren <jack.ren@intel.com>
Author-tracking-BZ: 52622
Currently, to activate OpenGL tracing, an application has to be
start with --opengl-trace option (or have a debug prop set).
This CL adds support for tracing an application which may already
be running. This is implemented as follows:
- DDMS initiates a JDWP message to the VM indicating that
opengl traces be enabled.
- When that message is received, a flag is set that indicates
that tracing should be enabled.
- The trace flag is checked during every eglSwap() operation,
and if it finds that tracing should be active and it isn't,
then it starts the tracing component.
Change-Id: I3347fe89fc06c7404d7aa9360f4b21e5bf36ebcb
The shell property debug.egl.trace can now be set to:
0
disables tracing
1
logs all GL calls
error
checks glGetError after every GL call, logs a stack trace on error
systrace
logs each GL call to systrace
Change-Id: I34a2a2d4e19c373fd9eaa1b0cd93e67c87378996
The debugging and tracing levels are initialized both during
early_egl_init and egl_display_t::initialize(). For apps this has
no effect since early_egl_init is called in the context of zygote.
This CL removes the unnecessary call from early_egl_init.
Change-Id: Ibc55c26228cd197f2e2623fbfb2d1a63d0722401
This change fixes a bug where initializing EGL multiple times (eglTerminate
followed by eglInitialize) would cause extensions to show up in the extension
string multiple times.
Change-Id: I707a3da62ed30ef13835087167f84a08bc6addd7
This change adds support for the EGL_ANDROID_native_fence_sync extension to the
Android EGL layer. It also fixes a couple minor issues with the extension spec.
Change-Id: Ic8829d21f37b701f33aa9c72c3d25e88e03fa3cd
eglGetRenderBufferANDROID was removed as it had no users. This commit
reintroduces this extensions as it's used by Qualcomm graphics HAL.
Change-Id: I493306830a0d4f8722a42bcc84fb49236afdabeb
Certain apps (e.g. chrome) seem to create contexts which are unused
for long periods of time. If tracing is stopped before those contexts
are used, then the debugger never gets to know that these contexts
were created. Flushing the trace after these calls ensures that
the debugger knows about all created/used contexts.
Change-Id: I01baa11aa56ac89eddce3c2851e4bf01076984d1
- Process is killed by system with SIGBUS signal if it writes
data to mapped sparse file on full filesystem.
- Allocate space using write() function instead of ftruncate()
to avoid creation of sparse files on full filesystem.
Catch write() errors to handle out-of-space case during allocation.
Bug: http://code.google.com/p/android/issues/detail?id=35376
Change-Id: Ifc366454f34e71a43a0973eda4f591a920ea3a14
Signed-off-by: Kirill Artamonov <kartamonov@nvidia.com>
this allows us to enable h/w acceleration on low-end
devices while keeping memory usage down.
Bug: 6557760
Change-Id: I8af2de3038dc2579360b8b73aa452cb7a0e506a9
Currently, gltrace offers very few trace collection options. As a
result, these options are encoded in a single integer. The trace
control task simply receives integers and interprets them as
commands.
This patch changes the control protocol to first receive the
command length followed by the actual command itself. This allows
for future flexibility to provide enable other commands.
Change-Id: Id5f56c80a025bbbe7613ab4457e092732e7d9dc9
Hibernating EGL takes a long time (>100 ms) and blocks all other
rendering. During window animations, the outgoing activity begins
hibernation before the animation stops, causing visible stutter.
Hibernation is still available by setting 'BOARD_ALLOW_EGL_HIBERNATION
:= true' in the devices BoardConfig.mk
Change-Id: Iab4e00723a1adcd97481e81b2efdc821b3e9712f
When eglTerminate() is called with a window surface still exists, a
deadlock would occur since egl_display_t::terminate() holds a lock
while destroying the window surface, which calls
onWindowSurfaceDestroyed() which attempts to take the same lock.
This change refactors the hibernation code and data into a separate
object with its own lock, separate from the egl_display_t lock. This
avoids the deadlock and better encapsulates the hibernation logic.
The change also fixes a bug discovered incidentally while debugging:
hibernating after calling eglTerminate() succeeds, but will cause
awakens from subsequent eglInitialize() to fail. We will no longer
hibernate a terminated display.
Change-Id: If55e5bb603d4f8953babc439ffc8d8a60af103d9
If the EGL implementation supports the EGL_IMG_hibernate_process
extension, use it to hibernate (and hopefully release memory or other
resources) when the process isn't actively using EGL or OpenGL ES. The
idleness heuristic used in this change is:
(a) Wake up when entering any EGL API call, and remain awake for the
duration of the call.
(b) Do not hibernate when any window surface exists; this means the
application is very likely in the foreground.
(c) Do not hibernate while any context is made current to a thread.
The app may be using a client API without the EGL layer knowing,
so it is not safe to hibernate.
(d) Only check these conditions and attempt to hibernate after a
window surface is destroyed or a thread's context is detached. By
not attempting to hibernate at the end of every EGL call, we avoid
some transient wakeups/hibernate cycles when the app is mostly idle,
or is starting to become active but hasn't created its window
surface yet.
On a Galaxy Nexus, hibernating frees 1567 VM pages from the process.
Both hibernating and waking can take anywhere from 30ms to over 100ms
-- measurements have been very inconsistent.
Change-Id: Ib555f5d9d069aefccca06e8173a89625b5f32d7e
This is in preparation for a change that will hibernate the underlying
EGL when idle. Instead of a bare egl_display_t*, get_display() now
returns a egl_display_ptr, which acts like a smart pointer. The
"wakecount" counter managed by the smart pointer isn't used for
anything in this change. It will be used to make sure we don't
hibernate when any thread is in an EGL call, without having to hold a
mutex for the duration of the call.
Change-Id: Iee52f3549a51162efc3800e1195d3f76bba2f2ce
This change adds a debug option to EGL to use an EGLSyncKHR each frame to
determine when the GPU finishes rendering the frame.
Change-Id: I09ce071db904b44f07ca814c586c291c8b59385a
This patch adds a function setGlDebugLevel() to libEGL to enable
GL tracing. This will be used by the Java layer to add an option
to "am start" that can enable tracing for a particular application.
Change-Id: Ie1dbdd550f502df8633553595cb33ee9d9ae44e1
This patch enables tracing of vertex attribute data that
is specified using glVertexAttribPointer().
At the time the glVertexAttribPointer() call is made, we
only receive a pointer in client space, without any indication
of the size (# of attributes). This size is known only at
the time of the glDraw() call.
This patch generates a new message glVertexAttribPointerData()
when a draw call is issued that contains the vertex attribute
data.
A glDrawArrays() call directly gives the size of data to copy.
A glDrawElements() call gives the indices to copy. In such a
case, all data between the min & max indices drawn are copied
and sent to the host. To support glDrawElements() with an
element array buffer, this patch also adds state that maintains
a copy of all element array buffers.
Change-Id: I434da794a0aa9ada8e7474e219ffb1d79b183ecf
Currently, the trace API passes the pointers that need to be
patched up via 32 bit integers. Such code will not be 64 bit safe.
This patch sends all pointers in a separate array of pointers
for the fixup calls to read from.
Change-Id: If975333f11a6f6f9a74fba57de328affaed452a5
this simplify our EGL wrapper implementation a lot.
This wrapping is no longer needed now that we can only
support a single underlaying EGL implementation.
Change-Id: I8213df7ac69daac447f1fe6e37044b78aac4e9a9
This patch updates the trace information for all glUniform*() calls
to have the right data (the actual uniforms that are passed).
In addition, as soon as a program is linked, information regarding
all the active attributes and uniforms is passed on to the debugger.
Change-Id: Icfbc6722789b42c413a845cf546577fa6de7da2b
this is possible now that we support only a single
EGL implementation. this allows a large code simplification.
Change-Id: I7a6b9db4c5d60f4407c6061e7a68729af63d5242
from now on, the system can only have one EGL
implementation. this means the software and h/w renderer
cannot be used at the same time on a device. Of course, the
h/w renderer is always prefered; in its absence we
default to the software renderer.
Change-Id: Ib579f58055dd0ce4c4a99144131efa11c16ca3d3
this was introduced in a recent change. eglMakeCurrent can
end up calling eglDestroyImageKHR via ANativewWindow::disconnect
when the consumer is in the same process.
we make sure we don't hold the lock while this is happening.
Change-Id: Id17fe4fd76eecf5f962cefb9aa32be41fc1b042d
it would happen when a context was made non-current, in this
case we would call the implementation's eglMakeCurrent() which
would succeed, if we're rescheduled at that point, another
eglMakeCurrent() could make that context current to another thread,
however, when we came back to the original call we would
overwrite egl_context_t internal state.
this is fixed by moving the critical section under
egl_display_t's lock.
Change-Id: I743c85696e13263d3a9570824940263df0caacdc
when increasing MAX_NUMBER_OF_GL_EXTENSIONS to 256
we also needed to create all the corresponding wrappers.
Change-Id: I90edaaf0885ccdfab48e7a1396bcf88e039cfb25
This extension is always added to the GL_EXTENSIONS
extension string for the current GL context, regardless
of if it's supported by the h/w driver.
The extension itself will be handled by GLES_trace (eventually),
when GLES_trace is not enabled, it'll result to a no-op.
If the h/w implementation has this extension, we'll call that version
instead of our dummy version.
Change-Id: Ie5dd3387c4d45cd5ed5f03b73bda6045620a96bc
Export trace information via abstract Unix Domain Socket (UDS).
This allows tracing of applications without INTERNET permission,
and should be faster as well.
Change-Id: Iabb67fcc2bc2484afd8128af07dca723b81c52c6
Currently users do not have control over the amount of data
captured during tracing. This patch adds 3 settings that users
can enable/disable at runtime:
- capture framebuffer on eglSwap() calls
- capture framebuffer on glDraw*() calls
- capture texture data passed to glTexImage*() calls
Disabling these options when not needed signficantly decreases
the size of the trace file, and reduces performance overhead for
the running application.
These settings are stored in the per process GLTraceState.
A separate thread listens for commands from the host, and updates
the state based on the user commands.
Change-Id: Ic4518b94e8bcbc5330ac7138153721caa98b365d
This patch adds two improvements:
1. Protobuf messages are buffered and sent in chunks.
2. Multiple EGL contexts are handled properly: Corresponding
to each EGLContext, a GLTraceContext with a unique ID is created.
On eglMakeCurrent, the appropriate GLTraceContext is set and is
used while tracing subsequent GL Calls in that thread.
Change-Id: I34076376d3e5af205c87c7396ea47659844abd6e
Currently, gltrace always attaches the contents of the currently
bound framebuffer. This patch changes it to attach the contents
of FB0 on eglSwap, and the currently bound framebuffer for the
glDraw* calls.
Change-Id: Ice0520d45d75638fe61cd91149df773074216510
Currently, the contents of the FrameBuffer are sent by encoding
them as the last argument to the function call. As a result, it is
not possible to know if a message has the framebuffer encoded in it
without looking at the function type.
This patch modifies the protobuf definition to include a separate
optional framebuffer message.
Change-Id: Ief3a6950052d927ca0743e729457435b48c25a92
This patch provides a framework for tracing GLES 1.0 and 2.0
functions. It is missing a lot of features, but here are the
things it accomplishes:
- Stop building the glesv2dbg library, and build the
glestrace library instead.
- Replace the hooks for glesv2dbg with the ones for glestrace.
- Add the basics for the trace library. Currently, this
traces all GL functions, but not all required data is
sent for all the functions. As a result, it will not
be possible to reconstruct the entire GL state on the
host side.
The files gltrace.pb.* and gltrace_api.* are both generated
using the tools/genapi.py script.
Change-Id: Id60a468f7278657f008bc6ea1df01f9bdfecfdd3
This change explicitly sets swap interval 1 on the window when an
EGLSurface is created to render to it.
Change-Id: I91eb29dbee3ae4a55076b921f084d503fbe94e03
This change introduces two new BoardConfig variables to control the size
limits of the EGL blob cache. MAX_EGL_CACHE_ENTRY_SIZE is the size
limit for values inserted into the cache, and MAX_EGL_CACHE_SIZE is the
size limit for all entries in the entire cache (including both keys and
values). If either of these BoardConfig variables are not defined then
a default size limit is used instead.
Change-Id: I6703d93f966b6389c6499f23d841e42339f9c9d7
- don't advertise extensions that are not supported
by any implementation
- remove EGL_ANDROID_swap_rectangle which is not
implemented by anybody and confuses people
- add some comments about mandatory extensions
Bug: 5428001
Change-Id: Id8dc48116ac1d1eb79ec9ef55d03e29d4257c1f3
the code that validated EGL objects dereferenced the object
to access its EGLDisplay -- needed for validation (!).
This was wrong for two reasons, first we dereferenced the object
before validating it (potentially leading to a crash), secondly
we didn't validate that the object existed in the right EGLDisplay.
We now use the EGLDisplay passed by the user API.
Change-Id: I66f9e851d4f8507892a6b1fee3065f124c4e7138
