When updateTexImage acquires a buffer but then aborts (due to an error
or the buffer being rejected), it releases the newly-acquired buffer.
It was passing the buffer slot's fences to releaseBuffer, even though
they hadn't been created after the acquire yet. This wasn't a bug,
since the fences would be cleared just after the buffer slot was last
released, but explicitly passing null fences makes this clearer.
Change-Id: I087f2ec3fd02c40f57782c1fca24eb9567e2943d
both API and implementation will change, this is just a prototype
intended to show feasability.
SurfaceFlinger is passed an ISurfaceTexture through a new
callback, it is in turn used to create an EGLSurface which
surfaceflinger will draw into in addition to the main screen.
Change-Id: Id0bbb0b854bb7bae44d57246a90b65d4567f9a21
Hand merge from ics-aah
> Utils: Fix a bug in the linear transformation code.
>
> Fix a bug where an incorrect result would be computed if you used the
> linear transformation code to do a reverse transformation (from B's
> domain into A's domain) when the scaler fraction was negative.
>
> Change-Id: I8e5f109314d235a177ab41f65d3c4cd08cff78be
> Signed-off-by: John Grossman <johngro@google.com>
Change-Id: Id90e18f685c61c1a89fd91c32adcf01363b3e8f3
Signed-off-by: John Grossman <johngro@google.com>
When the app_process is shutting down the main thread will close the
binder fd while pool threads are executing an ioctl (in
IPCThreadState::stopProcess called by AppRuntime::onStarted in
app_main.c).
The binder driver will then return all pending calls in ioctl
without any error and with a command. One of the threads gets a
BR_SPAWN_LOOPER which will create a new thread (the other thread
gets a BR_NOOP). This new thread then calls
vm->AttachCurrentThread. Usually this results in a log entry with
"AndroidRuntime: NOTE: attach of thread 'Binder Thread #3' failed",
but sometimes it also causes a SIGSEGV. This depends on the timing
between the new thread an the main thread that calls DestroyJavaVM
(in AndroidRuntime::start).
If IPCThreadState.cpp is compiled with "#define LOG_NDEBUG 0" the
pool thread will loop and hit the
ALOG_ASSERT(mProcess->mDriverFD >= 0) in
IPCThreadState::talkWithDriver.
Crashes like this has been seen when running the am command and
other commands that use the app_process.
This fix makes sure that any command that is received when the driver
fd is closed are ignored and IPCThreadState::talkWithDriver instead
returns an error which will cause the pool thread to exit and detach
itself from the vm. A check to avoid calling ioctl to a fd with -1
was also added in IPCThreadState::threadDestructor.
Another solution might be to change the binder driver so that it
returns an error when the fd is closed (or atleast not a
BR_SPAWN_LOOPER command). It might also be possible to call exit(0)
which is done when System.exit(0) is called from java.
Change-Id: I3d1f0ff64896c44be2a5994b3a90f7a06d27f429
ISurfaceTexture::dequeueBuffer now returns the buffer's fence for the
client to wait on. For BufferQueue, this means passing it through
Binder so it can be returned to the SurfaceTextureClient. Now
SurfaceTextureClient is responsible for waiting on the fence in
dequeueBuffer instead of BufferQueue: one step closer to the goal.
Change-Id: I677ae758bcd23acee2d784b8cec11b32cccc196d
After a HWC set, each SurfaceFlinger Layer retrieves the release fence
HWC returned and gives it to the layer's SurfaceTexture. The
SurfaceTexture accumulates the fences into a merged fence until the
next updateTexImage, then passes the merged fence to the BufferQueue
in releaseBuffer.
In a follow-on change, BufferQueue will return the fence along with
the buffer slot in dequeueBuffer. For now, dequeueBuffer waits for the
fence to signal before returning.
The releaseFence default value for BufferQueue::releaseBuffer() is
temporary to avoid transient build breaks with a multi-project
checkin. It'll disappear in the next change.
Change-Id: Iaa9a0d5775235585d9cbf453d3a64623d08013d9
FramebufferNativeWindow::dequeueBuffer now waits for the next buffer
to be non-front in addition to being free.
Change-Id: I991f154958cc6b488b1241aba83d1f95a0513b3c
This change updates the uses of ANativeWindow to use the new ANW functions that
accept and return Sync HAL fence file descriptors.
Change-Id: I3ca648b6ac33f7360e86754f924aa072f95242f6
SF now has its own implementation of ANW for the
framebuffer and it uses BufferQueue. FramebufferNativeWindow
is now only used by stand-alone apps.
Change-Id: Iddeb24087df62bd92b0f78e391dda9b97ddc859c
The Java implementation of writing the RPC response header
calculates the length of the header including the 4 bytes
specifying the header length but the native implementation
excludes the 4 bytes specifying the length from the header
length.
The native implementation has been aligned to the Java impl.
Change-Id: I325bf272a63152d8fded4cf4e51a906b5a9bfe19
SurfaceFlinger will no longer directly synchronize with early suspend.
Instead, PowerManagerService will synchronize with SurfaceFlinger to
ensure that a black frame has been drawn on the display, and then
trigger all early suspend handlers.
Change-Id: I07acdd628440d23fdb69db94319ec5d65d3f4919
SurfaceFlinger is using this new feature to reject buffers that
don't have the right size.
Bug: 6498869
Change-Id: I8a7250a47db6c082a357b703feb3b9d0fc8d3443
The host-accelerated GL ES driver doesn't do the implicit
synchronization required when fence sync isn't used.
Bug: 6515813
Change-Id: I6a667f2db6b519f3557b5abda78775f767841dae
this prevents SurfaceTexture producer to hang in dequeueBuffer()
when something goes wrong in the consumer.
only the consumer gets an error and the current frame is kept
instead of the new one. the producer is unaware of the problem.
Bug: 6476587
Change-Id: Ie6db5526632aabc3e60229b93dfe29c19491ade4
would happen when vectors are copied and new items is
added in both vectors. we didn't duplicate the underlying
storage when adding items in vectors.
Bug: 6515797
Change-Id: If544c07d96c05821e088d7f2c9b5736f7e306c31
SurfaceTexture would only create an EGLImage for a buffer slot when
BufferQueue returns a GraphicBuffer, i.e. either the slot was acquired
for the first time ever, or the buffer for the slot was reallocated.
But the EGLImage may also need to be re-created for a
previously-acquired buffer if the slot's EGLImage was destroyed during
detachFromContext(); in this case BufferQueue won't return a
GraphicBuffer since SurfaceTexture already has a reference to the
correct buffer.
Bug: 6461693
Change-Id: Ib95d0d757192efe336c5fda0866f857481a6617d
Aimed for use cases where gralloc buffers need to be consumed by CPU
users, such as camera image data streams.
The CpuConsumer is a synchronous queue, which exposes raw pointers to
the underlying graphics buffers to applications. Multiple buffers may
be acquired at once, up to the limit set at time of construction.
Change-Id: If1d99f12471438e95a69696e40685948778055fd
Previous kernels have allowed opening the CPU frequency file regardless
whether the CPU is up or not. This fixes some log spam on one device
with dynamic hot plug CPU feature, which does not allow opening the CPU
frequency file if CPU is down.
Also, since the file descriptors are global and have long lives, add
the close-on-exec flag.
Change-Id: Ia14a2b9e20038dfb96a573920176a47a96bd3f5a
This change removes the setPostTransformCrop function from
SurfaceTextureClient. It also includes a small logging fix in BufferQueue.
Bug: 6299171
Change-Id: Ifd0ed05b95dad6085e7a8267fda4d69b76ea3bad
This change adds a crop rectangle specified in window coordinates to the layer
state. The all window pixels outside this crop rectangle are treated as though
they were fully transparent. This change also adds the plumbing necessary for
WindowManager to set that crop.
Change-Id: I582bc445dc8c97d4c943d4db8d582a6ef5a66081
This change updates some of the SurfaceTextureClient and BufferQueue logging
and dumping to include the crop, transform and scaling mode. It also removes
the uses of the NO_SCALE_CROP scaling mode enum, which was added by accident in
a previous change.
Change-Id: I62912716a1e48885fb22f12b92678aa13f10fcd9
Bug: 6470541
Every IBinder object can accept a new transaction to tell it that
it might want to reload system properties, and in the process
anyone can register a callback to be executed when this happens.
Use this to reload the trace property.
This is very much ONLY for debugging.
Change-Id: I55c67c46f8f3fa9073bef0dfaab4577ed1d47eb4
This change reduces the shrinking of the crop rectangle in some cases. It adds
a way to inform the SurfaceTexture that its texture will be used without
bilinear interpolation, and uses knowledge of the pixel format to avoid
shrinking unecessarily.
Change-Id: I72365f39f74ecb7fcc51b4cf42f2d0fa97727212
This change adds support for specifying a crop rectangle to a
SurfaceTextureClient that is in post-transformed coordinate space.
Change-Id: I247901de343e71b32850f7ae3bac62dfa612ad3d
Bug: 6299171
This change adds a method to Rect to transform a rectangle by a graphics HAL
transform.
Change-Id: Ic0d0988e731bdb5662faee41a5927b1242891658
Bug: 6299171
Fixed two issues about casting between int and pointer which
cause compilation error in 64-bit mode (eg. for 64-bit emulator)
Change-Id: I5ca9a933a239db73e56d6f27de05aa42bd06a650
This change adds the plumbing to SurfaceTextureClient, BufferQueue, and
SurfaceTexture to get the active rectangle passed to the ANativeWindow to
the buffer consumer.
Change-Id: I35da0889b266327ebb079b6a7136fa3e2e8b00e6
Fixed AttachToContextSucceeds test SurfaceTexture test by
not deleting the current texture when cleaning up memory.
Bug: 6363186
Change-Id: Ib886bfe6bb86e25ea4825d18b5008afba6b33eae
indeed, connect and queueBuffer return the same data, so it's
easier to have them use the same protocol.
Change-Id: I4f9fa3be0a80c9ab0a7a4039b282ae843aab02e1
Remove C++ APIs androidSetThreadSchedulingGroup and
androidGetThreadSchedulingGroup, and the ANDROID_TGROUP_* constants.
Former callers of these should now use the C APIs set_sched_policy and
get_sched_policy, and the SP_* constants.
Note: debug.sys.noschedgroups is not supported by the C APIs,
this needs to be discussed.
Change-Id: I32bbfc539ef4090faf9ef0320380e8cca9eae07c
This change fixes an issue where we were sometimes setting the SurfaceTexture's
EGLDisplay to EGL_NO_DISPLAY in detachFromContext, and then subsequently
abandoning the texture. Abandoning while in the detached state would result in
the eglDestroyImageKHR calls failing, which resulted in a memory leak.
Bug: 6302694
Change-Id: I24c1de0dac029a83c7508075fb8aaeaed96a14ea
This change makes SurfaceTexture include an offset for all sides of the crop
region when cropping. This keeps the image centered, to minimize the visual
changes when switching between the texture transform matrix-based cropping and
something that does proper cropping (e.g. HWComposer).
Change-Id: I541d3046fd92e49221b488444df36d490924d1c5
- it returned an empty rect when the region was empty, instead
of returning an empty list of rect.
- also fixed an infinite loop when boolean_operation was given
an empty list of rects
Change-Id: I62225c7dcd2832025bb8f12e6cb3762f2a7b36cb
SurfaceTexture will modify the crop rect so it matches
the desired output aspect ratio when the scaling
mode is NATIVE_WINDOW_SCALING_MODE_CROP. Added a test
for this new scaling mode.
Change-Id: I60f24dcbc294b65cd10a393d9e27d40f07d27bb6
Add a new API native_window_set_buffers_user_dimensions to allow native
applications to override the native window size for the default buffer size.
This has lower precedence than the existing
native_window_set_buffers_dimensions and allows the two to co-exist.
Change-Id: Ie73590e1c94ef0dadbce500bd0941dfabbcace3c
BufferQueue is now more flexible as it can be used
by SurfaceMediaSource in addition to SurfaceTexture.
Change-Id: I4222be8918d63372c44fcd412d9ad241c6a3eeb9
BufferQueue's disconnect could race with updateTexImage
where invalid buffers could be released. Additionally
fixed similar bug with setBufferCount. Tests were added
to stress the disconnect mechanism.
Change-Id: I9afa4c64f3e025984e8a9e8d924852a71d044716
This change adds the detachFromContext and attachToContext methods to
SurfaceTexture. These methods allow the SurfaceTexture to switch from
one consumer GLES context to another. This change also includes a few
cleanups to the error return codes in updateTexImage.
Change-Id: I0df1eb599aa7b6f58f07431f242f8f09269559ed
collapse setCrop, setTransform and setScalingMode to queueBuffer()
this ends up simplifying things quite a bit and reducing the numnber
of IPC needed per frame.
Change-Id: I3a13c07603abe4e76b8251e6380b107fde22e6d9
pdk builds don't have libandroid.so available. Remove it from
projects that aren't actually using it.
Change-Id: I100dc84aaf8160c1062a122ef6d5aec0534d73fa
The DummyConsumer is a consumer that can connect to BufferQueue
that does nothing. It is required as BufferQueue checks
if a consumer is connected. Also fixes a bug where SurfaceTexture
was reusing old texture slots.
Bug: 6172739
Change-Id: I5e7a118dd4d07807ba4c763200242b5ee7d3412b
ZipUtils is needed by build/tools, move it from libandroidfw
(frameworks/base) to libutils (frameworks/native).
Change-Id: Ib8c41134ccdec6d6aa8d49860f8dddef49a24783
This change adds a check to SurfaceTexture::updateTexImage to verify
that the current GL context is the same as the one that was used for
previous updateTexImage calls.
Change-Id: If02d2f787bcfdb528046dc9ddf6665f8a90e1bf4
This change adds a new callback for BufferQueue consumers to be notified
when the BufferQueue frees some or all of its buffers. This is needed
to retain SurfaceTexture behavior where all buffers would be freed when
the producer disconnects. This change also modifies the
SurfaceTextureGLToGLTest.EglDestroySurfaceUnrefsBuffers test to catch
when the buffers are not freed.
The implementation is a little complicated because it needs to avoid
circular sp<> references across what will be a binder interface (so wp<>
can't be used directly). It also needs to avoid the possibility of
locking the BufferQueue and consumer (e.g. SurfaceTexture) mutexes in
the wrong order.
This change also includes a few additional fixes and test cleanups.
Change-Id: I27b77d0af15cb4b135f4b63573f634f5f0da2182
It's needed to build four shared libraries in 64-bit for 64-bit
emulator with "-gpu on"
lib64OpenglRender.so
lib64EGL_translator.so
lib64GLES_CM_translator.so
lib64GLES_V2_translator.so
Change-Id: Ia6c05b23df1e9cd9e7f2e94e4cd5bde4be5d336b
- condition wasn't signaled if an error happened between acquire and release
- also replace signal with broadcasts
Bug: 6109450
Change-Id: I8ac03c7eca35c9cc04a00ef7fad36bb9cb3fcef6
This is used by the Java trace wrappers to avoid JNI overhead
when trace tags are disabled.
Also added a new tag for the input subsystem and view hierarchy.
Change-Id: Ia04a507e42a37b1d3bbb19859e6c07a92f4fe9aa
As part of scheduling policy cleanup, remove or isolate
all references to the scheduling policy APIs.
Change-Id: Ia1ea2fe711a399039f25217309e061267744b856
prctl(PR_SET_NAME) limits to 15 characters. Before we had names like
"Binder Thread #" and the counter was cut off :-( Also remove redundant
"thread" at end of name; it's always a thread.
Change-Id: I1f99c2730ba0787ed9b59c15914356cddf698e2f
MemoryHeapPmem is not used any longer. PMEM is not a supported
type of memory by the system anymore. a particular device might
use PMEM and need something like MemoryHeapPmem, in this case this
should be implemented in device specific code (HAL).
This will most likely break older no longer supported targets.
Change-Id: I434e4291219950018de8b793b0403bb2d92dd5cc
1. Use "%zu" instead of "%d" for size_t in printf
2. Variable precision specifier (eg. "%.*s") in printf should be of type int.
(iov_len is size_t which is 64-bit when compiled with -m64)
3. Use PRId64 instead of "%lld" to print variables of type int64_t
Change-Id: I2be40a6514b5dffa0038d62b9bccc3401b8756e6
Refactored SurfaceTexture and BufferQueue such that share
no protected members. Created an consumer facing interface
for BufferQueue in preparation of connecting SurfaceTexture
and BufferQueue through a binder.
Change-Id: I938e63e085128148c58d0e26c7213b30145c109f
re-add support for pixelformats L_8, LA_88 and RGB_332 in libui
for backward compatibility.
This may or may not fix 6058926
Bug: 6049685
Change-Id: Ic1b8b4cc994522f7fe664da64c0ef76b98bc6d53
Refactored SurfaceTexture and BufferQueue such that share
no protected members. Created an consumer facing interface
for BufferQueue in preparation of connecting SurfaceTexture
and BufferQueue through a binder.
Change-Id: Iff55e740e36a7f70c9f7a17ee7a5af38e3d21f0f
This does not actually change the framebuffer format. It merely
fakes this format to surfaceflinger so that when it creates
framebuffer surfaces it will use this format. It's really a giant
HACK to allow interworking with buggy gralloc+GPU driver
implementations. You should *NEVER* need to set this for shipping
devices.
Change-Id: I03eeb5b4d72838ef219df386ecc489fc20ab9cc7
Signed-off-by: Dima Zavin <dima@android.com>
Instead of sending finished signals immediately when appending to
a batch, record the chain of sequence numbers that were part of
the batch and then send finished signals all at once when done.
This change helps the dispatcher keep track of the true state
of the application and can improve ANR detection slightly.
This is part of a series of changes to improve input system pipelining.
Bug: 5963420
Change-Id: I463c2221e2aa8fdf1c3d670c18e39e59ab69b0db
To support this feature, the input dispatcher now allows input
events to be acknowledged out-of-order. As a result, the
consumer can choose to defer handling an input event from one
device (because it is building a big batch) while continuing
to handle input events from other devices.
The InputEventReceiver now sends a notification when a batch
is pending. The ViewRoot handles this notification by scheduling
a draw on the next sync. When the draw happens, the InputEventReceiver
is instructed to consume all pending batched input events, the
input event queue is fully processed (as much as possible),
and then the ViewRoot performs traversals as usual.
With these changes in place, the input dispatch latency is
consistently less than one frame as long as the application itself
isn't stalled. Input events are delivered to the application
as soon as possible and are handled as soon as possible. In practice,
it is no longer possible for an application to build up a huge
backlog of touch events.
This is part of a series of changes to improve input system pipelining.
Bug: 5963420
Change-Id: I42c01117eca78f12d66d49a736c1c122346ccd1d
This change enables the input dispatcher to send multiple touch
events to an application without waiting for them to be acknowledged.
Essentially this is a variation on the old streaming optimization
but it is much more comprehensive. Event dispatch will stall as
soon as 0.5sec of unacknowledged events are accumulated or a
focused event (such as a key event) needs to be delivered.
Streaming input events makes a tremendous difference in application
performance. The next step will be to enable batching of input
events on the client side once again.
This is part of a series of changes to improve input system pipelining.
Bug: 5963420
Change-Id: I025df90c06165d719fcca7f63eed322a5cce4a78
Since we will not longer be modifying events in place, we don't need
to use an ashmem region for input. Simplified the code to instead
use a socket of type SOCK_SEQPACKET.
This is part of a series of changes to improve input system pipelining.
Bug: 5963420
Change-Id: I05909075ed8b61b93900913e44c6db84857340d8
Services now must explicitly opt in to being accessed by isolated
processes. Currently only the activity manager and surface flinger
allow this. Activity manager is needed so that we can actually
bring up the process; SurfaceFlinger is needed to be able to get the
display information for creating the Configuration. The SurfaceFlinger
should be safe because the app doesn't have access to the window
manager so can't actually get a surface to do anything with.
The activity manager now protects most of its entry points against
isolated processes.
Change-Id: I0dad8cb2c873575c4c7659c3c2a7eda8e98f46b0
This adds basic support for clip regions. It is currently disabled at compile
time. Enabling clip regions will require setting up a stencil buffer.
Change-Id: I638616a972276e38737f8ac0633692c3845eaa74
Switching activity stacks
Cache ContentProvider per user
Long-press power to switch users (on phone)
Added ServiceMap for separating services by user
Launch PendingIntents on the correct user's uid
Fix task switching from Recents list
AppWidgetService is mostly working.
Commands added to pm and am to allow creating and switching profiles.
Change-Id: I15810e8cfbe50a04bd3323a7ef5a8ff4230870ed
with this changes, SF transactions are handled as soon as possible
but do not trigger updates. the update is delayed until the next
vsync.
this allows us to work much better without requiring triple-buffering.
Change-Id: I1fa10794d0cf742129f0877698b7b1e1f2ec7401
In our current environment with very many translations, this can
save a lot of RAM -- for example over 200K in Gmail just by sorting
the strings in the Gmail .apk (not the framework).
Also add a new aapt command to print the contents of the resource
table string pool.
Change-Id: I1da037b3e2c377b890833ff57ab158965314ac48
This change enables the use of the EGL_KHR_fence_sync extension in
SurfaceTexture on omap4 and s5pc110 platforms.
Change-Id: Icad5245bab445413ffb8a7c823c296b678bf3250
This is needed when the parent or any other thread besides the child
needs access to the child's kernel tid.
Change-Id: Ib148505913eb78314cfd76657c30d7b20663dffd
