whether a physical sensor needed to be active or not was managed by
a simpe reference counter; unfortunatelly nothing prevented it to
get out of sync if a sensor was disabled more than once.
sensorservice already maintainted a list of all the "clients"
connected to a physical sensor; we now use that list to determine if
a sensor should be enabled. This can never be "out-of-sync" since
this is the only data structure linking a sensor to a user of that
sensor.
also removed the isEnabled() method, which was never used and
implemented wrongly (since it didn't take into account that a sensor
could be disabled for a client but not of another).
Change-Id: I789affb877728ca957e99f7ba749def37c4db1c7
Most accelerometers have 8-bits accuracy so we beed to
reject 48dB in thestop-band, which requires a 4-th order
filter at the cut-off frequency we're using.
Change-Id: Ic00421d38d751641f86b1f3ad7663e6b44a91198
- upadte documentation for rotation vector
- update method dealing with rotation vector to deal with 4 components
- virtual rotation-vector sensor reports all four components
- improve SensorManager documentation layout
Whent he 4-th component of the rotation-vector is present, we can save
a square-root when computing the quaternion or rotation matrix from it.
Change-Id: Ia84d278dd5f0909fab1c5ba050f8df2679e2c7c8
indeed, by construction of the rotation matrix, it is
guaranteed to have a length of 1.
moreover, the normalization code was missing a square-root,
fortunatelly, since the length is 1, this didn't cause any
damage (since sqrt(1) = 1).
Change-Id: I9facd668caaf5bb3bfccb139ab872f2bb2066365
Rework sensorservice to allow "virtual sensors", that is
sensors that report a synthetized value based on real sensors.
the main change to sensorservice is around managing which real
sensor need to be activated and which rate to use.
The logic for all this has been moved into SensorDevice, which
essentially wraps the sensor HAL but adds two features to it:
- it keeps track of which sensors need to be activated
- it keeps track of what rate needs to be used
For this purpose an "identity" is associated with each real sensor
activation, so we can track them.
On start-up we check for gravity, linear-acceleration and
rotation-vector sensors, if they're not present in the HAL, we
synthetize them in sensor-service.
Change-Id: I841db2c1b37ef127ed571efa21732ecc5adf1800
the per-connection state assumed the main sensorservice
lock was held during access. This is however not true while
pre-processing the events just before sending them to clients.
Therefore, there was a small window during which this state
could be modified while being used.
we now have an internal lock that protects this state.
Change-Id: I594680f20f09d6a4f1f38f093a1d3f650dcef1be
We only recorded the last received event (which is needed when a sensor
is activated on a connection) when there was some connection active.
This should fix an issue where sometimes the light sensor doesn't
return an event whent activated.
we also didn't need to hold the main lock while dispatching events
to clients.
Change-Id: I6c6386c040051ce205e3c0516c678e0603fa45e1
the increased maximum rate is needed for proper gyro integration, current gyro
parts can sample at up to 800Hz
Change-Id: Ide75f6d5bc7a0fdafeb2dafd72db39e7afb9e794
As part of this change, consolidated and cleaned up the Looper API so
that there are fewer distinctions between the NDK and non-NDK declarations
(no need for two callback types, etc.).
Removed the dependence on specific constants from sys/poll.h such as
POLLIN. Instead looper.h defines events like LOOPER_EVENT_INPUT for
the events that it supports. That should help make any future
under-the-hood implementation changes easier.
Fixed a couple of compiler warnings along the way.
Change-Id: I449a7ec780bf061bdd325452f823673e2b39b6ae
SensorService now correctly sends the last known
state of a sensor as soon as a new connection is made.
This fixes the issue where, for instance, an application
could wait a long time before getting the light or proximity
sensor initial state.
Change-Id: Ic41392f3626e26c4f15746c7e17c7ecd44bbb10b
remove old sensor service and implement SensorManager
on top of the new (native) SensorManger API.
Change-Id: Iddb77d498755da3e11646473a44d651f12f40281
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
The desc.txt file can now mark parts as 'must finish cleanly' by using
'c' as the part line prefix rather than 'p'. If so indicated, if the
bootanimation is asked to quit it will do so only after waiting to
finish that part.
I considered either making init.c service killing smarter or promoting
bootanim to be a bindable service with a requestExit method. However,
these changes are probably too big/risky given our ship date. So
I used a property as a mailbox between SurfaceFlinger and bootanim.
Bug: 6679877
Change-Id: Id7dca22caa50b450fff25ca94f7242d971034f41
When turning the screen off we could have 2 waiters on the
vsync condition: The main vsync waiter as well as one in
onScreenReleased(). We were only signaling the condition though,
so it it would be possible to wake onScreenReleased() without waking
the main vsync thread which would then be stuck in .wait().
We fix this by just using broadcast() when receiving a vsync event.
We also add a broadcast() to signal when the state of
mUseSoftwareVSync changes. This is important particularly for
the transition from hardware to software vsync because the main
vsync waiter might have observed mUseSoftwareVSync == false
and decided to block indefinitely pending a hardware vsync
signal that will never arrive.
Removed a potentially deadlocking wait for a signal in
onScreenReleased(). The function was trying to wait for the last
vsync event from the hardware to be delivered to clients but there
was no guarantee that another thread would signal it to wake up
again afterwards. (As far as I can tell, the only other other
thread that might wake it up at this point would be a client
application issuing a vsync request.) We don't really need to wait
here anyhow. It's enough to set the mUseSoftwareVSync flag,
wake up the thread loop and go. If there was a pending vsync
timestamp from the hardware, then the thread loop will grab
it and use it then start software vsync on the next iteration.
Bug: 6672102
Change-Id: I7c6abc23bb021d1dfc94f101bd3ce18e3a81a73e
When turning the screen off we could have 2 waiters on the
vsync condition: The main vsync waiter as well as one in
onScreenReleased(). We were only signaling the condition though,
so it it would be possible to wake onScreenReleased() without waking
the main vsync thread which would then be stuck in .wait().
We fix this by just using broadcast() when receiving a vsync event.
We also add a broadcast() to signal when the state of
mUseSoftwareVSync changes. This is important particularly for
the transition from hardware to software vsync because the main
vsync waiter might have observed mUseSoftwareVSync == false
and decided to block indefinitely pending a hardware vsync
signal that will never arrive.
Removed a potentially deadlocking wait for a signal in
onScreenReleased(). The function was trying to wait for the last
vsync event from the hardware to be delivered to clients but there
was no guarantee that another thread would signal it to wake up
again afterwards. (As far as I can tell, the only other other
thread that might wake it up at this point would be a client
application issuing a vsync request.) We don't really need to wait
here anyhow. It's enough to set the mUseSoftwareVSync flag,
wake up the thread loop and go. If there was a pending vsync
timestamp from the hardware, then the thread loop will grab
it and use it then start software vsync on the next iteration.
Bug: 6672102
Change-Id: I7c6abc23bb021d1dfc94f101bd3ce18e3a81a73e
