Commit Graph

872 Commits

Author SHA1 Message Date
Andy McFadden
c33f9a76b5 Merge "Pay attention to buffer timestamps" 2013-07-11 14:40:45 +00:00
Andy McFadden
1585c4d9fb Pay attention to buffer timestamps
When acquiring a buffer, SurfaceFlinger now computes the expected
presentation time and passes it to the BufferQueue acquireBuffer()
method.  If it's not yet time to display the buffer, acquireBuffer()
returns PRESENT_LATER instead of a buffer.

The current implementation of the expected-present-time computation
uses approximations and guesswork.

Bug 7900302

Change-Id: If9345611c5983a11a811935aaf27d6388a5036f1
2013-07-10 15:38:40 -07:00
Mathias Agopian
6b44267a3b fix SF buffer cropping
When a buffer had a crop (meaning its content is scaled to the window size)
and a window crop was defined, the resulting crop couldn't be expressed
properly because h/w composer's API was limited to integers, since
this is fixed in h/w composer 1.3, we take adventage of this to
make sure we get the correct crop.

this bug could result in the buffer being scaled by an incorrect ratio and
be slightly offset; moreover, it would produce different results from the
GL code path, which is always correct.

Change-Id: I8e20e00b6e26177d14f4ab4d2cd581e26c818892
2013-07-10 15:08:29 -07:00
Mathias Agopian
2ed0fe59b4 fix a bug where surfaceflinger and system_server could deadlock
because surfaceflinger handles screenshot in a different
    thread from the binder thread that requested it and because
    the IGraphicBufferProducer is a synchronous interface
    calling back into the system server; it is possible for
    the latter to run out of binder threads (b/c it holds
    a lock while calling into SF).

    The solution is to make sure all calls on IGraphicBufferProducer
    happen on the incoming binder thread. We achieve this by creating
    a IGBP wrapper which is given to the screenshot code.

    Bug: 8734824

Change-Id: I2be85660d9dc65d239d68f6d3ab3c973c13b34cc
2013-07-08 18:52:05 -07:00
Mathias Agopian
547e98f33c Merge "Refactor SF. Move all GL operations in their own class." 2013-07-08 22:50:01 +00:00
Mathias Agopian
ba02cd2f6c improve sensorservice's dumpsys
it now displays the reported trigger mode properly, as well as
the number and type of the last received data

Change-Id: I2ff64b32ab71f1332bc2e09671c8c02bb9550490
2013-07-08 15:44:41 -07:00
Mathias Agopian
cf90bad477 Merge "simplify some unnecessary complex code" 2013-07-08 22:34:55 +00:00
Mathias Agopian
fee2b463c5 Debug code for detecting all black pixels screenshots
Bug: 9120292
Change-Id: If60db32524db973bb1f905ba3cb415c2a1cd7e71
2013-07-08 22:28:37 +00:00
Mathias Agopian
7438fd1a01 simplify some unnecessary complex code
getSensorType() ran in O(n) instead of O(1). fix that.

Change-Id: Idcf29e46fc34db32604a0d8e5a9156486783b74f
2013-07-08 12:50:39 -07:00
Mathias Agopian
875d8e1323 Refactor SF. Move all GL operations in their own class.
this is the first step to add support for GLES 2.x, this
change breaks the dependency of SF on GLES 1.x by moving
all operation into their own class.

Bug: 8679321

Change-Id: I0d2741eca2cefe67dfd9cf837cac10c4d126928b
2013-07-03 14:39:27 -07:00
Lajos Molnar
98d3d6ec12 SurfaceFlinger: fix releaseBuffer in updateTexImage
Allow releasing a buffer acquired in the past.

Change-Id: I3f27f1caec109c1a47286fc4d2235c6e1afe9022
Signed-off-by: Lajos Molnar <lajos@google.com>
Bug: 9516405
2013-06-27 11:53:03 -07:00
Mathias Agopian
1eae0ee494 clean-up SurfaceFlinger a bit
- most methods on Layer didn't need to be virtual
- more consistency in naming drawing/current state

Change-Id: Ieb7b4951b40fc673b807994ed21ae4aea7281068
2013-06-06 13:06:10 -07:00
Mathias Agopian
2e3a7f1e3d Merge "Basic binary event-log helpers" 2013-06-06 00:01:15 +00:00
Mathias Agopian
7cc6df5957 fix a possible deadlock when removing a layer and destroying a client
generally the last reference to a Layer is released in commitTransaction()
with mStateLock held. Layer itself only holds weak references to Client,
however, ~Layer() briefly promotes this weak reference -- during that time
the all other strong references to that Client go away, ~Layer is left with
the last one... then hell breaks loose as ~Client is called, which in turn
needs to acquire mStateLock.

We fix this by holding a temporary copy of the drawing state during
the transaction so that the side-effects of copying the current
state into the drawing state are seen only after mStateLock has
been released.

Bug: 9106453
Change-Id: Ic5348ac12283500ead87286a37565e8da35f1db2
2013-06-05 14:38:23 -07:00
Mathias Agopian
85cce376cb Basic binary event-log helpers
EventLog can be used from SurfaceFlinger to log jank
statistics events to the binary event log.

Change-Id: If0fbbe7d7111f10957697166d37fd0c3418109bb
2013-06-05 13:37:49 -07:00
Mathias Agopian
b2397fff6f Merge changes Id5584bc1,I8ca826d0
* changes:
  revert parts of dc5b63e40, which made gyro drift estimation unstable
  use gyro timestamp directly in fusion
2013-05-30 21:30:17 +00:00
Mathias Agopian
24f9bea8eb revert parts of dc5b63e40, which made gyro drift estimation unstable
initialize the system covariance matrix with non-zero
    values for the gyro-bias part. this improves the initial
    bias estimation speed significantly.

the initial covariance matrix should be small because the drift
changes slowly. the real problem is that we're not starting with
a good estimate of the drift, which this algorithm relies on.

so with this revert, it'll take a while for the drift to be estimated
but it won't be unstable.

Change-Id: Id5584bc114a2390d507643b2451b2650c1b90721
2013-05-30 14:26:36 -07:00
Mathias Agopian
2e2a560c4b use gyro timestamp directly in fusion
we used to estimate the gyro rate and deduce the period from that
but it turns out this is causing problems.

Bug: 5192288
Change-Id: I8ca826d0e11e488587bcaa1720de99e92b82f191
2013-05-30 14:18:23 -07:00
Jesse Hall
29c3f35279 Prevent opaque windows from making framebuffer translucent
To keep the code readable now that we have four different texenv
configurations, this change separates the decisions about what
configuration to use from the GL calls to set up the configuration.

Bug: 8963244
Change-Id: Ia07a306a7809ba8f93493d0160ccbd509e948581
2013-05-24 10:51:41 -07:00
Lajos Molnar
c5d7b7d323 BufferQueue: track buffer-queue by instance vs. by reference
Instead of representing the buffer-queue as a vector of buffer
indices, represent them as a vector of BufferItems (copies).
This allows modifying the buffer slots independent of the queued
buffers.

As part of this change, BufferSlot properties that are only
been relevant in the buffer-queue have been removed.

Also, invalid scalingMode in queueBuffer now returns an error.

ConsumerBase has also changed to allow reuse of the same
buffer slots by different buffers.

Change-Id: If2a698fa142b67c69ad41b8eaca6e127eb3ef75b
Signed-off-by: Lajos Molnar <lajos@google.com>
Related-to-bug: 7093648
2013-05-23 22:17:54 +00:00
Dianne Hackborn
80d7fd86e0 Power manager now takes calling package name arg.
Change-Id: Iae57b8eb7abe4e66f0d2645db8cdf57acf7dcd3e
2013-05-20 16:50:10 -07:00
Mathias Agopian
2381033ad0 am e6eef6c1: am ea74d3b7: make the warning timout of Fence::waitForever() implicit and longer
* commit 'e6eef6c136b007f3a7e1e19884875b855c0c5f95':
  make the warning timout of Fence::waitForever() implicit and longer
2013-05-17 10:18:45 -07:00
Mathias Agopian
e6eef6c136 am ea74d3b7: make the warning timout of Fence::waitForever() implicit and longer
* commit 'ea74d3b78d607cde17790a7bb83e6f68ffd34cfd':
  make the warning timout of Fence::waitForever() implicit and longer
2013-05-17 10:16:51 -07:00
Mathias Agopian
0319306670 various fixes to the sensorservice
1) "google" sensors are now reporting AOSP as the vendor string
2) don't expose the system's sensor fusion if the HAL provides it
3) use uncalibrated gyro if availble for the system's sensor fusion

Change-Id: I25140436cdb29d55e39fd6fbbf8c44a410a83d5c
2013-05-16 18:47:21 -07:00
Mathias Agopian
ea74d3b78d make the warning timout of Fence::waitForever() implicit and longer
- timeout is now 3 seconds instead of 1
- simplifies the API a bit
- allows us to change/tweak this timeout globaly

Bug: 8988871

Change-Id: I8d3c6ec43a372f602fb3f29856710339f86c0ec9
2013-05-16 18:15:57 -07:00
Jesse Hall
59eaeed5fa am 28fc78bf: am e8fed71d: Merge "Abort surface composition if hw surface is invalid"
* commit '28fc78bf5fbb6288a27f3f25565a960a1873b0ef':
  Abort surface composition if hw surface is invalid
2013-05-13 23:31:54 -07:00
Jesse Hall
28fc78bf5f am e8fed71d: Merge "Abort surface composition if hw surface is invalid"
* commit 'e8fed71da2d3e03ef933d308520bab5b860f05aa':
  Abort surface composition if hw surface is invalid
2013-05-13 23:28:45 -07:00
Michael Chock
c8c7109619 Abort surface composition if hw surface is invalid
If an invalid display device is detected during surface composition
(e.g., a simulated secondary display is removed), abort the
composition.

Change-Id: Ia6afb2e287882d8ae0614eb25463d3f85b687adf
2013-05-13 13:26:19 -07:00
Mathias Agopian
f3e85d4327 take the "transparent region" into account for blending
until now it was only used to discard a layer entirely.
we're now reducing the size of the layer if it is still
visible, if possible.

this works for instance when a surfaceView is used and
only the menu bar is displayed over it.

Change-Id: I3f5527c5cd1e69ecc968272c8948f1513ada8c55
2013-05-10 18:35:43 -07:00
Mathias Agopian
6c7f25afb7 reformat Rect.{cpp|h}
Change-Id: I45b1f6646541a1abacce1e70df00a770e47b820e
2013-05-10 18:03:31 -07:00
Mathias Agopian
b6df7d0e4c make all libsurfaceflinger and libsensorservice symbols's visibility hidden
we only export the main entry-point.
this saves about 150KB.

Change-Id: I55eb2b6705386fdfa43860deb62f9cdd2a0982aa
2013-05-09 16:51:37 -07:00
Jesse Hall
c0db61966a am dda42952: am 64f3b2fd: Merge "Remove experimental HWC virtual display support" into jb-mr2-dev
* commit 'dda4295221f831560ca9d554cd9b7be1c672d0b6':
  Remove experimental HWC virtual display support
2013-05-08 15:05:15 -07:00
Jesse Hall
e737c11cd8 Remove experimental HWC virtual display support
Bug: 8384764
Change-Id: I97b52ed83ad85466bd91cb9291308994048568a1
2013-05-07 11:58:47 -07:00
Mathias Agopian
0aea53ff3f use a Framebuffer Object to render all screenshots
this allows us to render into a buffer with a
pixelformat of our own choice; this is much faster
on all platform.

Bug: 8582615
Change-Id: I61298fc8e43fa6f92044c5123955cb5c7897dab7
2013-04-29 21:06:06 +02:00
Mathias Agopian
3e25fd8609 Add a --color option to dumpsys SurfaceFlinger
colorize a bit the output of dumpsys SurfaceFlinger to
make it easier to read. Right now it will bold the
title of each section and use green for the name of
each layer.

Change-Id: I0d9f18d115401cb45109d244ef3a278481f68cc6
2013-04-22 21:51:28 +02:00
Mathias Agopian
74d211ae26 clean-up/simplify all dump() APIs
remove the scratch buffer parameter and use
String8::appendFormat() instead.

Change-Id: Ib96c91617c8e7292de87433d15cf6232b7d591b0
2013-04-22 21:51:28 +02:00
Mathias Agopian
bd1153382e allow dumpsys SurfaceFlinger in user builds from the shell user
Bug: 8659013
Change-Id: I1de43bb5808cb8c1103cfc73dca59ff70fc13f91
2013-04-18 16:41:04 -07:00
Jaikumar Ganesh
4c01b1ad80 Fix activation issue with auto disabled sensors.
Auto disabled sensors get auto disabled after trigger.
An activation after this wasn't working because the
state was not being reset.

b/8609561

Change-Id: If72c9f27345e91671d7ad0a7a066f6dc3d255b78
2013-04-17 14:58:22 -07:00
Mathias Agopian
48be323236 Merge "make sure to update the "current state" when we "fake" the transaction" into jb-mr2-dev 2013-04-13 00:27:32 +00:00
Jaikumar Ganesh
2ea0c67da6 Merge "Sensor changes." into jb-mr2-dev 2013-04-13 00:22:39 +00:00
Mathias Agopian
6c67f0fe45 make sure to update the "current state" when we "fake" the transaction
this mimics the code we used for the size. basically we need to
also update the "current state" because it gets copied to the
"drawing state" when a transaction occurs, and it would
"undo" our change.

Bug: 8511430
Bug: 8581533
Change-Id: I08c02abbf21b7f168f7124cd14ee717d7d3d502c
2013-04-12 16:58:11 -07:00
Jaikumar Ganesh
4342fdf14f Sensor changes.
1. Some sensors can wake up the AP. Add wakelocks.
2. Handle backward compatibility for rotation vector
   heading accuracy.
3. Cleanup auto disabled sensors.
4. Fix race condition between enable and dispatch.

Change-Id: I39dddf12e208d83cd288201986ee994312555820
2013-04-12 09:34:44 -07:00
Chet Haase
5e5bed60d0 Merge "DO NOT MERGE Make sure surfaces always have latest orientation info" into jb-mr2-dev 2013-04-12 14:11:54 +00:00
Mathias Agopian
180f10de6f Improve screenshot performance on some devices (DO NOT MERGE)
this affects devices that need a glReadPixels(). We use
a FBO instead of a GlConsumer as an intermediate render target, this
saves 2 calls to eglMakeCurrent().

On Galaxy Nexus this allows us to go from ~135ms to ~35ms for
recent's screenshots.

Bug: 8582615
Change-Id: I6b25291ecc235f1927579bbb2db3c731e985c6e8
2013-04-11 17:35:24 -07:00
Chet Haase
91d25932b6 DO NOT MERGE Make sure surfaces always have latest orientation info
When the screen is turned off, the current stack is set to -1. This causes
logic in iSurfaceFlinger's handleTransactionLocked() function to fail to
match the current stack, and the latest orientation is not set into the
layer. This causes BufferQueue, later, to potentially set an obsolete
transformHint on a created surface (such as in the case with ImageWallpaper's
Egl surface, in the bug below).

The fix is to note this situation and use a default value for the DisplayDevice,
which should have the current orientation information.

Issue #8508397 ImageWallpaper sometimes rendered in wrong orientation causing a ~30-40% drop in graphics performance

Change-Id: Ibae15d73b289a8343c67f4f6bb77fdf11dd95ee7
2013-04-11 17:06:38 -07:00
Mathias Agopian
d19e519f67 Merge "fix another bug where screenshots could end-up all black" into jb-mr2-dev 2013-04-11 04:43:49 +00:00
Mathias Agopian
9eb1f0558b fix another bug where screenshots could end-up all black
SF transactions were always handled on VSYNC which allowed
the screenshot to sneak-in between closing the transaction
and vsync (before it's latched), resulting in a screenshot
with the previous state.

we now always force transactions to happen immediately
before screenhots.

Bug: 7552304
Change-Id: I0afc86b7e8366173daff5b9988bbb4d2a0f43860
2013-04-10 16:27:17 -07:00
Ying Wang
8a0cb4ee0b Add liblog
Bug: 8580410
Change-Id: I2ca4fa00dde0b574df3619134836bcb9315bd506
2013-04-09 21:55:39 -07:00
Jesse Hall
1df59c93fe Merge "Tell HWComposer the dimensions of virtual displays" into jb-mr2-dev 2013-04-06 03:32:19 +00:00
Jesse Hall
d5e345b0cf Merge "Actually set the virtual display output buffer." into jb-mr2-dev 2013-04-06 03:31:49 +00:00