OSDN Git Service
Elliott Hughes [Tue, 28 Mar 2017 23:58:18 +0000 (23:58 +0000)]
Merge "Remove references to property size limits." am:
8f8daca7ab am:
75500cc6fd
am:
624f84956b
Change-Id: Ibddf32a6fbf37988755d08eef31135749af29089
Elliott Hughes [Tue, 28 Mar 2017 23:56:46 +0000 (23:56 +0000)]
Merge "Remove references to property size limits." am:
8f8daca7ab
am:
75500cc6fd
Change-Id: Idb75e757f75f4f390a4ef596f37bbcc7c854e73e
Elliott Hughes [Tue, 28 Mar 2017 23:55:17 +0000 (23:55 +0000)]
Merge "Remove references to property size limits."
am:
8f8daca7ab
Change-Id: I146d719f51f6337abe13fc96b7e99fd7c4aebdcf
Treehugger Robot [Tue, 28 Mar 2017 23:49:51 +0000 (23:49 +0000)]
Merge "Remove references to property size limits."
Elliott Hughes [Tue, 28 Mar 2017 21:55:31 +0000 (14:55 -0700)]
Remove references to property size limits.
Bug: http://b/
33926793
Test: "python systrace.py"
Change-Id: Ia11c160935b8601bc9c4d867abc57ccf37f84082
Jesse Hall [Tue, 28 Mar 2017 21:04:26 +0000 (21:04 +0000)]
Merge "Use gralloc usage conversion library" into oc-dev
am:
d787577ebf
Change-Id: I54219ad459b59ecd923814cf4ae7ea42dc6145c8
TreeHugger Robot [Tue, 28 Mar 2017 20:55:50 +0000 (20:55 +0000)]
Merge "Use gralloc usage conversion library" into oc-dev
Felipe Leme [Tue, 28 Mar 2017 20:12:31 +0000 (20:12 +0000)]
Temporarily skipping dump of bluetooth logs.
am:
3ae337c4a1
Change-Id: I93aa1aace4d2c23cd4de71455cb50bce4c0773ce
Felipe Leme [Tue, 28 Mar 2017 18:54:20 +0000 (11:54 -0700)]
Temporarily skipping dump of bluetooth logs.
Bug:
35998031
Test: manual verification
Change-Id: I0bdf6eb1546c0850b301441f48753b65fa7b1e5e
neo.he [Tue, 28 Mar 2017 18:42:58 +0000 (18:42 +0000)]
Merge changes If756ad2a,I19aeb7ac am:
15feba03ae am:
47609db709
am:
e59f1b0bbe
Change-Id: Ic743e74ef2560e1657c8f73750dc747e91d336f2
neo.he [Tue, 28 Mar 2017 18:38:27 +0000 (18:38 +0000)]
Merge changes If756ad2a,I19aeb7ac am:
15feba03ae
am:
47609db709
Change-Id: I5a7ceb58874d4432d51342d275610dc46a539a61
neo.he [Tue, 28 Mar 2017 18:34:28 +0000 (18:34 +0000)]
Merge changes If756ad2a,I19aeb7ac
am:
15feba03ae
Change-Id: Ib1842ab89d245fb5bde291a26c4952e0fc262aba
Treehugger Robot [Tue, 28 Mar 2017 18:25:37 +0000 (18:25 +0000)]
Merge changes If756ad2a,I19aeb7ac
* changes:
Add alpha channel for the EGLConfig
Change the conditions of layer as translucent
Jesse Hall [Thu, 23 Mar 2017 18:03:23 +0000 (11:03 -0700)]
Use gralloc usage conversion library
Converting from ..graphics.allocator@2.0 usage to gralloc0 usage isn't
quite as simple as ORing and truncating, which is what the default
implementation was doing. Switch to using library functions that do it
correctly.
Test: boot bullhead
Change-Id: Ibe644b53655e191b9bca9d619825a8918e8fc365
TreeHugger Robot [Tue, 28 Mar 2017 05:41:04 +0000 (05:41 +0000)]
Merge "Remove legacy cache clearing logic."
TreeHugger Robot [Tue, 28 Mar 2017 02:39:12 +0000 (02:39 +0000)]
Merge "Add stub for SensorManager::createEventQueue"
Jaesoo Lee [Fri, 24 Mar 2017 04:52:12 +0000 (13:52 +0900)]
fixed bug undefining LIKELY/UNLIKELY in math lib
Bug:
36275627
Test: Built and check the surfaceflinger compiled
Change-Id: I062b8c0db540443635ac746824edb295190235d0
TreeHugger Robot [Mon, 27 Mar 2017 23:16:23 +0000 (23:16 +0000)]
Merge "Disable BufferQueueInAnotherProcess test"
Jeff Sharkey [Mon, 27 Mar 2017 22:41:06 +0000 (16:41 -0600)]
Remove legacy cache clearing logic.
It was riddled with funky memory access, and it's been fully replaced
by the new V2 logic.
Test: builds, boots
Bug:
27948817,
36655947
Change-Id: Ic5f0636c414ccf5ccf56eb50fea321b6688a83a0
Mathias Agopian [Mon, 27 Mar 2017 22:03:23 +0000 (22:03 +0000)]
Merge changes Iad7bf035,I870e08c6
* changes:
VNDK ANativeWindow API - step 2
VNDK ANativeWindow API
Kalle Raita [Mon, 27 Mar 2017 21:11:54 +0000 (14:11 -0700)]
Disable BufferQueueInAnotherProcess test
The BufferQueueInAnotherProcess relies on executing before any test that
tries to open a BufferQueue in the local process. A recent test addition
broke the condition. Upcoming SF & BQ changes are likely to obsolete the
test, so won't spend time on fixing it. Disabling for now and the test
will be removed when the infrastructure changes land.
Test: libgui_test
Bug:
36592665
Change-Id: I648e197cab6bab3aed445390069265589a717bf1
Andreas Huber [Mon, 27 Mar 2017 21:12:41 +0000 (21:12 +0000)]
Merge "Enable --debug option for "lshal" command." am:
16175ca725 am:
674d3acd18
am:
829f3c6d85
Change-Id: I6751fec855783983b96672bec82ae9ba0c737a8b
Andreas Huber [Mon, 27 Mar 2017 21:10:09 +0000 (21:10 +0000)]
Merge "Enable --debug option for "lshal" command." am:
16175ca725
am:
674d3acd18
Change-Id: If87550eb68a9c8e280d8863063b4311a44ff6896
Andreas Huber [Mon, 27 Mar 2017 21:07:40 +0000 (21:07 +0000)]
Merge "Enable --debug option for "lshal" command."
am:
16175ca725
Change-Id: Ifa29afcb7677a3ec76a03d5f0365ba4de7ee4c9b
Andreas Huber [Mon, 27 Mar 2017 21:02:30 +0000 (21:02 +0000)]
Merge "Enable --debug option for "lshal" command."
Dan Stoza [Mon, 27 Mar 2017 19:34:43 +0000 (19:34 +0000)]
Merge "libgfx: Remove SafeLog"
Jiwen Cai [Mon, 27 Mar 2017 19:33:59 +0000 (19:33 +0000)]
Merge changes from topic 'libdvr'
* changes:
Add dvr_buffer_queue C API
Move DVR platform C API into libdvr
Mathias Agopian [Tue, 21 Mar 2017 01:07:26 +0000 (18:07 -0700)]
VNDK ANativeWindow API - step 2
The vndk API is a superset of the NDK API.
Prior to this, vendor would have access to
system/window.h which itself is a superset of
what they need, exposes too much of the internals.
With this change, system/window.h depends on
vndk/window.h which depends on
android/native_window.h
Test: compiled, booted Nexus 6P
Bug:
34453351
Change-Id: Iad7bf035143541843309437fe630cc603d96a76e
Mathias Agopian [Mon, 20 Mar 2017 22:48:44 +0000 (15:48 -0700)]
VNDK ANativeWindow API
The vndk API is a superset of the NDK API.
Prior to this, vendor would have access to
system/window.h which itself is a superset of
what they need, exposes too much of the internals.
With this change, system/window.h depends on
vndk/window.h which depends on
android/native_window.h
Test: compiled, booted Nexus 6P
Bug:
34453351
Change-Id: I870e08c652d90b3f2cf4ed7c7ca541ce2893f689
TreeHugger Robot [Mon, 27 Mar 2017 17:15:04 +0000 (17:15 +0000)]
Merge "VR: Rename HIDL service name from vr_hwcomposer to vr"
TreeHugger Robot [Mon, 27 Mar 2017 14:00:29 +0000 (14:00 +0000)]
Merge "Properly recover vr_wm from surface vr_flinger crashes."
TreeHugger Robot [Mon, 27 Mar 2017 09:02:31 +0000 (09:02 +0000)]
Merge "Update native ColorSpace API for libhwui"
Romain Guy [Mon, 27 Mar 2017 07:25:19 +0000 (00:25 -0700)]
Update native ColorSpace API for libhwui
Test: colorspace_test
Bug:
32984164
Change-Id: Id0728bfa90192173eb45c2d26a8b81ce970398d7
Alex Sakhartchouk [Sun, 26 Mar 2017 02:22:16 +0000 (22:22 -0400)]
Properly recover vr_wm from surface vr_flinger crashes.
This is addressing a number of issues:
1) surface_flinger_view_ was incorrectly reset
on de-allocation, which renders vr_wm incapable of seeing
new frames after a crash.
2) dvrGraphicsWaitNextFrame will return -1 if vr_flinger
crashes, which means our graphics context is no longer valid.
We need to de-allocate resources and wait for SF to
reconnect.
3) Displays needed be be de-allocated on SF detach because
they will return incorrect fences on SF re-connect.
4) display_client_ needs to be allocated on demand instead
of at initialization time because vr_flinger crash
invalidates it.
Bug:
36589793
Test: Crash vr_flinger, observe vr_wm working after SF restarts.
Change-Id: I72a79690ee5224e3326c7c571ae4e23797103e6d
Howard Chen [Sat, 25 Mar 2017 09:12:59 +0000 (17:12 +0800)]
This test is used to check if the kernel scheduler
1. supports priority inheritance.
2. utilize the sync flag.
It also make statistic on average, best and worst latency
as well as number of deadline miss with a configurable
deadline specified in us.
The output is in json format hopefully to make it easy
to integrated with present test framework.
usage:
./schd-dbg # test and output json
./schd-dbg -i 1 # number of iterations
./schd-dbg -pair 4 # number of process pairs
./schd-dbg -deadline_us 2500 # deadline in us
./schd-dbg -v # debug
Test: pixel
Change-Id: Idccf0d402dcd7b81c639020364b5ee32075e18cb
John Muir [Fri, 24 Mar 2017 23:32:44 +0000 (23:32 +0000)]
Merge "libdvrcommon: Remove revision code."
John Muir [Fri, 24 Mar 2017 21:43:39 +0000 (14:43 -0700)]
libdvrcommon: Remove revision code.
Move revision to libdvrplatform.
Bug:
36040117
Test: N/A.
Change-Id: I3f70509f9a694383bd7eb94173d5230bd8c66e51
Peng Xu [Fri, 17 Mar 2017 19:20:27 +0000 (12:20 -0700)]
Checks hidl service and remove hidl call at dump
Two things:
* check for hidl service availability before calling any hidl
functions.
* remove hidl dependency on dump to ensure dump will not cause
crash due to hidl failure.
Bug:
36073404
Test: check sensor running after reboot
Test: check dumpsys sensorservice works normally.
Change-Id: I8626635a143a9f07f263f85d95ab16749ea1e452
Alex Vakulenko [Fri, 24 Mar 2017 20:16:20 +0000 (20:16 +0000)]
Merge "Revert "Quick fix to correctly handle empty file/channel handles.""
Andreas Huber [Fri, 24 Mar 2017 20:14:11 +0000 (13:14 -0700)]
Enable --debug option for "lshal" command.
This calls into live binderized services and logs their "IBase::debug(...)"
output.
Bug:
36520620
Test: make, adb shell lshal, adb shell bugreport
Change-Id: Ida3b0c61eafd9f3de562a5df3af24bbaf383de22
Daniel Nicoara [Fri, 24 Mar 2017 19:15:50 +0000 (15:15 -0400)]
VR: Rename HIDL service name from vr_hwcomposer to vr
This is part of an effort to simplify service names.
Bug:
36564122
Test: Ran on device and verified SurfaceFlinger connects to vr_wm
Change-Id: I3e2a2b778f3f0d2066bf5b756263de4b238b2f83
Jiwen 'Steve' Cai [Thu, 23 Mar 2017 00:26:00 +0000 (17:26 -0700)]
Add dvr_buffer_queue C API
Bug:
36401767
Bug:
32213274
Bug:
36266201
Test: build and ran dvr_buffer_queue-test
Change-Id: Icd8d063bd8b7adb15bf67f7e543b87dad1c32a68
Jiwen 'Steve' Cai [Wed, 22 Mar 2017 23:59:53 +0000 (16:59 -0700)]
Move DVR platform C API into libdvr
Introduce a new static library build with Android.mk. The reasoning
behind this is:
1/ This library is going to consolidate all C headers and implementation
of the DVR platform library.
2/ No other internal system components need to depend on this library.
3/ This library is build with Android.mk so that it has depend on other
system components (such as libandroid_runtime) without worrying about
that Android.bp-based libs cannot depent and Androib.mk-based libraries.
4/ This library defines dvr_api.h header as the official definition of
DVR platform library headers. The header is moved from
'vendor/unbundled_google/packages/PrebuiltGoogleVr', which is not an
ideal place to hold the platform library anyway.
5/ If we ever decide to move these C wrappers out of frameworks/native,
this makes it easier to do so by bundling things now.
TODO: C APIs from libvrsensor and libvirtualtouchpadclient are not moved
into this bundle as some functions are still used by internal system
components. Ideally, these will be moved either to Google3 or vendor projects.
Test: Built and flash
Bug:
36563654
Change-Id: Ie5ce89faa70f86ac934a77d8f7d4fca7e94112ee
Alex Vakulenko [Fri, 24 Mar 2017 15:53:50 +0000 (15:53 +0000)]
Revert "Quick fix to correctly handle empty file/channel handles."
This reverts commit
e71fb7d2ab1118d30c5bec8a85114ac610844dc6.
This CL broke PDX unit tests
Bug:
36587839
Change-Id: I5b3cb857aacff337cb9a268a0415f8398303abe8
TreeHugger Robot [Fri, 24 Mar 2017 03:33:26 +0000 (03:33 +0000)]
Merge "media: mechanism to set/get vendor-extension configs for OMX component"
TreeHugger Robot [Fri, 24 Mar 2017 02:39:10 +0000 (02:39 +0000)]
Merge "Remove more EGL_ANDROID_create_native_client_buffer references"
Craig Donner [Fri, 24 Mar 2017 00:55:35 +0000 (17:55 -0700)]
Remove more EGL_ANDROID_create_native_client_buffer references
Bug:
35101209
Test: build
Change-Id: I8709b9881f0133756e2cfcfad3819d4032cfb03e
Yifan Hong [Thu, 23 Mar 2017 23:50:55 +0000 (16:50 -0700)]
Add stub for SensorManager::createEventQueue
It is not yet implemented.
Test: pass
Bug:
35219747
Change-Id: I8ef1627376ac88f552c20a2809416115ba342b82
TreeHugger Robot [Thu, 23 Mar 2017 23:26:14 +0000 (23:26 +0000)]
Merge "Implement an API for vrwm to query if the current vr app is active"
Tim Murray [Thu, 23 Mar 2017 23:09:08 +0000 (23:09 +0000)]
binder: remove cgroup interactions [ DO NOT MERGE ] am:
05f4679d03
am:
400e9a72fd -s ours
Change-Id: Ic31695b9c7a981341ab601638b68f0bb7ef99f28
Tim Murray [Thu, 23 Mar 2017 23:02:25 +0000 (23:02 +0000)]
binder: remove cgroup interactions [ DO NOT MERGE ]
am:
05f4679d03
Change-Id: Idba2f747151fc50866fb470681e50ba64b28b2d0
TreeHugger Robot [Thu, 23 Mar 2017 22:54:01 +0000 (22:54 +0000)]
Merge "Revert "atrace: Make atrace notify VR services of new trace sessions""
Fabien Sanglard [Thu, 23 Mar 2017 22:49:40 +0000 (22:49 +0000)]
Merge "Fix typo in Region to avoid unexpected values" am:
2cbfb0cda6 am:
7fc294e10e
am:
5288e22d00
Change-Id: Ib8095bd01becfacbbcecd0f4ec369ed933eff173
TreeHugger Robot [Thu, 23 Mar 2017 22:47:30 +0000 (22:47 +0000)]
Merge "Dumpstate: dump HAL native traces processes."
Fabien Sanglard [Thu, 23 Mar 2017 22:43:34 +0000 (22:43 +0000)]
Merge "Fix typo in Region to avoid unexpected values" am:
2cbfb0cda6
am:
7fc294e10e
Change-Id: I7feff39e1be80c69a744c4eefad835ba50796f4c
Fabien Sanglard [Thu, 23 Mar 2017 22:38:03 +0000 (22:38 +0000)]
Merge "Fix typo in Region to avoid unexpected values"
am:
2cbfb0cda6
Change-Id: I92a204fa492374c4c69a5d794d974b3cb8f454b3
Treehugger Robot [Thu, 23 Mar 2017 22:26:44 +0000 (22:26 +0000)]
Merge "Fix typo in Region to avoid unexpected values"
Albert Chaulk [Mon, 20 Mar 2017 17:03:39 +0000 (13:03 -0400)]
Implement an API for vrwm to query if the current vr app is active
The API returns if the client_order 0 layer is visible or not. This
aids in 2D app detection
Bug: None
Test: Manual with permissionsgen & calculator
Change-Id: Id988e61a9f93885e93c28ed83ffaffd84f3bdf15
John Reck [Thu, 23 Mar 2017 20:58:29 +0000 (20:58 +0000)]
Revert "atrace: Make atrace notify VR services of new trace sessions"
This reverts commit
213feb32c76bb4ed4262dad8c71f0377ee710fac.
Reason for revert: tracing & perf teams don't want to support PDX at this point in time.
Change-Id: I28b8a7c55c9d31c8469236fb60a6e034b1ef8af7
Mark Urbanus [Thu, 23 Mar 2017 20:58:06 +0000 (20:58 +0000)]
Merge "Enable Universal Bandwidth Compression for textures"
Mark Urbanus [Wed, 22 Mar 2017 20:11:51 +0000 (13:11 -0700)]
Enable Universal Bandwidth Compression for textures
We removed texture compression for VR display surfaces, which introduced memory
bandwidth issues on some QC platforms.
This patch reintroduces texture compression for QC devices compatible with
both gralloc0/gralloc1 implementations.
Bug:
36370337
Test: Validate VR comes up
Change-Id: Ieaaa56f67522c85913476ed0d8da04f8eedbce63
Steven Moreland [Wed, 22 Mar 2017 01:38:05 +0000 (18:38 -0700)]
Dumpstate: dump HAL native traces processes.
Test: hal service dumps in data/anr/bugreport-traces.txt after bugreport
Bug:
36414311
Change-Id: Ifca491d02f77bf4fb26450f06c2ea02769475786
Lajos Molnar [Fri, 17 Mar 2017 14:49:45 +0000 (07:49 -0700)]
media: mechanism to set/get vendor-extension configs for OMX component
Add OMX_CONFIG_ANDROID_VENDOR_EXTENSIONTYPE and
OMX_CONFIG_ANDROID_VENDOR_PARAMTYPE.
Android vendor extensions for OMX components have a name and one or
more parameter values - each with a string key - that are set together.
Support for not-set values is provided for vendor extensions in order
to not suddenly confuse applications with format keys that they were
not expecting.
Bug:
32746650
Change-Id: I21aad4a82d509fe2466764f65ee9e3b2d32f0842
TreeHugger Robot [Thu, 23 Mar 2017 18:01:59 +0000 (18:01 +0000)]
Merge "Re-add access of DisplayClient from vr_rm"
Stephen Kiazyk [Wed, 22 Mar 2017 20:27:43 +0000 (13:27 -0700)]
Re-add access of DisplayClient from vr_rm
This was removed in a previous commit because it relied on an in-flight
CL that had not yet landed. Now that it has landed, the change is back
in.
Bug:
36139334
Test: Compile and run VR app, confirm that touch region is correct
through hierarchy viewer.
Change-Id: I4ecc5e050fd896ba7442eac8b3a4e0b793915ed7
Tim Murray [Tue, 7 Mar 2017 17:31:09 +0000 (09:31 -0800)]
binder: remove cgroup interactions [ DO NOT MERGE ]
Having binder change cgroups on the fly is too expensive in the common case to be useful.
Test: builds
bug
34825105
Change-Id: Ia692c9d2ab261d852cb2fac8f27d6779d0e2b057
Courtney Goeltzenleuchter [Thu, 23 Mar 2017 16:04:18 +0000 (16:04 +0000)]
Merge "Check board config for wide-color color spaces"
TreeHugger Robot [Thu, 23 Mar 2017 16:04:02 +0000 (16:04 +0000)]
Merge "VR: Add dvr API for service apps connecting to VR HWC"
Courtney Goeltzenleuchter [Fri, 17 Mar 2017 22:16:48 +0000 (16:16 -0600)]
Check board config for wide-color color spaces
Check that board config allows wide-color color
spaces.
Test: make tests in libs/gui/tests/
Test: adb sync
Test: adb shell /data/nativetest/libgui_test/libgui_test --gtest_filter="SurfaceTest.GetWideColorSupport"
bug #
35996305
Change-Id: Ia5b8116bc25befb6491e72e67cd8ed3c90ca5698
TreeHugger Robot [Thu, 23 Mar 2017 15:44:14 +0000 (15:44 +0000)]
Merge "Quick fix to correctly handle empty file/channel handles."
TreeHugger Robot [Thu, 23 Mar 2017 15:42:22 +0000 (15:42 +0000)]
Merge "libpdx: Remove superfluous DispatchRemoteMethod overloads."
Naveen Kalla [Fri, 17 Mar 2017 01:17:25 +0000 (18:17 -0700)]
Add notification_title and notification_description to dumpstate
Initiator of dumpstate can set dumpstate.options.title and
dumpstate.options.description properties to notify dumpstate the details
that will be used to file a bug. dumpstate will send this information
along with the BUGREPORT_FINISHED intent.
Bug:
33561517
Test: Use the new API to invoke bugreport from TelephonyMonitor and
ensure that the bugreport is taken and title and description are
updated properly in the notification. Run dumpstate and
BugreportReceiver unit tests
Change-Id: Ia23a3ef4d5751be14f347ac38ef7c8a18f79799d
TreeHugger Robot [Thu, 23 Mar 2017 06:04:44 +0000 (06:04 +0000)]
Merge "Follow "atomic" to "group" refactoring."
Steven Moreland [Thu, 23 Mar 2017 04:00:59 +0000 (04:00 +0000)]
Merge "lshal: --init-vintf exclude system hals" am:
8db626d2f3 am:
34206360f2
am:
d1b6b40adb
Change-Id: Ie2d28a374a3fb18d969783394b80b3a0d6fe0869
Steven Moreland [Thu, 23 Mar 2017 03:58:27 +0000 (03:58 +0000)]
Merge "lshal: --init-vintf exclude system hals" am:
8db626d2f3
am:
34206360f2
Change-Id: I21d0b37f6942204c748017e03e8c9a0e892a39dc
Steven Moreland [Thu, 23 Mar 2017 03:55:27 +0000 (03:55 +0000)]
Merge "lshal: --init-vintf exclude system hals"
am:
8db626d2f3
Change-Id: Ibe4636eb13c9439d09f51948ca49921a1e209747
Treehugger Robot [Thu, 23 Mar 2017 03:47:57 +0000 (03:47 +0000)]
Merge "lshal: --init-vintf exclude system hals"
TreeHugger Robot [Thu, 23 Mar 2017 03:34:08 +0000 (03:34 +0000)]
Merge "Fix BlockUntilVSync() to actually block"
Jeff Sharkey [Thu, 23 Mar 2017 03:27:40 +0000 (21:27 -0600)]
Follow "atomic" to "group" refactoring.
Remove noisy logging about UIDs that are relying on default cache
quota of 64MiB.
Move away from yucky old statfs() and use statvfs() instead.
Test: /data/nativetest/installd_cache_test/installd_cache_test
Bug:
35812899,
35684969,
36482620
Change-Id: I3d68da97eac2ebcda489bdf9d27061cac5b3f7cc
Rob Carr [Thu, 23 Mar 2017 02:35:00 +0000 (02:35 +0000)]
Merge "SurfaceFlinger: Missed changes in HWC1. Fix build error."
Steven Moreland [Wed, 22 Mar 2017 21:22:55 +0000 (14:22 -0700)]
lshal: --init-vintf exclude system hals
--init-vintf is provided for convenience so that vendors can initialize
their vendor interface. Excluding system hals so that they aren't
included in the default vintf file that is generated. They are
maintained manually.
Test: Verified frameworks hals are no longer visible in init-vintf.
Bug:
34772739
Change-Id: Ie40c384334affcb3abee063d58000b808f04e234
Steven Thomas [Thu, 23 Mar 2017 01:45:31 +0000 (18:45 -0700)]
Fix BlockUntilVSync() to actually block
In my refactoring in ag/
2008740 I introduced a regression where blocking
until the next vsync would busy loop instead of actually blocking on an
event on the vsync fd. The problem was I changed the poll() call to
listen for POLLIN in addition to POLLPRI. While waiting for vsync events
we only want to listen for POLLPRI.
Bug:
36495351
Test: Confirmed with debug logs the busy loop bug and that we correctly
block with this patch applied.
Change-Id: Iec4951e014575a5d0ed3cfe3fc20ea91c67edf1b
Robert Carr [Thu, 23 Mar 2017 01:23:23 +0000 (18:23 -0700)]
SurfaceFlinger: Missed changes in HWC1. Fix build error.
Apparently HWC1 isn't in presubmit anymore.
Test: Manual
Change-Id: Id173d71be62b8a3c888943c57c0ae0ec511bc94b
Dmitry Shmidt [Thu, 23 Mar 2017 01:22:17 +0000 (01:22 +0000)]
Merge "dumpstate: Remove FWDUMP_bcmdhd support" am:
86388dbc9a am:
2d71f97dfb
am:
b09149f6ec -s ours
Change-Id: Iba7447b4f5eb32cbdd480b69fcc369d1ee1a3e3f
Adam Lesinski [Thu, 23 Mar 2017 01:12:09 +0000 (01:12 +0000)]
Merge "batteryservice: add scheduleUpdate binder method" am:
e007716bc4 am:
799bb3501a
am:
cd1d8c6b10
Change-Id: I6fcf1788701a7b3c9e481a2bdfbdd352cf24d598
Dmitry Shmidt [Thu, 23 Mar 2017 01:09:37 +0000 (01:09 +0000)]
Merge "dumpstate: Remove FWDUMP_bcmdhd support" am:
86388dbc9a
am:
2d71f97dfb
Change-Id: I1e3a7ae0e7855a8003f463586d368a81ad89be5f
Adam Lesinski [Thu, 23 Mar 2017 01:03:08 +0000 (01:03 +0000)]
Merge "batteryservice: add scheduleUpdate binder method" am:
e007716bc4
am:
799bb3501a
Change-Id: Ia1ef3d769c453bc58cc52b93df5d6f03e601d23e
Dmitry Shmidt [Thu, 23 Mar 2017 00:52:59 +0000 (00:52 +0000)]
Merge "dumpstate: Remove FWDUMP_bcmdhd support"
am:
86388dbc9a
Change-Id: Iadf37acb012f762919567354a7aede95f81b9eff
Adam Lesinski [Thu, 23 Mar 2017 00:48:42 +0000 (00:48 +0000)]
Merge "batteryservice: add scheduleUpdate binder method"
am:
e007716bc4
Change-Id: I8565c0502c3838d74fcd09ec75677e65cf85a0a9
Dmitry Shmidt [Thu, 23 Mar 2017 00:47:47 +0000 (00:47 +0000)]
Merge "dumpstate: Remove FWDUMP_bcmdhd support"
Treehugger Robot [Wed, 22 Mar 2017 23:53:59 +0000 (23:53 +0000)]
Merge "batteryservice: add scheduleUpdate binder method"
TreeHugger Robot [Wed, 22 Mar 2017 23:51:07 +0000 (23:51 +0000)]
Merge "BQ: Resend queued buffer handles after disconnect"
Rob Carr [Wed, 22 Mar 2017 23:44:12 +0000 (23:44 +0000)]
Merge changes Ic8309518,If3199d7c,Ie4d32162
* changes:
SurfaceFlinger: Inherit Alpha for child surfaces.
Add SurfaceFlinger test filter for presubmit.
Include finalCrop in setGeometryAppliesWithResize.
TreeHugger Robot [Wed, 22 Mar 2017 23:28:46 +0000 (23:28 +0000)]
Merge "vrwm: Add 2D Mode + control app change"
TreeHugger Robot [Wed, 22 Mar 2017 22:57:38 +0000 (22:57 +0000)]
Merge "dumpstate: Remove FWDUMP_bcmdhd support"
Martijn Coenen [Wed, 22 Mar 2017 22:14:05 +0000 (22:14 +0000)]
Merge changes from topic 'vendorbinder'
* changes:
Add vndservicemanager.
Allow ProcessState to use another binder dev node.
TreeHugger Robot [Wed, 22 Mar 2017 21:54:19 +0000 (21:54 +0000)]
Merge "Reduce vr_wm logging."
Albert Chaulk [Tue, 21 Mar 2017 21:27:53 +0000 (17:27 -0400)]
vrwm: Add 2D Mode + control app change
2D mode is similar to debug mode in that it will always show the
application layer if it exists, but unlike debug mode will strip any
system layers from the output and will attempt to recognize the VR
layers. The virtual display is always in 2D mode as it can only show
2D apps. The primary display can be switched by vr_wm_ctl
Bug: None
Test: Manual with permissionsgen and launching calculator app
Change-Id: I38d025c457b1fca9aeb47cb94ed481ac9ff084f9
Dan Stoza [Tue, 21 Mar 2017 20:43:22 +0000 (13:43 -0700)]
BQ: Resend queued buffer handles after disconnect
When the producer disconnects from a BufferQueue, all of its slots are
cleared and buffers in the queue are marked as stale. When the slots
are cleared, we notify the consumer that buffers were released so that
it can clear its cache of slot-to-buffer mappings.
However, when we were marking queued buffers as stale, we were not
clearing the bit that dictates whether the BQ needs to send a buffer
handle upon acquire, so in the case where the producer queues a buffer
and disconnects, and then the consumer acquires that queued buffer,
the consumer doesn't get the buffer handle, nor can it retrieve it from
its cache, because it has (or should have) already cleared it.
This change clears that bit upon disconnect.
Bug:
36453246
Test: New test in libgui_tests + Vulkan CTS test from bug
Change-Id: Ic27f119a4164878257c3841b4e1d421694159774
Robert Carr [Tue, 21 Mar 2017 17:41:29 +0000 (10:41 -0700)]
SurfaceFlinger: Inherit Alpha for child surfaces.
Necessary for various animations to work correctly and
for the general goal of treating child surfaces like
pixels in the parent surface.
Test: Includes new test in SurfaceFlinger_test
Bug: None yet ;)
Change-Id: Ic8309518bd327bbba6ad2f79f265e24cf37898e7