OSDN Git Service
smain@google.com [Fri, 8 Jul 2016 00:11:35 +0000 (17:11 -0700)]
update NDK doxygen templates to create DevSite-style HTML.
Previously, we had doxygen create .jd files with some syntax that
was parsed by the DocLava build tools when creating the final HTML
files for the website. But now that we're publishing these docs through
Piper, we don't need to run them through DocLava anymore. So with this
change, we now create .html files that then need to be copied to
google3/googledata/devsite/site-android/en/ndk/
where they can be immediately published via devsite once merged.
bug:
29943712
Change-Id: I8af5555f802e8c0122378bd504ad4308ff9aad56
Pablo Ceballos [Tue, 28 Jun 2016 00:56:53 +0000 (17:56 -0700)]
BQ: Add and expose a unique id
Bug
29422927
Change-Id: I80eab94f073ebc378302f00fa86a740c3643657e
Mark Salyzyn [Fri, 24 Jun 2016 21:06:15 +0000 (14:06 -0700)]
dumpstate: Also pick up /sys/fs/pstore/console-ramoops-0 for LAST KMSG
(cherry pick from commit
7d0a762ecaf9f4d005f0f6db913034c2e084d362)
Bug:
29639467
Change-Id: Iba0b1754eaebfb2b0624dcbef522b7d8456eeb50
Greg Kaiser [Thu, 23 Jun 2016 03:25:44 +0000 (03:25 +0000)]
Merge "Sensor: Make getId() more varied" into nyc-dev
Greg Kaiser [Tue, 21 Jun 2016 23:11:14 +0000 (16:11 -0700)]
Sensor: Make getId() more varied
Apps from different developers will now receive a different
ID for the same dynamic sensor. Additionally, all apps
will now receive a different/new ID for the same dynamic
sensor after a factory reset.
Bug:
28775590,
29547335
Change-Id: I368655b0e4bdc16736a7fd163ea3f7dad2886b3b
TreeHugger Robot [Wed, 22 Jun 2016 20:24:53 +0000 (20:24 +0000)]
Merge "Run iptables as root." into nyc-dev
Felipe Leme [Sat, 18 Jun 2016 00:37:13 +0000 (17:37 -0700)]
Run iptables as root.
iptables requires root access and su is not available on user builds,
hence it must be ran before dropping uid to shell user.
BUG:
29455997
Change-Id: I10c6337b411875b061ae9556b6e9c44f83301ae1
Greg Kaiser [Mon, 20 Jun 2016 16:24:51 +0000 (16:24 +0000)]
Merge "SensorList: Don't output UUIDs" into nyc-dev
Eino-Ville Talvala [Sat, 18 Jun 2016 02:10:11 +0000 (02:10 +0000)]
Merge "BufferQueueProducer: Signal consumer when a buffer is detached" into nyc-dev
Greg Kaiser [Fri, 17 Jun 2016 22:27:12 +0000 (15:27 -0700)]
SensorList: Don't output UUIDs
We don't want the UUIDs of dynamic sensors to be in the logs.
Bug:
28775590
Change-Id: I85ca83e0ad1a0a8ed6618999bd22d86d7ef972f9
Jesse Hall [Fri, 17 Jun 2016 20:54:53 +0000 (13:54 -0700)]
vulkan: Updated IHV documentation
Also deleted DevelopersGuide.pdf, which was very out of date and has
been superceded by public documentation on developer.android.com.
Change-Id: I7e47177191a773eea6db3ee2af0a287877cb8532
David Brazdil [Thu, 16 Jun 2016 13:11:05 +0000 (13:11 +0000)]
Merge "Revert "Attach profiles to bug reports"" into nyc-dev
David Brazdil [Thu, 16 Jun 2016 09:11:31 +0000 (09:11 +0000)]
Revert "Attach profiles to bug reports"
This reverts commit
0faf2b870f934c1cdd883d29642fd84da3b352e2.
Bug:
28610953
Bug:
29395357
Change-Id: I609fe344dfc61311e37b91b48480d9392764a598
Craig Donner [Mon, 13 Jun 2016 23:13:02 +0000 (23:13 +0000)]
Merge "Revert "GLConsumer: add build-time disable of gpu protected content"" into nyc-dev
Craig Donner [Mon, 13 Jun 2016 22:14:15 +0000 (22:14 +0000)]
Revert "GLConsumer: add build-time disable of gpu protected content"
This reverts commit
dea97b7aa2b5cfbb7d42f30bcbf86a01778b609b.
Bug:
28545166
Change-Id: I4b069e14c2b99e190ebdea6df1ed83827ffe17b6
Amith Yamasani [Mon, 13 Jun 2016 18:37:40 +0000 (18:37 +0000)]
Merge "Fixed file descriptor leak in IMemory" into nyc-dev
Pablo Ceballos [Mon, 13 Jun 2016 18:26:04 +0000 (18:26 +0000)]
Add FrameStats default constructor am:
9a8c42a72b am:
8cd7d257de am:
67105cc24d am:
b990d75315 am:
a1952f0a6d am:
81d678cf18 am:
ed968dc61c
am:
d1d1d4e922
Change-Id: I581c4df5e3866e8a2326830e6b9e6ca2ffb3bf1c
Pablo Ceballos [Mon, 13 Jun 2016 18:22:14 +0000 (18:22 +0000)]
Add FrameStats default constructor am:
9a8c42a72b am:
8cd7d257de am:
67105cc24d am:
b990d75315 am:
a1952f0a6d am:
81d678cf18
am:
ed968dc61c
Change-Id: Id30d35b89c46fda89990bed6acf932240bdb6890
Pablo Ceballos [Mon, 13 Jun 2016 18:18:09 +0000 (18:18 +0000)]
Add FrameStats default constructor am:
9a8c42a72b am:
8cd7d257de am:
67105cc24d am:
b990d75315 am:
a1952f0a6d
am:
81d678cf18
Change-Id: I24f1e2594fdbd2022a534332009cad9564becd54
Pablo Ceballos [Mon, 13 Jun 2016 18:13:16 +0000 (18:13 +0000)]
Add FrameStats default constructor am:
9a8c42a72b am:
8cd7d257de am:
67105cc24d am:
b990d75315
am:
a1952f0a6d
Change-Id: Icc23cc2a483b13c6ac2134008d553574fffe06f4
Pablo Ceballos [Mon, 13 Jun 2016 18:09:20 +0000 (18:09 +0000)]
Add FrameStats default constructor am:
9a8c42a72b am:
8cd7d257de am:
67105cc24d
am:
b990d75315
Change-Id: Ibc0caec8dc613a6701791f1a633f3793522615af
Pablo Ceballos [Mon, 13 Jun 2016 18:04:17 +0000 (18:04 +0000)]
Add FrameStats default constructor am:
9a8c42a72b am:
8cd7d257de
am:
67105cc24d
Change-Id: Icf85c944e561763b0a9c353f84d614a992168c20
Pablo Ceballos [Mon, 13 Jun 2016 18:00:25 +0000 (18:00 +0000)]
Add FrameStats default constructor am:
9a8c42a72b
am:
8cd7d257de
Change-Id: Ic0df6b15e28ac2acda566fa7f30c9460d36f2a00
Pablo Ceballos [Mon, 13 Jun 2016 17:56:06 +0000 (17:56 +0000)]
Add FrameStats default constructor
am:
9a8c42a72b
Change-Id: Ib78d68c71438391711ea3b91102408d5ad0b966b
Craig Donner [Mon, 13 Jun 2016 16:36:51 +0000 (16:36 +0000)]
Merge "Fix assignment of 565 format when creating native client buffers." into nyc-dev
Craig Donner [Sat, 11 Jun 2016 00:20:15 +0000 (17:20 -0700)]
Fix assignment of 565 format when creating native client buffers.
Change-Id: Ia379f469a6e18e1f31748dc97aa596d03e507f84
Fixes:
29279664
Felipe Leme [Fri, 10 Jun 2016 23:56:33 +0000 (16:56 -0700)]
Fixed -t order (should be *after* dumpsys :-)
BUG:
29042642
Change-Id: I1b88634b0152f849a57f8e82adc82bdc5e609db7
Eino-Ville Talvala [Fri, 10 Jun 2016 21:21:02 +0000 (14:21 -0700)]
BufferQueueProducer: Signal consumer when a buffer is detached
When a buffer is detached, it should be cleared from the consumer-side
cache as well.
Not doing so can leak buffers no longer owned by the queue for extended
lengths of time, depending on the exact usage patterns of the producer
and consumer.
Bug:
28695173
Change-Id: I2c618be51f2c827d92d0e19b92a2d5befda87c38
TreeHugger Robot [Fri, 10 Jun 2016 20:40:20 +0000 (20:40 +0000)]
Merge "Set version to 1.0" into nyc-dev
Felipe Leme [Fri, 10 Jun 2016 00:19:18 +0000 (00:19 +0000)]
Merge "Expose more useful functions that can be used by dumpstate_board()." into nyc-dev
John Eckerdal [Thu, 21 Apr 2016 13:04:14 +0000 (15:04 +0200)]
Fixed file descriptor leak in IMemory
Include dup in critical section to avoid multiple dup calls from
different threads.
The leak happens more frequently since patch in frameworks/av -
6773d4777f4ccbbe6377e4ae1b42c117066ae6ba when recording video since it
will access IMemory from CameraUtils::isNativeHandleMetadata
Bug:
28924660
Author: Henrik Antonsson <henrik.antonsson@sonymobile.com>
Change-Id: I769a914c3adc505b522659a0c43719fb2922e6fd
TreeHugger Robot [Wed, 8 Jun 2016 23:35:11 +0000 (23:35 +0000)]
Merge "Improved error handling." into nyc-dev
Felipe Leme [Wed, 8 Jun 2016 18:11:01 +0000 (11:11 -0700)]
Improved error handling.
adb bugreport now properly handles stderr.
BUG:
29161586
Change-Id: I7a59d1feadda3d160eae6fe447a7a3c7e0843cbb
TreeHugger Robot [Wed, 8 Jun 2016 18:14:00 +0000 (18:14 +0000)]
Merge "Explicitly sets dumpsys timeout." into nyc-dev
Felipe Leme [Wed, 8 Jun 2016 16:51:29 +0000 (09:51 -0700)]
Explicitly sets dumpsys timeout.
The run_command() function takes a timeout as argument, and dumpsys can
also optionally takes a timeout with the -t argument, hence it makes
sense to pass the same value for both (otherwise dumpsys will timeout in
10s).
BUG:
29042642
Change-Id: Ib6e987d94e966686bab71223d7f910af9e549814
David Brazdil [Wed, 8 Jun 2016 15:36:45 +0000 (15:36 +0000)]
Merge "Attach profiles to bug reports" into nyc-dev
David Brazdil [Fri, 3 Jun 2016 13:40:44 +0000 (14:40 +0100)]
Attach profiles to bug reports
Includes the entire '/data/misc/profiles' folder in the bug report
ZIP file if not user build.
Bug:
28610953
Change-Id: I99d46f024e5557cf94e6045a51c46a605a00c52b
TreeHugger Robot [Tue, 7 Jun 2016 20:42:03 +0000 (20:42 +0000)]
Merge "SurfaceFlinger: Add mode to apply position with resize." into nyc-dev
Janis Danisevskis [Tue, 7 Jun 2016 19:23:45 +0000 (19:23 +0000)]
Merge "Fix Parcelable::readParcelableVector<T>" into nyc-dev
Robert Carr [Sun, 15 May 2016 18:27:47 +0000 (11:27 -0700)]
SurfaceFlinger: Add mode to apply position with resize.
For some cases, like scaled windows with shadows,
we need to be able to apply the position concurrent with
window resize. This is because the scaling of the shadows causes
the top left coordinate of the non shadow surface content
to change before and after the resize.
Bug:
28899837
Change-Id: I522eacfbbcd79707dc1e5ab71901a263b3004ba9
Calvin On [Fri, 3 Jun 2016 22:17:07 +0000 (15:17 -0700)]
Expose more useful functions that can be used by dumpstate_board().
BUG:
28640932
Fixes:
28867848
Change-Id: I38123e5b348f4560212559176eb60dbad4f22364
David Sehr [Mon, 6 Jun 2016 20:16:10 +0000 (20:16 +0000)]
Merge "Enable profman pretty printing" into nyc-dev
Pablo Ceballos [Mon, 6 Jun 2016 19:32:15 +0000 (19:32 +0000)]
Merge "SF: Also edit currentState in doTransaction" into nyc-dev
David Sehr [Thu, 2 Jun 2016 17:42:12 +0000 (10:42 -0700)]
Enable profman pretty printing
Bug:
28748264
Change-Id: Ia625a7976d9b95ec3b7e1d03af39af9f39015f3b
Mike Stroyan [Mon, 6 Jun 2016 16:29:41 +0000 (16:29 +0000)]
Merge "libgui: Check result in hook_dequeueBuffer_DEPRECATED" into nyc-dev
Craig Donner [Mon, 6 Jun 2016 16:18:23 +0000 (16:18 +0000)]
Merge "Allow setting alpha to 0 when creating native client buffers." into nyc-dev
Mike Stroyan [Fri, 3 Jun 2016 18:43:26 +0000 (12:43 -0600)]
libgui: Check result in hook_dequeueBuffer_DEPRECATED
Surface::hook_dequeueBuffer_DEPRECATED calls c->dequeueBuffer without
checking the returned result value. It should check for failing
result and return on error without using the values that
c->dequeueBuffer did not set.
Bug:
28692406
Change-Id: I10f9a23ad1052c5149040124366e0eb91b616267
Pablo Ceballos [Fri, 3 Jun 2016 00:46:05 +0000 (17:46 -0700)]
SF: Also edit currentState in doTransaction
Bug
28899837
Change-Id: I46c8301549c627c5d1c7efa661055aba787a33d6
Craig Donner [Sat, 4 Jun 2016 00:54:25 +0000 (17:54 -0700)]
Allow setting alpha to 0 when creating native client buffers.
Note that RGBA sizes are checked later in the function anyway, so there was no
reason to disallow 0, but it did preclude setting alpha to 0 for 565 buffers.
Fixes:
29126799
Change-Id: Iba11649976221ab385227fc022ae74b14c342203
Janis Danisevskis [Fri, 20 May 2016 17:10:04 +0000 (18:10 +0100)]
Fix Parcelable::readParcelableVector<T>
Parcelable::readParcelableVector<T> was apparently never
used. It now is and needed some adjustments to actually
work.
Bug:
22914603
Change-Id: I58ead3b3956299d549cff245588f56f3220563d3
Jesse Hall [Fri, 3 Jun 2016 19:39:49 +0000 (12:39 -0700)]
libvulkan: reduce log level for swapchain usage errors
Bug:
26927424
Change-Id: Id9346356436663db83a2d6fa8fefc4ea24d7854b
Jesse Hall [Thu, 2 Jun 2016 23:57:02 +0000 (23:57 +0000)]
Merge "libvulkan: Slightly better handling of swapchain re-creation" into nyc-dev
Chia-I Wu [Thu, 2 Jun 2016 22:36:26 +0000 (22:36 +0000)]
Merge "vulkan: remove EnumerateDeviceLayerProperties internally" into nyc-dev
Yin-Chia Yeh [Thu, 2 Jun 2016 18:12:21 +0000 (18:12 +0000)]
Merge "Camera: Add media/camera NDK documents" into nyc-dev
Jesse Hall [Tue, 31 May 2016 05:40:14 +0000 (22:40 -0700)]
libvulkan: Slightly better handling of swapchain re-creation
Previously we would fail vkCreateSwapchainKHR if
VkSwapchainCreateInfo::oldSwapchain was used, because we were unable
to dequeue all buffers for the new swapchain from an already-used
native window.
Now we disconnect and re-connect to the native window in order to
fully reset state, and allow us to dequeue all buffers. Additionally,
we tag the old swapchain as out-of-date, so future image acquires and
presents will fail with VK_ERROR_OUT_OF_DATE_KHR on that swapchain.
This is less than ideal, but better than what we had, and the best we
have time to do for N.
Bug:
26927424
Change-Id: Ifaa5048376f72a63ecb1dca3d1ff85dbee2c24d0
Pablo Ceballos [Thu, 2 Jun 2016 17:14:00 +0000 (17:14 +0000)]
Merge "BQ: Don't return fences from dequeue in shared buffer mode" into nyc-dev
Pablo Ceballos [Thu, 2 Jun 2016 17:10:06 +0000 (17:10 +0000)]
Merge "Push empty sync transaction before taking screenshot in tests" into nyc-dev
Calin Juravle [Thu, 2 Jun 2016 10:10:35 +0000 (10:10 +0000)]
Merge "Create primary profile file together with app data folders" into nyc-dev
Chia-I Wu [Thu, 2 Jun 2016 03:25:47 +0000 (11:25 +0800)]
vulkan: remove EnumerateDeviceLayerProperties internally
We never use it. Since it is also deprecated, we will never need it.
Bug:
27910383
Change-Id: If4e556bb69890153be2987b165482e43ab06a890
Pablo Ceballos [Thu, 2 Jun 2016 01:53:40 +0000 (18:53 -0700)]
Push empty sync transaction before taking screenshot in tests
Because of ag/
1043009 it's now necessary to push an empty sync
transaction before taking a screenshot to ensure that all transactions
have been applied.
Bug
29074285
Change-Id: I9f1b4a5b6b1f39e1ff26d536817aef6f40865177
Pablo Ceballos [Wed, 1 Jun 2016 22:03:21 +0000 (15:03 -0700)]
BQ: Don't return fences from dequeue in shared buffer mode
Bug
29007505
Change-Id: I112ae6bd64ddc9a923ad9080e85a71adb96146af
David Sehr [Tue, 17 May 2016 23:06:22 +0000 (16:06 -0700)]
installd support for profman --dump-info-for
Bug:
28748264
Change-Id: Idcd3ed86294c1d0c75c26ca938b88744a25e387a
Mark Salyzyn [Tue, 31 May 2016 19:44:18 +0000 (19:44 +0000)]
Merge "svcmgr: don't send reply to one-way txn (part deux)" into nyc-dev
Felipe Leme [Thu, 19 May 2016 23:30:15 +0000 (16:30 -0700)]
Set version to 1.0
BUG:
27366368
Change-Id: Iac8ed413d41d0c26a39343025920f433d5d7ddba
Erik Kline [Mon, 30 May 2016 02:55:44 +0000 (11:55 +0900)]
Use explicit path for /vendor/xbin/wlutil
At this time /vendor/xbin is not in root's path.
Bug:
28850734
Change-Id: Ib6c2b6f8b0285854670c843dcc437cfcc84484a8
Chia-I Wu [Fri, 27 May 2016 08:22:49 +0000 (08:22 +0000)]
Merge "vulkan: fix crash when only non-global layers are enabled" into nyc-dev
Yin-Chia Yeh [Fri, 27 May 2016 06:07:15 +0000 (23:07 -0700)]
Camera: Add media/camera NDK documents
Include media/camera ndk headers to codegen source.
Add docs/images directory for image links in generated documents.
Bug:
27102995
Change-Id: I9c75e2ceba3458c9962c26cf786e1402fd963696
Chia-I Wu [Fri, 27 May 2016 02:18:25 +0000 (10:18 +0800)]
vulkan: fix crash when only non-global layers are enabled
We fail to set up get_instance_proc_addr_ and get_device_proc_addr_ when
there are layers but only non-global ones.
Bug:
28990561
Change-Id: I82704f946deb5a76640f9dfb792e6fd118c78aa7
Erik Kline [Fri, 27 May 2016 02:29:19 +0000 (11:29 +0900)]
Add "ip maddr" to dumpstate output for debugging.
Bug:
28966611
Change-Id: I8fdab94fbd41448e52f2fed406e8f43edf7f18ad
Pablo Ceballos [Thu, 26 May 2016 22:35:55 +0000 (15:35 -0700)]
Add FrameStats default constructor
Bug
28592402
Change-Id: I857e46c9ab3ffae0d96923d665d13a4128a6cafa
Mark Salyzyn [Wed, 25 May 2016 20:14:10 +0000 (13:14 -0700)]
svcmgr: don't send reply to one-way txn (part deux)
Do not split transaction into two when sending synchronous responses.
Partially revert changes associated with binder_send_reply from
previous incantations in commit
51f592c7bad13d3d8e8f7cfec25edfef42901802
that resulted in a performance regression.
Signed-off-by: Mark Salyzyn <salyzyn@google.com>
Bug:
27111612
Bug:
28696050
Bug:
28946245
Change-Id: I12edc10a2d1408a71507a7983afa7f17b5a6a81b
Michael Wright [Thu, 26 May 2016 12:49:53 +0000 (13:49 +0100)]
Mark the multiwindow divider as a trusted overlay.
Having it be untrusted breaks the permissions granting dialog in
multiwindow mode.
Bug:
28828041
Change-Id: I9193dbce7315ebaa205673edd4f406d0b72665c8
Calin Juravle [Tue, 24 May 2016 14:33:12 +0000 (15:33 +0100)]
Create primary profile file together with app data folders
Bug:
28785119
Bug:
28625993
Change-Id: I12c0de85bc3dc35199acf32c10fe69f93e76d7fc
TreeHugger Robot [Tue, 24 May 2016 23:15:00 +0000 (23:15 +0000)]
Merge "SF: Force empty sync transactions to apply anyway" into nyc-dev
Robert Carr [Tue, 24 May 2016 18:41:28 +0000 (11:41 -0700)]
SF: Force empty sync transactions to apply anyway
Following the change to not flush transactions in the screen shot
path, this enables the use of empty synchronous transactions as a
flush mechanism for previous asynchronous transactions.
Bug:
27098060
Change-Id: I282fdfffb72db148edcbf1ded65b15816fc714a2
Dan Stoza [Tue, 24 May 2016 18:01:24 +0000 (18:01 +0000)]
Merge "SF: Remove forced transaction on screenshot" into nyc-dev
Dan Stoza [Tue, 24 May 2016 16:55:43 +0000 (16:55 +0000)]
Merge "HWC2: Fix breakage from header change" into nyc-dev
Lajos Molnar [Tue, 24 May 2016 14:02:01 +0000 (07:02 -0700)]
stagefright: define mpeg2 HighP level constant
Bug:
28671284
Change-Id: Ic97fd9c3cb4ce04f9cfcd18a4daf68db3349694b
Chia-I Wu [Mon, 23 May 2016 07:31:02 +0000 (15:31 +0800)]
vulkan: advertise VK_EXT_debug_report
We use VK_EXT_debug_report to report most messages and we should always
advertise the extension. This is complicated however because the HAL may
also advertise the extension.
Do the minimal and cache whether the HAL supports VK_EXT_debug_report in
class Hal. More refactorings will be required if we want to be more
flexible.
Bug:
28753291
Change-Id: I0298b2c3024e2ed059411128dbbe264c0b289a4b
Chia-I Wu [Mon, 23 May 2016 02:47:57 +0000 (10:47 +0800)]
vulkan: add a wrapper for hwvulkan_device_t
Add class Hal as a wrapper for hwvulkan_device_t.
Bug:
28753291
Change-Id: I9612ee5515c7f2266cfc1ef0c8583fa98c82dd86
Chia-I Wu [Mon, 23 May 2016 23:36:52 +0000 (23:36 +0000)]
Merge "vulkan: do not use exported functions internally" into nyc-dev
TreeHugger Robot [Mon, 23 May 2016 20:01:25 +0000 (20:01 +0000)]
Merge "Expose functions used to add entries to the zip file." into nyc-dev
Felipe Leme [Thu, 19 May 2016 23:18:17 +0000 (16:18 -0700)]
Expose functions used to add entries to the zip file.
Currently only dumpstate.cpp is using the new functions and variables
used to add an entry to the zip file so they're defined locally, but
they should be exposed through the .h so they can be used by the
device-specific dumpstate_board() implementations.
This CL moves the following definitions to the .h:
- bugreport_dir
- ZIP_ROOT_DIR
- add_zip_entry()
- add_zip_entry_from_fd()
It also refactored functions that were taking a specific file location
to use bugreport_dir instead, to avoid pollution on main().
BUG:
28867848
Change-Id: Ic5faa203b0863bc950d640f22fb91977e52b6bfc
Lajos Molnar [Sat, 21 May 2016 02:38:40 +0000 (19:38 -0700)]
omx: add asString for new enumerations and fix some issues
Bug:
28671284
Change-Id: I54e98d727f974544d51b2a14b565ad24912ab242
Lajos Molnar [Mon, 23 May 2016 13:17:25 +0000 (13:17 +0000)]
Merge "omx: add missing MPEG4 levels" into nyc-dev
Chia-I Wu [Fri, 20 May 2016 08:15:06 +0000 (16:15 +0800)]
vulkan: do not use exported functions internally
Our vkGet*ProcAddr uses the exported functions. They will break when any
of the exported functions are overridden (e.g., through LD_PRELOAD).
Unexport and move all exported functions to vulkan::api namespace.
Re-export them by having vkFoo as a wrapper to vulkan::api::Foo.
Another option is to re-export vulkan::api::Foo by having vkFoo as an
alias using __attribute__((alias)). That results in smaller binaries.
But we will not be able to catch mismatches between vulkan.h and
vulkan.api.
To avoid future breakage, define VK_NO_PROTOTYPES for all files except
api_gen.cpp.
Bug:
28886971
Change-Id: I08fde7ebb247f8c7e040ccf812b40b02094d3c7f
Jesse Hall [Sat, 21 May 2016 21:21:42 +0000 (21:21 +0000)]
Merge "Ensure memory ordering around libagl and EGL refcount operations" into nyc-dev
Jesse Hall [Fri, 20 May 2016 17:47:07 +0000 (10:47 -0700)]
Ensure memory ordering around libagl and EGL refcount operations
The android_atomic_inc/android_atomic_dec functions don't impose
sufficient memory ordering. Using them for object refcounting could
allow an object to be destroyed prior to writes by a different thread
being visible.
Bug:
28820690
Change-Id: Ie018091035174255a22ebc52852528cdaec2d648
Dan Stoza [Fri, 20 May 2016 21:02:39 +0000 (14:02 -0700)]
HWC2: Fix breakage from header change
Updates the shim and adapter to handle the header changes introduced
by "HWC2: Add format to VD create, surface damage" in libhardware.
Bug:
22767098
Change-Id: I8142301c121d9b5602a4455f3fbadefefd685cf0
Jesse Hall [Fri, 20 May 2016 14:13:52 +0000 (07:13 -0700)]
vulkan: Update vulkan.h to 1.0.13
vulkan.h and vk_platform.h imported from
github.com:KhronosGroup/Vulkan-Docs @
2656f459333b3a1dc63619a9ebd83490eea22e93
Updated vulkan.api to reflect the changes in vulkan.h. There were no
changes to any of the code we generate. I didn't update the
vulkan_h.tmpl, since we're not currently using it and it was already
out of date.
Bug:
28878216
Change-Id: I71676683148fae7ad8799bd39386a931319aa395
Jesse Hall [Fri, 20 May 2016 13:58:01 +0000 (13:58 +0000)]
Merge "libvulkan: Don't try to load compressed/unaligned libraries from APK" into nyc-dev
Wei Liu [Thu, 19 May 2016 20:59:01 +0000 (13:59 -0700)]
Fix the DUMPSYS skip.
b/
28869921
Change-Id: Idc108b2d3cfd237f459467a3c399a0ba5e248d24
Jesse Hall [Thu, 19 May 2016 17:58:35 +0000 (10:58 -0700)]
libvulkan: Don't try to load compressed/unaligned libraries from APK
Bug:
28825642
Change-Id: I46ea3a54010cccf9e23696a4aff52a42a31d92b1
Lajos Molnar [Thu, 19 May 2016 22:27:31 +0000 (15:27 -0700)]
omx: add missing MPEG4 levels
Bug:
28671284
Change-Id: I96b50580cc29c11732fd2b496ba6d832b7067256
Chia-I Wu [Wed, 18 May 2016 22:47:45 +0000 (22:47 +0000)]
Merge "vulkan: fix double dlopen for layer libraries" into nyc-dev
Mike Stroyan [Wed, 18 May 2016 18:07:00 +0000 (18:07 +0000)]
Merge "Preserve parameters to getProcAddress functions" into nyc-dev
Pablo Ceballos [Wed, 18 May 2016 17:59:02 +0000 (17:59 +0000)]
Merge "SF: Fix arbitrary rotation with inverse display transform" into nyc-dev
Chia-I Wu [Wed, 18 May 2016 00:50:15 +0000 (08:50 +0800)]
vulkan: fix double dlopen for layer libraries
Bug:
28826014
Change-Id: I10693426f987e63bb490ac8d1f4445ce2d195d8d
Pablo Ceballos [Tue, 17 May 2016 20:34:45 +0000 (13:34 -0700)]
SF: Fix arbitrary rotation with inverse display transform
Bug
28043535
Change-Id: I8b939e17f91cf7e5b617bbcceb9df4efcc9d0161
Mike Stroyan [Tue, 17 May 2016 20:03:49 +0000 (14:03 -0600)]
Preserve parameters to getProcAddress functions
The contents of register r0 parameter were overwritten after commit
4690754. That used an "r" register to hold larger offsets, but doing
that overwrote r0 before it was passed on to the next call.
Applying the offset as two parts allows the total largest offset of
3588+1020 to be reached without using a register to hold that value.
Use an "addne" instruction to add the _api offset of up to 1020 within
the extensions array.
Use a "ldrne" offset to add the offset to the start of the extensions
array.
Bug:
26541117
Change-Id: Ie19e4aa584c9fa6fdec4de18e858777ac59d8dbd