OSDN Git Service
Andreas Gampe [Mon, 28 Sep 2015 15:55:01 +0000 (08:55 -0700)]
Installd: Fully decouple pre- and post-bootcomplete properties
Do not fall back to dalvik.vm.dex2oat-threads if
dalvik.vm.boot-dex2oat-threads is not set. Also do not ever use
the value of dalvik.vm.image-dex2oat-threads. Instead use the
default behavior of dex2oat in that situation.
Change-Id: I42e60996341414b95b8df2d243d5e07f7f5ff9b6
Todd Kennedy [Fri, 25 Sep 2015 21:41:37 +0000 (21:41 +0000)]
Merge "Pass dexopt flags as integer"
Todd Kennedy [Fri, 25 Sep 2015 14:47:47 +0000 (07:47 -0700)]
Pass dexopt flags as integer
Instead of using a series of booleans, create a single flags integer
that contains all of the dexopt options.
Change-Id: I35542aa73ca57b0e765d19b1339b1429849c1ae8
Todd Kennedy [Thu, 24 Sep 2015 20:35:07 +0000 (20:35 +0000)]
Merge "Installd: Take boot status as dexopt parameter"
Andreas Gampe [Mon, 21 Sep 2015 20:21:30 +0000 (13:21 -0700)]
Installd: Take boot status as dexopt parameter
Expect the boot status explicitly as a parameter so that we do not
have to rely on dev.bootcomplete, which isn't meaningfully set
when the device needs the decryption screen on boot.
Bug:
23898216
(cherry picked from commit
72ebebed876a62e719b098e43d9d516361bde029)
Change-Id: I9b34298caf70b1e5d40970cc0d04c469016a80a7
Dan Austin [Wed, 23 Sep 2015 16:47:59 +0000 (16:47 +0000)]
Merge "Enable integer overflow sanitization in libinput."
Dan Austin [Tue, 22 Sep 2015 21:33:57 +0000 (14:33 -0700)]
Enable integer overflow sanitization in libinput.
Bug:
24171356
Change-Id: I1412b78171c60b0252560956a672f2d92a7011e6
Dan Austin [Wed, 23 Sep 2015 16:36:28 +0000 (16:36 +0000)]
Merge "Eliminate multiple benign overflow conditions."
Sergio Giro [Wed, 23 Sep 2015 16:22:55 +0000 (16:22 +0000)]
Merge "Remove usage of SharedBuffer"
Dan Austin [Wed, 23 Sep 2015 14:45:51 +0000 (14:45 +0000)]
Merge "Enable multiple benign overflow conditions."
Dan Austin [Wed, 23 Sep 2015 14:45:25 +0000 (14:45 +0000)]
Merge "Eliminate benign overflow condition triggered upon loop termination in Input.cpp"
Dan Austin [Tue, 22 Sep 2015 21:28:07 +0000 (14:28 -0700)]
Eliminate multiple benign overflow conditions.
In InputTransport.cpp, there are multiple loops in which loop
termination occurs when the value becomes zero. These termination
conditions are all written value-- > 0, which, since value is
unsigned, result in an unsigned integer overflow when value is 0.
These loops were refactored to eliminate these conditions.
Bug:
24171356
Change-Id: Ie135c4306d1f2cef2778e295242305ed5139221a
Dan Austin [Tue, 22 Sep 2015 21:32:03 +0000 (14:32 -0700)]
Enable multiple benign overflow conditions.
In VelocityTracker.cpp, there are multiple loops in which loop
termination occurs when the value becomes zero. These termination
conditions are all written value-- > 0, which, since value is
unsigned, result in an integer overflow condition when value is 0.
These loop were refactored to eliminate these conditions.
Bug:
24171356
Change-Id: Ie44940cfef8a122ee1aff59c021274ba39a368bf
Dan Austin [Tue, 22 Sep 2015 21:22:41 +0000 (14:22 -0700)]
Eliminate benign overflow condition triggered upon loop termination in Input.cpp
In readFromParcel, a while loop is terminated when sampleCount = 0. The
decrement operation that was here would decrease sampleCount, an unsigned value,
below 0, triggering an unsigned integer overflow. The while loop was refactored
to eliminate this condition.
Bug:
24171356
Change-Id: I7669f54a41d11548b33e322b025431c6f6038952
Nick Kralevich [Mon, 21 Sep 2015 21:42:06 +0000 (21:42 +0000)]
Merge "constify generated JNINativeMethod fn ptr tables"
Daniel Micay [Mon, 21 Sep 2015 17:17:57 +0000 (13:17 -0400)]
constify generated JNINativeMethod fn ptr tables
Change-Id: Icbad34e298f6d8e60ef12e9a91200994249fe2f5
Pablo Ceballos [Fri, 11 Sep 2015 21:37:39 +0000 (14:37 -0700)]
Remove usage of SharedBuffer
- Remove getSharedBuffer() from Region
- Don't use SharedBuffer for memory management in HWCLayerVersion1,
instead keep shallow copies of the Regions.
Bug
23962051
(cherry picked from commit
d814cf2a3e3a2fdb73efa80539fe8af0a93da1db)
Change-Id: I8fa17beed57936648c7b4defc9219dff1d5b337f
Evgenii Stepanov [Fri, 18 Sep 2015 20:23:35 +0000 (20:23 +0000)]
Merge "asan: update condition to work with multiple SANITIZE_TARGET values."
Ivan Krasin [Fri, 18 Sep 2015 19:03:44 +0000 (12:03 -0700)]
asan: update condition to work with multiple SANITIZE_TARGET values.
The goal is to enable SANITIZE_TARGET='address coverage', which
will be used by LLVMFuzzer.
Bug:
22850550
Change-Id: Iaa300aa20798f2e9aa613842b463edf8aeebe3a7
Andreas Gampe [Thu, 17 Sep 2015 22:36:50 +0000 (22:36 +0000)]
Merge "Installd: Allow different behavior before bootcomplete"
Andreas Gampe [Mon, 14 Sep 2015 23:33:11 +0000 (16:33 -0700)]
Installd: Allow different behavior before bootcomplete
Check dev.bootcomplete in dex2oat(). Use the information for
two changes.
Only switch to the background when we're post bootcomplete.
This will ensure better utilization after upgrades.
Add a second dex2oat-threads property that is used pre
bootcomplete. A separation of these phases allows using less
cores when the device is up, freeing up resources for other
purposes, e.g., avoid jank. The precedence during boot is the
boot property, or the image property if the former doesn't
exist, or the default property as a fallback.
Bug:
23898216
Bug:
24004256
(cherry picked from commit
21948005f0294a64a42932b21a33ae75295dc3d9)
Change-Id: I5063f3fc4b437cbe88c4e94584e01c1c78eccc4d
Daniel Erat [Wed, 16 Sep 2015 02:45:23 +0000 (02:45 +0000)]
Merge "Update libpowermanager to expose enum and export headers."
Daniel Erat [Tue, 15 Sep 2015 16:50:17 +0000 (10:50 -0600)]
Update libpowermanager to expose enum and export headers.
Move libpowermanager's transaction ID enum (needed when
implementing BnPowerManager::onTransact()) into
IPowerManager.h and update the library's Android.mk file to
export its headers.
Bug:
22122485
Change-Id: Ie97198c3d2ebe8e880125e19a001ef7162009858
Daniel Erat [Tue, 15 Sep 2015 14:21:30 +0000 (14:21 +0000)]
Merge "Remove BnServiceManager."
Daniel Erat [Mon, 14 Sep 2015 22:29:54 +0000 (16:29 -0600)]
Remove BnServiceManager.
As far as I can tell (e.g. https://goo.gl/nxpFBZ from 2009),
BnServiceManager is unused. It's only relevant for
libbinder-based C++ daemons implementing the
IServiceManager.aidl interface. service_manager.c rolls its
own binder code instead.
Change-Id: I837a897d8a201b04bf3ff65c87b0d35d10196199
Dan Austin [Thu, 10 Sep 2015 22:02:03 +0000 (22:02 +0000)]
Merge "Benign unsigned integer overflow in Parcel"
Dan Austin [Thu, 10 Sep 2015 21:58:27 +0000 (21:58 +0000)]
Merge "Enable fsanitize integer in ui"
Dan Austin [Thu, 10 Sep 2015 20:46:02 +0000 (13:46 -0700)]
Benign unsigned integer overflow in Parcel
The realloc case in continueWrite did not update the
gParcelGlobalAllocCount value when an allocation occurred.
In addition, there are conditions that could cause the
gParcelGlobalAllocCount value to be decremented below 0, resulting
in a benign unsigned integer overflow that can cause corrupted values
to be returned through system profiling mechanisms.
BUG:
23972600
Change-Id: Ibe07db91a811a04b486760eb78d81c926ba8503d
Dan Austin [Wed, 9 Sep 2015 14:39:25 +0000 (14:39 +0000)]
Merge "Enable fsanitize integer in binder"
Dan Austin [Tue, 8 Sep 2015 23:18:13 +0000 (16:18 -0700)]
Enable fsanitize integer in binder
Enables checks for undefined or suspicious integer behavior. If
suspicious behavior is detected, the process is aborted instead of
potentially resulting in a security bug, such as #
20674694 and
#
19573085.
Bug:
23836686
Change-Id: I14d6d565afcb8a6435fd5a8e8175e00bbe5d9b6c
Dan Austin [Tue, 8 Sep 2015 21:53:00 +0000 (14:53 -0700)]
Enable fsanitize integer in ui
Enables checks for undefined or suspicious integer behavior. If
suspicious behavior is detected, the process is aborted instead of
potentially resulting in a security bug.
Bug:
23834892
Change-Id: I9111bce2ebb36e045f4da1dbf78f90f04a0aa81c
Tom Cherry [Thu, 3 Sep 2015 21:52:44 +0000 (21:52 +0000)]
Merge "move init.trace.rc to frameworks/native"
Tom Cherry [Thu, 3 Sep 2015 21:15:24 +0000 (14:15 -0700)]
move init.trace.rc to frameworks/native
init.trace.rc will be renamed to atrace.rc and use the LOCAL_INIT_RC
mechanism to be included on /system appropriately.
Bug
23186545
Change-Id: Ic5833b4bb6e6364e4fef6ab4882ae75f8f3856e9
Yasuhiro Matsuda [Wed, 2 Sep 2015 12:54:23 +0000 (12:54 +0000)]
Merge "Add a tag for tracing system server."
Dan Willemsen [Mon, 31 Aug 2015 19:30:20 +0000 (19:30 +0000)]
Merge "Whitelist windows module"
Johan Euphrosine [Sat, 29 Aug 2015 04:15:16 +0000 (04:15 +0000)]
Merge "include/android/sensor: add enum for LINEAR_ACCELERATION"
Elliott Hughes [Sat, 29 Aug 2015 03:37:09 +0000 (03:37 +0000)]
Merge "ignore SIGPIPE signal in native services"
Young-Ho Cha [Fri, 28 Aug 2015 10:55:18 +0000 (19:55 +0900)]
ignore SIGPIPE signal in native services
When run dump method via dumpsys command with invalid pipe fd, native
services can crash with SIGPIPE.
So ignore SIGPIPE signal.
Bug: https://code.google.com/p/android/issues/detail?id=184099
Change-Id: Icb15e0dff34a245c08c1b4c4bed51992ae382da4
Signed-off-by: Young-Ho Cha <ganadist@gmail.com>
Johan Euphrosine [Fri, 21 Aug 2015 01:13:43 +0000 (18:13 -0700)]
include/android/sensor: add enum for LINEAR_ACCELERATION
Change-Id: I6e387f905ba87bde810952daf939ba0f529d18b8
Dan Willemsen [Fri, 28 Aug 2015 21:09:44 +0000 (14:09 -0700)]
Whitelist windows module
Bug:
23566667
Change-Id: I4fa1243421885a0c1a8febbbcefb0d7e9cdcc99b
Ruchi Kandoi [Tue, 25 Aug 2015 23:45:51 +0000 (23:45 +0000)]
Merge "BatteryProperties: Adds batteryCycleCount, batteryCurrent, batteryFullCharge fields in the the battery properties."
Ruchi Kandoi [Thu, 13 Aug 2015 00:49:42 +0000 (17:49 -0700)]
BatteryProperties: Adds batteryCycleCount, batteryCurrent,
batteryFullCharge fields in the the battery properties.
Bug:
22981343
Change-Id: I426b4b7d1eb544d2b1d97d87ae1eba0fdf35c503
Signed-off-by: Ruchi Kandoi <kandoiruchi@google.com>
Jesse Hall [Sat, 22 Aug 2015 02:40:45 +0000 (02:40 +0000)]
Merge "libagl: Fix buffer read overrun in eglCreatePbufferSurface"
Tom Cherry [Fri, 21 Aug 2015 17:42:30 +0000 (17:42 +0000)]
Merge "bundle init.rc contents with its service"
Jesse Hall [Fri, 21 Aug 2015 14:41:46 +0000 (07:41 -0700)]
libagl: Fix buffer read overrun in eglCreatePbufferSurface
The code was assuming that EGL_NONE==0, which isn't true.
Bug:
23403170
Change-Id: Ic1eccdef086b4d610bd78dbb6b2ae883c91dc322
Elliott Hughes [Thu, 20 Aug 2015 01:07:16 +0000 (01:07 +0000)]
Merge "binder: Change from C11 <stdatomic.h> to C++11 <atomic>."
Tom Cherry [Fri, 14 Aug 2015 20:01:23 +0000 (13:01 -0700)]
bundle init.rc contents with its service
Bug:
23186545
Change-Id: I2a70a39122b146d5bbc9b55156e769782e011bff
Bailey Forrest [Wed, 19 Aug 2015 00:15:10 +0000 (17:15 -0700)]
binder: Change from C11 <stdatomic.h> to C++11 <atomic>.
Change-Id: Ib831a9ccfcc7089940d37883d860574359c270e2
Riley Andrews [Mon, 17 Aug 2015 20:58:26 +0000 (20:58 +0000)]
Merge "binder: Add a throughput performance test for binder."
Riley Andrews [Tue, 11 Aug 2015 20:19:53 +0000 (13:19 -0700)]
binder: Add a throughput performance test for binder.
Change-Id: I04801cd505a439307f2a963f6267db1831b3829f
Riley Andrews [Wed, 12 Aug 2015 23:36:31 +0000 (23:36 +0000)]
Merge "Tracing for binder locks/transactions"
Elliott Hughes [Wed, 12 Aug 2015 22:30:07 +0000 (22:30 +0000)]
Merge "Lose HAVE_ANDROID_OS from frameworks/native."
Elliott Hughes [Wed, 12 Aug 2015 22:27:47 +0000 (15:27 -0700)]
Lose HAVE_ANDROID_OS from frameworks/native.
Change-Id: I8e6af2a46a9d875192fc7957ada9b5f66d84bad6
Nick Kralevich [Mon, 10 Aug 2015 20:53:31 +0000 (20:53 +0000)]
Merge "More O_DIRECTORY | O_NOFOLLOW changes"
Nick Kralevich [Mon, 10 Aug 2015 20:43:00 +0000 (13:43 -0700)]
More O_DIRECTORY | O_NOFOLLOW changes
Fix a few other spots where we're not using O_NOFOLLOW
with O_DIRECTORY.
Add O_CLOEXEC for good measure.
https://code.google.com/p/android/issues/detail?id=181945
Bug:
23044050
Change-Id: Ib535cbba29cdd587443ea9e1e75831f57fd7a93c
Nick Kralevich [Mon, 10 Aug 2015 19:48:39 +0000 (19:48 +0000)]
Merge "installd _delete_dir_contents: add O_NOFOLLOW"
Nick Kralevich [Mon, 10 Aug 2015 19:28:22 +0000 (12:28 -0700)]
installd _delete_dir_contents: add O_NOFOLLOW
Add O_NOFOLLOW to make sure we're not following symlinks.
https://code.google.com/p/android/issues/detail?id=181945
Bug:
23044050
Change-Id: I3408309fd401aefb6ebecb2a052bba198eed7633
Yasuhiro Matsuda [Tue, 30 Jun 2015 16:46:25 +0000 (01:46 +0900)]
Add a tag for tracing system server.
This change is needed for
https://android-review.googlesource.com/#/c/157562/
BUG:
21739901
Change-Id: Ic4616065749236fbde9263b8f1898c434913564a
Todd Kennedy [Mon, 3 Aug 2015 17:44:44 +0000 (17:44 +0000)]
Merge "Add a trace constant for PackageManager"
Todd Kennedy [Fri, 31 Jul 2015 21:36:20 +0000 (14:36 -0700)]
Add a trace constant for PackageManager
Change-Id: Icd9afc86c88296919e4750d71cb0aef314194bee
Andreas Gampe [Mon, 3 Aug 2015 16:55:45 +0000 (16:55 +0000)]
Merge "Native/opengl: Remove unused variables from test"
Andreas Gampe [Mon, 3 Aug 2015 15:57:34 +0000 (08:57 -0700)]
Native/opengl: Remove unused variables from test
For build-system CFLAGS clean-up.
Bug:
18632512
(cherry picked from commit
a75f729650b18478e3ff822dbf686406ed7f7169)
Change-Id: Ibe2d59c7e0677f02e4bf5214d34655cd300d5f15
Mark Salyzyn [Fri, 31 Jul 2015 22:10:51 +0000 (22:10 +0000)]
Merge "dumpstate: add logcat -b all -S"
Mark Salyzyn [Thu, 30 Jul 2015 21:57:09 +0000 (14:57 -0700)]
dumpstate: add logcat -b all -S
- For those that are proud to call themselves logspam police
- Every time someone declares that something is too chatty, or that
they lost their logs because they were declared too chatty and
their associated logs were aggresively pruned we generally
ask them to report the logger statistics to pinpoint what software
product that is the elephant in the room.
- Every time we want to spawn a new 'stop being so spammy' bug spawned
off a bugreport collected for another purpose, we *wish* we had the
logger statistics to help add gravitas to the claim that some piece
of software is the top, or near the top, polluter.
Change-Id: Ifae33cd21d0ae2917a3b4381502d723725b1701c
Yasuhiro Matsuda [Thu, 30 Jul 2015 07:42:08 +0000 (07:42 +0000)]
Merge "Enable atrace to read categories from file."
Nick Kralevich [Sun, 26 Jul 2015 04:26:11 +0000 (04:26 +0000)]
Merge "dumpstate: remove unnecessary mkdir"
Nick Kralevich [Sun, 26 Jul 2015 04:13:49 +0000 (21:13 -0700)]
dumpstate: remove unnecessary mkdir
/data/anr is now created in init.rc. This code is unnecessary.
Bug:
22385254
Change-Id: I4ea9c311d52eee7fb6b90235fbd12a4c9364d71f
Andreas Gampe [Fri, 17 Jul 2015 01:23:10 +0000 (01:23 +0000)]
Merge "Installd: Add a swap override flag"
Andreas Gampe [Thu, 16 Jul 2015 22:55:41 +0000 (15:55 -0700)]
Installd: Add a swap override flag
Add dalvik.vm.dex2oat-swap system property to override a default
decision.
Bug:
20658562
(cherry picked from commit
c968c0175e967e39e72f557b5e014b9575ba4727)
Change-Id: I34368c6e435d1a9ceec20a0bf1c8c6213e527f5e
Evgenii Stepanov [Tue, 14 Jul 2015 23:47:59 +0000 (23:47 +0000)]
Merge "With SANITIZE_TARGET, load EGL libs from the instrumented path."
Elliott Hughes [Thu, 9 Jul 2015 17:16:58 +0000 (17:16 +0000)]
Merge "Binder on Windows? Unlikely."
Elliott Hughes [Thu, 9 Jul 2015 17:04:34 +0000 (10:04 -0700)]
Binder on Windows? Unlikely.
Change-Id: I36c1bee4ed33b9b0d24109a6dbaefa0548dd5aa8
Evgenii Stepanov [Wed, 8 Jul 2015 22:49:52 +0000 (15:49 -0700)]
With SANITIZE_TARGET, load EGL libs from the instrumented path.
Bug:
21785137
Change-Id: Iac56419a9ca776c5704bed44d3f0a1a1edd9d172
Narayan Kamath [Wed, 1 Jul 2015 13:03:55 +0000 (13:03 +0000)]
Merge "Installd: Fix delete_code_cache() clearing the wrong cache dir"
Hao Li [Thu, 25 Jun 2015 02:31:56 +0000 (14:31 +1200)]
Installd: Fix delete_code_cache() clearing the wrong cache dir
Fix a typo in delete_code_cache() introduced by commit
c03de09, which
causes the function to empty /data/data/<package_name>/cache/ rather
than /data/data/<package_name>/code_cache/.
The following CTS test case should now pass:
com.android.cts.appsecurity.SplitTests#testClearCodeCache
Change-Id: I2d9452d70ed39b3ea29ad65fb35e788c2fef5100
Signed-off-by: Hao Li <hao.i.li@intel.com>
Signed-off-by: Yu Ning <yu.ning@intel.com>
Dan Stoza [Tue, 30 Jun 2015 17:11:08 +0000 (17:11 +0000)]
Merge "sf: Initialize EventThread before creating HWC"
John Reck [Mon, 29 Jun 2015 14:17:43 +0000 (14:17 +0000)]
Merge "Fix async_stop"
John Reck [Fri, 15 May 2015 17:00:34 +0000 (10:00 -0700)]
Fix async_stop
Change-Id: I68a258cf71cfcd999bbd568ab3c96255c11bec41
(cherry picked from commit
2c237ee1d5961b9d11618e0c5a97b3bb7fb694b2)
Yasuhiro Matsuda [Mon, 29 Jun 2015 10:20:39 +0000 (19:20 +0900)]
Enable atrace to read categories from file.
Add -f option to atrace, which enables atrace to use the
categories written in the specified file.
BUG:
21739901
Change-Id: I2cf1f277280434b28c426c68e2864b62aebd20cc
Scott Bauer [Mon, 8 Jun 2015 23:32:36 +0000 (16:32 -0700)]
Tracing for binder locks/transactions
Added logic to pull traces from the binder driver.
Change-Id: I9aa00b4650fc3f90d7bd55328dfef11e2514f7b4
Saurabh Shah [Tue, 16 Jun 2015 00:55:51 +0000 (17:55 -0700)]
sf: Initialize EventThread before creating HWC
Once HWC is created, it could use any of the provided hooks, which
could lead to a crash if the EventThread (handler) isn't initialized
prior to creating HWC.
Change-Id: I5ea35fe9bcb150fb74aae1295b798bd787ad6cee
Mark Salyzyn [Mon, 15 Jun 2015 14:27:10 +0000 (14:27 +0000)]
Merge "dumpstate: modify logcat timeout to compensate for storage speed"
Mark Salyzyn [Fri, 12 Jun 2015 19:28:24 +0000 (12:28 -0700)]
dumpstate: modify logcat timeout to compensate for storage speed
Discover the logcat buffer size, divide by the worst storage
speed in kb/s, and multiply by ten to estimate the timeout in
ms for each named buffer. If that value is less than 20 seconds,
then select 20 seconds. Apply this to each logcat request.
Bug:
21555259
Change-Id: Id36aebff83cb06e94350e4a84fa266b66a8444fc
Dan Stoza [Fri, 12 Jun 2015 16:47:21 +0000 (16:47 +0000)]
Merge "sf: Fix incorrect state reporting in dumpsys"
Manoj Kumar AVM [Thu, 11 Jun 2015 21:18:14 +0000 (14:18 -0700)]
sf: Fix incorrect state reporting in dumpsys
Dumpsys utility accesses layer compositionType variable
with out proper protection. These variables are modified
during hwc_prepare call. Existing HAL lock protection is not
sufficient to address this issue. Failure to do this will
result in incorrect state reporting in dumpsys.
A new displayLock mutex in HWComposer will be used in both
dumpsys and draw calls to ensure correct state is accessed.
Change-Id: I8a57de59525adc0e089b3bed95c067c01e42b666
Dan Stoza [Thu, 11 Jun 2015 19:31:43 +0000 (19:31 +0000)]
Merge "sf: Add a NULL check in getDisplayConfigs"
Tatenda Chipeperekwa [Wed, 29 Oct 2014 23:47:19 +0000 (16:47 -0700)]
sf: Add a NULL check in getDisplayConfigs
Validate the display binder by adding a NULL check in getDisplayConfigs.
This will prevent a false match if the caller queries the display
configs for an inactive display (whose binder is NULL by default).
Without this change we might end up attempting to index the display
config array, which is unpopulated for inactive displays, and this will
result in a crash. (See getDisplayInfo in SurfaceComposerClient.cpp for
an example of this scenario)
Change-Id: Ib32a7dc8378d3438df0dba1ecd608bbcfc837717
Dan Stoza [Thu, 11 Jun 2015 17:34:15 +0000 (17:34 +0000)]
Merge "GLConsumer: Fix crop math"
Dan Stoza [Tue, 9 Jun 2015 22:05:23 +0000 (15:05 -0700)]
GLConsumer: Fix crop math
When we have excess pixels that need to be removed from (for example)
the left and right sides, we currently do something like:
left += excess / 2;
right -= excess / 2;
If excess is odd, however, this removes 1 too few pixels since the odd
pixel gets rounded down twice. This changes the math to effectively:
left += excess / 2;
right -= (excess - excess / 2);
Which removes the correct total number of pixels.
Bug:
19611086
Change-Id: I8d1ad9fe7ba67c149794c148663d12acbccccef0
(cherry picked from commit
ec4cb387502d5bd09e49b55605cf679fe3a0207a)
Dan Stoza [Wed, 10 Jun 2015 18:10:07 +0000 (18:10 +0000)]
Merge "SF: Swap w/h when capturing rotated screenshots"
Dan Stoza [Tue, 9 Jun 2015 23:44:40 +0000 (16:44 -0700)]
SF: Swap w/h when capturing rotated screenshots
Swaps width and height when capturing a screenshot that is rotated by
90 or 270 degrees.
Bug:
8433742
Change-Id: Ibf0b604f541e3cc271e56fe0ad04dc366beb5d79
(cherry picked from commit
3502416204d9dbd905012ee586d8bd145323809f)
Dan Stoza [Tue, 9 Jun 2015 20:28:59 +0000 (20:28 +0000)]
Merge "SF: Don't crop surface damage against viewport"
Dan Stoza [Fri, 5 Jun 2015 02:40:03 +0000 (19:40 -0700)]
SF: Don't crop surface damage against viewport
Apply the display transform to surface damage, but don't intersect it
with the viewport.
Bug:
21486007
Change-Id: I94e697cea52096a960385d5f3b89609fdcf30a02
(cherry picked from commit
17833a1b0d70cd5cea342cadea9fdbbf16748bc6)
Dan Stoza [Tue, 9 Jun 2015 18:34:10 +0000 (18:34 +0000)]
Merge "libgui/SurfaceFlinger: Add getConsumerName"
Dan Stoza [Mon, 8 Jun 2015 16:32:50 +0000 (09:32 -0700)]
libgui/SurfaceFlinger: Add getConsumerName
Adds a getConsumerName method to IGraphicBufferProducer and Surface.
Currently, the name is cached inside of IGBP and is update on connect
and dequeueBuffer, which should be good enough for most uses.
Bug:
6667401
Change-Id: I22c7881d778e495cf8276de7bbcd769e52429915
(cherry picked from commit
c6f30bdee1f634eb90d68cb76efe935b6535a1e8)
Dan Stoza [Tue, 9 Jun 2015 18:23:48 +0000 (18:23 +0000)]
Merge "libgui: Add generation numbers to BufferQueue"
Dan Stoza [Tue, 2 Jun 2015 22:45:22 +0000 (15:45 -0700)]
libgui: Add generation numbers to BufferQueue
This change allows producers to set a generation number on a
BufferQueue. This number will be embedded in any new GraphicBuffers
created in that BufferQueue, and attempts to attach buffers which have
a different generation number will fail.
It also plumbs the setGenerationNumber method through Surface, with the
additional effect that any buffers attached to the Surface after
setting a new generation number will automatically be updated with the
new number (as opposed to failing, as would happen on through IGBP).
Bug:
20923096
Change-Id: I32bf726b035f99c3e5834beaf76afb9f01adcbc2
(cherry picked from commit
812ed0644f8f8f71ca403f4e5793f0dbc1fcf9b2)
Nick Kralevich [Wed, 3 Jun 2015 13:33:36 +0000 (13:33 +0000)]
Merge "Call selinux_status_open() prior to use of selinux_status_updated()."
Stephen Smalley [Wed, 3 Jun 2015 13:25:37 +0000 (09:25 -0400)]
Call selinux_status_open() prior to use of selinux_status_updated().
The selinux_status_*() interfaces require calling selinux_status_open()
first. Since this was not being called by servicemanager, the
selinux_status_updated() call was always failing and servicemanager
was not reloading service_contexts upon a policy reload.
Change-Id: I6ac62e114bbca99c5f0999feb918fcde15009274
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Dan Stoza [Mon, 1 Jun 2015 21:59:20 +0000 (21:59 +0000)]
Merge "libgui: Propagate error by clearing sp<GB>"