OSDN Git Service

android-x86/frameworks-native.git
8 years agoMerge "Fix the execution point of onFrameAvailable/onFrameReplaced callbacks"
Dan Stoza [Tue, 8 Dec 2015 17:05:25 +0000 (17:05 +0000)]
Merge "Fix the execution point of onFrameAvailable/onFrameReplaced callbacks"

8 years agoMerge "Replace libGLESv3 symlink with shared library"
Dimitry Ivanov [Mon, 7 Dec 2015 19:00:58 +0000 (19:00 +0000)]
Merge "Replace libGLESv3 symlink with shared library"

8 years agoFix the execution point of onFrameAvailable/onFrameReplaced callbacks
Christian Poetzsch [Mon, 7 Dec 2015 13:36:22 +0000 (13:36 +0000)]
Fix the execution point of onFrameAvailable/onFrameReplaced callbacks

In a4650a5 the concept of a maximum frame number allowance for the consumer was
introduced. A call to acquireBuffers will only return buffers when their frame
number is less-than-or-equal-to this maximum frame number. When SurfaceFlinger
is the consumer, this maximum  frame number is calculated in the
onFrameAvailable/onFrameReplaced callbacks. These callbacks are called when a
new buffer is dequeued by the application. The problem is that these callbacks
are called _after_ the fence wait which is used to throttle the frame
production of client apps. When the previous frame needs a long time to draw,
those waits can potentially be a long time. As a result SurfaceFlinger won't do
any composition with the new frame until the wait is over.

Normally this isn't a big problem because there is a queue of buffers for
SurfaceFlinger to work with. However, this changes massively when a client app
is using a swap interval of zero. In this case, a new frame will instantly
replace the previous queued frame. However, SurfaceFlinger doesn't know this
until the onFrameReplaced callback gets called - which is delayed by the fence
wait. If the timing is bad, SurfaceFlinger never gets a chance to pick up a new
frame to do the composition with.

We see this behaviour on our TC development system (slow GPU) with legacy
on-screen benchmarks. Such apps are using a swap interval of zero and sometimes
frames don't get updated for several seconds. This behaviour can be also seen
on a Nexus5, although it isn't as obvious as on our TC.

The fix in this cl is to move the EGL throttling to the end of the queueBuffers
function. This ensures that if a frame gets replaced in the queue, all
consumers who installed the callbacks, get called in a timely fashion.

Change-Id: I36e9ecda162150f41e97d4fb7437963a3d86b371
Signed-off-by: Christian Poetzsch <christian.potzsch@imgtec.com>
8 years agoMerge "Revert "libbinder: Don't redefine B_PACK_CHARS""
Bart Sears [Sat, 5 Dec 2015 03:38:59 +0000 (03:38 +0000)]
Merge "Revert "libbinder: Don't redefine B_PACK_CHARS""

8 years agoRevert "libbinder: Don't redefine B_PACK_CHARS"
Bart Sears [Sat, 5 Dec 2015 03:38:34 +0000 (03:38 +0000)]
Revert "libbinder: Don't redefine B_PACK_CHARS"

This reverts commit 3eab1b40acbcc2e82b3e90da26706129019eccf9.

Change-Id: Iedc99961c6014029722f814e621970c1732f9092

8 years agoMerge "Track rename from base/ to android-base/."
Elliott Hughes [Sat, 5 Dec 2015 01:33:52 +0000 (01:33 +0000)]
Merge "Track rename from base/ to android-base/."

8 years agoMerge "libbinder: Don't redefine B_PACK_CHARS"
Christopher Wiley [Sat, 5 Dec 2015 00:15:12 +0000 (00:15 +0000)]
Merge "libbinder: Don't redefine B_PACK_CHARS"

8 years agolibbinder: Don't redefine B_PACK_CHARS
Christopher Wiley [Fri, 4 Dec 2015 22:15:27 +0000 (14:15 -0800)]
libbinder: Don't redefine B_PACK_CHARS

This macro is defined in linux/binder.h which is *sometimes* included
before binder/IBinder.h depending on codepath.

Bug: 25868260
Test: Compiles

Change-Id: Ied1e6658936668c16cbdacab3a3205107c907655

8 years agoTrack rename from base/ to android-base/.
Elliott Hughes [Fri, 4 Dec 2015 23:39:32 +0000 (15:39 -0800)]
Track rename from base/ to android-base/.

Change-Id: I62c18891ffe319d20ff084f5ca2aa0c038c23cd3

8 years agoMerge "libbinder: Handle transaction failures correctly"
Christopher Wiley [Thu, 3 Dec 2015 21:17:42 +0000 (21:17 +0000)]
Merge "libbinder: Handle transaction failures correctly"

8 years agolibbinder: Handle transaction failures correctly
Christopher Wiley [Mon, 23 Nov 2015 00:29:58 +0000 (16:29 -0800)]
libbinder: Handle transaction failures correctly

Java code expects status_t != OK to be caught at the JNI level in
android_util_Binder.cpp (see signalExceptionForError).  We were
incorrectly mapping this kind of failure to a special exception type
and writing that exception type to parcels.

Instead, refuse to write EX_TRANSACTION_FAILED to a parcel and return
the status value instead.

While here, remove non-trivial constructors to push authors toward the
more explicit factory methods.  Remove getException() and push authors
toward using the simpler getter methods. Fix minor camelCase issues.

Bug: 25615695
Test: system/tools/aidl integration tests still pass

Change-Id: I7cad3ac8ae8300b5ac0b466606f4934d01e503c5

8 years agoMerge "bug#460672 framebuffer surface can't keep triple buffer if define NUM_FRAMEBUF...
Dan Stoza [Wed, 2 Dec 2015 18:27:05 +0000 (18:27 +0000)]
Merge "bug#460672 framebuffer surface can't keep triple buffer if define NUM_FRAMEBUFFER_SURFACE_BUFFERS := 3"

8 years agobug#460672 framebuffer surface can't keep triple buffer if define NUM_FRAMEBUFFER_SUR...
james.zhang [Tue, 1 Dec 2015 09:55:11 +0000 (17:55 +0800)]
bug#460672 framebuffer surface can't keep triple buffer if define NUM_FRAMEBUFFER_SURFACE_BUFFERS := 3

[root cause  ] framebuffer surface is in the same process of surface flinger
[changes     ] call IGraphicBufferProducer::allocateBuffers on producer inside the DisplayDevice constructor to force-allocate all three buffers.
[side effects] none
[self test   ]<tshark, enter and exit mainmenu>: ok
[reviewers   ] zhongjun.chen
[change_type ] AOB --> google_original
[tag_product ] common

Change-Id: Ideb75d3aa1ed23b2fa4d263a120ea3ca39d01361

8 years agoMerge "Atrace: support streaming data to stdout."
Martijn Coenen [Mon, 30 Nov 2015 18:51:11 +0000 (18:51 +0000)]
Merge "Atrace: support streaming data to stdout."

8 years agoAtrace: support streaming data to stdout.
Martijn Coenen [Thu, 26 Nov 2015 09:00:55 +0000 (10:00 +0100)]
Atrace: support streaming data to stdout.

Change-Id: I8534a9c69f25ecdd82230a5d8b4b8d1c4d58439e

8 years agoMerge "Add support for file descriptors and vectors thereof"
Casey Dahlin [Mon, 23 Nov 2015 22:22:45 +0000 (22:22 +0000)]
Merge "Add support for file descriptors and vectors thereof"

8 years agoAdd support for file descriptors and vectors thereof
Casey Dahlin [Mon, 23 Nov 2015 21:24:23 +0000 (13:24 -0800)]
Add support for file descriptors and vectors thereof

We use ScopedFd from nativehelper to manage the lifetime.

Test: Passes AIDL unit and integration tests. Full build completes and
      runs.
Bug: 25242023
Signed-off-by: Casey Dahlin <sadmac@google.com>
Change-Id: I22c87ed1cb3bd4bf667b372ea4a2e9fcd4dd986a

8 years agoMerge "libbinder: Add support for C++ Parcelable"
Christopher Wiley [Mon, 23 Nov 2015 18:14:26 +0000 (18:14 +0000)]
Merge "libbinder: Add support for C++ Parcelable"

8 years agoMerge "libbinder: Move vector writing templates to header"
Christopher Wiley [Fri, 20 Nov 2015 22:25:48 +0000 (22:25 +0000)]
Merge "libbinder: Move vector writing templates to header"

8 years agoMerge "Add support for new sched kernel tracepoint"
Riley Andrews [Thu, 19 Nov 2015 22:05:29 +0000 (22:05 +0000)]
Merge "Add support for new sched kernel tracepoint"

8 years agoAdd support for new sched kernel tracepoint
Riley Andrews [Thu, 19 Nov 2015 21:31:17 +0000 (13:31 -0800)]
Add support for new sched kernel tracepoint

This tracepoint gives systrace the means to display the reason
for uninterruptible sleep.

8 years agolibbinder: Add support for C++ Parcelable
Christopher Wiley [Thu, 19 Nov 2015 14:49:05 +0000 (06:49 -0800)]
libbinder: Add support for C++ Parcelable

Bug: 23600712
Test: compiles, integration tests for Java/C++ compatibility pass

Change-Id: I0919571919a3633350169ed5668bbb000da9691c

8 years agolibbinder: Move vector writing templates to header
Christopher Wiley [Thu, 19 Nov 2015 14:42:40 +0000 (06:42 -0800)]
libbinder: Move vector writing templates to header

This allows us to reuse them to read and write vectors of parcelables.

Bug: 23600712
Test: Compiles, aidl integration tests pass

Change-Id: Ibbfc2f158b6b3000b13f35f3c09a0cd3741e8128

8 years agoReplace libGLESv3 symlink with shared library
Dmitriy Ivanov [Wed, 18 Nov 2015 22:57:49 +0000 (14:57 -0800)]
Replace libGLESv3 symlink with shared library

Because of the link to libGLESv2.so, libGLESv3.so had
incorrect dt_soname, which led to inability to use
in in linker public namespaces.

This change replaces symlink with shared library.

Bug: http://b/25754041
Change-Id: I3743c81f05bc6caf746e40ef7531723e97632a70

8 years agoMerge "Revert "Add support for unique_fds and vectors of file descriptors""
Casey Dahlin [Wed, 18 Nov 2015 21:33:48 +0000 (21:33 +0000)]
Merge "Revert "Add support for unique_fds and vectors of file descriptors""

8 years agoRevert "Add support for unique_fds and vectors of file descriptors"
Casey Dahlin [Wed, 18 Nov 2015 21:30:02 +0000 (21:30 +0000)]
Revert "Add support for unique_fds and vectors of file descriptors"

This reverts commit 9626d88882972d45576dcceedc037d2e9843196b.

Change-Id: I6121f388d17c6f2d0cf6f31bc42f0804dd72aba2

8 years agoMerge "Add support for unique_fds and vectors of file descriptors"
Casey Dahlin [Wed, 18 Nov 2015 20:54:52 +0000 (20:54 +0000)]
Merge "Add support for unique_fds and vectors of file descriptors"

8 years agoAdd support for unique_fds and vectors of file descriptors
Casey Dahlin [Thu, 5 Nov 2015 20:17:29 +0000 (12:17 -0800)]
Add support for unique_fds and vectors of file descriptors

Change-Id: Ic61e69d9db438b4686032459c896aed2a482c9a5
Test: Unit tests pass
Bug: 25242023
Signed-off-by: Casey Dahlin <sadmac@google.com>
8 years agoMerge "libbinder: Add binder::Status type"
Christopher Wiley [Wed, 18 Nov 2015 16:18:30 +0000 (16:18 +0000)]
Merge "libbinder: Add binder::Status type"

8 years agolibbinder: Add binder::Status type
Christopher Wiley [Mon, 9 Nov 2015 23:06:15 +0000 (15:06 -0800)]
libbinder: Add binder::Status type

This object implements equivalent functionality to the Java logic which
serializes and re-throws exceptions from services.

Bug: 25615695
Test: Integration test for generated AIDL code reveals this to work
      correctly.

Change-Id: Ic80c9def0dd232582ea9c49717ec50894af6bfc2

8 years agoMerge "use negative mask"
Todd Kennedy [Tue, 17 Nov 2015 21:48:31 +0000 (21:48 +0000)]
Merge "use negative mask"

8 years agouse negative mask
Todd Kennedy [Mon, 16 Nov 2015 22:41:36 +0000 (14:41 -0800)]
use negative mask

the test should be for _invalid_ bits; as it's written,
it tests for valid bits

Change-Id: Id5e42e865449a7877fe58aca9fb47675fa7452c9

8 years agoMerge "Remove obsolete profile handling from installd."
Calin Juravle [Mon, 16 Nov 2015 11:07:38 +0000 (11:07 +0000)]
Merge "Remove obsolete profile handling from installd."

8 years agoMerge "Use templates for write*Vector methods"
Casey Dahlin [Fri, 13 Nov 2015 22:15:01 +0000 (22:15 +0000)]
Merge "Use templates for write*Vector methods"

8 years agoUse templates for write*Vector methods
Casey Dahlin [Fri, 13 Nov 2015 21:46:29 +0000 (13:46 -0800)]
Use templates for write*Vector methods

Change-Id: I86223bd54940ebe3e8f090b7786ddfa3ee57195e
Test: AIDL integration tests still pass
Bug: None
Signed-off-by: Casey Dahlin <sadmac@google.com>
8 years agoMerge "Move dumpstate to top -H rather than top -t."
Elliott Hughes [Fri, 13 Nov 2015 20:23:22 +0000 (20:23 +0000)]
Merge "Move dumpstate to top -H rather than top -t."

8 years agoMove dumpstate to top -H rather than top -t.
Elliott Hughes [Fri, 13 Nov 2015 19:32:48 +0000 (11:32 -0800)]
Move dumpstate to top -H rather than top -t.

Change-Id: Ibc26b51e92425b2de7870a7a2890fc3566909cda

8 years agoMerge "libbinder: Return UNEXPECTED_NULL when appropriate"
Christopher Wiley [Thu, 12 Nov 2015 22:25:04 +0000 (22:25 +0000)]
Merge "libbinder: Return UNEXPECTED_NULL when appropriate"

8 years agolibbinder: Return UNEXPECTED_NULL when appropriate
Christopher Wiley [Tue, 10 Nov 2015 17:44:30 +0000 (09:44 -0800)]
libbinder: Return UNEXPECTED_NULL when appropriate

Return a meaningful and consistent error code when libbinder encounters
a null reference that it didn't expect.  This allows us to detect these
situations and cause a corresponding NullPointerException to be thrown
on the remote callsite.

While here, consolidate places we read vector<T> from a parcel into a
single templated version.  This dramatically reduces code duplication.

Bug: 25617251
Test: Integration tests asserting that vectors serialize correctly
      passes.  Integration test asserting that we throw
      NullPointerException on seeing a null reference also passes.

Change-Id: Ia59c264884ec89539c1b99cd7bd28ac201b5b7ac

8 years agoMerge "Allow read access to /proc entries for other UIDs"
Nick Kralevich [Mon, 9 Nov 2015 20:05:20 +0000 (20:05 +0000)]
Merge "Allow read access to /proc entries for other UIDs"

8 years agoAllow read access to /proc entries for other UIDs
Nick Kralevich [Sun, 8 Nov 2015 01:05:41 +0000 (17:05 -0800)]
Allow read access to /proc entries for other UIDs

Allow dumpstate (aka adb bugreport), servicemanager, and
surfaceflinger to access /proc entries associated with other UIDs.

Bug: 23310674
Change-Id: I385dcf0db3376ba979409cbe6fe1a468d36237ef

8 years agoMerge "Add support for reading/writing a vector of binders"
Casey Dahlin [Thu, 5 Nov 2015 00:12:57 +0000 (00:12 +0000)]
Merge "Add support for reading/writing a vector of binders"

8 years agoRemove obsolete profile handling from installd.
Calin Juravle [Wed, 4 Nov 2015 14:47:37 +0000 (14:47 +0000)]
Remove obsolete profile handling from installd.

Bug: 24698874
Bug: 17173268
Change-Id: I140d70262f8f8fa7afd52c3cf63b4abaf401a479

8 years agoMerge "Add ipi tracing to irq tracing."
Riley Andrews [Wed, 4 Nov 2015 00:29:45 +0000 (00:29 +0000)]
Merge "Add ipi tracing to irq tracing."

8 years agoAdd support for reading/writing a vector of binders
Casey Dahlin [Tue, 3 Nov 2015 21:50:37 +0000 (13:50 -0800)]
Add support for reading/writing a vector of binders

Change-Id: Iaa8da704b2ae3c1ca5456177441a335991b40e8a
Test: unit tests pass
Bug: 24470786
Signed-off-by: Casey Dahlin <sadmac@google.com>
8 years agoAdd ipi tracing to irq tracing.
Riley Andrews [Tue, 3 Nov 2015 05:01:34 +0000 (21:01 -0800)]
Add ipi tracing to irq tracing.

8 years agoMerge "Fix bug in byte vector serialization"
Christopher Wiley [Mon, 2 Nov 2015 23:58:23 +0000 (23:58 +0000)]
Merge "Fix bug in byte vector serialization"
am: fe98e8d6e5

* commit 'fe98e8d6e53ffef172c4571e6c408c4801c3e96c':
  Fix bug in byte vector serialization

8 years agoMerge "Fix bug in byte vector serialization"
Christopher Wiley [Mon, 2 Nov 2015 23:53:16 +0000 (23:53 +0000)]
Merge "Fix bug in byte vector serialization"

8 years agoMerge "Further reduce duplication in bugreport ps output."
Elliott Hughes [Mon, 2 Nov 2015 22:45:36 +0000 (22:45 +0000)]
Merge "Further reduce duplication in bugreport ps output."
am: 670508d09a

* commit '670508d09acf64fad83d0c7ac90330d3db89c861':
  Further reduce duplication in bugreport ps output.

8 years agoMerge "Further reduce duplication in bugreport ps output."
Elliott Hughes [Mon, 2 Nov 2015 22:43:23 +0000 (22:43 +0000)]
Merge "Further reduce duplication in bugreport ps output."

8 years agoFurther reduce duplication in bugreport ps output.
Elliott Hughes [Mon, 2 Nov 2015 22:16:09 +0000 (14:16 -0800)]
Further reduce duplication in bugreport ps output.

PROCESSES and PROCESSES AND THREADS can be combined. There's nothing in
the former that isn't in the latter.

Change-Id: I105a6d8d35488236ef6dbf4107496ea7e5ad70ea

8 years agoFix bug in byte vector serialization
Christopher Wiley [Sat, 31 Oct 2015 20:22:15 +0000 (13:22 -0700)]
Fix bug in byte vector serialization

Byte vectors are used by the generated C++ code as the representation
of Java byte[].  This type is serialized as a packed byte array on
the java side.

Bug: 25012838
Test: integration tests for byte[] show this type crossing language
      boundaries correctly.

Change-Id: I8bb1b8ffcb77ced44f99f6b370226a32694f7df1

8 years agoMerge "Reduce duplication in bugreport ps output."
Elliott Hughes [Fri, 30 Oct 2015 23:44:50 +0000 (23:44 +0000)]
Merge "Reduce duplication in bugreport ps output."
am: 8ebad13b34

* commit '8ebad13b340921a40a367c9f1808d0c26147e2a6':
  Reduce duplication in bugreport ps output.

8 years agoMerge "Reduce duplication in bugreport ps output."
Elliott Hughes [Fri, 30 Oct 2015 23:39:29 +0000 (23:39 +0000)]
Merge "Reduce duplication in bugreport ps output."

8 years agoReduce duplication in bugreport ps output.
Elliott Hughes [Fri, 30 Oct 2015 23:17:49 +0000 (16:17 -0700)]
Reduce duplication in bugreport ps output.

PROCESSES AND THREADS and PROCESSES (SELINUX LABELS) can be combined.
I suspect we can just remove PROCESSES too, since it's a strict
subset of the other information, but one step at a time. (The worry
is that someone's actually parsing this output, but it would be good
to find out who's doing that as soon as possible so we can work with
them as we transition to a more complete ps implementation.)

Change-Id: I5f1085f8853f7a988f7edcaeefdd7b63175c1473

8 years agoMerge "Add readStrongBinder that takes an interface"
Casey Dahlin [Wed, 28 Oct 2015 23:48:12 +0000 (23:48 +0000)]
Merge "Add readStrongBinder that takes an interface"
am: 71b70a244c

* commit '71b70a244c56719d30654b7123a944d971eb8d02':
  Add readStrongBinder that takes an interface

8 years agoMerge "Add readStrongBinder that takes an interface"
Casey Dahlin [Wed, 28 Oct 2015 23:41:58 +0000 (23:41 +0000)]
Merge "Add readStrongBinder that takes an interface"

8 years agoAdd readStrongBinder that takes an interface
Casey Dahlin [Wed, 28 Oct 2015 01:33:56 +0000 (18:33 -0700)]
Add readStrongBinder that takes an interface

We use a template function to let us pass arbitrary interfaces, all of which
can be cast to IBinder.

Change-Id: Iadf21c495cde43e8a5adb85a49e6592196f401ff
Test: unit tests
Bug: 23600713
Signed-off-by: Casey Dahlin <sadmac@google.com>
8 years agoMerge "libbinder: Fix out of bounds in readBoolVector"
Christopher Wiley [Wed, 28 Oct 2015 18:20:29 +0000 (18:20 +0000)]
Merge "libbinder: Fix out of bounds in readBoolVector"
am: 1f76049c95

* commit '1f76049c955c605b61ef0d923582c740c325b1a0':
  libbinder: Fix out of bounds in readBoolVector

8 years agoMerge "libbinder: Fix out of bounds in readBoolVector"
Christopher Wiley [Wed, 28 Oct 2015 18:14:01 +0000 (18:14 +0000)]
Merge "libbinder: Fix out of bounds in readBoolVector"

8 years agolibbinder: Fix out of bounds in readBoolVector
Christopher Wiley [Tue, 27 Oct 2015 23:33:47 +0000 (16:33 -0700)]
libbinder: Fix out of bounds in readBoolVector

Bug: 25012838
Test: This code now works at runtime.

Change-Id: I3d6b2677f11276e6be4c51451ca36572a4ee573d

8 years agoMerge "surfaceflinger: Set max acquired buffer count"
Dan Stoza [Mon, 26 Oct 2015 18:02:31 +0000 (18:02 +0000)]
Merge "surfaceflinger: Set max acquired buffer count"
am: af5c3b1d6a

* commit 'af5c3b1d6a9cc1d36b526291391009b2a4e097ff':
  surfaceflinger: Set max acquired buffer count

8 years agoMerge "surfaceflinger: Set max acquired buffer count"
Dan Stoza [Mon, 26 Oct 2015 17:58:53 +0000 (17:58 +0000)]
Merge "surfaceflinger: Set max acquired buffer count"

8 years agoMerge "libbinder: Stub out checkPermissions() on Brillo."
Daniel Erat [Thu, 22 Oct 2015 01:40:35 +0000 (01:40 +0000)]
Merge "libbinder: Stub out checkPermissions() on Brillo."
am: 0d6d2917cb

* commit '0d6d2917cb22ba09da632552c7074d3aa110fbdb':
  libbinder: Stub out checkPermissions() on Brillo.

8 years agoMerge "libbinder: Stub out checkPermissions() on Brillo."
Daniel Erat [Thu, 22 Oct 2015 01:38:14 +0000 (01:38 +0000)]
Merge "libbinder: Stub out checkPermissions() on Brillo."

8 years agolibbinder: Stub out checkPermissions() on Brillo.
Daniel Erat [Tue, 13 Oct 2015 21:29:32 +0000 (15:29 -0600)]
libbinder: Stub out checkPermissions() on Brillo.

When BRILLO is defined at build-time, avoid checking
permissions with ActivityManager (which isn't built for
Brillo).

Bug: 24907738
Change-Id: I8e277fe45d73d551e3a8ab41bd9dfa2535967522

8 years agoMerge "Fix benign integer overflow in printHexDump"
Dan Austin [Wed, 21 Oct 2015 20:01:45 +0000 (20:01 +0000)]
Merge "Fix benign integer overflow in printHexDump"
am: b36c7d0b6e

* commit 'b36c7d0b6e793fae2d27b43d0022981c0d366d55':
  Fix benign integer overflow in printHexDump

8 years agoMerge "Fix benign integer overflow in printHexDump"
Dan Austin [Wed, 21 Oct 2015 19:53:56 +0000 (19:53 +0000)]
Merge "Fix benign integer overflow in printHexDump"

8 years agoFix benign integer overflow in printHexDump
Dan Austin [Wed, 21 Oct 2015 18:28:59 +0000 (11:28 -0700)]
Fix benign integer overflow in printHexDump

There is an instance where an unsigned integer is multiplied by -1.
Refactor the code to eliminate this operation.

Bug: 25085348
Change-Id: I9c5dc1cc22a27dc998ae8eeacc9f889373d993cd

8 years agoMerge "Fix build breakage due to extraneous dereference"
Casey Dahlin [Tue, 20 Oct 2015 23:50:31 +0000 (23:50 +0000)]
Merge "Fix build breakage due to extraneous dereference"
am: a763da3f1f

* commit 'a763da3f1fc010576b4807fa9b050c0720f81d00':
  Fix build breakage due to extraneous dereference

8 years agoMerge "Fix build breakage due to extraneous dereference"
Casey Dahlin [Tue, 20 Oct 2015 23:27:30 +0000 (23:27 +0000)]
Merge "Fix build breakage due to extraneous dereference"

8 years agoFix build breakage due to extraneous dereference
Casey Dahlin [Tue, 20 Oct 2015 23:26:23 +0000 (16:26 -0700)]
Fix build breakage due to extraneous dereference

Change-Id: If84d8b68a6af1e04a6085351655e588ebef65664

8 years agoMerge "Add read/write methods for vectors"
Casey Dahlin [Tue, 20 Oct 2015 23:07:29 +0000 (23:07 +0000)]
Merge "Add read/write methods for vectors"
am: d83a34d8c7

* commit 'd83a34d8c7fa4be9bd2fa3da4d3ca2c16febd651':
  Add read/write methods for vectors

8 years agoMerge "Add read/write methods for vectors"
Casey Dahlin [Tue, 20 Oct 2015 23:00:12 +0000 (23:00 +0000)]
Merge "Add read/write methods for vectors"

8 years agoAdd read/write methods for vectors
Casey Dahlin [Tue, 20 Oct 2015 01:12:18 +0000 (18:12 -0700)]
Add read/write methods for vectors

Also a readString16 with the expected semantics.

Change-Id: Idbeaff1b4337089b2a5a7124f77226a693b2be4f
Test: Unit tests pass
Bug: 25012838
Signed-off-by: Casey Dahlin <sadmac@google.com>
8 years agoam 4e6d8c5b: Merge "Add methods to Parcel for bool and char and byte"
Casey Dahlin [Fri, 16 Oct 2015 20:30:50 +0000 (13:30 -0700)]
am 4e6d8c5b: Merge "Add methods to Parcel for bool and char and byte"

* commit '4e6d8c5b9d51f11db7e8dbe91971c8b46beae5c0':
  Add methods to Parcel for bool and char and byte

8 years agoMerge "Add methods to Parcel for bool and char and byte"
Casey Dahlin [Fri, 16 Oct 2015 20:13:00 +0000 (20:13 +0000)]
Merge "Add methods to Parcel for bool and char and byte"

8 years agoAdd methods to Parcel for bool and char and byte
Casey Dahlin [Thu, 15 Oct 2015 22:44:59 +0000 (15:44 -0700)]
Add methods to Parcel for bool and char and byte

We lift these to int, but handling the details of that here will be better
than having AIDL generate casting code all the time.

Test: unit tests pass
Bug: 24981507
Change-Id: If6aa33ab5fe365dd237136a95c5d9a702af41d15
Signed-off-by: Casey Dahlin <sadmac@google.com>
8 years agoam 8e0c88ea: Merge "servicemanager: log pid and uid on selinux denial"
Nick Kralevich [Sat, 10 Oct 2015 20:11:05 +0000 (20:11 +0000)]
am 8e0c88ea: Merge "servicemanager: log pid and uid on selinux denial"

* commit '8e0c88ea489a7b395a3d349f81000fb2639bf448':
  servicemanager: log pid and uid on selinux denial

8 years agoMerge "servicemanager: log pid and uid on selinux denial"
Nick Kralevich [Sat, 10 Oct 2015 20:07:22 +0000 (20:07 +0000)]
Merge "servicemanager: log pid and uid on selinux denial"

8 years agoam 52b25375: Merge "dumpstate: add -v printable to logcat"
Mark Salyzyn [Fri, 9 Oct 2015 21:25:32 +0000 (21:25 +0000)]
am 52b25375: Merge "dumpstate: add -v printable to logcat"

* commit '52b2537550a6fc43587980887858c7e7150255b0':
  dumpstate: add -v printable to logcat

8 years agoMerge "dumpstate: add -v printable to logcat"
Mark Salyzyn [Fri, 9 Oct 2015 21:20:05 +0000 (21:20 +0000)]
Merge "dumpstate: add -v printable to logcat"

8 years agodumpstate: add -v printable to logcat
Mark Salyzyn [Fri, 9 Oct 2015 21:02:07 +0000 (14:02 -0700)]
dumpstate: add -v printable to logcat

In the case where binary data lands in an utf-8 log buffer,
ensure that the content is made printable.

Bug: 23517551
Change-Id: I41696372d07228875d01a19f0c7ab2cb7494b7ef

8 years agoam ae0405d6: Merge "dumpstate: execute procrank using su"
Nick Kralevich [Thu, 8 Oct 2015 00:19:53 +0000 (00:19 +0000)]
am ae0405d6: Merge "dumpstate: execute procrank using su"

* commit 'ae0405d607f9ed08d708ee990716a7881a6cfd4d':
  dumpstate: execute procrank using su

8 years agoMerge "dumpstate: execute procrank using su"
Nick Kralevich [Thu, 8 Oct 2015 00:08:35 +0000 (00:08 +0000)]
Merge "dumpstate: execute procrank using su"

8 years agodumpstate: execute procrank using su
Nick Kralevich [Wed, 7 Oct 2015 23:38:42 +0000 (16:38 -0700)]
dumpstate: execute procrank using su

procrank is a setuid binary which only exists on userdebug/eng
builds. Instead of executing the setuid binary, run the binary using
the su command. This eliminates one more setuid binary, and allows
the tightening of the SELinux policy.

Bug: 18342188
Change-Id: I81d06196a29489feaefb2fea0cf5c4ff1e273d2f

8 years agoservicemanager: log pid and uid on selinux denial
William Roberts [Fri, 2 Oct 2015 05:02:15 +0000 (22:02 -0700)]
servicemanager: log pid and uid on selinux denial

The audit logs for servicemanager were missing the pid and uid of the source.
This is useful for debugging.

Before:
... SELinux : avc:  denied  { find } for service=android.security.keystore scontext=u:r:system_app:s0 ...
After:
... SELinux : avc:  denied  { find } for service=android.security.keystore pid=1252 uid=1000 scontext=u:r:system_app:s0 ...

Change-Id: Id1a6f38b99f11a31315439620ead2f01108b18b2
Signed-off-by: William Roberts <william.c.roberts@intel.com>
8 years agoam ddf5f250: am 05c86225: do not dumpstate raft BUG: 24159112
Erik Wolsheimer [Fri, 2 Oct 2015 17:43:26 +0000 (17:43 +0000)]
am ddf5f250: am 05c86225: do not dumpstate raft BUG: 24159112

* commit 'ddf5f2509b48ee85dc352ee245f34a9d6307be39':
  do not dumpstate raft BUG: 24159112

8 years agoam 05c86225: do not dumpstate raft BUG: 24159112
Erik Wolsheimer [Fri, 2 Oct 2015 17:37:05 +0000 (17:37 +0000)]
am 05c86225: do not dumpstate raft BUG: 24159112

* commit '05c862252d5021f55a69281473c67d76e7097422':
  do not dumpstate raft BUG: 24159112

8 years agodo not dumpstate raft
Erik Wolsheimer [Fri, 2 Oct 2015 17:02:24 +0000 (10:02 -0700)]
do not dumpstate raft
BUG: 24159112

Change-Id: I344aa73808992b5f84cb0536201a3ab2eb6f2b3c

8 years agoam c9b39239: am 6c51a07d: am de18f6c3: InputResampling: Don\'t extrapolate for very...
Andrew de los Reyes [Fri, 2 Oct 2015 08:26:24 +0000 (08:26 +0000)]
am c9b39239: am 6c51a07d: am de18f6c3: InputResampling: Don\'t extrapolate for very low frame rates.

* commit 'c9b392395f3eada767cf9d1dce1eb40930ca1b95':
  InputResampling: Don't extrapolate for very low frame rates.

8 years agoam 6c51a07d: am de18f6c3: InputResampling: Don\'t extrapolate for very low frame...
Andrew de los Reyes [Fri, 2 Oct 2015 02:22:46 +0000 (02:22 +0000)]
am 6c51a07d: am de18f6c3: InputResampling: Don\'t extrapolate for very low frame rates.

* commit '6c51a07d30a3c4bed9675da88c5f6f24a741d9a2':
  InputResampling: Don't extrapolate for very low frame rates.

8 years agoam de18f6c3: InputResampling: Don\'t extrapolate for very low frame rates.
Andrew de los Reyes [Fri, 2 Oct 2015 02:12:59 +0000 (02:12 +0000)]
am de18f6c3: InputResampling: Don\'t extrapolate for very low frame rates.

* commit 'de18f6c32add6fb22065807a00ddc88b363df527':
  InputResampling: Don't extrapolate for very low frame rates.

8 years agoInputResampling: Don't extrapolate for very low frame rates.
Andrew de los Reyes [Thu, 1 Oct 2015 22:57:25 +0000 (15:57 -0700)]
InputResampling: Don't extrapolate for very low frame rates.

In very low framerate situations, extrapolation is generally going to
either cause no benefit or make a mistake. We can safely turn it off
with no user-visible negative impact.

BUG=https://buganizer.corp.google.com/u/0/issues/24550942
TEST=Scrolled very slowly and saw mispredictions on Angler. With change,
saw the log message that the mispredictions were suppressed.

Change-Id: Ic9747d3ff098d7918047ada2ed1c2d21282c65b0

8 years agoam caf1bb7f: am bd55551d: Merge "Add more methods to IPowerManager."
Daniel Erat [Wed, 30 Sep 2015 21:33:29 +0000 (21:33 +0000)]
am caf1bb7f: am bd55551d: Merge "Add more methods to IPowerManager."

* commit 'caf1bb7f2bcfc9183208a360e51f9d54268265b0':
  Add more methods to IPowerManager.

8 years agoam bd55551d: Merge "Add more methods to IPowerManager."
Daniel Erat [Wed, 30 Sep 2015 21:26:39 +0000 (21:26 +0000)]
am bd55551d: Merge "Add more methods to IPowerManager."

* commit 'bd55551d80832f62931ccb2d54d9a239822eeb20':
  Add more methods to IPowerManager.

8 years agoMerge "Add more methods to IPowerManager."
Daniel Erat [Wed, 30 Sep 2015 21:21:23 +0000 (21:21 +0000)]
Merge "Add more methods to IPowerManager."

8 years agoAdd more methods to IPowerManager.
Daniel Erat [Wed, 30 Sep 2015 20:49:15 +0000 (14:49 -0600)]
Add more methods to IPowerManager.

Add the full list of transaction IDs from IPowerManager.aidl
to IPowerManager.h and make BpPowerManager support calling
goToSleep, reboot, shutdown, and crash. These are currently
needed by or likely to be needed by Brillo.

Bug: 22122485
Change-Id: I19abd3587c9d53b28ec150210e07f97517ee4ff4

8 years agoam b450d9d8: am 77f147ff: Merge "frameworks/native: remove superfluous mention of...
Sergio Giro [Wed, 30 Sep 2015 20:16:55 +0000 (20:16 +0000)]
am b450d9d8: am 77f147ff: Merge "frameworks/native: remove superfluous mention of SharedBuffer in Region.h"

* commit 'b450d9d8940f7406def0b1ccd7c139bb7203115f':
  frameworks/native: remove superfluous mention of SharedBuffer in Region.h

8 years agoam 77f147ff: Merge "frameworks/native: remove superfluous mention of SharedBuffer...
Sergio Giro [Wed, 30 Sep 2015 20:10:52 +0000 (20:10 +0000)]
am 77f147ff: Merge "frameworks/native: remove superfluous mention of SharedBuffer in Region.h"

* commit '77f147ff566d350dc23c3a9807ea1262f056ff2f':
  frameworks/native: remove superfluous mention of SharedBuffer in Region.h