OSDN Git Service

android-x86/frameworks-base.git
9 years agoMerge "Stop using apache http to download Gps XTRA data."
Narayan Kamath [Mon, 27 Oct 2014 16:16:10 +0000 (16:16 +0000)]
Merge "Stop using apache http to download Gps XTRA data."

9 years agoMerge "PreInitializeNativeBridge only if its Available"
Calin Juravle [Mon, 27 Oct 2014 15:16:03 +0000 (15:16 +0000)]
Merge "PreInitializeNativeBridge only if its Available"

9 years agoPreInitializeNativeBridge only if its Available
Calin Juravle [Fri, 24 Oct 2014 15:16:49 +0000 (16:16 +0100)]
PreInitializeNativeBridge only if its Available

Also, assert that we always have a private application directory for non
system_server case.

Change-Id: I8397d7f7b04cde0bf65b60ca415bf3f4b349783c

9 years agoStop using apache http to download Gps XTRA data.
Narayan Kamath [Fri, 24 Oct 2014 12:23:45 +0000 (13:23 +0100)]
Stop using apache http to download Gps XTRA data.

Also, general cleanup of unused parameters.

Change-Id: Iebd1aa05de9cd676b0b6d08c99df8c15b8e8473c

9 years agoMerge "Migrate off timeFormat12 and timeFormat24."
Elliott Hughes [Thu, 23 Oct 2014 18:21:56 +0000 (18:21 +0000)]
Merge "Migrate off timeFormat12 and timeFormat24."

9 years agoMigrate off timeFormat12 and timeFormat24.
Elliott Hughes [Thu, 23 Oct 2014 18:06:43 +0000 (11:06 -0700)]
Migrate off timeFormat12 and timeFormat24.

libcore now offers a wider variety of 12-/24-hour time formats,
so be more specific about which one we want here.

(cherry-pick of 85f60d3a03b5b5d9a0e8b8a138eb85a6b53a1eca.)

Bug: 10361358
Change-Id: I846ab7a6f84cd49e876ad21e9366aff1600e0530

9 years agoMerge "Doesn't re-sort app windows when move task to top"
Craig Mautner [Wed, 22 Oct 2014 21:19:43 +0000 (21:19 +0000)]
Merge "Doesn't re-sort app windows when move task to top"

9 years agoDoesn't re-sort app windows when move task to top
tingna_sung [Mon, 20 Oct 2014 07:38:21 +0000 (15:38 +0800)]
Doesn't re-sort app windows when move task to top

To avoid app window isn't moved by new task order when move task to top

Change-Id: I5f8a901bb2f20a163b7fa0a59d1dc58c1fca5f1f
Signed-off-by: tingna_sung <tingna_sung@htc.com>
9 years agoMerge "Avoid to skip app window due to un-assigned sending to bottom state while...
Craig Mautner [Tue, 21 Oct 2014 19:24:17 +0000 (19:24 +0000)]
Merge "Avoid to skip app window due to un-assigned sending to bottom state while app window sort"

9 years agoMerge "Include <time.h> for clock_nanosleep."
Elliott Hughes [Tue, 21 Oct 2014 17:49:02 +0000 (17:49 +0000)]
Merge "Include <time.h> for clock_nanosleep."

9 years agoInclude <time.h> for clock_nanosleep.
Elliott Hughes [Tue, 21 Oct 2014 17:41:33 +0000 (10:41 -0700)]
Include <time.h> for clock_nanosleep.

Change-Id: If3098fa03003bce5d06d60e9129c10cce3c850bb

9 years agoAvoid to skip app window due to un-assigned sending to bottom state while app window...
tingna_sung [Mon, 6 Oct 2014 08:36:40 +0000 (16:36 +0800)]
Avoid to skip app window due to un-assigned sending to bottom state while app window sort

should assign value of WindowToken.sendingToBottom according to the state of this token is moved to top or bottom
to avoid app window is skiped incorrectly due to sending to bottom state is un-assigned while app window sort
and finding focus window.

Change-Id: I190838d1d01ca229155d5176930178ab41335556
Signed-off-by: tingna_sung <tingna_sung@htc.com>
9 years agoMerge "More LOCAL_ADDITIONAL_DEPENDENCIES."
Dan Albert [Tue, 21 Oct 2014 16:03:10 +0000 (16:03 +0000)]
Merge "More LOCAL_ADDITIONAL_DEPENDENCIES."

9 years agoMore LOCAL_ADDITIONAL_DEPENDENCIES.
Dan Albert [Tue, 21 Oct 2014 15:58:39 +0000 (08:58 -0700)]
More LOCAL_ADDITIONAL_DEPENDENCIES.

Change-Id: I9bf20507442b0722e93446fdb32d8236be974e28

9 years agoMerge "Switch to libsigchain.a"
Dmitriy Ivanov [Tue, 21 Oct 2014 03:14:27 +0000 (03:14 +0000)]
Merge "Switch to libsigchain.a"

9 years agoMerge "Fix an overflow found by ASAN."
Dan Albert [Mon, 20 Oct 2014 21:48:43 +0000 (21:48 +0000)]
Merge "Fix an overflow found by ASAN."

9 years agoFix an overflow found by ASAN.
Dan Albert [Mon, 20 Oct 2014 21:44:39 +0000 (14:44 -0700)]
Fix an overflow found by ASAN.

Sometimes len is 0.

Change-Id: I031000188edc57663a5519b00cbbb751600d32f4

9 years agoSwitch to libsigchain.a
Dmitriy Ivanov [Mon, 20 Oct 2014 21:14:39 +0000 (14:14 -0700)]
Switch to libsigchain.a

 Use LOCAL_WHOLE_STATIC_LIBRARIES to instruct static linker
 to export symbols from the static library.

Change-Id: I465806dcffc777a682523f74ae6f7ab61aaa6962

9 years agoMerge "Replacing FloatMath native implementation with calls to Math"
Neil Fuller [Mon, 20 Oct 2014 10:08:10 +0000 (10:08 +0000)]
Merge "Replacing FloatMath native implementation with calls to Math"

9 years agoReplacing FloatMath native implementation with calls to Math
Nick Kralevich [Sat, 18 Oct 2014 21:10:46 +0000 (14:10 -0700)]
Replacing FloatMath native implementation with calls to Math

On modern versions of Android running in AOT mode
FloatMath is slower than Math. Calls to Math.sqrt(),
etc. are replaced by intrinsics which can be as small
as a single CPU opcode.

When running in interpreted mode the new
implementation is unfortunately slower, but I'm
judging this acceptable and likely to be improved
over time. This change saves a small amount of native
code.

Example timings:

Mako AOSP AOT:

Method: Original / New / Direct call to Math
ceil: 596ns / 146.ns / 111ns
sqrt: 694ns / 56ns / 25ns

Mako AOSP interpreted:

Method: Original / New / Direct call to Math
ceil: 1900ns / 2307ns / 1485ns
sqrt: 1998ns / 2603ns / 1788ns

Other calls Mako AOT:

Method: Original / New
cos: 635ns / 270ns
exp: 566ns / 324ns
floor: 604ns / 150ns
hypot: 631ns / 232ns
pow: 936ns / 643ns
sin: 641ns / 299ns

The advice to use Math directly, in preference to
FloatMath, is still good. FloatMath will be deprecated
separately.

Bug: https://code.google.com/p/android/issues/detail?id=36199
Change-Id: If07fcbd78543d13bc6d75f9743f999860e8d58d7

9 years agoMerge "Revert "Replacing FloatMath native implementation with calls to Math""
Elliott Hughes [Sat, 18 Oct 2014 20:43:07 +0000 (20:43 +0000)]
Merge "Revert "Replacing FloatMath native implementation with calls to Math""

9 years agoRevert "Replacing FloatMath native implementation with calls to Math"
Elliott Hughes [Sat, 18 Oct 2014 20:41:27 +0000 (20:41 +0000)]
Revert "Replacing FloatMath native implementation with calls to Math"

This reverts commit 32b88b48daa7383880088246d7222dd93cf55285.

Change-Id: I40ea17f3a68f2b53eb013aebd3ba1becdc724f94

9 years agoMerge "Removing some more FloatMath references"
Neil Fuller [Fri, 17 Oct 2014 14:48:16 +0000 (14:48 +0000)]
Merge "Removing some more FloatMath references"

9 years agoMerge "Clean up unused parameter warnings."
Dan Albert [Fri, 17 Oct 2014 00:03:05 +0000 (00:03 +0000)]
Merge "Clean up unused parameter warnings."

9 years agoClean up unused parameter warnings.
Dan Albert [Thu, 16 Oct 2014 23:55:29 +0000 (16:55 -0700)]
Clean up unused parameter warnings.

Change-Id: Ifbf310a0c10bd6afc3099bcad1026990bd1e6a22

9 years agoMerge "Replacing FloatMath native implementation with calls to Math"
Neil Fuller [Thu, 16 Oct 2014 08:16:23 +0000 (08:16 +0000)]
Merge "Replacing FloatMath native implementation with calls to Math"

9 years agoMerge "Fix log call output format error. (DO NOT MERGE)"
Chih-Hung Hsieh [Wed, 15 Oct 2014 17:03:06 +0000 (17:03 +0000)]
Merge "Fix log call output format error. (DO NOT MERGE)"

9 years agoFix log call output format error. (DO NOT MERGE)
Chih-Hung Hsieh [Wed, 8 Oct 2014 17:50:23 +0000 (10:50 -0700)]
Fix log call output format error. (DO NOT MERGE)

Cherry picked from googleplex-android c/445200.

BUG: 17916851
Change-Id: If4077e94d6dc7ef31ecf09cbcf41091c7e30268f

9 years agoMerge "Fix Time docs to document various error cases"
Neil Fuller [Wed, 15 Oct 2014 15:56:05 +0000 (15:56 +0000)]
Merge "Fix Time docs to document various error cases"

9 years agoFix Time docs to document various error cases
Neil Fuller [Tue, 30 Sep 2014 10:47:19 +0000 (11:47 +0100)]
Fix Time docs to document various error cases

Bug: https://code.google.com/p/android/issues/detail?id=76386
Change-Id: Ibb71405b70453d5abaf45426ec4e36261b6695cf

9 years agoReplacing FloatMath native implementation with calls to Math
Neil Fuller [Mon, 29 Sep 2014 16:56:01 +0000 (17:56 +0100)]
Replacing FloatMath native implementation with calls to Math

On modern versions of Android running in AOT mode
FloatMath is slower than Math. Calls to Math.sqrt(),
etc. are replaced by intrinsics which can be as small
as a single CPU opcode.

When running in interpreted mode the new
implementation is unfortunately slower, but I'm
judging this acceptable and likely to be improved
over time. This change saves a small amount of native
code.

Example timings:

Mako AOSP AOT:

Method: Original / New / Direct call to Math
ceil: 596ns / 146.ns / 111ns
sqrt: 694ns / 56ns / 25ns

Mako AOSP interpreted:

Method: Original / New / Direct call to Math
ceil: 1900ns / 2307ns / 1485ns
sqrt: 1998ns / 2603ns / 1788ns

Other calls Mako AOT:

Method: Original / New
cos: 635ns / 270ns
exp: 566ns / 324ns
floor: 604ns / 150ns
hypot: 631ns / 232ns
pow: 936ns / 643ns
sin: 641ns / 299ns

The advice to use Math directly, in preference to
FloatMath, is still good. FloatMath will be deprecated
separately.

Bug: https://code.google.com/p/android/issues/detail?id=36199
Change-Id: I8d1947d88b3c576643138b1df589fb9da7c1ab88

9 years agoRemoving some more FloatMath references
Neil Fuller [Fri, 3 Oct 2014 08:38:24 +0000 (09:38 +0100)]
Removing some more FloatMath references

See frameworks/base commit 33253a4baa6279f81a73425b49dfb6abe5f5416e
for details.

Bug: https://code.google.com/p/android/issues/detail?id=36199
Change-Id: I46d4ee4c4be7972e3bcc6782fb50f024b6fff1ee

9 years agoMerge "Link sigchain to main executable"
Dmitriy Ivanov [Tue, 14 Oct 2014 17:37:34 +0000 (17:37 +0000)]
Merge "Link sigchain to main executable"

9 years agoLink sigchain to main executable
Dmitriy Ivanov [Tue, 7 Oct 2014 20:34:42 +0000 (13:34 -0700)]
Link sigchain to main executable

 Also export necessary symbols to preempt calls from
 libart.so

Bug: 15345057
Bug: 15426766
Change-Id: If0b7f59a59824e30aa0c33dad76c7a44932180de

9 years agoMerge "Track change to private libcore data structure."
Elliott Hughes [Thu, 9 Oct 2014 18:08:57 +0000 (18:08 +0000)]
Merge "Track change to private libcore data structure."

9 years agoTrack change to private libcore data structure.
Elliott Hughes [Wed, 8 Oct 2014 23:07:33 +0000 (16:07 -0700)]
Track change to private libcore data structure.

Bug: 15476051
Change-Id: I4e28ecfa3a1d03c21f9bbde7f3daf79a79d96fe6

9 years agoMerge "AudioService: Fix monitorRotation for landscape applications"
Glenn Kasten [Thu, 2 Oct 2014 16:19:12 +0000 (16:19 +0000)]
Merge "AudioService: Fix monitorRotation for landscape applications"

9 years agoMerge "Switch from FloatMath -> Math and Math.hypot where possible"
Neil Fuller [Thu, 2 Oct 2014 10:14:18 +0000 (10:14 +0000)]
Merge "Switch from FloatMath -> Math and Math.hypot where possible"

9 years agoMerge "Frameworks/base: Fix more aapt issues"
Andreas Gampe [Thu, 2 Oct 2014 06:58:38 +0000 (06:58 +0000)]
Merge "Frameworks/base: Fix more aapt issues"

9 years agoFrameworks/base: Fix more aapt issues
Andreas Gampe [Thu, 2 Oct 2014 02:07:51 +0000 (19:07 -0700)]
Frameworks/base: Fix more aapt issues

Change-Id: I6da7dc674e9f9e3b0886a3cfd1e3194d6b5ac948

9 years agoMerge "Frameworks/base: Fix AAPT warnings"
Andreas Gampe [Wed, 1 Oct 2014 23:34:31 +0000 (23:34 +0000)]
Merge "Frameworks/base: Fix AAPT warnings"

9 years agoFrameworks/base: Fix AAPT warnings
Andreas Gampe [Wed, 1 Oct 2014 03:55:57 +0000 (20:55 -0700)]
Frameworks/base: Fix AAPT warnings

Turn on -Wall -Werror. Fix warnings.

Change-Id: I287fb3c1e851c654479bcf9ea8c73bd354a6b2a1

9 years agoMerge "Use std::sort instead of qsort_r wrapper."
Dan Albert [Wed, 1 Oct 2014 20:30:51 +0000 (20:30 +0000)]
Merge "Use std::sort instead of qsort_r wrapper."

9 years agoUse std::sort instead of qsort_r wrapper.
Dan Albert [Wed, 1 Oct 2014 18:34:17 +0000 (11:34 -0700)]
Use std::sort instead of qsort_r wrapper.

The HAVE_(BSD|GNU)_QSORT_R in AndroidConfig.h has never worked, and we
should just be using something standard anyway.

Change-Id: I784d6212f6c890a11c4af2f83d1ce2d279708652

9 years agoSwitch from FloatMath -> Math and Math.hypot where possible
Neil Fuller [Wed, 1 Oct 2014 10:55:10 +0000 (11:55 +0100)]
Switch from FloatMath -> Math and Math.hypot where possible

The motivation is an API change: FloatMath is going to be
deprecated and/or removed. Performance is not the goal of
this change.

That said...

Math is faster than FloatMath with AOT compilation.

While making the change, occurances of:

{Float}Math.sqrt(x * x + y * y) and
{Float}Math.sqrt({Float}Math.pow(x, 2) + {Float}Math.pow(y, 2))

have been replaced with:

{(float)} Math.hypot(x, y)

Right now there is no runtime intrinsic for hypot so is not faster
in all cases for AOT compilation:

Math.sqrt(x * x + y * y) is faster than Math.hypot(x, y) with
AOT, but all other combinations of FloatMath, use of pow() etc.
are slower than hypot().

hypot() has the advantage of being self documenting and
could be optimized in future. None of the behavior differences
around NaN and rounding appear to be important for the cases
looked at: they all assume results and arguments are in range
and usually the results are cast to float.

Different implementations measured on hammerhead / L:

AOT compiled:

[FloatMath.hypot(x, y)]
benchmark=Hypot_FloatMathHypot} 633.85 ns; σ=0.32 ns @ 3 trials

[FloatMath.sqrt(x*x + y*y)]
benchmark=Hypot_FloatMathSqrtMult} 684.17 ns; σ=4.83 ns @ 3 trials

[FloatMath.sqrt(FloatMath.pow(x, 2) + FloatMath.pow(y, 2))]
benchmark=Hypot_FloatMathSqrtPow} 1270.65 ns; σ=12.20 ns @ 6 trials

[(float) Math.hypot(x, y)]
benchmark=Hypot_MathHypot} 96.80 ns; σ=0.05 ns @ 3 trials

[(float) Math.sqrt(x*x + y*y)]
benchmark=Hypot_MathSqrtMult} 23.97 ns; σ=0.01 ns @ 3 trials

[(float) Math.sqrt(Math.pow(x, 2) + Math.pow(y, 2))]
benchmark=Hypot_MathSqrtPow} 156.19 ns; σ=0.12 ns @ 3 trials

Interpreter:

benchmark=Hypot_FloatMathHypot} 1180.54 ns; σ=5.13 ns @ 3 trials
benchmark=Hypot_FloatMathSqrtMult} 1121.05 ns; σ=3.80 ns @ 3 trials
benchmark=Hypot_FloatMathSqrtPow} 3327.14 ns; σ=7.33 ns @ 3 trials
benchmark=Hypot_MathHypot} 856.57 ns; σ=1.41 ns @ 3 trials
benchmark=Hypot_MathSqrtMult} 1028.92 ns; σ=9.11 ns @ 3 trials
benchmark=Hypot_MathSqrtPow} 2539.47 ns; σ=24.44 ns @ 3 trials

Bug: https://code.google.com/p/android/issues/detail?id=36199
Change-Id: I06c91f682095e627cb547d60d936ef87941be692

9 years agoMerge "Use strerror(3) when reporting zygote failures."
Elliott Hughes [Tue, 30 Sep 2014 17:03:37 +0000 (17:03 +0000)]
Merge "Use strerror(3) when reporting zygote failures."

9 years agoUse strerror(3) when reporting zygote failures.
Elliott Hughes [Tue, 30 Sep 2014 15:49:01 +0000 (08:49 -0700)]
Use strerror(3) when reporting zygote failures.

Bug: 17700596
Change-Id: I39a72ab6f19063de3c90c9fccdb749dc9a86324e

9 years agoMerge "Frameworks/base: Fix services/jni warnings"
Andreas Gampe [Mon, 29 Sep 2014 18:10:26 +0000 (18:10 +0000)]
Merge "Frameworks/base: Fix services/jni warnings"

9 years agoMerge "Allow AOSP to boot."
Dan Albert [Thu, 25 Sep 2014 18:50:18 +0000 (18:50 +0000)]
Merge "Allow AOSP to boot."

9 years agoAllow AOSP to boot.
Dan Albert [Thu, 25 Sep 2014 18:47:09 +0000 (11:47 -0700)]
Allow AOSP to boot.

HAVE_OOM_ADJ is really __linux__, and this code should never be run on
another platform.

Bug: 17641445
Change-Id: Ib8174106b6d03c9c2429794a4f582c09cb308646

9 years agoFrameworks/base: Fix services/jni warnings
Andreas Gampe [Thu, 25 Sep 2014 04:39:34 +0000 (21:39 -0700)]
Frameworks/base: Fix services/jni warnings

Update code for unused variables.

Ignore warnings generated by pulling in Skia.

Change-Id: I5b33c9b003c90bcabcea0761304c4c99cf5e39ec

9 years agoMerge "Frameworks/base: Early init native bridge"
Andreas Gampe [Wed, 24 Sep 2014 21:59:50 +0000 (21:59 +0000)]
Merge "Frameworks/base: Early init native bridge"

9 years agoMerge "Fix build. Cherry-pick 2393c364b0f727c6b904febba32ad2ddc4142c96 - DO NOT MERGE"
Bill Yi [Wed, 24 Sep 2014 16:46:40 +0000 (16:46 +0000)]
Merge "Fix build. Cherry-pick 2393c364b0f727c6b904febba32ad2ddc4142c96 - DO NOT MERGE"

9 years agoFix build. Cherry-pick 2393c364b0f727c6b904febba32ad2ddc4142c96 - DO NOT MERGE
Bill Yi [Wed, 24 Sep 2014 16:17:57 +0000 (09:17 -0700)]
Fix build. Cherry-pick 2393c364b0f727c6b904febba32ad2ddc4142c96 - DO NOT MERGE

Change-Id: I30305a3f821b4f75f5037ebf11b74cfb8f2c8bb2

9 years agoAudioService: Fix monitorRotation for landscape applications
Jon Eklund [Thu, 23 Jan 2014 23:53:48 +0000 (17:53 -0600)]
AudioService: Fix monitorRotation for landscape applications

Current implmntation only sends rotation updates on orientation
changes, so does not handle direct 0<->180 or 90<->270 transitions.

Update rotation based on an OrientationEventListener instead of
Intent.ACTION_CONFIGURATION_CHANGED

Bug 17606902

Change-Id: I01dfcd1c587f5b2e8a96365c2389782ad77936ef

9 years agoFrameworks/base: Early init native bridge
jgu21 [Wed, 10 Sep 2014 10:55:07 +0000 (06:55 -0400)]
Frameworks/base: Early init native bridge

Add the app directory to the arguments for starting a process.
Add a check for NeedsNativeBridge and a call to PreInitializeBridge
in the native fork code.

Change-Id: I0b93da93251c6b4638de786bf98cf99df07c3fc2

9 years agoMerge "Broadcast an intent to mediascanner after executing screencap"
Marco Nelissen [Mon, 22 Sep 2014 14:34:01 +0000 (14:34 +0000)]
Merge "Broadcast an intent to mediascanner after executing screencap"

9 years agoMerge "Remove MALLOC_LEAK_CHECK references."
Christopher Ferris [Sat, 20 Sep 2014 01:23:26 +0000 (01:23 +0000)]
Merge "Remove MALLOC_LEAK_CHECK references."

9 years agoRemove MALLOC_LEAK_CHECK references.
Christopher Ferris [Sat, 20 Sep 2014 01:08:29 +0000 (18:08 -0700)]
Remove MALLOC_LEAK_CHECK references.

Bug: 17378595
Change-Id: Ie7f8af88c7f3b0c3dc2c2e75bbaaa9e6a81269ae

9 years agoMerge "Remove MALLOC_LEAK_CHECK references."
Christopher Ferris [Sat, 20 Sep 2014 00:42:22 +0000 (00:42 +0000)]
Merge "Remove MALLOC_LEAK_CHECK references."

9 years agoMerge "Use constants instead of resources for ellipsis."
Elliott Hughes [Fri, 19 Sep 2014 22:02:45 +0000 (22:02 +0000)]
Merge "Use constants instead of resources for ellipsis."

9 years agoUse constants instead of resources for ellipsis.
Igor Viarheichyk [Thu, 11 Sep 2014 22:07:05 +0000 (15:07 -0700)]
Use constants instead of resources for ellipsis.

All supported locales use only U+2025 and U+2026 to represent
ellipses, and it will unlikely change in future. Given translated
resources are inconsistent and often use three dots it is safer
to use constants instead of resources.

Change-Id: I51a6cb903f62f739fbadd6b78e5765c0028d641a

9 years agoRemove MALLOC_LEAK_CHECK references.
Christopher Ferris [Fri, 19 Sep 2014 19:24:59 +0000 (12:24 -0700)]
Remove MALLOC_LEAK_CHECK references.

Bug: 17378595
Change-Id: Ib9680ef97dfea6bc20d96f0bed90f4553186315c

9 years agoMerge "Clean up makefile cruft."
Dan Albert [Thu, 11 Sep 2014 23:26:26 +0000 (23:26 +0000)]
Merge "Clean up makefile cruft."

9 years agoClean up makefile cruft.
Dan Albert [Thu, 11 Sep 2014 23:20:16 +0000 (16:20 -0700)]
Clean up makefile cruft.

The build system already takes care of all these things.

Change-Id: Idef80b554e07cef5e5ed2adf1b5740a76f13ef6d

9 years agoMerge "Clean up test makefiles."
Dan Albert [Thu, 11 Sep 2014 23:21:13 +0000 (23:21 +0000)]
Merge "Clean up test makefiles."

9 years agoClean up test makefiles.
Dan Albert [Tue, 9 Sep 2014 01:41:01 +0000 (18:41 -0700)]
Clean up test makefiles.

The build system takes care of linking the correct STL and gtest
libraries for you, and specifying them manually confuses the build
system when using libc++.

Change-Id: I9f76ab26a63ace51293614cfb5ca002f37438e02

9 years agoMerge "Make a field final."
Ian Rogers [Thu, 11 Sep 2014 20:01:21 +0000 (20:01 +0000)]
Merge "Make a field final."

9 years agoMerge "Don't check a staticly allocated array for null."
Dan Albert [Wed, 10 Sep 2014 17:31:55 +0000 (17:31 +0000)]
Merge "Don't check a staticly allocated array for null."

9 years agoDon't check a staticly allocated array for null.
Dan Albert [Wed, 10 Sep 2014 15:17:26 +0000 (08:17 -0700)]
Don't check a staticly allocated array for null.

Change-Id: I5568174f1a9d53078c2169cfd3401e43d88a7643

9 years agoMerge "Don't check for null references."
Dan Albert [Tue, 9 Sep 2014 16:25:41 +0000 (16:25 +0000)]
Merge "Don't check for null references."

9 years agoMerge "Use char16_t for char things."
Dan Albert [Tue, 9 Sep 2014 04:30:34 +0000 (04:30 +0000)]
Merge "Use char16_t for char things."

9 years agoMerge "Don't pollute the namespace for two uses."
Dan Albert [Tue, 9 Sep 2014 04:15:29 +0000 (04:15 +0000)]
Merge "Don't pollute the namespace for two uses."

9 years agoMerge "Make string literal concatenation play nice with C++11."
Dan Albert [Tue, 9 Sep 2014 03:58:57 +0000 (03:58 +0000)]
Merge "Make string literal concatenation play nice with C++11."

9 years agoDon't pollute the namespace for two uses.
Dan Albert [Tue, 9 Sep 2014 01:44:12 +0000 (18:44 -0700)]
Don't pollute the namespace for two uses.

Clang complains about ambiguity between std::hash and ::hash (locally
defined). There's no need to pull in the whole namespace for two uses of
std::map anyway.

Change-Id: Icbaa7ebbaad05999988784ad34662c721fb12b29

9 years agoMake string literal concatenation play nice with C++11.
Dan Albert [Tue, 9 Sep 2014 01:38:35 +0000 (18:38 -0700)]
Make string literal concatenation play nice with C++11.

In C++11 mode, "foo"MACRO_THAT_EXPANDS_TO_STRING gets lexed as a user
defined literal. Add space around the macro.

Change-Id: I24fbe1fda483db4a8c606e80dbe17430ad070dc2

9 years agoDon't check for null references.
Dan Albert [Tue, 9 Sep 2014 01:34:29 +0000 (18:34 -0700)]
Don't check for null references.

The address of a reference will never be null in well formed C++ code.
The function returning the value being checked simply returns a
dereferenced pointer (which is just going to segfault if it is null!).

Change-Id: I3a4eb6adfc003c63ebf21b81d46088d52154d99d

9 years agoUse char16_t for char things.
Dan Albert [Tue, 9 Sep 2014 01:28:00 +0000 (18:28 -0700)]
Use char16_t for char things.

When compiling in C++ mode, the compiler will complain about conversions
from uint16_t to char16_t. Be consistent in using char16_t for strings.

Change-Id: I052b6176ced635162920b31560052d9a64f92764

9 years agoMerge "Small documentation fixes across many files."
Elliott Hughes [Mon, 8 Sep 2014 18:44:14 +0000 (18:44 +0000)]
Merge "Small documentation fixes across many files."

9 years agoMerge "Fix use of %llx for sscanf 64 bit values."
Christopher Ferris [Sat, 6 Sep 2014 16:55:49 +0000 (16:55 +0000)]
Merge "Fix use of %llx for sscanf 64 bit values."

9 years agoMerge "Add new header to support 32 bit/64 bit dumps."
Christopher Ferris [Sat, 6 Sep 2014 01:19:02 +0000 (01:19 +0000)]
Merge "Add new header to support 32 bit/64 bit dumps."

9 years agoAdd new header to support 32 bit/64 bit dumps.
Christopher Ferris [Tue, 2 Sep 2014 19:49:32 +0000 (12:49 -0700)]
Add new header to support 32 bit/64 bit dumps.

Bug: 17319589

(cherry picked from commit 0d3c045a0d48193150689e5924bd04c511380a8a)

Change-Id: I79aa035098115de7b02171a9432ff7e9393328ab

9 years agoSmall documentation fixes across many files.
Mark Doliner [Wed, 27 Aug 2014 23:04:02 +0000 (16:04 -0700)]
Small documentation fixes across many files.

Change-Id: I3e8787ce4bc6018ea1dc9aef2a2cd4e0a8dde663

9 years agoMerge "Frameworks/base: Add native bridge post-fork initialization"
Andreas Gampe [Thu, 4 Sep 2014 21:47:12 +0000 (21:47 +0000)]
Merge "Frameworks/base: Add native bridge post-fork initialization"

9 years agoMerge "android_media_MediaMuxer_setLocation: amended signature."
Narayan Kamath [Thu, 4 Sep 2014 11:06:47 +0000 (11:06 +0000)]
Merge "android_media_MediaMuxer_setLocation: amended signature."

9 years agoandroid_media_MediaMuxer_setLocation: amended signature.
Robert Shih [Wed, 5 Feb 2014 18:42:01 +0000 (10:42 -0800)]
android_media_MediaMuxer_setLocation: amended signature.

Change parameter `nativeObject` from type `jint` to `jlong` to match its JNI
signature.

Bug: 12890910

(cherry picked from commit 3cb78498d3f664f541ba7d28f4543cf8c12733f1)

Reported-By: ashok.bhat@arm.com, marcus.oakland@arm.com
Change-Id: I624dfb41485be823d31797514664d3a5f9e52eb0

9 years agoFrameworks/base: Add native bridge post-fork initialization
Andreas Gampe [Wed, 3 Sep 2014 04:23:06 +0000 (21:23 -0700)]
Frameworks/base: Add native bridge post-fork initialization

Change-Id: I5a20de1cb68dd1802937b369b14c50c9c1031c67

9 years agoFix use of %llx for sscanf 64 bit values.
Christopher Ferris [Thu, 4 Sep 2014 02:59:08 +0000 (19:59 -0700)]
Fix use of %llx for sscanf 64 bit values.

Change-Id: Ie7a0e243dfcf24722783fe7d697a1412de7efe68

9 years agoMerge "Revert "SELinuxMMAC unit tests""
Nick Kralevich [Wed, 3 Sep 2014 20:04:26 +0000 (20:04 +0000)]
Merge "Revert "SELinuxMMAC unit tests""

9 years agoRevert "SELinuxMMAC unit tests"
Nick Kralevich [Wed, 3 Sep 2014 20:03:17 +0000 (20:03 +0000)]
Revert "SELinuxMMAC unit tests"

Broken build.

frameworks/base/services/tests/servicestests/src/com/android/server/pm/SELinuxMMACTests.java:89: error: constructor PackageParser in class PackageParser cannot be applied to given types;
        PackageParser packageParser = new PackageParser(archiveFilePath);
                                      ^
  required: no arguments
  found: String
  reason: actual and formal argument lists differ in length
frameworks/base/services/tests/servicestests/src/com/android/server/pm/SELinuxMMACTests.java:93: error: method parsePackage in class PackageParser cannot be applied to given types;
        PackageParser.Package pkg = packageParser.parsePackage(sourceFile,
                                                 ^
  required: File,int
  found: File,String,DisplayMetrics,int
  reason: actual and formal argument lists differ in length
frameworks/base/services/tests/servicestests/src/com/android/server/pm/SELinuxMMACTests.java:99: error: incompatible types
        boolean savedCerts = packageParser.collectCertificates(pkg, 0);
                                                              ^
  required: boolean
  found:    void
Note: Some input files use or override a deprecated API.
Note: Recompile with -Xlint:deprecation for details.
3 errors
make: *** [out/target/common/obj/APPS/FrameworksServicesTests_intermediates/classes-full-debug.jar] Error 41

This reverts commit 2f446561f2f6d1b320b34432854d2f55e5b56f9e.

Change-Id: I0aa992d7b9eccce7e38c430a8a96084a6ee99866

9 years agoMerge "SELinuxMMAC unit tests"
Nick Kralevich [Wed, 3 Sep 2014 17:48:31 +0000 (17:48 +0000)]
Merge "SELinuxMMAC unit tests"

9 years agoSELinuxMMAC unit tests
Robert Craig [Tue, 2 Sep 2014 16:41:42 +0000 (12:41 -0400)]
SELinuxMMAC unit tests

Added unit tests to help with future integration and regression
testing.

Change-Id: I5e7d709e80ccbbe1dde1dc26b3e9a155f5009ad1
Signed-off-by: rpcraig <rpcraig@tycho.ncsc.mil>
9 years agoMerge "Add testing api to SELinuxMMAC code."
Nick Kralevich [Tue, 2 Sep 2014 16:52:13 +0000 (16:52 +0000)]
Merge "Add testing api to SELinuxMMAC code."

9 years agoAdd testing api to SELinuxMMAC code.
Robert Craig [Tue, 2 Sep 2014 11:16:52 +0000 (07:16 -0400)]
Add testing api to SELinuxMMAC code.

Added needed api for local unit testing.

Change-Id: If52555ecb4f0e6bc06e153b6a17e2c706b11253e
Signed-off-by: rpcraig <rpcraig@tycho.ncsc.mil>
9 years agoMerge "Update WebViewFactory.java to compile in AOSP."
Marcin Kosiba [Mon, 1 Sep 2014 08:50:25 +0000 (08:50 +0000)]
Merge "Update WebViewFactory.java to compile in AOSP."

9 years agoMerge "Load the WebView Java code via createPackageContext."
Marcin Kosiba [Mon, 1 Sep 2014 08:50:08 +0000 (08:50 +0000)]
Merge "Load the WebView Java code via createPackageContext."

9 years agoMerge "Remove WebView library preloading mechanism."
Marcin Kosiba [Mon, 1 Sep 2014 08:48:48 +0000 (08:48 +0000)]
Merge "Remove WebView library preloading mechanism."

9 years agoMerge "Fixing android.text.format.Time for non-English locales"
Neil Fuller [Fri, 29 Aug 2014 08:27:29 +0000 (08:27 +0000)]
Merge "Fixing android.text.format.Time for non-English locales"

9 years agoFixing android.text.format.Time for non-English locales
Neil Fuller [Wed, 27 Aug 2014 16:15:09 +0000 (17:15 +0100)]
Fixing android.text.format.Time for non-English locales

The test fixes for bug 17262063 showed up a real issue for
non-English locales with the Time.format() method:
If the Android string resources that contain the pattern use
non-ASCII characters then a '?' would be output instead of
those characters.

For example, in France the pattern for '%c' includes a 'à'
(a with a grave accent) and Japan includes 日.

The problem was due to converting the pattern to bytes using
the US_ASCII character set, which turns non-ASCII characters
into '?'. The code has been changed to use char throughout
and avoid bytes.

Internal documentation has been improved.

Some calls to modifyAndAppend() have been replaced with a
direct call to outputBuilder.append() because the
modify step is guaranteed to a no-op for the literals given.

The formatter has been changed to use Locale.US because it
is only used for outputting numbers. It has been renamed
to make this more obvious and the locale field has been
removed.

Bug: 17262063
Change-Id: I32b92f7f7e3e6931d3514d87f1d9a38f136d4021

9 years agoMerge "Remove obsolete pruneDexCache"
Brian Carlstrom [Thu, 28 Aug 2014 23:45:09 +0000 (23:45 +0000)]
Merge "Remove obsolete pruneDexCache"