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
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
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
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
Added unit tests to help with future integration and regression
testing.
Change-Id: I5e7d709e80ccbbe1dde1dc26b3e9a155f5009ad1
Signed-off-by: rpcraig <rpcraig@tycho.ncsc.mil>
The flag is used to enforce --interpret-only flag when running dex2oat.
Bug: 12457423
(cherry picked from commit 11e08c010a42390803b65eaa698ca0e768aed896)
Change-Id: I215339527e998b24e274c8df42a5024839e6a9fa
This reverts commit 5073145f85c8197ec4387d5e2599442a7291ca9c.
I spoke to Dianne and she said that if this is making a problem go away then it is only hiding it and there is a deeper problem we need to fix.
I think if killedByApp is true then we should set app and app.thread to null. I'll look at it further when I have more time.
Change-Id: I44c29e15ca654666440472eab2d2af1c5af80b9a
Symptom: No activity was started for startActivity, low repro rate.
Root Cause: Starting activity on a process which was killed by ActivityManagerService.killUnneededProcessLocked() and before receiving AppDeathRecipient of it.
Solution: Choose the flow of starting a new process if ProcessRecord.killedByAm is true.
Change-Id: Ida2639b0fb4631222fc92d65aadc9fd2da637b45
This complements https://android-review.googlesource.com/#/c/103231/
where the instruction set was mapped just for comptilation. The same
should have been done for move,remove and getSize.
The list of dex code ISAs is alo de-duped.
Bug: 16185267
Change-Id: I8fe453a800812e382e8f41b5f7922997aa9c18a9
In the presence of a native bridge it is more efficient to compile the
dex directly to the native ISA than to use the shared library ISA as a
reference.
This can be achieve by configuring the readonly system properties to map
between the .so ISA and the desired dex code .ISA (e.g.
ro.dalvik.vm.isa.ISA1=ISA2).
Bug: 16185267
Change-Id: I50baa7b37e1465b9adf72d6f6b96f526a08d59c7
Sympton:
During testing, skip kill native crash process manually because it will continue to die by default.
Root Cause:
Large process may take some time to do coredump.In auto test, crash process will be killed immediately that results incomplete coredump file.
Solution:
If the tester (IActivityController) will handle app crash event,
Do not kill native crashed process if the rom is debuggable.
Change-Id: Ia360af147d694125d440e5ba2f958c4759a50494
Sympton:
Next activity only adds to history but does not launch/resume then results ANR.
Root Cause:
In a rare timing, some windows are switched at the same time,
it will cause some finishing records on the top temporarily,
then set startIt to false that skip to resume the real top activty.
Solution:
If all activities in a task are finishing, do not use it to check.
The behavior/checking is the same concept as in JellyBean:
// If starting in an existing task, find where that is...
boolean startIt = true;
for (int i = NH-1; i >= 0; i--) {
ActivityRecord p = mHistory.get(i);
if (p.finishing) { // <--
continue;
}
Change-Id: I9d81a7b5182400c52e173da23eee61c74692beee
When switching to a newly created user, the user may face this bug when he opens the RecentTask screen.
A possible bug scenario is described as follows:
A user id of a removed user may be recycled when created a new user.
However, mRecentTasks is not correctly controlled so that old information may still remain and be possibly mapped to wrong user.
This patch prevents this bug by explicitly removing old information in mRecentTasks when removing existing user.
Change-Id: I1874dbd604598a5d740ae1e034981e21214c15c6
Signed-off-by: Eunae Kim <eunae.kim@lge.com>
If an application that has a broadcast receiver is killed
during broadcasting, thread variable of ProcessRecord becomes null
so that IIntentReceiver#performReceive() is called
in BroadcastQueue#performReceiveLocked(). But if binder driver has not
noticed the death of the application yet, it can't throw
DeadObjectException. After that, binder driver notices. But it can't
notify DeadObjectException to the caller because performReceive() is
async call. So broadcasting keeps on waiting for finishing
performReceive() until timeout.
This change checks the death of the application before calling
performReceive() and skips broadcasting to the receiver
if the application has already died.
Change-Id: Ifa02b8b1a7e7b6fd314de90fedff5b7a5326825d
