OSDN Git Service

android-x86/dalvik.git
14 years agoam 8673ee75: am 9a35ef1b: Guarantee deflaters get cleaned up, even if subclassed.
Jesse Wilson [Thu, 15 Oct 2009 20:45:21 +0000 (13:45 -0700)]
am 8673ee75: am 9a35ef1b: Guarantee deflaters get cleaned up, even if subclassed.

Merge commit '8673ee75052e7398aa6f5925e968d6a0b2419be8'

* commit '8673ee75052e7398aa6f5925e968d6a0b2419be8':
  Guarantee deflaters get cleaned up, even if subclassed.

14 years agoam 9a35ef1b: Guarantee deflaters get cleaned up, even if subclassed.
Jesse Wilson [Thu, 15 Oct 2009 20:36:21 +0000 (13:36 -0700)]
am 9a35ef1b: Guarantee deflaters get cleaned up, even if subclassed.

Merge commit '9a35ef1b4b196f0479054993ab0eabac416c5214' into eclair-mr2-plus-aosp

* commit '9a35ef1b4b196f0479054993ab0eabac416c5214':
  Guarantee deflaters get cleaned up, even if subclassed.

14 years agoMerge change I581bed77
Android (Google) Code Review [Thu, 15 Oct 2009 19:04:24 +0000 (15:04 -0400)]
Merge change I581bed77

* changes:
  Add a memory barrier.

14 years agoGuarantee deflaters get cleaned up, even if subclassed.
Jesse Wilson [Thu, 15 Oct 2009 18:46:52 +0000 (11:46 -0700)]
Guarantee deflaters get cleaned up, even if subclassed.

See bug 2186927

14 years agoMerge change I11b31b39
Android (Google) Code Review [Thu, 15 Oct 2009 18:28:09 +0000 (14:28 -0400)]
Merge change I11b31b39

* changes:
  Fix icu4jni Resources ("Locale") to not expose its internals.

14 years agoAdd a memory barrier.
Andy McFadden [Thu, 15 Oct 2009 18:24:54 +0000 (11:24 -0700)]
Add a memory barrier.

Looks like an appropriate place for one.

14 years agoFix icu4jni Resources ("Locale") to not expose its internals.
Elliott Hughes [Wed, 14 Oct 2009 22:40:27 +0000 (15:40 -0700)]
Fix icu4jni Resources ("Locale") to not expose its internals.

We shouldn't expose internal arrays without copying.

Bug: 2102273

14 years agoam c1d492ca: merge from open-source master
Jean-Baptiste Queru [Wed, 14 Oct 2009 23:12:50 +0000 (16:12 -0700)]
am c1d492ca: merge from open-source master

Merge commit 'c1d492ca536a3acf4156d06ce01c932b8b762cb9'

* commit 'c1d492ca536a3acf4156d06ce01c932b8b762cb9':

14 years agomerge from open-source master
Jean-Baptiste Queru [Wed, 14 Oct 2009 23:02:09 +0000 (16:02 -0700)]
merge from open-source master

14 years agoFix build.
Elliott Hughes [Wed, 14 Oct 2009 19:12:11 +0000 (12:12 -0700)]
Fix build.

Add missing files from https://android-git.corp.google.com/g/29998.

14 years agoMerge change Ic8d6b3cb
Android (Google) Code Review [Wed, 14 Oct 2009 18:42:05 +0000 (14:42 -0400)]
Merge change Ic8d6b3cb

* changes:
  Fix Constructor and Method to not expose their internals.

14 years agoMerge change Ia0e7b60d
Android (Google) Code Review [Wed, 14 Oct 2009 17:48:59 +0000 (13:48 -0400)]
Merge change Ia0e7b60d

* changes:
  Remove incorrect definition of nested local variables.

14 years agoMerge change I0e52b194
Android (Google) Code Review [Wed, 14 Oct 2009 17:12:35 +0000 (13:12 -0400)]
Merge change I0e52b194

* changes:
  Bug fixing for NumberFormat and BigDecimal.

14 years agoFix build.
Elliott Hughes [Wed, 14 Oct 2009 17:07:42 +0000 (10:07 -0700)]
Fix build.

Work around apicheck bug.

14 years agoMerge change I3a0b1f8c
Android (Google) Code Review [Wed, 14 Oct 2009 15:44:10 +0000 (11:44 -0400)]
Merge change I3a0b1f8c

* changes:
  Update libcore/security to Harmony r823222.

14 years agoMerge change Ic89dfdd8
Android (Google) Code Review [Wed, 14 Oct 2009 14:23:05 +0000 (10:23 -0400)]
Merge change Ic89dfdd8

* changes:
  Bump "public" version from 1.1.0 to 1.2.0.

14 years agoMerge change I0d6fff2d
Android (Google) Code Review [Wed, 14 Oct 2009 14:22:50 +0000 (10:22 -0400)]
Merge change I0d6fff2d

* changes:
  Avoid JDWP hang on nested thread suspension.

14 years agoMerge change I9297389b
Android (Google) Code Review [Wed, 14 Oct 2009 07:42:03 +0000 (03:42 -0400)]
Merge change I9297389b

* changes:
  Udating luni to Harmony r823222.

14 years agoRemove incorrect definition of nested local variables.
Ben Cheng [Wed, 14 Oct 2009 06:09:01 +0000 (23:09 -0700)]
Remove incorrect definition of nested local variables.

14 years agoUpdate libcore/security to Harmony r823222.
Elliott Hughes [Sat, 10 Oct 2009 00:21:46 +0000 (17:21 -0700)]
Update libcore/security to Harmony r823222.

Squashed commit of the following:

commit e3083dde77b71fa817c8a52d9edafcb6325cfc25
Author: Elliott Hughes <enh@google.com>
Date:   Fri Oct 9 13:10:11 2009 -0700

    security_dalvik

commit 0fc0101e5dbfb3e3044702579ab8087a4e07984a
Author: Elliott Hughes <enh@google.com>
Date:   Fri Oct 9 13:09:48 2009 -0700

    security_527399

14 years agoFix Constructor and Method to not expose their internals.
Elliott Hughes [Wed, 14 Oct 2009 01:18:21 +0000 (18:18 -0700)]
Fix Constructor and Method to not expose their internals.

We shouldn't expose internal arrays without copying. Behavior confirmed
by testing against RI.

Also connect up the dalvik ThreadsTest test that had fallen by the wayside.

Bug: 2102273

14 years agoBump "public" version from 1.1.0 to 1.2.0.
Andy McFadden [Tue, 13 Oct 2009 23:19:22 +0000 (16:19 -0700)]
Bump "public" version from 1.1.0 to 1.2.0.

Also, send correct version in JDWP VirtualMachine.Version packet.

14 years agoAvoid JDWP hang on nested thread suspension.
Andy McFadden [Tue, 13 Oct 2009 23:04:31 +0000 (16:04 -0700)]
Avoid JDWP hang on nested thread suspension.

With jdb, you can suspend a thread multiple times.  If you try to
execute a method -- which requires resuming a thread that has been
stopped at a breakpoint -- the VM currently only does a single "resume",
which means the thread is still suspended, and the JDWP thread hangs
waiting for it.

This adds a check to prevent the hang.

For bug 2183735.

14 years agoMerge change Ied79ff0c
Android (Google) Code Review [Tue, 13 Oct 2009 20:45:17 +0000 (16:45 -0400)]
Merge change Ied79ff0c

* changes:
  Accept JitOff as a legal state in the debug interpreter when JIT is configured.

14 years agoAccept JitOff as a legal state in the debug interpreter when JIT is configured.
Ben Cheng [Tue, 13 Oct 2009 20:26:40 +0000 (13:26 -0700)]
Accept JitOff as a legal state in the debug interpreter when JIT is configured.

14 years agoAdd JDWP ReferenceType.ClassObject handler.
Andy McFadden [Tue, 13 Oct 2009 20:06:03 +0000 (13:06 -0700)]
Add JDWP ReferenceType.ClassObject handler.

For bug 2157236.

14 years agoMerge change I60282050
Android (Google) Code Review [Tue, 13 Oct 2009 19:24:25 +0000 (15:24 -0400)]
Merge change I60282050

* changes:
  Add implementation of JDWP ClassType.NewInstance

14 years agoam fef7556f: am 3238cc2a: am 32f89821: Merge change 26044 into donut-gms
Brett Chabot [Tue, 13 Oct 2009 18:57:48 +0000 (11:57 -0700)]
am fef7556f: am 3238cc2a: am 32f89821: Merge change 26044 into donut-gms

Merge commit 'fef7556f058d0d17b30180c03387558bc59d3105'

* commit 'fef7556f058d0d17b30180c03387558bc59d3105':
  Remove tests.api.java.io.FileTest#test_delete as known failure.

14 years agoam 4ebfaec4: am 98f0612b: am c4d33d9d: Fix ClassLoaderTest to work on DEXPREOPT builds.
Brett Chabot [Tue, 13 Oct 2009 18:57:43 +0000 (11:57 -0700)]
am 4ebfaec4: am 98f0612b: am c4d33d9d: Fix ClassLoaderTest to work on DEXPREOPT builds.

Merge commit '4ebfaec453ae79e3ac1d7ff30cbc9f2d3beef2d1'

* commit '4ebfaec453ae79e3ac1d7ff30cbc9f2d3beef2d1':
  Fix ClassLoaderTest to work on DEXPREOPT builds.

14 years agoAdd implementation of JDWP ClassType.NewInstance
Andy McFadden [Fri, 2 Oct 2009 22:56:26 +0000 (15:56 -0700)]
Add implementation of JDWP ClassType.NewInstance

Add a handler for the ClassType.NewInstance request, and some bits of
plumbing to go with it.

For bug 2157236.

14 years agoMerge change I8f61592d
Android (Google) Code Review [Mon, 12 Oct 2009 21:51:09 +0000 (17:51 -0400)]
Merge change I8f61592d

* changes:
  Set the debug interpreter entry point properly on the self-verification path.

14 years agoBug fixing for NumberFormat and BigDecimal.
Jesse Wilson [Mon, 12 Oct 2009 21:07:16 +0000 (14:07 -0700)]
Bug fixing for NumberFormat and BigDecimal.

  1. Fixed the bug that DecimalFormat does not handle multiplier.
  2. Fixed the bug that DecimalFormat does not handle precision.

This is a copy of the original Eclair change,
  https://android-git.corp.google.com/g/26297

Bug: 1897917.

14 years agoMerge change I6618182f
Android (Google) Code Review [Mon, 12 Oct 2009 21:17:00 +0000 (17:17 -0400)]
Merge change I6618182f

* changes:
  Fix CoreTestRunner to not run filtered-out tests.

14 years agoSet the debug interpreter entry point properly on the self-verification path.
Ben Cheng [Mon, 12 Oct 2009 20:46:55 +0000 (13:46 -0700)]
Set the debug interpreter entry point properly on the self-verification path.

Also fix the encoding for SFP/DFP register names to make self-verification
happy on FP benchmarks.

14 years agoFix CoreTestRunner to not run filtered-out tests.
Elliott Hughes [Mon, 12 Oct 2009 21:08:50 +0000 (14:08 -0700)]
Fix CoreTestRunner to not run filtered-out tests.

https://android-git.corp.google.com/g/29284 broke test filtering by accidentally
passing the unfiltered test suite to super.doRun.

14 years agoMerge change I1acdca65
Android (Google) Code Review [Mon, 12 Oct 2009 20:54:08 +0000 (16:54 -0400)]
Merge change I1acdca65

* changes:
  Fix memory corruption.

14 years agoMerge change Iff8a3e91
Android (Google) Code Review [Mon, 12 Oct 2009 20:43:15 +0000 (16:43 -0400)]
Merge change Iff8a3e91

* changes:
  Adding timeouts to HttpsURLConnectionTest.

14 years agoAdding timeouts to HttpsURLConnectionTest.
Jesse Wilson [Mon, 12 Oct 2009 19:58:01 +0000 (12:58 -0700)]
Adding timeouts to HttpsURLConnectionTest.

See bug http://b/issue?id=2180571

14 years agoFix memory corruption.
Jesse Wilson [Mon, 12 Oct 2009 19:31:09 +0000 (12:31 -0700)]
Fix memory corruption.

14 years agoam e652b123: Make sure res is NULL if getaddrinfo fails.
Lorenzo Colitti [Mon, 12 Oct 2009 18:35:56 +0000 (11:35 -0700)]
am e652b123: Make sure res is NULL if getaddrinfo fails.

Merge commit 'e652b123d2d65d40e5f51c0145cc453ea88631dc'

* commit 'e652b123d2d65d40e5f51c0145cc453ea88631dc':
  Make sure res is NULL if getaddrinfo fails.

14 years agoAdding a timeout for standard out as well as standard error.
Jesse Wilson [Sat, 10 Oct 2009 04:53:23 +0000 (21:53 -0700)]
Adding a timeout for standard out as well as standard error.

14 years agoUdating luni to Harmony r823222.
Jesse Wilson [Fri, 9 Oct 2009 22:26:15 +0000 (15:26 -0700)]
Udating luni to Harmony r823222.

Highlights:
 - InputStream.skip concurrency issue
 - "better" messages in bound exceptions for streams and arrays
 - prefer fewer writes to underlying streams (using byte[] buffers)
 - Rename subclasses to not reuse names from their superclasses
 - PlatformAddressFactory.allocMap bugfix

Plus some spelling fixes, style fixes, serial version UIDs and other
boilerplate improvements.

14 years agoMerge change I164613f2
Android (Google) Code Review [Sat, 10 Oct 2009 01:21:46 +0000 (21:21 -0400)]
Merge change I164613f2

* changes:
  Rewrite Support_Exec to support timeouts on waiting processes.

14 years agoRewrite Support_Exec to support timeouts on waiting processes.
Jesse Wilson [Sat, 10 Oct 2009 00:41:52 +0000 (17:41 -0700)]
Rewrite Support_Exec to support timeouts on waiting processes.

Also rewriting SupportExec to use ProcessBuilder rather
than Runtime.exec(). Changed callers to use the ProcessBuilder
directly rather than calling-through chained methods.

14 years agoMerge change Ic98d470a
Android (Google) Code Review [Fri, 9 Oct 2009 23:51:32 +0000 (19:51 -0400)]
Merge change Ic98d470a

* changes:
  Remove the intermediate representation in getNetworkInterfaces.

14 years agoMerge change I6239d93b
Android (Google) Code Review [Fri, 9 Oct 2009 22:47:36 +0000 (18:47 -0400)]
Merge change I6239d93b

* changes:
  Fix the docs for the no-arg Random() constructor to better reflect reality, and make the constructors more resilient with respect to subclassing by using System.identityHashCode() instead of Object.hashCode(), and also by not calling overridable methods.

14 years agoFix the docs for the no-arg Random() constructor to better reflect reality,
Dan Bornstein [Fri, 9 Oct 2009 20:09:34 +0000 (13:09 -0700)]
Fix the docs for the no-arg Random() constructor to better reflect reality,
and make the constructors more resilient with respect to subclassing by
using System.identityHashCode() instead of Object.hashCode(), and also by
not calling overridable methods.

Change-Id: I6239d93bb46876ef1c4a5e155a6dc1ac6fab4eae

14 years agoMerge change I0e98511f
Android (Google) Code Review [Fri, 9 Oct 2009 20:27:42 +0000 (16:27 -0400)]
Merge change I0e98511f

* changes:
  Do not initialize the JIT if it is not the designated execution mode.

14 years agoMerge change I98acebf9
Android (Google) Code Review [Fri, 9 Oct 2009 20:16:49 +0000 (16:16 -0400)]
Merge change I98acebf9

* changes:
  Disable SMALL_BUFFERS and CUTTHROUGH, as SSLSession test fails with these flags.

14 years agoDo not initialize the JIT if it is not the designated execution mode.
Ben Cheng [Fri, 9 Oct 2009 18:55:30 +0000 (11:55 -0700)]
Do not initialize the JIT if it is not the designated execution mode.

14 years agoMinor tweak for compatibility with BSD: Use MAP_ANON instead of MAP_ANONYMOUS.
Dan Bornstein [Thu, 8 Oct 2009 00:33:46 +0000 (17:33 -0700)]
Minor tweak for compatibility with BSD: Use MAP_ANON instead of MAP_ANONYMOUS.

The former is deprecated on Linux, but it is equivalent, and
MAP_ANONYMOUS isn't valid at all on BSD. (The alternative for BSD
compatibility is to use #ifdefs, but that seems gratuitously
heavyweight to me.)

Change-Id: I9c0b2cc78c4d7edc633487d5b0dde5b2f928247e

14 years agoMerge change I5ccdf0be
Android (Google) Code Review [Fri, 9 Oct 2009 16:49:15 +0000 (12:49 -0400)]
Merge change I5ccdf0be

* changes:
  Fix for "Compiler thread shutdown should switch out of VMWAIT"

14 years agoDisable SMALL_BUFFERS and CUTTHROUGH, as SSLSession test fails with these flags.
Nagendra Modadugu [Fri, 9 Oct 2009 01:34:48 +0000 (18:34 -0700)]
Disable SMALL_BUFFERS and CUTTHROUGH, as SSLSession test fails with these flags.

14 years agoSquashed commit of the following:
Elliott Hughes [Thu, 8 Oct 2009 23:56:37 +0000 (16:56 -0700)]
Squashed commit of the following:

commit 07d78447c89a11265bf909ab6bcc315c1a784281
Author: Elliott Hughes <enh@google.com>
Date:   Thu Oct 8 16:38:26 2009 -0700

    text_dalvik

commit c390506ce060c705b6c1b04fb1e737617de1bd8a
Author: Elliott Hughes <enh@google.com>
Date:   Thu Oct 8 16:38:22 2009 -0700

    text_802921

14 years agoFix for "Compiler thread shutdown should switch out of VMWAIT"
Ben Cheng [Thu, 8 Oct 2009 23:09:49 +0000 (16:09 -0700)]
Fix for "Compiler thread shutdown should switch out of VMWAIT"

14 years agoTools to push code back upstream.
Jesse Wilson [Thu, 8 Oct 2009 07:25:33 +0000 (00:25 -0700)]
Tools to push code back upstream.

14 years agoMerge change I685130ae
Android (Google) Code Review [Thu, 8 Oct 2009 18:29:37 +0000 (14:29 -0400)]
Merge change I685130ae

* changes:
  JNI Delete* and Release* _are_ allowed while an exception is pending.

14 years agoam 3238cc2a: am 32f89821: Merge change 26044 into donut-gms
Brett Chabot [Thu, 8 Oct 2009 18:18:20 +0000 (11:18 -0700)]
am 3238cc2a: am 32f89821: Merge change 26044 into donut-gms

Merge commit '3238cc2a901ed6e0e3bb11fa601913c2ed09ddb7' into eclair-plus-aosp

* commit '3238cc2a901ed6e0e3bb11fa601913c2ed09ddb7':
  Remove tests.api.java.io.FileTest#test_delete as known failure.

14 years agoam 98f0612b: am c4d33d9d: Fix ClassLoaderTest to work on DEXPREOPT builds.
Brett Chabot [Thu, 8 Oct 2009 18:18:10 +0000 (11:18 -0700)]
am 98f0612b: am c4d33d9d: Fix ClassLoaderTest to work on DEXPREOPT builds.

Merge commit '98f0612b22111dd9e2d23b171b9059582cc44991' into eclair-plus-aosp

* commit '98f0612b22111dd9e2d23b171b9059582cc44991':
  Fix ClassLoaderTest to work on DEXPREOPT builds.

14 years agoam 3238cc2a: am 32f89821: Merge change 26044 into donut-gms
Brett Chabot [Thu, 8 Oct 2009 17:55:14 +0000 (10:55 -0700)]
am 3238cc2a: am 32f89821: Merge change 26044 into donut-gms

Merge commit '3238cc2a901ed6e0e3bb11fa601913c2ed09ddb7' into eclair-mr2

* commit '3238cc2a901ed6e0e3bb11fa601913c2ed09ddb7':
  Remove tests.api.java.io.FileTest#test_delete as known failure.

14 years agoam 98f0612b: am c4d33d9d: Fix ClassLoaderTest to work on DEXPREOPT builds.
Brett Chabot [Thu, 8 Oct 2009 17:55:09 +0000 (10:55 -0700)]
am 98f0612b: am c4d33d9d: Fix ClassLoaderTest to work on DEXPREOPT builds.

Merge commit '98f0612b22111dd9e2d23b171b9059582cc44991' into eclair-mr2

* commit '98f0612b22111dd9e2d23b171b9059582cc44991':
  Fix ClassLoaderTest to work on DEXPREOPT builds.

14 years agoMerge change I9c147b84
Android (Google) Code Review [Thu, 8 Oct 2009 17:50:00 +0000 (13:50 -0400)]
Merge change I9c147b84

* changes:
  Fix various bugs found when debugger is attached to the VM.

14 years agoam 32f89821: Merge change 26044 into donut-gms
Brett Chabot [Thu, 8 Oct 2009 17:44:24 +0000 (10:44 -0700)]
am 32f89821: Merge change 26044 into donut-gms

Merge commit '32f8982116e5155bc3c75c86e3c861cab1a848be' into eclair

* commit '32f8982116e5155bc3c75c86e3c861cab1a848be':
  Remove tests.api.java.io.FileTest#test_delete as known failure.

14 years agoam c4d33d9d: Fix ClassLoaderTest to work on DEXPREOPT builds.
Brett Chabot [Thu, 8 Oct 2009 17:44:19 +0000 (10:44 -0700)]
am c4d33d9d: Fix ClassLoaderTest to work on DEXPREOPT builds.

Merge commit 'c4d33d9d89444d86808be7072e0548007d98a49f' into eclair

* commit 'c4d33d9d89444d86808be7072e0548007d98a49f':
  Fix ClassLoaderTest to work on DEXPREOPT builds.

14 years agoFixing the JUnit runner to output Ant-compatible XML.
Jesse Wilson [Thu, 8 Oct 2009 00:20:15 +0000 (17:20 -0700)]
Fixing the JUnit runner to output Ant-compatible XML.

We can use this with our continuous build to track test
failure changes over time.

14 years agoFix various bugs found when debugger is attached to the VM.
Ben Cheng [Wed, 7 Oct 2009 23:41:46 +0000 (16:41 -0700)]
Fix various bugs found when debugger is attached to the VM.

See b/2161257 for details.

14 years agoJNI Delete* and Release* _are_ allowed while an exception is pending.
Elliott Hughes [Wed, 7 Oct 2009 23:33:23 +0000 (16:33 -0700)]
JNI Delete* and Release* _are_ allowed while an exception is pending.

See "Exceptions" in our own documentation:
http://android.git.kernel.org/?p=platform/dalvik.git;a=blob_plain;f=docs/jni-tips.html;hb=HEAD

This is also true of the RI, though the spec sometimes implies otherwise.
Here's the canonical reference:
http://java.sun.com/docs/books/jni/html/design.html#2193

14 years agoFix several leaks in OpenSSL JNI cipher suites code.
Elliott Hughes [Wed, 7 Oct 2009 22:04:52 +0000 (15:04 -0700)]
Fix several leaks in OpenSSL JNI cipher suites code.

I first spotted the missing ReleaseStringUTFChars, but then noticed all
the duplication in the cipher suites functions, and noticed that no
copy appeared to be completely correct. The factored-out replacements
shouldn't leak, and should check all error conditions.

14 years agoMerge change I492dd3f7
Android (Google) Code Review [Wed, 7 Oct 2009 06:51:47 +0000 (02:51 -0400)]
Merge change I492dd3f7

* changes:
  Enable SMALL_BUFFERS and HANDSHAKE_CUTTHROUGH for SSL connections.

14 years agoEnable SMALL_BUFFERS and HANDSHAKE_CUTTHROUGH for SSL connections.
Nagendra Modadugu [Tue, 6 Oct 2009 22:52:53 +0000 (15:52 -0700)]
Enable SMALL_BUFFERS and HANDSHAKE_CUTTHROUGH for SSL connections.

14 years agoRemove duplication between readv and writev.
Elliott Hughes [Tue, 6 Oct 2009 22:31:35 +0000 (15:31 -0700)]
Remove duplication between readv and writev.

This also means we don't pin the three int arrays for the duration
of the syscall.

14 years agoDon't use GetPrimitiveArrayCritical around network operations.
Elliott Hughes [Tue, 6 Oct 2009 21:08:53 +0000 (14:08 -0700)]
Don't use GetPrimitiveArrayCritical around network operations.

This code seems to have been using GetPrimitiveArrayCritical to
work around the fact that it was supply JNI_ABORT on release,
where it really wanted GetByteArrayElements and to keep its
changes (since the whole point of the byte[] was to keep state
between calls).

14 years agoRemove the intermediate representation in getNetworkInterfaces.
Elliott Hughes [Tue, 6 Oct 2009 20:01:12 +0000 (13:01 -0700)]
Remove the intermediate representation in getNetworkInterfaces.

This gets rid of a lot of unnecessary complexity by going straight from
the kernel's representation to the Java representation, without the useless
C structs in between.

14 years agoMerge change Ia4c68278
Android (Google) Code Review [Tue, 6 Oct 2009 18:27:47 +0000 (14:27 -0400)]
Merge change Ia4c68278

* changes:
  Remove dead code.

14 years agoMerge change I40159068
Android (Google) Code Review [Tue, 6 Oct 2009 17:36:11 +0000 (13:36 -0400)]
Merge change I40159068

* changes:
  Use icu4jni_error more consistently.

14 years agoRemove dead code.
Elliott Hughes [Tue, 6 Oct 2009 17:18:34 +0000 (10:18 -0700)]
Remove dead code.

14 years agoMake sure res is NULL if getaddrinfo fails.
Lorenzo Colitti [Fri, 2 Oct 2009 19:45:21 +0000 (12:45 -0700)]
Make sure res is NULL if getaddrinfo fails.

14 years agoMerge change I7aabd791
Android (Google) Code Review [Tue, 6 Oct 2009 00:15:01 +0000 (20:15 -0400)]
Merge change I7aabd791

* changes:
  Refactor java_net_NetworkInterface.cpp for clarity.

14 years agoUse icu4jni_error more consistently.
Elliott Hughes [Mon, 5 Oct 2009 23:01:35 +0000 (16:01 -0700)]
Use icu4jni_error more consistently.

In BidiWrapperInterface.c, replace check_fail with icu4jni_error for better
error reporting.

In CollationInterface.c, make sure we call icu4jni_error *after* calling
the function that may or may not return an error.

14 years agoRefactor java_net_NetworkInterface.cpp for clarity.
Elliott Hughes [Sat, 3 Oct 2009 00:20:00 +0000 (17:20 -0700)]
Refactor java_net_NetworkInterface.cpp for clarity.

All I wanted to do was remove the SOCKERR_ mess, but things got out of hand.
There's still plenty of stuff that could be cleaned up, but nothing that can't
wait. I'm only doing this to make things easier for the IPv6 changes.

This patch:
* removes the SOCKERR_ mess which was causing vague "Operation failed"
IOExceptions; we now use the actual errno value of whatever caused the failure.
* simplifies memory management so we don't have to do lots of cleanup on
every exit from each function.
* simplified fd management for the same reason.
* reduces the use of the intermediate structure, going straight to Java objects
in places. (This is the unfinished work: we should get rid of the intermediate
representation. It's not useful.)
* replaces structInToJavaAddress and structInToInetAddress with a call to the
better socketAddressToInetAddress from OSNetworkSystem.cpp.
* uses more intention-revealing variable and function names.
* removes attempts to throw exceptions when perfectly good exceptions have
already been thrown.
* removes filtering of interfaces that aren't up. The RI doesn't do this
filtering, there's no reason to do it, and it's always going to be a race
condition anyway.
* removes untrue comments.
* removes a lot of inexplicably empty lines.

14 years agoSwitch to TEMP_FAILURE_RETRY now it's moved to bionic.
Elliott Hughes [Mon, 5 Oct 2009 20:48:44 +0000 (13:48 -0700)]
Switch to TEMP_FAILURE_RETRY now it's moved to bionic.

(The same EINTR_RETRY macro that used to be here has moved to <unistd.h>, so
we don't need another copy. EINTR_RETRY is a nicer name, I think, but
TEMP_FAILURE_RETRY is the GNU-compatible name.)

14 years agoMerge change Idc84bb2b
Android (Google) Code Review [Fri, 2 Oct 2009 20:19:46 +0000 (16:19 -0400)]
Merge change Idc84bb2b

* changes:
  Fix trace builder to not confuse self-verification.

14 years agoFix trace builder to not confuse self-verification.
Ben Cheng [Fri, 2 Oct 2009 19:58:52 +0000 (12:58 -0700)]
Fix trace builder to not confuse self-verification.

14 years agoMake sure res is NULL if getaddrinfo fails.
Lorenzo Colitti [Fri, 2 Oct 2009 19:45:21 +0000 (12:45 -0700)]
Make sure res is NULL if getaddrinfo fails.

14 years agoMerge change I4bf8618f
Android (Google) Code Review [Fri, 2 Oct 2009 19:37:32 +0000 (15:37 -0400)]
Merge change I4bf8618f

* changes:
  Add back missing shared libraries used by executable in link command.  These libraries are linked though dependencies of other libraries.  While that works, it is not the right thing to do.

14 years agoAdd back missing shared libraries used by executable in link command. These
Doug Kwan [Fri, 2 Oct 2009 18:21:07 +0000 (11:21 -0700)]
Add back missing shared libraries used by executable in link command.  These
libraries are linked though dependencies of other libraries.  While that works,
it is not the right thing to do.

14 years agoRemove adb networking support from dalvik/libcore.
Elliott Hughes [Wed, 30 Sep 2009 20:26:11 +0000 (13:26 -0700)]
Remove adb networking support from dalvik/libcore.

There's more to be removed from other projects, but this should probably
come out first.

Bug: 1122968

14 years agoMerge change Id5ccf4bc
Android (Google) Code Review [Fri, 2 Oct 2009 14:52:14 +0000 (10:52 -0400)]
Merge change Id5ccf4bc

* changes:
  Catch bad getStackTrace() calls.

14 years agoMerge change I9d8f66aa
Android (Google) Code Review [Fri, 2 Oct 2009 00:37:18 +0000 (20:37 -0400)]
Merge change I9d8f66aa

* changes:
  Switch java_net_NetworkInterface.c to C++.

14 years agoSwitch java_net_NetworkInterface.c to C++.
Elliott Hughes [Fri, 2 Oct 2009 00:20:14 +0000 (17:20 -0700)]
Switch java_net_NetworkInterface.c to C++.

(I have a bigger change to make to this file, but the web claims that it's
best to keep renames in minimally small patches so git doesn't get confused.)

14 years agoCatch bad getStackTrace() calls.
Andy McFadden [Fri, 2 Oct 2009 00:10:47 +0000 (17:10 -0700)]
Catch bad getStackTrace() calls.

Normally Throwable.getStackTrace() can't be called until after there's a
stack trace, but with a debugger you can call it during construction.
We weren't expecting that and the VM went boom.

14 years agoMerge change I3bc71eab
Android (Google) Code Review [Fri, 2 Oct 2009 00:07:17 +0000 (20:07 -0400)]
Merge change I3bc71eab

* changes:
  :: is a valid local address on a newly-created socket if IPv6 is preferred.

14 years ago:: is a valid local address on a newly-created socket if IPv6 is preferred.
Lorenzo Colitti [Thu, 1 Oct 2009 19:24:40 +0000 (12:24 -0700)]
:: is a valid local address on a newly-created socket if IPv6 is preferred.

14 years agoam ae8d0254: am 0c0cfd5c: Merge change I96cd6f5b into eclair
Lorenzo Colitti [Thu, 1 Oct 2009 20:47:10 +0000 (13:47 -0700)]
am ae8d0254: am 0c0cfd5c: Merge change I96cd6f5b into eclair

Merge commit 'ae8d0254dd69392dae9c922462ca2a01944b6eef'

* commit 'ae8d0254dd69392dae9c922462ca2a01944b6eef':
  Make the socket test accept ::1 as a valid result for

14 years agoam 1507c7ce: am f94c5d77: More multicast fixes.
Lorenzo Colitti [Thu, 1 Oct 2009 20:47:04 +0000 (13:47 -0700)]
am 1507c7ce: am f94c5d77: More multicast fixes.

Merge commit '1507c7ce4fc88f7bdf5e89ba10839809693eeefb'

* commit '1507c7ce4fc88f7bdf5e89ba10839809693eeefb':
  More multicast fixes.

14 years agoam 003d7f8d: am e851c562: Merge change Iea75a523 into eclair
Lorenzo Colitti [Thu, 1 Oct 2009 20:46:58 +0000 (13:46 -0700)]
am 003d7f8d: am e851c562: Merge change Iea75a523 into eclair

Merge commit '003d7f8da66ec53b5dbece36308197dc7a932e3e'

* commit '003d7f8da66ec53b5dbece36308197dc7a932e3e':
  Fixes for socket options on multicast sockets.

14 years agoMerge change I2cdad38a
Android (Google) Code Review [Thu, 1 Oct 2009 20:35:06 +0000 (16:35 -0400)]
Merge change I2cdad38a

* changes:
  Update archive to Harmony r820767.

14 years agoUpdate archive to Harmony r820767.
Jesse Wilson [Thu, 1 Oct 2009 20:16:51 +0000 (13:16 -0700)]
Update archive to Harmony r820767.

14 years agoam aa4320a4: am 1131d3da: Merge change Ie2e14dde into eclair
Jesse Wilson [Thu, 1 Oct 2009 19:22:38 +0000 (12:22 -0700)]
am aa4320a4: am 1131d3da: Merge change Ie2e14dde into eclair

Merge commit 'aa4320a4fb27684bd70cfb4d82feca3845ad272e'

* commit 'aa4320a4fb27684bd70cfb4d82feca3845ad272e':
  Remove @KnownFailure tags for tests that pass.

14 years agoam 2b791cc2: am 910363bd: Merge change Ib5dc3f1f into eclair
Jesse Wilson [Thu, 1 Oct 2009 19:22:26 +0000 (12:22 -0700)]
am 2b791cc2: am 910363bd: Merge change Ib5dc3f1f into eclair

Merge commit '2b791cc256e80f93fef7db2157f1cee202133201'

* commit '2b791cc256e80f93fef7db2157f1cee202133201':
  Fixing a lame test that was relying on hashmap iteration order.