OSDN Git Service
Jesse Wilson [Tue, 9 Nov 2010 21:43:01 +0000 (13:43 -0800)]
Don't kill the instrumentation test runner with an uncaught exception.
Change-Id: Idfb25df214cfeeb32b32d7cdb046954511cf755d
http://b/
3171231
Brian Carlstrom [Mon, 1 Nov 2010 21:19:59 +0000 (14:19 -0700)]
Avoid races between OpenSSLSocketImpl I/O and close()
The previous change:
commit
5f2e6872311240319509aed64d9f58cd5b64719b
Author: Brian Carlstrom <bdc@google.com>
Date: Mon Aug 23 14:06:51 2010 -0700
SSLSocket.read should throw SocketException not NullPointerException
added checkOpen() to throw SocketException instead of
NullPointerException, but there was still a race between read/write on
one thread and close on another that could allow a
NullPointerException to escape. This change moves checkOpen() calls to
be protected by the existing writeLock/readLock/handshakeLock
synchronzied blocks to avoid this case.
byte buffer error checking for read/write is also moved into the to
lock region to preserve compatability as measured by the test:
libcore.javax.net.ssl.SSLSocketTest#test_SSLSocket_close
Bug:
3153162
Dan Bornstein [Sat, 30 Oct 2010 00:58:50 +0000 (17:58 -0700)]
Remove a pointless test, which just tested a test support class.
Change-Id: Ifef2b25500474f7e3b18ff97e7831717f2c9b391
Dan Bornstein [Fri, 29 Oct 2010 22:16:17 +0000 (15:16 -0700)]
Replace the implementation of this class.
Use the Apache Harmony version.
Change-Id: I5b9d6fac10dcaf20e16fccb73cfc2ae7110a683d
Andy McFadden [Wed, 27 Oct 2010 18:26:29 +0000 (11:26 -0700)]
Set capabilities sooner.
Bug
3135433.
Change-Id: Ief58bdfff2cb82d7824ef796c2c1031f525f6ee9
Brian Carlstrom [Tue, 26 Oct 2010 17:19:04 +0000 (10:19 -0700)]
Merge "Remove Elliptic Curve CA certificate" into gingerbread
Brian Carlstrom [Tue, 26 Oct 2010 07:04:35 +0000 (00:04 -0700)]
Remove Elliptic Curve CA certificate
Bug:
3131996
Change-Id: I6fc19c87c489905802607aaf587cd551ef9cb09e
Nick Kralevich [Mon, 25 Oct 2010 23:18:05 +0000 (16:18 -0700)]
Merge "fix operator precedence bug when calculating bits." into gingerbread
Jesse Wilson [Fri, 22 Oct 2010 23:17:54 +0000 (16:17 -0700)]
Merge "Documenting WiFi sign in and SSL fallback with HttpURLConnection." into gingerbread
Jesse Wilson [Fri, 22 Oct 2010 22:50:18 +0000 (15:50 -0700)]
Documenting WiFi sign in and SSL fallback with HttpURLConnection.
Change-Id: Ibb0f86aefd4d4db8d1560608f62735a357da73ea
http://b/
2860493
Nick Kralevich [Wed, 20 Oct 2010 20:53:55 +0000 (13:53 -0700)]
fix operator precedence bug when calculating bits.
Change-Id: I521cdb82f7ad9b2d5ec983716770dac218c49478
Jesse Wilson [Fri, 22 Oct 2010 17:56:04 +0000 (10:56 -0700)]
Merge "Improving the documentation around SecureRandom.setSeed()." into gingerbread
Jesse Wilson [Thu, 21 Oct 2010 23:51:56 +0000 (16:51 -0700)]
Improving the documentation around SecureRandom.setSeed().
Change-Id: Ica6f2814702a90faf96da6fb96de9f7d295102a7
http://b/
3036439
Brian Carlstrom [Thu, 21 Oct 2010 20:50:16 +0000 (13:50 -0700)]
Merge "Restore md2WithRSAEncryption version of Version Class 3 CA" into gingerbread
Brian Carlstrom [Thu, 21 Oct 2010 20:21:14 +0000 (13:21 -0700)]
Restore md2WithRSAEncryption version of Version Class 3 CA
Restoring the md2WithRSAEncryption version of
C=US, O=VeriSign, Inc., OU=Class 3 Public Primary Certification Authority
for HttpsURLConnection compatibility.
Bug:
3119154
Change-Id: I7a0ec4f3a1448987dcdd0ba1d0692a87ce4b3861
Jesse Wilson [Thu, 21 Oct 2010 20:09:21 +0000 (13:09 -0700)]
Merge "Remove obsolete test from AllTests group." into gingerbread
Jesse Wilson [Thu, 21 Oct 2010 20:00:00 +0000 (13:00 -0700)]
Remove obsolete test from AllTests group.
http://b/
3102252
Change-Id: If9a1b1291e6241153b1cb732dc62f2f157b4acfe
Jesse Wilson [Thu, 21 Oct 2010 18:43:04 +0000 (11:43 -0700)]
Merge "Remove tests for SecurityManager." into gingerbread
Jesse Wilson [Thu, 21 Oct 2010 17:16:40 +0000 (10:16 -0700)]
Remove tests for SecurityManager.
Many of the changes are to org.apache.harmony.* tests that we've copied
from upstream. We'll remove these duplicated tests completely once we
have the upstream Harmony tests properly integrated into CTS.
Change-Id: Idd6cfbeba5345156b207b9433928ee37b662c395
http://b/
3102252
Brian Carlstrom [Thu, 21 Oct 2010 00:51:17 +0000 (17:51 -0700)]
Fix SamplingProfiler for use in zygote
Two small fixes to fix SamplingProfilerIntegration:
- Fix name of Timer thread to match native code, allowing
"set prop persist.sampling_profiler 1"
to profile startup of apps.
- clone newThreads array since ThreadSet may reuse it internally
(which is the case in SamplingProfilerIntegration's use of
SamplingProfiler.newThreadGroupTheadSet)
Bug:
3117365
Change-Id: Ic8c2779b7961a39b87494253bb99c56e6e35ae15
Jesse Wilson [Tue, 19 Oct 2010 22:13:56 +0000 (15:13 -0700)]
Merge "Test that strings interned manually and then later loaded as literals maintain reference equality." into gingerbread
Jesse Wilson [Tue, 19 Oct 2010 20:50:18 +0000 (13:50 -0700)]
Test that strings interned manually and then later loaded as literals maintain reference equality.
http://b/
3098960
Change-Id: Ie045df6b1b70eac21ce0a1a19be97692f1278e6f
Dan Bornstein [Sat, 16 Oct 2010 00:27:26 +0000 (17:27 -0700)]
Clean up the construction-time arrangements.
There was all sorts of code here that pointlessly pushed the bulk of
the constructor operation off to instance methods, which (a) could
cause confusion when there's bona fide trouble -- that is, it's better
to throw in the constructor -- and (b) didn't really help with
efficiency in any realistic scenario -- that is, while the original
code nominally avoided extra work until first use, in practice first
use occurs immediately after construction (and in the same thread,
even).
Now, all the one-time work happens in the constructor, and we get to
avoid asking "am I set up yet?" before doing anything substantive.
While I was in the territory, I also did a bit of more minor code
hygiene maintenance.
Change-Id: I41225d373e3ab481b6bc3da5b74bde89c4fd895e
Jesse Wilson [Thu, 14 Oct 2010 19:14:54 +0000 (12:14 -0700)]
Merge "Tests for HTTP bugs using HTTPS through a proxy." into gingerbread
Jesse Wilson [Thu, 14 Oct 2010 19:12:09 +0000 (12:12 -0700)]
Merge "Fix HTTPS hostname verification to use the URL host, not the proxy host." into gingerbread
Jesse Wilson [Thu, 14 Oct 2010 18:31:46 +0000 (11:31 -0700)]
Tests for HTTP bugs using HTTPS through a proxy.
Change-Id: I01b470204d4da46016794ab7040d5751e7116958
http://b/
3097277
http://b/
3097518
Jesse Wilson [Thu, 14 Oct 2010 18:40:15 +0000 (11:40 -0700)]
Fix HTTPS hostname verification to use the URL host, not the proxy host.
Change-Id: I1a0de6f650b89ce0eed908567a04d2e86f76f889
http://b/
3097277
Jesse Wilson [Thu, 14 Oct 2010 17:02:56 +0000 (10:02 -0700)]
Merge "Fix Class.getMethods() for visibility bridge methods." into gingerbread
Jesse Wilson [Thu, 14 Oct 2010 07:57:17 +0000 (00:57 -0700)]
Fix Class.getMethods() for visibility bridge methods.
We had a bug where we couldn't find synthetic methods added by
subclasses to boost visibility of their superclass methods.
The old logic appeared to work fine with the covariant return
type synthetic methods, but broke down on these visibility
synthetic methods.
For reference, the general rules are:
- when there is a synthetic method for a covariant return type,
include it in getMethods() but prefer the natural method on
getMethod() calls.
- when there is a synthetic method for a visibility boost, prefer
it over the superclass method on getMethods() and getMethod()
calls.
Change-Id: Ied1f9fc9892d211647da87959746839f7e0e1140
http://b/
2908173
Brian Carlstrom [Wed, 13 Oct 2010 18:31:56 +0000 (11:31 -0700)]
Change SSLParametersImpl.getDefaultTrustManager to not throw checked exceptions
Change-Id: Id5a042873acc0a8185567ca18ce009c06e54f38d
Brian Carlstrom [Wed, 13 Oct 2010 16:50:52 +0000 (09:50 -0700)]
SSLParameters.getDefaultTrustManager() should lazily initialize its value
Make SSLParametersImpl's defaultKeyManager, defaultTrustManager,
defaultSecureRandom, and defaultParameters all use the single check
idiom for initialization. Move such initialization for
defaultKeyManager and defaultTrustManager out of SSLParametersImpl
constructor into static functions, replacing original
getDefaultTrustManager simple accessor with code that performs lazy
initialization.
luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLParametersImpl.java
dirrect -> direct
luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLParametersImpl.java
luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLSocketImpl.java
hanshake -> handshake
luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLRecordProtocol.java
luni/src/main/java/org/apache/harmony/xnet/provider/jsse/HandshakeProtocol.java
luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLSocketImpl.java
Bug:
2954292
Change-Id: I19bae541613666903b57fccf3e8bfef65b74d6cf
Jesse Wilson [Thu, 7 Oct 2010 23:19:40 +0000 (16:19 -0700)]
Drop a proper finalizer in RandomAccessFile to catch resource leaks.
Submitted to dalvik-dev as https://android-git.corp.google.com/g/72685
http://b/
3015023
Change-Id: I30635aab91774f5eaecd354e8418ba6401705757
Jesse Wilson [Thu, 7 Oct 2010 18:37:31 +0000 (11:37 -0700)]
Test what happens to RandomAccessFile when it is leaked.
Change-Id: I6b5044a5c3375c9e9456b9c72d897ecd2ccd7e26
http://b/
3015023
Elliott Hughes [Thu, 7 Oct 2010 18:09:58 +0000 (11:09 -0700)]
java.io.File javadoc improvements.
Bug: http://code.google.com/p/android/issues/detail?id=5929
Change-Id: Ie24c1060a968e3da611aef000ff29762f6f7101a
Elliott Hughes [Wed, 6 Oct 2010 22:51:17 +0000 (15:51 -0700)]
Merge "Fix and improve Matcher.group(int) documentation." into gingerbread
Elliott Hughes [Wed, 6 Oct 2010 18:13:52 +0000 (11:13 -0700)]
Fix and improve Matcher.group(int) documentation.
Bug: http://code.google.com/p/android/issues/detail?id=11723
Change-Id: Idd1c9009f1267ca7624ba8492a162bd72232f2cc
Brian Carlstrom [Tue, 5 Oct 2010 21:40:42 +0000 (14:40 -0700)]
Merge "Favor Harmony's CertificateFactory.X509 over BouncyCastle's" into gingerbread
Brian Carlstrom [Tue, 5 Oct 2010 04:57:48 +0000 (21:57 -0700)]
Favor Harmony's CertificateFactory.X509 over BouncyCastle's
Bug:
2225935
Change-Id: Ibca92c9fa34fc539ebb8ea86fb0ced62e3dbe5de
Brian Carlstrom [Mon, 4 Oct 2010 23:32:48 +0000 (16:32 -0700)]
Merge "Test updates for new SecretKeyFactory.PBKDF2WithHmacSHA1 support" into gingerbread
Brian Carlstrom [Mon, 4 Oct 2010 19:38:59 +0000 (12:38 -0700)]
Test updates for new SecretKeyFactory.PBKDF2WithHmacSHA1 support
Bug:
3059950
Change-Id: I24546cb9e38b17ea615e36de3606ec6d373df594
Jesse Wilson [Mon, 4 Oct 2010 21:33:36 +0000 (14:33 -0700)]
Merge "Remove logging tests duplicated from Harmony" into gingerbread
Jesse Wilson [Mon, 4 Oct 2010 21:19:52 +0000 (14:19 -0700)]
Merge "Delete test for the AES algorithm parameter generator, which we no longer support." into gingerbread
Jesse Wilson [Mon, 4 Oct 2010 20:18:03 +0000 (13:18 -0700)]
Delete test for the AES algorithm parameter generator, which we no longer support.
The RI does not include an algorithm parameter generator for AES.
http://b/
3061194
BouncyCastle dropped support for AES on February 18 2007:
http://www.bouncycastle.org/viewcvs/viewcvs.cgi/java/crypto/src/org/bouncycastle/jce/provider/BouncyCastleProvider.java.diff?r1=1.38&r2=1.39
http://www.bouncycastle.org/viewcvs/viewcvs.cgi/java/crypto/src/org/bouncycastle/jce/provider/JDKAlgorithmParameterGenerator.java.diff?r1=1.3&r2=1.4
Unfortunately, their checkin comment isn't intention-revealing:
"fixed implictyCa to work with KeyFactory
added SEED/Camllia support to CMS
added further ISO9797 mac support
further refactoring of JCE symmetric cipher code
Change-Id: I9aa045e1b5c8103cdf736efb1eb927c1acc747ae
Jesse Wilson [Mon, 4 Oct 2010 18:10:54 +0000 (11:10 -0700)]
Fix a test with broken expectations for localized arabic exponent.
Also tagging the RI as failing this test, and fixing a broken bug number
in another test.
Change-Id: I800e82832050d45df6853fd7d3c5bc14e60842d3
http://b/
3060661
Brian Carlstrom [Mon, 4 Oct 2010 17:52:11 +0000 (10:52 -0700)]
Merge "Update TestUtils certificates to fix CertPathBuilder1Test and CertPathTest" into gingerbread
Brian Carlstrom [Mon, 4 Oct 2010 05:20:54 +0000 (22:20 -0700)]
Update TestUtils certificates to fix CertPathBuilder1Test and CertPathTest
Bug:
2322662
Change-Id: I8ad9a91f4095807bd710045eef3a97a86b560f49
Jesse Wilson [Sun, 3 Oct 2010 18:55:43 +0000 (11:55 -0700)]
Remove math tests duplicated from Harmony.
Change-Id: Ia7b24e2f7637b8036b646182cc0411ea83a4b09b
Jesse Wilson [Sat, 2 Oct 2010 21:54:19 +0000 (14:54 -0700)]
Remove logging tests duplicated from Harmony
Change-Id: I057552b555ee891810d4577b7359ed309d5c1611
Jesse Wilson [Sat, 2 Oct 2010 00:23:26 +0000 (17:23 -0700)]
Scrub test failures in java.text.
Change-Id: Icee3cd8bd3b8fa4b3902435c66021cabb7fc322d
Brian Carlstrom [Fri, 1 Oct 2010 22:26:53 +0000 (15:26 -0700)]
Merge "SSL* AppData should not hold onto JNI global references" into gingerbread
Brian Carlstrom [Fri, 1 Oct 2010 02:22:21 +0000 (19:22 -0700)]
SSL* AppData should not hold onto JNI global references
Summary:
NativeCrypto.SSL_do_handshake stored JNI global references in its
AppData instance for use in upcalls from OpenSSL that invoke Java
callbacks. However, one of the references was to the
SSLHandshakeCallbacks which in the common case of OpenSSLSocketImpl is
the OpenSSLSocketImpl instance itself. This meant that if code dropped
the OpenSSLSocketImpl without closing (such as Apache HTTP Client),
the instances would never be collected, and perhaps more importantly,
file descriptors would not be closed.
The fix is to pass in the objects required during a callback in all
downcalls to SSL_* methods that could result in a callback and clear
them on return. The existing code already did this for the JNIEnv*, so
that code was expanded to handle setting the jobjects as well.
Details:
In the native code used to extract the FileDescriptor object from a
Socket on the call to NativeCrypto.SSL_do_handshake. However, since we
need this for every read and write operations, we now do this in Java
to avoid the repeated overhead. NativeCrypto.SSL_do_handshake now
takes a FileDescriptor, which it extracted from the Socket the
convenience function using NativeCrypto.getFileDescriptor(Socket)
luni/src/main/java/org/apache/harmony/xnet/provider/jsse/NativeCrypto.java
In addition to tracking changes to pass FileDescriptor and
SSLHandshakeCallbacks, removed final uses of getFieldId since the
code no longer needs to extract FileDescriptors itself
luni/src/main/native/NativeCrypto.cpp
The Socket field used to be non-null in the wrapper case and null in
the non-wrapper case. To simplify things a bit, "socket == this" in
the non-wrapper case. The socket field is now also final and joined by
a final FileDescriptor field.
luni/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java
Updated NativeCryptoTest to track FileDescriptor and
SSLHandshakeCallbacks by expanding the Hooks.afterHandshake to provide
them. Also changed to add a 5 second timeout to many test cases.
luni/src/test/java/org/apache/harmony/xnet/provider/jsse/NativeCryptoTest.java
Bug:
2989218
Change-Id: Iccef92b59475f3c1929e990893579493ece9d442
Elliott Hughes [Fri, 1 Oct 2010 18:01:23 +0000 (11:01 -0700)]
Remove an out-of-date expectation.
(You have to remove all the SecurityManager cruft from this SecurityManager
test to see that this expectation is bogus.)
Bug:
1677221
Change-Id: I3907e0c105523195669c32d2aad7e8133f2e7f12
Elliott Hughes [Fri, 1 Oct 2010 17:35:20 +0000 (10:35 -0700)]
Merge "Better detail messages for System.loadLibrary UnsatisfiedLinkErrors." into gingerbread
Jesse Wilson [Fri, 1 Oct 2010 00:11:24 +0000 (17:11 -0700)]
Include a message when a task is rejected from an executor.
Browser folks have been seeing this exception in their logcat logs
and it would be handy to give them some insight on why the pool
rejected the task.
Change-Id: I7257b3b174e6e2342e63238a542095bb3f7845f2
http://b/issue?id=
3023092
Elliott Hughes [Thu, 30 Sep 2010 22:52:41 +0000 (15:52 -0700)]
Better detail messages for System.loadLibrary UnsatisfiedLinkErrors.
This has a companion change in dalvik:
https://android-git.corp.google.com/g/71344
Bug:
3044042
Change-Id: Ia665ee59adf1ae1dbb45ba95988355839f4b0f23
Jesse Wilson [Thu, 30 Sep 2010 22:13:08 +0000 (15:13 -0700)]
Merge "Don't explode if flush() is called when an HTTP upload stream is closed." into gingerbread
Jesse Wilson [Thu, 30 Sep 2010 21:55:32 +0000 (14:55 -0700)]
Merge "Don't use the decompression-buffer as the skip buffer in InflaterInputStream." into gingerbread
Jesse Wilson [Thu, 30 Sep 2010 21:51:31 +0000 (14:51 -0700)]
Don't explode if flush() is called when an HTTP upload stream is closed.
Streams like BufferedOutputStream always call flush() even if it is
a no-op. These flushes must be permitted for compatibility with apps
that don't call close() until after the response has been retrieved.
Change-Id: Ic5c86ab6050c6c4d166c44ae5b4fc7a1688e7e45
http://b/
3038470
Brian Muramatsu [Thu, 30 Sep 2010 21:50:54 +0000 (14:50 -0700)]
Merge "Add More Support_Locale Conditions to ScannerTest" into gingerbread
Elliott Hughes [Thu, 30 Sep 2010 21:06:57 +0000 (14:06 -0700)]
Merge "Fix File.getCanonicalizePath to resolve symbolic links." into gingerbread
Elliott Hughes [Wed, 29 Sep 2010 23:20:03 +0000 (16:20 -0700)]
Fix File.getCanonicalizePath to resolve symbolic links.
This requires fixing a few tests that don't cope with the fact that
all our temporary files are created on /sdcard, which is a symbolic
link to /mnt/sdcard, meaning that the canonical path of any temporary
file is different from the file's absolute path.
Bug:
3047893
Change-Id: I02245a290b6d2962fb1dd4d2faba30d9aa7168e0
Jesse Wilson [Thu, 30 Sep 2010 02:15:50 +0000 (19:15 -0700)]
Don't use the decompression-buffer as the skip buffer in InflaterInputStream.
This has bad consequences for the GzipInputStream subclass, which can't
use the input buffer as a scratch space because it needs to CRC all of
the decompressed data.
Change-Id: I310261dc8ef7a5082768cc6d6a2eff4ef77f5a21
http://b/
3042574
Brian Carlstrom [Thu, 30 Sep 2010 17:23:58 +0000 (10:23 -0700)]
Merge "CA certificate update" into gingerbread
Jesse Wilson [Wed, 29 Sep 2010 21:34:56 +0000 (14:34 -0700)]
Add expectations for DecimalFormat test failures.
Change-Id: Iee4dffb4f3dded37866f35f0100d905553b9b3ca
Jesse Wilson [Wed, 29 Sep 2010 22:10:59 +0000 (15:10 -0700)]
Merge "Fix problems with OldSimpleDateFormatTest.java." into gingerbread
Jesse Wilson [Wed, 29 Sep 2010 21:17:23 +0000 (14:17 -0700)]
Fix problems with OldSimpleDateFormatTest.java.
This changes the single 'y' documentation to point out our behavior.
http://b/
2788874
It also isolates two other failures for week-in-year and timezone
formatting.
http://b/
1613709 Minimal days in first week is broken
http://b/
3049014 Custom SimpleTimeZones not honored by SimpleDateFormat
Change-Id: Ic632bf70076fe0dfd376317379dc363269d5b2df
Elliott Hughes [Wed, 29 Sep 2010 21:06:43 +0000 (14:06 -0700)]
Fix IP_MULTICAST_IF getsockopt.
Also clean out test cruft so we can run on the RI, and remove
SecurityManager cruft so we can run more tests. Our failures
now match the RI's (but I'm not planning on fixing the tests
right now, nor have I checked the upstream tests).
Bug:
3045865
Change-Id: I326fff9753faa88eacb20a04ffd2f2e84e76259a
Brian Carlstrom [Wed, 29 Sep 2010 18:35:44 +0000 (11:35 -0700)]
CA certificate update
Added SHA1 : B5:1C:06:7C:EE:2B:0C:3D:F8:55:AB:2D:92:F4:FE:39:D4:E7:0F:0E
Subject : C=US, ST=Arizona, L=Scottsdale, O=Starfield Technologies, Inc., CN=Starfield Root Certificate Authority - G2
Added SHA1 : 92:5A:8F:8D:2C:6D:04:E0:66:5F:59:6A:FF:22:D8:63:E8:25:6F:3F
Subject : C=US, ST=Arizona, L=Scottsdale, O=Starfield Technologies, Inc., CN=Starfield Services Root Certificate Authority - G2
Bug:
2994278
Change-Id: I29ab0219a8d5ba04b6812330516a781b9e25a929
Jesse Wilson [Wed, 29 Sep 2010 17:19:40 +0000 (10:19 -0700)]
Merge "Support creating unbound server sockets in DefaultServerSocketFactory." into gingerbread
Jesse Wilson [Wed, 29 Sep 2010 16:59:08 +0000 (09:59 -0700)]
Support creating unbound server sockets in DefaultServerSocketFactory.
Change-Id: I34106bb55ad99a134b4aae4e24e61b59c0aaa967
http://b/
2587385
Brian Carlstrom [Wed, 29 Sep 2010 04:58:53 +0000 (21:58 -0700)]
Merge "Scrub missing calls to super.finalize()" into gingerbread
Brian Carlstrom [Wed, 29 Sep 2010 00:18:08 +0000 (17:18 -0700)]
Scrub missing calls to super.finalize()
Bug:
3024226
Change-Id: I6642cb9d4929ba72244529efe4ebdfa595ae4fa7
Jesse Wilson [Wed, 29 Sep 2010 04:10:51 +0000 (21:10 -0700)]
Merge "Tests for ServerSocketFactory.getDefault()." into gingerbread
Jesse Wilson [Wed, 29 Sep 2010 02:36:39 +0000 (19:36 -0700)]
Tests for ServerSocketFactory.getDefault().
Motivated by http://b/
2587385
Change-Id: If30898859f869c88342d1069a2425575752ebf6e
Jesse Wilson [Wed, 29 Sep 2010 00:48:27 +0000 (17:48 -0700)]
Merge "Fix a problem where URL.equals() was returning true when both hosts couldn't resolve." into gingerbread
Jesse Wilson [Wed, 29 Sep 2010 00:31:05 +0000 (17:31 -0700)]
Fix a problem where URL.equals() was returning true when both hosts couldn't resolve.
See http://b/
3045007
Change-Id: I83f2a0d8888dd30aaf6099049ce8008487d4337d
Jesse Wilson [Tue, 28 Sep 2010 22:13:55 +0000 (15:13 -0700)]
Fix problems where knownfailures has drifted out of date from our suite.
http://b/
3045163.
Change-Id: I9c9c6d4429f09d5f5cd92447de9a3c9d67294286
Jesse Wilson [Tue, 28 Sep 2010 20:40:14 +0000 (13:40 -0700)]
Remove java.net tests that are redundant with Harmony.
Change-Id: I01d6b3633a6de8a075982975d4cbf6c08e757490
Jesse Wilson [Tue, 28 Sep 2010 20:14:24 +0000 (13:14 -0700)]
Merge "Remove java.text tests duplicated from Harmony" into gingerbread
Elliott Hughes [Tue, 28 Sep 2010 16:55:50 +0000 (09:55 -0700)]
Merge "Add submitter's test for a publicly-reported SimpleTimeZone.clone bug I already fixed." into gingerbread
Elliott Hughes [Tue, 28 Sep 2010 16:53:41 +0000 (09:53 -0700)]
Add submitter's test for a publicly-reported SimpleTimeZone.clone bug I already fixed.
Bug: http://code.google.com/p/android/issues/detail?id=11542
Change-Id: Ic13e6083ec26e9d8725e94294f7cd69902e69dd8
Brian Carlstrom [Tue, 28 Sep 2010 04:06:09 +0000 (21:06 -0700)]
Merge "CA certificate update" into gingerbread
Jesse Wilson [Tue, 28 Sep 2010 02:08:51 +0000 (19:08 -0700)]
Merge "Update expectations for tests failing only on the host." into gingerbread
Jesse Wilson [Tue, 28 Sep 2010 01:05:32 +0000 (18:05 -0700)]
Update expectations for tests failing only on the host.
Change-Id: Ieb30ef5b4607f7dfed0b873a10e3dcb31139141d
Jesse Wilson [Tue, 28 Sep 2010 02:04:21 +0000 (19:04 -0700)]
Remove java.text tests duplicated from Harmony
Change-Id: Ib20c8956be045158225d08327781a5e16ca08689
Elliott Hughes [Mon, 27 Sep 2010 23:53:25 +0000 (16:53 -0700)]
Improve DateFormat.parse's exception detail.
Bug:
3041734
Change-Id: I73e9d2aefecc5fe67415eb3a6586294f8dbbcb82
Elliott Hughes [Mon, 27 Sep 2010 23:41:44 +0000 (16:41 -0700)]
Somewhat perfunctory javadoc for the new Java 6 JDBC APIs.
I feel dirty.
Bug:
3038452
Change-Id: I7ecd24b09c4e37a5ad29de4c6ba50a6e151a1296
Elliott Hughes [Sun, 26 Sep 2010 23:59:04 +0000 (16:59 -0700)]
Don't wrap select(2) in TEMP_FAILURE_RETRY.
Luckily, since the caller has to cope with us saying "we're not connected yet,
try again", we can just go that route.
I've also added another plausible possible exception to one of the tests, seen
once when testing on a host dalvikvm.
Bug:
3022824
Change-Id: Ia155f86f9f7b9b0a54e58c55f668faaebcbda55e
Jesse Wilson [Sun, 26 Sep 2010 20:56:10 +0000 (13:56 -0700)]
Merge "Remove FormatterTest tests duplicated with Harmony." into gingerbread
Jesse Wilson [Sun, 26 Sep 2010 20:51:17 +0000 (13:51 -0700)]
Remove FormatterTest tests duplicated with Harmony.
Change-Id: I9c95aac22b9df7efff6595e94700a17cfc5d3d7b
Brian Carlstrom [Sun, 26 Sep 2010 17:34:41 +0000 (10:34 -0700)]
Merge "Fix OpenSSLSessionImpl.getValueNames regression" into gingerbread
Jesse Wilson [Sat, 25 Sep 2010 15:53:24 +0000 (08:53 -0700)]
Attach bugs to failures in MathTest and StrictMathTest.
Change-Id: Ifd0cf35e6e3dee4fe512db962db5f44c6e614af4
Brian Carlstrom [Sat, 25 Sep 2010 05:53:46 +0000 (22:53 -0700)]
Fix OpenSSLSessionImpl.getValueNames regression
In
e32b21f14d52bac429a9c54fe031f9e92c911d64, the code was converted to
use Objects.equals. However, because of a typo, an autoboxed Boolean
was passed instead of an AccessControlContext. I reviewed the rest of
the original change to make sure there were no other instances of this
regression.
Also cleaned up the SSLSessionTest (fixing two broken tests
test_getLocalPrincipal and test_getPeerPrincipal) and fixed a
whitespace issue in AccessControlContext.
Change-Id: Icaee8a0c2f5f527bea7a80037fe3f99c509d9f42
Jesse Wilson [Sat, 25 Sep 2010 00:29:36 +0000 (17:29 -0700)]
Attach charset tests to their tracking bugs.
Change-Id: I598a6b34a662e528f9b03a41f419bf7e47755ce2
Brian Carlstrom [Fri, 24 Sep 2010 23:54:34 +0000 (16:54 -0700)]
Merge "OpenSSLSocketImpl should not call NativeCrypto.SSL_set_client_CA_list with an empty array" into gingerbread
Brian Carlstrom [Fri, 24 Sep 2010 23:50:35 +0000 (16:50 -0700)]
OpenSSLSocketImpl should not call NativeCrypto.SSL_set_client_CA_list with an empty array
Bug:
3034616
Change-Id: Ib39ebfa737910f0ebce5ac2ad87715579bd7aa3d
Jesse Wilson [Fri, 24 Sep 2010 23:42:25 +0000 (16:42 -0700)]
Merge "Rely on the test runner to ensure a pristine VM." into gingerbread
Jesse Wilson [Fri, 24 Sep 2010 23:03:12 +0000 (16:03 -0700)]
Rely on the test runner to ensure a pristine VM.
Change-Id: I6f5bfad6f861eb7b398ed7d86747d66cea4f2343
http://b/issue?id=
2660429
Elliott Hughes [Fri, 24 Sep 2010 18:12:49 +0000 (11:12 -0700)]
Throw SocketTimeoutException rather than SocketException if connect times out.
Bug:
3032900
Change-Id: I5da70fc55f2366a504b2492f2bc7f2269a57ef14
Jesse Wilson [Fri, 24 Sep 2010 16:23:44 +0000 (09:23 -0700)]
Merge "Add expectations for some tests failing in Hudson." into gingerbread