- 09 Jul, 2010 2 commits
-
-
Elliott Hughes authored
(The ArrayIndexOutOfBoundsException for System.arraycopy already talks about 'dst' and 'dstPos'.) Change-Id: Iba9415dd4a9ec3b457938ea4469b4a0024bab6e4
-
Brian Carlstrom authored
Summary: Goal here was to just make most tests faster by only having TestKeyStore create RSA keys by default. However, when I did that SSLEngineTest#test_SSLEngine_clientAuth started working, so I ended up investigating a much deeper issue with DSA client authentication against an RSA SSLEngine server. Details: Changed the TestKeyStore.get singleton to only contain RSA keys. TestKeyStore.create now requires the caller enumerate what keys they want if they need more than that or an alternative. support/src/test/java/javax/net/ssl/TestKeyStore.java Changed test_SSLSocket_getSupportedCipherSuites_connect to explicitly request RSA and DSA keys since it needs both to try connecting all possible cipher suites. luni/src/test/java/javax/net/ssl/SSLSocketTest.java Fixing SSLEngine client authentication when server uses RSA but client uses DSA Fixed java.net.ssl.SSLEngineTest#test_SSLEngine_clientAuth expectations/knownfailures.txt Added CiperSuite.authType field which contains the algorithm name such as RSA, DSA, DH, that the client will use to authenticate the server. Like the cipherName, hmacName, and hashName, this is logically derivable from the the CiperSuite.KEY_EXCHANGE_*, but we remember it to avoid repeatedly doing large cascading "if" tests to determine which key algorithm should be used for each case. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/CipherSuite.java Fixed a number of client certificate authentication bugs in SSLEngine - Changed ClientHandshakeImpl's in the SSL/Tls Certificate message code to mirror ServerHandshakeImpl's implementation to properly use chooseEngineClientAlias in the SSLEngine case. - Changed to use the client certifcates key algorithm for computing the signature for the SSL/TLS CertificateVerify message. Previously we used the cipher suites negoitated key exchange method, but if the client may select a certificate with a different algorithm if the server provides a CA for another algorithm. - Also changed to use CipherSuite.isAnonymous in two places rather than the inlined equivalent. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/ClientHandshakeImpl.java Fixed client authentication to use the client's certificate (not the server's) to do verify the CertificateVerify message signature. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/ServerHandshakeImpl.java Fixed bug in DigitalSignature which did not Signature.update in verifySignature, so it could never have properly authenticated DSA signatures. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/DigitalSignature.java Added CertificateMessage getAuthType convenience luni/src/main/java/org/apache/harmony/xnet/provider/jsse/CertificateMessage.java Made CertificateRequest certificate_authorities final, found we were double allocating it luni/src/main/java/org/apache/harmony/xnet/provider/jsse/CertificateRequest.java Cleaning up imports of HandshakeProtocol while working on its subclasses. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/HandshakeProtocol.java Cleaned up while looking at X509KeyManager implementations while debugging. support/src/test/java/org/apache/harmony/xnet/tests/support/X509KeyManagerImpl.java Change-Id: I74b98754c11000cbfea416f1571c380c9c67abf3
-
- 29 Jun, 2010 1 commit
-
-
Brian Carlstrom authored
The following new benchmarks where tested with the below changes: - DigestBenchmark - MessageDigestBenchmark - SSLSocketBenchmark - SignatureBenchmark Fix package name of OpenSSLProvider luni/src/main/java/java/security/security.properties Restore Java (vs OpenSSL) SSLSocket wrappers on SSLEngine for benchmarking luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLServerSocketFactoryImpl.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLServerSocketImpl.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLSocketFactoryImpl.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLSocketImpl.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLSocketInputStream.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLSocketOutputStream.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLSocketWrapper.java Restore HandshakeProtocol.socketOwner code for SSLSocket to function luni/src/main/java/org/apache/harmony/xnet/provider/jsse/ClientHandshakeImpl.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/HandshakeProtocol.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/ServerHandshakeImpl.java Remove unneeded OpenSSLMessageDigestJDK.getInstance since these are registered via OpenSSLProvider and SHA224 which is not part of the RI. We had already removed the BouncyCastle version of this. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLMessageDigestJDK.java luni/src/test/java/tests/targets/security/AllTests.java luni/src/test/java/tests/targets/security/MessageDigestTestSHA224.java luni/src/test/java/tests/targets/security/SignatureTestSHA224withRSA.java Change-Id: I7daae7f0d9f50acad6df9157eac1b0133af83062
-
- 26 Jun, 2010 1 commit
-
-
Brian Carlstrom authored
external/bouncycastle - Change to be the primary build for bouncycastle sources (as opposed to part of libcore) - Moved OpenSSLMessageDigest from libcore to OpenSSLDigest It uses NativeCrypto API from core, but implements a bouncycastle specific interface - restored registration of bouncycastle MessageDigests for SHA-1, SHA-256, MD5 OpenSSLProvider versions take precedence, but explicit provider of "BC" allows choice - enabled native versions of SHA-384 and SHA-512 - pruned MD4 implementation frameworks/base - frameworks and CoreTests modules now depend on bouncycastle - update preloades classes for NativeBN package change - moved CryptoTest to libcore libcore - core now builds without bouncycastle sources - core-tests, core-tests-support, core-tests-supportlib now depend on bouncycastle - removed libcore/openssl directory, moving NativeBN to java/math - minor cleanup of Provider, Security, Services style while working on ProviderTest - added new OpenSSLProvider registered as first provider to have priority over the others to ensure our native implementations are used - moved BouncyCastle to have priority as a provider over Harmony - JarVerifier and JarUtils now implicitly use OpenSSLMessageDigest - Cleanedup OpenSSLSignature, implementation needs to be finished to move to OpenSSLProvider - To avoid using PEMWriter from BouncyCastle, NativeCrypto now takes binary encoded certs and keys This is more efficient as well avoiding the base64 decode/encode of the binary data - removed SHA-224 to match the RI packages/apps/CertInstaller - CertificateInstaller module now depends on bouncycastle this is the only app to depend on bouncycastle system/core - updated BOOTCLASSPATH Change-Id: I6205366b12baec4331b4a76e2c85d8324bf64b2c
-
- 22 Jun, 2010 1 commit
-
-
Brian Carlstrom authored
Change-Id: I65292321560c9f4551dc79fc7c6795f093638bbf
-
- 18 Jun, 2010 1 commit
-
-
Jesse Wilson authored
Moving TimeZoneTest to OldTimeZoneTest and removing test methods that are duplicated between libcore and Harmony. Also adding Objects.equals() to make implementing this easy, and removing redundant time zone tests. I did a few searches to find candidate code that could take advantage of this new utility method and adopted it there. Change-Id: I133298f1b36d755bd35c1ad0dc0ab366fd164270
-
- 11 Jun, 2010 1 commit
-
-
Elliott Hughes authored
As explained in the bug, I don't think we can/should fix this potential native crash, but we can and should improve the documentation to explain how you're _supposed_ to use MessageDigest. Bug: http://code.google.com/p/android/issues/detail?id=8709 Change-Id: I1cbab5995e5673d5386e21270ac52b6f90b9f421
-
- 03 Jun, 2010 1 commit
-
-
Elliott Hughes authored
Change-Id: I6905802cfe7cda73bb6f52fe0cc79767b4645e82
-
- 28 May, 2010 2 commits
-
-
Jesse Wilson authored
Highlights: code was moved from SSLContextImpl to its superclass. took X500Principal code from Harmony Tested with Harmony's tests.api.javax.security.auth.x500.X500PrincipalTest. Change-Id: I89b46d4b47e692a5461916cca972e05de95f3280
-
Brian Carlstrom authored
Conflicts: JavaLibrary.mk luni/src/main/java/java/lang/System.java luni/src/main/java/org/apache/harmony/luni/internal/net/www/protocol/http/HttpURLConnectionImpl.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java luni/src/test/java/java/net/URLConnectionTest.java support/src/test/java/tests/TestSuiteFactory.java x-net/src/main/native/org_apache_harmony_xnet_provider_jsse_NativeCrypto.cpp Change-Id: I1038f749c8c9bd640aae7ca96627810936454883
-
- 27 May, 2010 2 commits
-
-
Brian Carlstrom authored
Change-Id: I470c929f67ecaffa91d5a67c87f1ed5358cfd84c
-
Brian Carlstrom authored
Summary: - RI 6 support for javax.net.ssl - SSLEngine fixes based on new SSLEngineTest - fix Cipher.checkMode bug recently introduced in dalvik-dev Details: Fix Cipher.checkMode that was preventing most javax.net.ssl tests from working luni/src/main/java/javax/crypto/Cipher.java RI 6 has introduced the concept of a "Default" SSLContext. This is accessed via SSLContext.getDefault() and also SSLContext.getInstance("Default"). Harmony had its own DefaultSSLContext but it was not created via an SSLContextSpi. It also was a single shared instance whereas the new RI6 Default SSLContext shares internal SSLSessionContext instances between different Default SSLContexts. Refactored the old code into an SSLContextImpl subclass that allows it to be created via SSLContext.getInstance. SSLContextImpl ensures that we only ever create one set of SSLSessionContext instances for the Default context. luni/src/main/java/javax/net/ssl/DefaultSSLContext.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/DefaultSSLContextImpl.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLContextImpl.java Added SSLContext.getDefault and SSLContext.setDefault luni/src/main/java/javax/net/ssl/SSLContext.java Replace dependencies of old DefaultSSLContext with use of SSLContext.getDefault luni/src/main/java/javax/net/ssl/SSLServerSocketFactory.java luni/src/main/java/javax/net/ssl/SSLSocketFactory.java Register "SSLContext.Default" as DefaultSSLContextImpl class for SSLContext.getInstance() luni/src/main/java/org/apache/harmony/xnet/provider/jsse/JSSEProvider.java Added constant for new "Default" standard name and added it to SSL_CONTEXT_PROTOCOLS. New tests based on SSL_CONTEXT_PROTOCOLS made it clear that neither Android or RI support SSLv2 so removed it from SSL_CONTEXT_PROTOCOLS and SSL_SOCKET_PROTOCOLS. Added constant for TLS as well which was previously scattered all over tests. Remove SSLv2Hello from SSL_SOCKET_PROTOCOLS for Android since with OpenSSL disablign SSLv2 means you can not use SSLv2Hello either. support/src/test/java/javax/net/ssl/StandardNames.java Added tests for SSLContext.getDefault and SSLContext.setDefault. Changed existing tests to work on all protocols including new "Default". luni/src/test/java/javax/net/ssl/SSLContextTest.java RI 6 has introduced the notion of SSLParameters which encapsulate SSL the handshake parameters of desired cipher suites, protocols, and client authentication requirements. The main new class SSLParameters is basically just a bag of fields with accessors and a couple simple constructors. The only things of note are that it clones all String arrays on input and output and the setters for the two boolean fields ensure that only one is true at a time. luni/src/main/java/javax/net/ssl/SSLParameters.java Added SSLContext.getDefaultSSLParameters and SSLContext.getSupportedSSLParameters which simply delegate to the SSLContextSpi. luni/src/main/java/javax/net/ssl/SSLContext.java Added abstract SSLContextSpi.engineGetDefaultSSLParameters and SSLContext.engineGetSupportedSSLParameters. luni/src/main/java/javax/net/ssl/SSLContextSpi.java Added engineGetDefaultSSLParameters and engineGetSupportedSSLParameters implementation. The RI documents in SSLContextSpi that these are implemented by default by creating a socket via the SSLContext's SocketFactory and asking for the enabled/supported cipher suites and protocols respectively, so that is what is done. The doc mentions throwing UnsupportedOperationException if there is a problem, so we do that as well. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLContextImpl.java Added {SSLEngine,SSLSocket}.{getSSLParameters,setSSLParameters} which are analogous. luni/src/main/java/javax/net/ssl/SSLEngine.java luni/src/main/java/javax/net/ssl/SSLSocket.java Added SSLParametersTest luni/src/test/java/javax/net/ssl/SSLParametersTest.java luni/src/test/java/javax/net/ssl/AllTests.java Added SSLContext.get{Default,Supported}SSLParameters tests luni/src/test/java/javax/net/ssl/SSLContextTest.java Added SSLSocket.{getSSLParameters,setSSLParameters} tests and added some extra asserts to test_SSLSocketPair_create based on experience with test_SSLEnginePair_create. luni/src/test/java/javax/net/ssl/SSLSocketTest.java Dummy implementation of new SSLContextSpi for test classes. support/src/test/java/org/apache/harmony/security/tests/support/MySSLContextSpi.java support/src/test/java/org/apache/harmony/xnet/tests/support/MySSLContextSpi.java Other minor RI 6 API changes: RI 6 removed Serializable from HandshakeCompletedEvent and SSLSessionBindingEvent luni/src/main/java/javax/net/ssl/HandshakeCompletedEvent.java luni/src/main/java/javax/net/ssl/SSLSessionBindingEvent.java RI 6 added generic types to the KeyStoreBuilderParameters List constructor and accessor as well as to SSLSessionContext.getIds. Fixed tests to compile with generic types. luni/src/main/java/javax/net/ssl/KeyStoreBuilderParameters.java luni/src/main/java/javax/net/ssl/SSLSessionContext.java luni/src/test/java/tests/api/javax/net/ssl/KeyStoreBuilderParametersTest.java SSLEngine improvements. Since I was changing SSLEngine, I wrote an SSLEngineTest based on my SSLSocketTest to do some simply sanity checking. It expose a number of issues. I've fixed the small ones, marked the rest as known failures. Renamed some TLS_ cipher suites to SSL_ to match JSSE standard names. These were all old suites no longer supported by RI or OpenSSL which is why they were missed in an earlier cleanup of this type in this class. Also fixed SSLEngine supported cipher suites list not to include SSL_NULL_WITH_NULL_NULL which is not a valid suite to negotiate. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/CipherSuite.java SSLEngine instances can have null host values, which caused a NullPointerException in the ClientSessionContext implementation. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/ClientSessionContext.java SSLEngine tests were failing because SSLParameters was throwing NullPointerException instead of IllegalArgument exception on null element values. Fixed null pointer message style while I was here. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLParameters.java Fixed SSLEngine instances to default to server mode like RI luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLContextImpl.java Fixed KEY_TYPES based on SSLEngine implementation. Removed dead code NativeCrypto.getEnabledProtocols which was recently made obsolete. Cleaned up null exception messages to follow our convention. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/NativeCrypto.java Added SSLEngineTest which parallels SSLSocketTest in its coverage. Similarly added TestSSLEnginePair which loosely parallels TestSSLSocketPair. luni/src/test/java/javax/net/ssl/SSLEngineTest.java luni/src/test/java/javax/net/ssl/AllTests.java support/src/test/java/javax/net/ssl/TestSSLEnginePair.java SSLEngineTest betters exposed the differences between SSLSocket and SSLEngine supported cipher suites. StandardNames now has an CIPHER_SUITES_SSLENGINE definition which denotes what is missing and what is extra and why in the SSLEngine implementation. support/src/test/java/javax/net/ssl/StandardNames.java Created StandardNames.assert{Valid,Supported}{CipherSuites,Protocols} to factor out some code test code that is also used by new tests. support/src/test/java/javax/net/ssl/StandardNames.java luni/src/test/java/javax/net/ssl/SSLSocketFactoryTest.java luni/src/test/java/javax/net/ssl/SSLSocketTest.java Remove SSLSocketTest known failure and add new SSLEngineTest known failures expectations/knownfailures.txt SSL_OP_NO_TICKET change was recently merged from master which required some fixes. For the moment, sslServerSocketSupportsSessionTickets always returns false. support/src/test/java/javax/net/ssl/TestSSLContext.java Fixed flakey test_SSLSocket_HandshakeCompletedListener which had a race because the client thread look in the server session context for an session by id potentially before the server thread had a chance to store its session. Made noticable because of SSL_OP_NO_TICKET recently merged from master (before this code path was host only, not device) luni/src/test/java/javax/net/ssl/SSLSocketTest.java Fix checkjni issue where we need to check for pending exception in OpenSSL callback. Possibly introduced by recent merge of SSL_OP_NO_TICKET from master. luni/src/main/native/org_apache_harmony_xnet_provider_jsse_NativeCrypto.cpp Expectation updates Remove SSLSocketTest known failure and add new SSLEngineTest known failures expectations/knownfailures.txt Tag test_SSLSocket_getSupportedCipherSuites_connect as large expectations/taggedtests.txt Misc changes: opening brace on wrong line luni/src/main/java/org/apache/harmony/xnet/provider/jsse/ServerSessionContext.java Long line cleanup while debugging luni/src/main/java/org/apache/harmony/xnet/provider/jsse/HandshakeProtocol.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLServerSocketFactoryImpl.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketFactoryImpl.java support/src/test/java/javax/net/ssl/TestKeyStore.java Removed bogus import luni/src/test/java/javax/net/ssl/SSLSessionContextTest.java Comment clarify while debugging luni/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java Ctor -> Constructor in comment luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLEngineImpl.java Fixed naming of SocketTest_Test_create to TestSocketPair_Create to match renamed classes luni/src/test/java/javax/net/ssl/SSLSocketTest.java Change-Id: I99505e97d6047eeabe4a0b93202075a0b2d486ec
-
- 20 May, 2010 1 commit
-
-
Brian Carlstrom authored
Enable Diffie-Hellman cipher suites in NativeCrypto (and in StandardNames to match for testing). This means we now have the same default cipher suite list as RI 5. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/NativeCrypto.java support/src/test/java/javax/net/ssl/StandardNames.java Enabling DH made it obvious that the RI check for enable cipher suites on SSLServerSocket.accept was not as stringent as first thought. Apparently they don't care if all enabled cipher suites have certificates/keys, just that at least one of them will work, even if its anonymous. Factored out the logic to check this into checkEnabledCipherSuites for clarity along with the supporting checkForPrivateKey. Also only check if the socket is in server mode, since its fine to have nothing configured for server acting as a client for handshake purposes. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLServerSocketImpl.java The real work to enable Diffie-Hellman was to use SSL_CTX_set_tmp_dh_callback to set a callback to get DH parameters. There are two ways to create the parameters. The first is to use DH_generate_parameters_ex which is very slow (minutes) as is recommended as install time option. The second is to use DSA_generate_parameters_ex followed by DSA_dup_DH, which is faster for a single call, but must be done every time, so slower overall. We currently take the second approach to just have DH working. luni/src/main/native/org_apache_harmony_xnet_provider_jsse_NativeCrypto.cpp Changed ephemeral RSA keys to be stored per SSL in AppData, not in a static global. luni/src/main/native/org_apache_harmony_xnet_provider_jsse_NativeCrypto.cpp Fix LS_ to TLS_ typo in commented out constant. Removed easy to miss wrapping in array definition. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/CipherSuite.java Renamed CipherSuites defaultPretendant to defaultCipherSuites which led to renaming the CipherSuites constants to follow the coding style. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/CipherSuite.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/ClientHandshakeImpl.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/DigitalSignature.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLServerSocketImpl.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLParameters.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/ServerHandshakeImpl.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/ServerKeyExchange.java Change-Id: Ia38de48cabb699b24fe6e341ba79f34e3da8b543
-
- 19 May, 2010 1 commit
-
-
Brian Carlstrom authored
Make CipherSuite static fields final (noticed because I tried to use some in a switch statement). Also renamed "cuites*" to "suites*" and fixed UNKNOUN to UNKNOWN luni/src/main/java/org/apache/harmony/xnet/provider/jsse/CipherSuite.java SSLServerSocket now matches the RI behavior of throwing an SSLException for missing keys for non-anonymous cipher suites. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLServerSocketImpl.java Fixed one KnownFailure luni/src/test/java/javax/net/ssl/SSLSocketTest.java Change-Id: I1ccbf93cfc5aa5951b1f33881446d93c380b6e68
-
- 18 May, 2010 2 commits
-
-
Brian Carlstrom authored
SSLSession.getPeerPort is supposed to return -1 when the port is undefined so now we initialize it to that value. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSessionImpl.java Avoid creating InetAddress to store the OpenSSLSessionImplWrapper host and port arguments since it was causing an exception on an port value of -1 and was just used to go back to the original host and port when creating the SSLSession, which is allowed to return a port value of -1. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java Remove last of KnownFailures for SSLSocketFactory luni/src/test/java/javax/net/ssl/SSLSocketFactoryTest.java Update classpath for newly seperated out junit jars run-core-tests Change-Id: I646a8f23c3d6ae01f1dd38e40bc9c32d436e6254
-
Brian Carlstrom authored
Client certificates should only be set into the SSL* when requested by the server so that after the handshake is completed the client can inspect its SSLSession to see what certificate if any was requested. Previously the value was always non-null even if the server didn't request the certificate. - Created RAND_seed and RAND_load_file out of the NativeCrypto.SSL_new - NativeCrypto.SSL_new now simply performs SSL_new and does not deal with certificates, private keys, or random seeds. - Removed helper version of NativeCrypto.SSL_new Moved code to OpenSSLSocketImpl.setCertificate - Created SSL_use_certificate, SSL_use_PrivateKey, SSL_check_private_key from SSL_new. These are used not just on server handshake but also via clientCertificateRequested callback. - Merged CertificateChainVerifier and HandshakeCompletedCallback into new SSLHandshakeCallbacks while adding new clientCertificateRequested callback from OpenSSL C code to Java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/NativeCrypto.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java luni/src/main/native/org_apache_harmony_xnet_provider_jsse_NativeCrypto.cpp In addition to supporting NativeCrypto.java changes, also changed to_SSL_CTX and to_SSL_SESSION to allow null checking and throwing NullPointerException. Changed these and to_SSL to log exception on JNITrace, taking these logs out of individual functions. There were a lot of null checks missing previously, mostly in to_SSL_SESSION cases. luni/src/main/native/org_apache_harmony_xnet_provider_jsse_NativeCrypto.cpp All KnownFailures now fixed. luni/src/test/java/javax/net/ssl/SSLSessionTest.java Three more KnownFailures now fixed. luni/src/test/java/javax/net/ssl/SSLSocketTest.java Change-Id: Iddcd5512e8395d947d3b894f03e3a059e63afe8a
-
- 17 May, 2010 2 commits
-
-
Brian Carlstrom authored
Change text names of Harmony CipherSuite's (used by SSLEngine and some places with OpenSSL code) to match JSSE names. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/CipherSuite.java Added StandardName constant for SSL_NULL_WITH_NULL_NULL support/src/test/java/javax/net/ssl/StandardNames.java Marked test as working with above fix, changed to use newly defined constant. luni/src/test/java/javax/net/ssl/SSLSessionTest.java Change-Id: Id48d2adcbbff71306296f1fdf8ff970c618fdcc6
-
Brian Carlstrom authored
Added new test_SSLSocket_getSupportedCipherSuites_connect to make sure all cipher suites we claim work actually do. It clearly exposed that although a large number of cipher suites are supported by libssl.so, they are not properly wired up into the OpenSSL JSSE implementation. In particular Elliptic Curve has been disabled in our version Bouncy Castle does not work. In addition Diffie-Hellman does not work because we need to further integration work with OpenSSL via SSL_set_tmp_dh_callback or SSL_set_tmp_dh. Finally, SSL_RSA_EXPORT_WITH_RC4_40_MD5 doesn't work but that is being left as KnownFailure for more immediate cleanup based on ServerHandshakeImpl's handling of KeyExchange_RSA_EXPORT as part of having OpenSSL call us back for certificates dynamically. luni/src/test/java/javax/net/ssl/SSLSocketTest.java Refactored TestSSLContext.createKeyStore to create TestKeyStore which now factors out TestSSLContext.createKeys from the old createKeyStore method, which allows createKeys to be called multiple times for different key algorithms (for example DSA in addition to RSA). Also added a reusable singleton instance to cut down on test execution time. support/src/test/java/javax/net/ssl/TestKeyStore.java Removed publicAlias/privateAlias from TestSSLContext since we now include both RSA and DSA key pairs in they KeyStore by default. Added TestSSLContext.assertCertificateInKeyStore methods to help tests the previously used the alias fields fields. TestSSLContext.create API changed as well since the alias names are no longer required. TestSSLContext.createClient now needs to iterate over all server certificates when setting up its TrustManager instead of just grabbing one by alias name. support/src/test/java/javax/net/ssl/TestSSLContext.java luni/src/test/java/javax/net/ssl/SSLContextTest.java luni/src/test/java/javax/net/ssl/SSLSessionTest.java luni/src/test/java/javax/net/ssl/SSLSocketTest.java TestSSLSocketPair.connect now allows optional inclusion of server cipher suite list. support/src/test/java/javax/net/ssl/TestSSLSocketPair.java luni/src/test/java/javax/net/ssl/SSLSessionContextTest.java Turning off Elliptic Curve and Diffie-Hellman which are not currently working. Updating test expectations to match. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/NativeCrypto.java support/src/test/java/javax/net/ssl/StandardNames.java Turn on registration of ECDSA and DSA since this part is currently functional (and excercised by TestKeyStore.create()) luni/src/main/java/org/bouncycastle/x509/X509Util.java Improve logging by including SSL pointer in error messages, which makes it easier to relate these errors to JNI_TRACE messages. luni/src/main/native/org_apache_harmony_xnet_provider_jsse_NativeCrypto.cpp Change-Id: I014d001a6a21a46c360678a346d3a3c8232f4d53
-
- 14 May, 2010 3 commits
-
-
Elliott Hughes authored
-
Elliott Hughes authored
Responsibility is split between Socket, SocketImpl, PlainSocketImpl, SocketChannel, SocketChannelImpl, and SocketChannelImpl.SocketAdapter, and we need to keep them synchronized. Our hands are somewhat tied by the fact that the RI exposed way too much. I think, now I understand the relationships a bit better, that we can probably rewrite this cluster of classes to be simpler, but I don't want to bite off more than I can chew right now, and this does fix the known problems. This patch also makes us more compatible with the RI by making getLocalAddress after the socket has been closed return the address we used. By strange coincidence, harmony addressed this at the same time I was looking at it (see http://svn.apache.org/viewvc?rev=944119&view=rev) but I feel they're going in the wrong direction and making the relationships even more complicated. I have run their new tests in addition to my own, though. Bug: 1952042 Bug: http://code.google.com/p/android/issues/detail?id=1933 Bug: http://code.google.com/p/android/issues/detail?id=3123 Change-Id: Icb7793fb5d868e0d1f1b8b3d5da88c32fb973744
-
Brian Carlstrom authored
Summary: - Switch to using JSSE cipher suite names - SSLSessionContext implementation cleanup - Updated tests Details: Switch to using JSSE cipher suite names - We maintain backward compatability for enabling cipher suites using OpenSSL names for old code that did so without checking for the presence of the names in the supported list. - We now have a well defined list of the supported cipher suites which are sorted in priority order as specified in JSSE documentation so that callers doing: s.setEnabledCipherSuites(s.getSupportedCipherSuites()) will get something reasonable. - We now have a default cipher suite list that is chose to match RI behavior and priority, not based on OpenSSLs default and priorities. Details: - Added NativeCrypto OPENSSL_TO_STANDARD and STANDARD_TO_OPENSSL mapping between naming conventions. STANDARD_TO_OPENSSL is a LinkedHashMap so enumerating it gives the proper order for SUPPORTED_CIPHER_SUITES. - SSL_get_ciphers and SSL_set_cipher_list are removed, we now use our own SSL_set_cipher_lists (defined seperately in external/openssl/patches/jsse.patch) to set the set and order of cipher suites. SSL_CTX_get_ciphers is also removed because we no longer rely on the OpenSSL for the default cipher suites behavior. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/NativeCrypto.java luni/src/main/native/org_apache_harmony_xnet_provider_jsse_NativeCrypto.cpp Add cipherSuite and protocol field caches for native values, mapping the cipherSuite to a JSSE name from the OpenSSL name returned by SSL_SESSION_cipher. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSessionImpl.java Fixed a long standing bug where we reused sessions found in the client host/port cache even if the old protocol and cipher suite where no longer compatible with what was specified by setEnabledCipherSuites and setProtocols. Also fixed a recently introduced bug where lastAccessedTime was being set on a cached session even if it was not reused, found by fixed the above. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java Move most of SSLSessionContext implementation from subclasses to AbstractSessionContext. This was primarily to align the implementations of how different sessions id for the same host and port were handled for RI compatability. client subclasses now focuses on handling its host/port based cache and both deal with their own persistent cache details. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/AbstractSessionContext.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/ClientSessionContext.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/ServerSessionContext.java Tests Added some variants of assertSSLSessionContextSize to simplify tests code. Broke test_SSLSessionContext_setSessionCacheSize_oneConnect out of test_SSLSessionContext_setSessionCacheSize_dynamic. Renamed test_SSLSessionContext_setSessionCacheSize_basic to test_SSLSessionContext_setSessionCacheSize_noConnect to match name of _oneConnect. _dynamic was cleaned up a bit as getting it working was the only goal of this change list. Fixed to filter SSL_RSA_EXPORT_ ciphers since our test certificate key length is too long for those. Lower test requirement to 3 unique cipher suites. luni/src/test/java/javax/net/ssl/SSLSessionContextTest.java Added checks that cipher suites and protocols have standard names. luni/src/test/java/javax/net/ssl/SSLSessionTest.java Removing known failures related to cipher suite naming. Fixed bug of using assertNotNull instead of assertTrue. Added extra size/length check which would have found the assertNotNull/assertTrue issue. luni/src/test/java/javax/net/ssl/SSLSocketFactoryTest.java luni/src/test/java/javax/net/ssl/SSLSocketTest.java Fixing test the explicitly worked around broken cipher suite naming. luni/src/test/java/tests/api/javax/net/ssl/SSLSessionTest.java Updated standard cipher suites to RI 6 list, which also now specifies ordering, which we now align with. support/src/test/java/javax/net/ssl/StandardNames.java Unrelated Remove more now obsolete jars from the test classpath run-core-tests Change-Id: I45c274a9327c9a1aeeccb39ecaf5a3fbe2903c8f
-
- 13 May, 2010 1 commit
-
-
Elliott Hughes authored
Gentlemen, you may now set your editors to "strip trailing whitespace"... Change-Id: I85b2f6c80e5fbef1af6cab11789790b078c11b1b
-
- 11 May, 2010 2 commits
-
-
Brian Carlstrom authored
Fix getIds Enumeration to filter invalid sessions. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/AbstractSessionContext.java Implement SSLSessionContext.setSessionTimeout to remove newly invalid sessions as specified by the RI documentation. getSession interfaces now filters invalid sessions from results. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/ClientSessionContext.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/ServerSessionContext.java Added OpenSSLSocketImpl.creationTime instance field cache to avoid repeated native calls since this is now used for all isValid tests. Fixed broken isValid implementation: - compared seconds to milliseconds - direction of comparison backwards - used last accessed time instead creation time as clarified in RI 7 documentation. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSessionImpl.java Unrelated Replace java.io.* java.util.* imports with properly expanded versions: luni/src/main/java/org/apache/harmony/xnet/provider/jsse/AbstractSessionContext.java Change-Id: Ib02218df414f014f1d260f7acc067e5647fb700b
-
Brian Carlstrom authored
Add an explicit null check to ensure failure on a null argument to getSession to match the RI luni/src/main/java/org/apache/harmony/xnet/provider/jsse/ClientSessionContext.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/ServerSessionContext.java Remove KnownFailures resolved by above fix as well as clarifiying SSL session cache expections on Android vs the RI. The KnownFailures were also hiding some latent issues to do SSL session tickets, so fixed those up as well. luni/src/test/java/javax/net/ssl/SSLSessionContextTest.java Added constants for expected SSL session cache behavior for RI vs Android support/src/test/java/javax/net/ssl/TestSSLContext.java Change-Id: Ic6285192cf76c0a5c3fa45a24eaa504ed0babff5
-
- 10 May, 2010 1 commit
-
-
Brian Carlstrom authored
Moved initialization of SSLContextImpl clientSessionContext and serverSessionContext from engineInit time (in SSLParameters constructor) to constructor time, making them final. This is to fix javax.net.ssl.SSLContextTest which was failing because it tried to access this before init was called, which worked fine on the RI. The SSLParameters now simply takes the preallocated session contexts as arguments. SSLParameters.getDefault() now needs to create its own session contexts when an SSLContext is not used, which is how Harmony does it. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLContextImpl.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLParameters.java Removed KnownFailure from SSLContextTest as its 100% working. luni/src/test/java/javax/net/ssl/SSLContextTest.java Changed persistentCache fields of ClientSessionContext and ServerSessionContext from final to private and added a public setter. This replaces passing the persistentCache implementation in via the constructor. For momentarily backward compatibility with frameworks/base, the now deprecated 5 argument engineInit method now uses these setters for backward compatability. The SSLParameters previously took these persistent caches as arguments in order to pass them to the session context contructors, but as SSLParameters no longer creates these, they are no longer relevant. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/ClientSessionContext.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/ServerSessionContext.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLContextImpl.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLParameters.java While moving the call of the AbstractSessionContext constructor from SSLParameters to SSLContextImpl after removing the persistent cache arguments, I realized there was no longer any reason to take any arguments. I pushed the initization of sslCtxNativePointer to the point of declaration. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/AbstractSessionContext.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/ClientSessionContext.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/ServerSessionContext.java Change-Id: Ied2903a2f369bf4e521e702bf58f32f21cb97d17
-
- 07 May, 2010 1 commit
-
-
Brian Carlstrom authored
Change-Id: I7925a11870231b85abbd8f754f5edd34835605a9
-
- 06 May, 2010 1 commit
-
-
Brian Carlstrom authored
Summary: - Finished consolidating OpenSSL native code into NativeCrypto - fixing local vs global ref bug with AppData Added new ScopedGlobalRef as part of this fix - fixed many historical memory leaks identified during code review - fixed lack of error checking on allcoation with OpenSSL *_new routines - Added to_SSL_CTX and to_SSL_SESSION to match to_SSL (renamed from getSslPointer) - Replaced most uses of GetByteArrayElements with ScopedByteArray (including cases where we we using ReleaseByteArrayElements(..,...,0) instead of JNI_ABORT) - Replaced uses of GetStringUTFChars with ScopedUtfChars Details: Finished consolidating OpenSSL native code into NativeCrypto OpenSSLSocketImpl NativeCrypto --------------------------------------- nativeread SSL_read_byte nativeread SSL_read nativewrite SSL_write_byte nativewrite SSL_write nativeinterrupt SSL_interrupt nativeclose SSL_shutdown nativeverifysignature verifysignature Also removed dead code that was wrapping SSL_get1_session luni/src/main/java/org/apache/harmony/xnet/provider/jsse/NativeCrypto.java luni/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java luni/src/main/native/org_apache_harmony_xnet_provider_jsse_NativeCrypto.cpp Fixed NativeCrypto_SSL_write and NativeCrypto_d2i_SSL_SESSION to use JNI_ABORT on release to avoid copy back of unchanged data (via ScopedByteArray). luni/src/main/native/org_apache_harmony_xnet_provider_jsse_NativeCrypto.cpp While running the usual tests: adb shell run-core-tests tests.xnet.AllTests javax.net.ssl.AllTests there was an abort from the JNI checking because in the recent handshaking change, local refs were kept in AppData and then reused in later calls. Added new ScopedGlobalRef to handle the book keeping of this. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java include/ScopedGlobalRef.h Fixed various leaks on old error paths spotted by reviewer. luni/src/main/native/org_apache_harmony_xnet_provider_jsse_NativeCrypto.cpp Tracking move of verifySignature, a non-SSL bit of code that was lurking in OpenSSLSocketImpl luni/src/main/java/org/apache/harmony/security/provider/cert/X509CertImpl.java Change-Id: If1e409782bc99dc684039cfe3f53f8244e29346e
-
- 05 May, 2010 1 commit
-
-
Brian Carlstrom authored
OpenSSLSessionImpl NativeCrypto ------------------------------------------------------- getId SSL_SESSION_session_id getPeerCertificatesImpl SSL_SESSION_get_peer_cert_chain getCreationTime SSL_SESSION_get_time getProtocol SSL_SESSION_get_version getCipherSuite SSL_SESSION_cipher freeImpl SSL_SESSION_free getEncoded i2d_SSL_SESSION initializeNativeImpl d2i_SSL_SESSION Change-Id: I4538df52280266711986a577b14868af3ea0ed62
-
- 04 May, 2010 3 commits
-
-
Brian Carlstrom authored
Following up on feedback from earlier change https://android-git.corp.google.com/g/50435 Added new test_SSLSocket_startHandshake_noClientCertificate to make sure handshaking works when no client certificates are present after issues raised by hwu during code review. luni/src/test/java/javax/net/ssl/SSLSocketTest.java Improve TestSSLContext.create* options - added javadoc comments to help distinguish different versions - fixed bug of not passing in keyStorePassword in create() - added new createClient(server) method to create a TestSSLContext that trusts the provided server TestSSLContext's certificate for use by test_SSLSocket_startHandshake_noClientCertificate - made createKeyStore optionally create a more minimal keystore if aliases are not present support/src/test/java/javax/net/ssl/TestSSLContext.java Fixed argument names in SSL_*_mode methods names as pointed out by hwu luni/src/main/java/org/apache/harmony/xnet/provider/jsse/NativeCrypto.java Added comment to explain purpose of OpenSSLSessionImpl.resetId. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSessionImpl.java Two changes to OpenSocketImpl - Added logging on runtime exception catch around HandshakeCompletedListener execution to closely mirror RI behavior. - Cleaned up peerCertificate check to not just be on the client path. luni/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java Addressed enh's comments about using clearEnv and when to delete AppData luni/src/main/native/org_apache_harmony_xnet_provider_jsse_NativeCrypto.cpp Change-Id: I34f54e3e41a5d53d81fdc22aa34ca4de4ee9826f
-
Peter Hallam authored
Merge xml except xmlpull and kxml into luni
-
Brian Carlstrom authored
Summary: - SSLSocket.startHandshake now generalized to handle both client and server handshaking as well as client/server role reversal - handshake_cutthrough.patch is properly integrated with support delayed handshake completion now integrated with delayed updates to session cache and callbacks to HandshakeCompletedListeners - Many fixes to SSLSession, which is the end product of the handshake - Generally more RI and SSLEngine compliant behavior. - More native code deletion through unification of client/server handshake, unification of client/server certificate chain verification, etc. More native code moved from various OpenSSL classes to cleaner NativeCrypto interfaces that more directly mirror the OpenSSL interfaces. Details: Delay SSL_new call until handshake time when we know for sure whether the OpenSSLSocket will be used in client or server mode and we can allocate the SSL_new from the apppriate client or server SSL_CTX used for session caching. Now that no SSL is allocated for an OpenSSLServerSocketImpl, store enabledProtocols and enabledCipherSuites in instance String arrays. Use new NativeCrypto.checkEnabled* methdods for argument validation. OpenSSLServerSocketImpl passes these enabled arrays to a new OpenSSLSocket constructor during accept(). Removed finalizer from OpenSSLServerSocketImpl since it no longer has any native storage and socket is already closed by PlainSocketImpl finalizer. X-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLServerSocketImpl.java x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java OpenSSLSocket major overhaul to properly implement handshaking including switching client and server roles and session ID caching with handshake_cutthrough.patch. - now implements NativeCrypto.HandshakeCompletedListeners for properly timed callback when handshake_cutthrough.patch delays handshake completion until first SSLSocket.getInputStream() read. - similar enabledProtocols/enabledCipherSuites changes as OpenSSLServerSocketImpl since we need to store the state somewhere other than an openssl SSL struct until we are sure if we are doing a client or server handshake. - added handshake completed field so that startHandshake can tell if handshake was completed during SSL_do_handshake or will be completed later by a call to HandshakeCompletedCallback.handshakeCompleted. - removed nativegetsession as the equivalent value is now returned by SSL_do_handshake - removed nativecipherauthenticationmethod as the value is now passed to verifyCertificateChain - startHandshake is now a wrapper that forces a fully synchronous handshake - startHandshake(boolean) is the the most changed method in this changelist, combinding both the old startHandshake logic, but also the OpenSSLSocketImpl.accept code as well. Notable differences from the old code: * now responsible for SSL_new * single code path for client/server handshaking dealing with SSLSession caching * now handles server certificate requests previously in OpenSSLServerSocketImpl, since a client can request to act like a server and therefore need to be able to make suck demands on its peer. * supports turning off handshake_cutthrough at a callers request via explicit call to startHandshake() * certificate verification happens during an upcall from openssl during SSL_do_handshake to verifyCertificateChain for both client and server cases. previously there was not quite right upcall support on the server side and post-handshake checking on the client, which did not allow for a proper alert to be sent to the peer informing them of the issue, which the RI and SSLEngine code do. * Similarly, setEnableSessionCreation(false) did not send an alert to the peer as the RI and SSLEngine code in the client case. In the server case, nothing was previously done. * The use of local certificates was not determined from introspecting the SSL struct post-handshake. This is now partially implemented and will be completed in a later change. - SSLSocket.{shutdownInput,shutdownOutput} are now restored to the proper behavior of throwing UnsupportedOperationException. - Gutted OpenSSLSocketImpl finalizer. The comment explains in detail the trouble of having the finalizer do anything more than touch its the instances own state due to unpredictable order of finalization and the future possability of parallel finalization. x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java SSLSession fixes - Made OpenSSLSessionImpl.sessionContext non-final so it could be nulled by SSLSession.invalidate to match RI behavior. - As noted in AbstractSessionContext discussion, removed OpenSSLSessionImpl constructor that took SSLParameters, instead we take the possibly null localCertificates directly. OpenSSLSessionImpl.getLocalCertificates now simply returns the localCertificates member variable instead of incorrectly trying to query the KeyManager for certificates that may not have been used. - OpenSSLSessionImpl now caches its native ID to avoid numerious native calls but also now provides as resetId which will update the cache when a delayed handshake happens due to the handshake_cutthrough.patch - Fixed bug in getPeerPrincipal that it wasn't calling getPeerCertificates to initialize peerCertificates field. - freeImpl is now 'public static' in preparation for move to NativeCrypto. x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSessionImpl.java The old SSLSessionImpl class that is still used for representing the invalid session now returns isValid => false and getProtocol => "NONE" to match the RI. x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLSessionImpl.java NativeCrypto improvements - Adding NativeCrypto.SSL_{get,set,clear}_mode similar to NativeCrypto.SSL_{get,set,clear}_options along with SSL_MODE_HANDSHAKE_CUTTHROUGH constant which is used to explicitly disable/enable the Android handshake_cutthrough.patch behavior. - Added missing NativeCrypto.SSL_clear_options and used to properly implement NativeCrypto.setEnabledProtocols. - Added NativeCrypto.checkEnabledProtocols and NativeCrypto.checkEnabledCipherSuites helpers to implement exception compatability with the RI. While some of this code is refactored from existing NativeCrypto code, it is now also used by OpenSSLServerSocketImpl and OpenSSLSocketImpl which maintain their own String[]s of what is enabled until startHandshake time. (see below) - Changed NativeCrypto.findSuite to use foreach style loop for clarity. - Moved OpenSSLServerSocketImpl nativesetclientauth and SSL_VERIFY_* constants to NativeCrypto.SSL_set_verify - Added NativeCrypto.SSL_set_session based on part of old OpenSSLSocketImpl.nativeconnect - Added NativeCrypto.SSL_set_session_creation_enabled to properly implement SSLSocket.setEnableSessionCreation(false) which uses new external/openssl/patches/jsse.patch functionality. - New NativeCrypto.SSL_do_handshake consolidates OpenSSLSocketImpl.{nativeconnect, nativeaccept} while properly implementing SSLSocket.setUseClientMode(false) for clients and SSLSocket.setUseClientMode(true) for servers. - New NativeCrypto.SSL_get_certificate is determine if local certificate requested by peer. While functional, currently NativeCrypto.SSL_new always sets a value via SSL_use_certificate instead of relying on a callback set via SSL_CTX_set_client_cert_cb. - Changed NativeCrypto.CertificateChainVerifier.verifyCertificateChain to throw a checked CertificateException to match TrustManager.{checkServerTrusted, checkClientTrusted}. It also takes an authMethod so avoid the need to call the old OpenSSLSocketImpl.nativecipherauthenticationmethod. - Added NativeCrypto.HandshakeCompletedCallback which has its handshakeCompleted method called from OpenSSL when the now delayed handshake_cutthrough.patch handshake is completed so SSLSession caching can be delayed until a session ID is available and to provide a better time for HandshakeCompletedListeners to be notified. x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/NativeCrypto.java x-net/src/main/native/org_apache_harmony_xnet_provider_jsse_NativeCrypto.cpp Some other changes specific to the naitve side of the code - Added JNITRACE calls (enabled at compile time with JNI_TRACE) for future debugging. - throw SSLException subclass of IOException instead IOException itself for better RI compatability x-net/src/main/native/org_apache_harmony_xnet_provider_jsse_NativeCrypto.cpp - changed from old struct app_data to new class AppData at enh's request Remove dubious usage of SSLParameters within AbstractSessionContext to pass through to OpenSSLSessionImpl constructor for use in calling getLocalCertificates for sessions created from a byte array with AbstractSessionContext.toSession. Our AbstractSessionContext.toBytes doesn't currently include the local certificates in its output, so it cannot be expected to have in toSession. x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/AbstractSessionContext.java x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/ClientSessionContext.java x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/ServerSessionContext.java x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLParameters.java Test maintenance openssl 1.0.0 adds support for RFC 4507 session tickets which remove the need for server side session state. These tests needed to be updated for this new behavior. If IS_RI is true, they still follow the old behavior. luni/src/test/java/javax/net/ssl/SSLSessionContextTest.java luni/src/test/java/javax/net/ssl/SSLSessionTest.java luni/src/test/java/javax/net/ssl/SSLSocketTest.java Update KnownFailures and add specific comments at point of failure about what remains to be fixed. luni/src/test/java/javax/net/ssl/SSLSessionTest.java Added tests to cover the use of standard cipher suite names. Historically Android has used OpenSSL string constants for cipher suite names, but JSSE actually specifies supported and expected names. luni/src/test/java/javax/net/ssl/SSLSocketFactoryTest.java luni/src/test/java/javax/net/ssl/SSLSocketTest.java Create new support/src/test/java/javax/net/ssl with old Helper support code pulled from javax.net.ssl tests: SSLContextTest.Helper -> TestSSLContext SSLSocketTest.Helper -> TestSSLSocketPair SSLSessionTest.Helper -> TestSSLSessions Also added new StandardNames here, which contains a collection of expected constants for test validation. luni/src/test/java/javax/net/ssl/SSLContextTest.java luni/src/test/java/javax/net/ssl/SSLSocketTest.java luni/src/test/java/javax/net/ssl/SSLSessionTest.java support/src/test/java/javax/net/ssl/TestSSLContext.java support/src/test/java/javax/net/ssl/TestSSLSocketPair.java support/src/test/java/javax/net/ssl/TestSSLSessions.java support/src/test/java/javax/net/ssl/StandardNames.java Removed some now fixed KnownFailures and unneeded !IS_RI code. Marked some [Un]KnownFailures where exceptions are thrown and visible in the output but aren't correctly causing the test to fail. Fixed assertNonNull to assertTrue in test_SSLSocketTest_Test_create. Added stress_test_SSLSocketTest_Test_create to track down test flakiness, leading to rewrite of SSLSocket finalization. luni/src/test/java/javax/net/ssl/SSLSocketTest.java Reenable javax.net.ssl.AllTests now that it is does not hang luni/src/test/java/tests/AllTests.java Improve error messages while debugging overflow problem. Added new assert when debugging new RFC 4507 behavior. Removed KnownFailure annotation for now working test case. x-net/src/test/java/tests/api/javax/net/ssl/SSLSessionTest.java Client code changes Now that startHandshake implies synchronous vs Android's default async handshake, remove unneeded explict calls to SSLSocket.startHandshake luni/src/main/java/org/apache/harmony/luni/internal/net/www/protocol/http/HttpConnection.java Removed IBM 1.4.x codepath that involved startHandshake x-net/src/main/java/javax/net/ssl/DefaultHostnameVerifier.java Unrelated Remove unneed SSLSocket.setUseClientMode while removing unneeded SSLSocket.startHandshake luni/src/main/java/org/apache/harmony/luni/internal/net/www/protocol/http/HttpConnection.java Removed warnings due to now missing modules in classpath run-core-tests Change-Id: I6e149ae259b3feccdfb0673209c85cfeb60befc8
-
- 27 Apr, 2010 1 commit
-
-
Peter Hallam authored
-
- 06 Apr, 2010 2 commits
-
-
Brian Carlstrom authored
Summary: b/1758225: Revisit OpenSSL locking Removed the locking original put in to address b/1678800 which had been causing problems for the HeapWorker thread which was timing out waiting for the lock in the finalizers while other threads were connecting. b/1678800: Reliability tool: Crash in libcrypto @ https://opac.ntu.ac.uk Properly fixed the native crash by avoid sharing SSL_SESSION objects between SSL_CTX objects Testing: - adb shell run-core-tests --verbose tests.xnet.AllTests - adb shell run-core-tests --verbose javax.net.ssl.AllTests - Test app that reloads https://opac.ntu.ac.uk Details: Each AbstractSessionContext now has an associated SSL_CTX, referenced through the sslCtxNativePointer. SSL_CTX on the native side defines the scope of SSL_SESSION caching, and this brings the Java SSLSessionContext caching into alignment with the native code. OpenSSLSessionImpl now uses AbstractSessionContext instead of SSLSessionContext for access to the underlying SSL_CTX. x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/AbstractSessionContext.java x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/ClientSessionContext.java x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSessionImpl.java x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLParameters.java x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/ServerSessionContext.java Added AbstractSessionContext.putSession so OpenSSLSocketImpl/OpenSSLSessionImpl can directly assign to the current AbstractSessionContext (whether it be a ClientSessionContext or a ServerSessionContext) without casting. x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/AbstractSessionContext.java x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/ClientSessionContext.java x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/ServerSessionContext.java Cleaning up use of SSL_CTX and SSL instances in SSLSocket/SSLServerSocket implementation The major change is that openssl SSL instances are allocated for the life of the matching Java object, replacing the SSL_CTX and the SSL objects that had previously been allocated only starting at handshake time. We should never have been sharing SSL_SESSION instances between SSL_CTX instances, which was the source of the native crashes dating back to cupcake which the OpenSSLSocket.class locking had been preventing. - NativeCrypto now has better defined and independant wrappers on openssl functionality. A followon checkin should move the remaining openssl JNI code here with the intent of being able to write and end-to-end test of the openssl code using NativeCrypto without the JSSE implementation classes. The following gives a list of the new native functions with a mapping to the old implementation code. The new code has a more functional style where SSL_CTX and SSL instances are passed and returned as arguments, not extracted from Java instances SSL_CTX_new OpenSSLSocketImpl.nativeinit, OpenSSLServerSocketImpl.nativeinit, SSLParameters.nativeinitsslctx SSL_CTX_get_ciphers_list OpenSSLSocketImpl.nativeGetEnabledCipherSuites SSL_CTX_free OpenSSLSocketImpl.nativefree, OpenSSLServerSocketImpl.nativefree SSL_new OpenSSLSocketImpl.nativeinit, OpenSSLSocketImpl.init, OpenSSLServerSocketImpl.nativeinit, OpenSSLServerSocketImpl.init SSL_get_options OpenSSLSocketImpl.nativesetenabledprotocols SSL_set_options OpenSSLSocketImpl.nativesetenabledprotocols SSL_get_ciphers OpenSSLSocketImpl.nativeGetEnabledCipherSuites SSL_set_cipher_list OpenSSLSocketImpl.nativeSetEnabledCipherSuites SSL_free OpenSSLSocketImpl.nativefree, OpenSSLServerSocketImpl.nativefree - While the focus in NativeCrypto is on native code, it also contains some helpers/wrappers especially for code that doesn't depend on specific SSL_CTX, SSL instances or that needs to do massaging of data formats between Java and OpenSSL. Some of these had previously been duplicated in the client and server versions of the code. For example: getSupportedCipherSuites OpenSSLSocketImpl.nativegetsupportedciphersuites, OpenSSLServerSocketImpl.nativegetsupportedciphersuites getSupportedProtocols OpenSSLSocketImpl.getSupportedProtocols, OpenSSLServerSocketImpl.getSupportedProtocols getEnabledProtocols OpenSSLSocketImpl.getEnabledProtocols,OpenSSLServerSocketImpl.getEnabledProtocols setEnabledProtocols OpenSSLSocketImpl.setEnabledProtocols setEnabledCipherSuites OpenSSLSocketImpl.setEnabledCipherSuites - Moved JNI initialization from OpenSSLSocketImpl to NativeCrypto which is the future home of all the openssl related native code. clinit OpenSSLSocketImpl.nativeinitstatic - NativeCrypto.CertificateChainVerifier is a new interface to decouple callbacks from openssl from a specific dependence on a OpenSSLSocketImpl.verify_callback method. Changed to return boolean instead of int. - Renamed OpenSSLSocketImpl.ssl to OpenSSLSocketImpl.sslNativePointer for consistency - Changed OpenSSLSocketImpl nativeconnect, nativegetsslsession, nativecipherauthenticationmethod, nativeaccept, nativeread, nativewrite, nativeinterrupt, nativeclose, nativefree to take arguments instead of inspect object state in preparation for moving to NativeCrypto - other notable NativeCrypto changes included * adding SSL_SESSION_get_peer_cert_chain, SSL_SESSION_get_version, and SSL_get_version (and get_ssl_version) which are "missing methods" in openssl * ssl_msg_callback_LOG callback and get_content_type for handshake debugging * removing jfieldID's for our classes now that we pass in values in arguments * changed aliveAndKicking to be volative since we poll on it to communicate between threads * changed from C style declarations at beginning of block to C++ at first use on methods with major changes * stop freeing SSL instances on error, only SSL_clear it * improved session reuse logging when reproducing b/1678800 * change verify_callback to return verifyCertificateChain result x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/NativeCrypto.java x-net/src/main/native/org_apache_harmony_xnet_provider_jsse_NativeCrypto.cpp x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLServerSocketImpl.java x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketFactoryImpl.java When we accept a server socket, we pass the existing SSL state instance from the server socket to the newly accepted socket via the constructor where it is copied with SSL_dup, instead of through both the constructor and later the accept method. x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLServerSocketImpl.java Cleaned up nativesetclientauth from using SSL_CTX to SSL, passing ssl as argument in preparation for future movement to NativeCrypto. x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLServerSocketImpl.java Removed ssl_op_no cache for rarely used enabled protocol methods so that code could more easily be shared in NativeCrypto between client and server. x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLServerSocketImpl.java x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java Changed public getId, getCreationTime, getPeerCertificates, getCipherSuite, getProtocol from being instance methods that looked at the OpenSSLSessionImpl object state to be static mthods that take the native pointers as arguments in preparation for moving to NativeCrypto. Rename session -> sslSessionNativePointer for consistency. Inlined initializeNative, which wasn't really the native code. x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSessionImpl.java Removed lock on OpenSSLSocketImpl.class lock from around OpenSSLSocketImpl's use of nativeconnect, nativegetsslsession, and nativecipherauthenticationmethod as well as OpenSSLSessionImpl's use of freeImpl, fixing b/1758225: Revisit OpenSSL locking x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSessionImpl.java Unrelated changes Removed unused ssl_ctx, nativeinitsslctx, getSSLCTX x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/SSLParameters.java Fix bug in both putSession implementations where we cached sessions with zero length id. Also change indexById to pass in id in client implementation. x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/ClientSessionContext.java x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/ServerSessionContext.java Make sure we clone SSLParameters passed to the SSLSocketFactory and SSLServerSocketFactory so that muting the client instance does not change the server instance and vice versa. Explicitly set setUseClientMode(false) on the server SSLParameters. These changes are to bring things more into alignment with the original harmony classes which properly support client/server role switching during handshaking. x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLServerSocketFactoryImpl.java Make locks object fields final x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java Moved updateInstanceCount(1) logic and sslParameters assignment to init method x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java Changed getCachedClientSession to respect getUseClientMode x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java Spelling of listensers to listeners in javadoc x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java Spelling SSLInputStream to SSLOutputStream in comment x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java Changed shutdownInput and shutdownOutput to call to the underlying socket x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java Set sslNativePointer to 0 when freeing underlying SSL object x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java Removed IOException logging in getSession, which is expected to simply return SSL_NULL_WITH_NULL_NULL when there are problems. x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/OpenSSLSocketImpl.java Disabled "Using factory" message on successful creation of SocketFactory which was a bit noisy running tests. However, added logging in failure case including the related exception: x-net/src/main/java/javax/net/ssl/SSLSocketFactory.java Disabled logging of OpenSSL session deallocation x-net/src/main/native/org_apache_harmony_xnet_provider_jsse_NativeCrypto.cpp Register SSLContextImpl as a source of SSL and SSL3 SSLContexts, not just TLS and TLSv1. x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/JSSEProvider.java Fix whitespace in comment x-net/src/main/java/org/apache/harmony/xnet/provider/jsse/CertificateRequest.java Change-Id: I99975ae22599c7df0d249fa013ae7ea7c9c08051
-
Elliott Hughes authored
Change-Id: Ia95af76e2995ce7fb0778b020baf2882a8b0a3dd
-
- 05 Apr, 2010 1 commit
-
-
Elliott Hughes authored
This method causes a lot of confusion, and we can do a lot better. (Ideally, the API would either not exist or be something like "public boolean ready()".) I've removed poor-quality documentation overrides too, so the full documentation is visible in most places. (InflaterInputStream is an obvious exception.) Also, to a lesser extent, improve the InputStream.skip documentation. Change-Id: I6d6cd788e6a32ad4a2613d1e381610f1ad8575fe
-
- 22 Mar, 2010 1 commit
-
-
Elliott Hughes authored
I've fixed a few typos, and removed a few of the more egregiously nonsensical or incorrect comments that were nearby. Change-Id: I35851baebd532f949cc269f4738a26eeb9b6e697
-
- 05 Mar, 2010 2 commits
-
-
Costin Manolache authored
Merge commit '1a29c735752f4082c8e32347f4a6b10c4fdeb1f5' into dalvik-dev * commit '1a29c735752f4082c8e32347f4a6b10c4fdeb1f5': Fix server side SSLEngine ServerKeyExchange signature.
-
Costin Manolache authored
Code using SSLEngine for non-blocking SSL can't talk with openssl as a client, since the signature is computed on different content (and openssl checks it, unlike java). The fix is to use strip the 0x00 prefix when signing - like it is done when generating the message, refactored both to use a common method. We also include the length in the signature, it was also missing.
-
- 28 Feb, 2010 1 commit
-
-
Brian Carlstrom authored
-
- 25 Feb, 2010 1 commit
-
-
Brian Carlstrom authored
At enh's suggestion, I'm consoldiating the OpenSSL related native code into a single wrapper class NativeCrypto. This changes is the firs step, combining the cpp code into a single NativeCode.cpp. The next step will involved introducting a single SSL_CTX in NativeCode and cleaning up SSL_CTX use. As part of this, I'll start moving the native wrappers to from various OpenSSL*.java classes into NativeCode.
-