GitLab

This is derived from costin's change Ib18da136cb628515d6909c438cd0809452d7058a.
It moves the protocols data to the AppData's callbacks so the memory can be
released when the handshake completes.

Change-Id: Id61feaa6f28250e393f5c8093688b099e92dce9c

Bug: 6168278

Change-Id: I240d2cbc91f616fd486efc5203e2221c9896d90f

Change-Id: I4c46e643647440fb3606179f773cd202a38a36ce

This allows OpenSSL ENGINE to be used for RSA and DSA private key
operations.

Also add in support for directly passing an OpenSSLKey to the
OpenSSLSocketImpl in case we are using ENGINEs.

Change-Id: Ia31735109052a13e421900b69ba5de13bbce0f6f

Bug: 5835156

(cherry picked from commit 943036db28577070ae5b5fed5051a2e47d95f6ea)

Change-Id: I7a1a9d7ae7003e54847fb2b0e99cc560e1fe34f0

Bug: http://b/6152698
Change-Id: I5b0bd61aeaf8d27c0087210859de1fd2cbcf88d7

Added    SHA1      : 5F:3B:8C:F2:F8:10:B3:7D:78:B4:CE:EC:19:19:C3:73:34:B9:C7:74
         Subject   : C=JP, O=SECOM Trust Systems CO.,LTD., OU=Security Communication RootCA2

Bug: http://code.google.com/p/android/issues/detail?id=26616
Change-Id: I9e37c787029a84b19a725773f30b9d8bc6842ebf

The address can be null for SSLSockets that have not been connected.

Bug: 5835165

(cherry picked from commit cb047c49abcf3b7b5c231b68431c291fe2d81b52)

Change-Id: I12eb92ab0cdb42b89333361a485979c48365d5da

Before the code would squash excecptions into null values, then turn
the values back into KeyManagementException. Now the code preservers
the underlying exception as the cause of the KeyManagementException.

(cherry picked from commit b6d100ca03dd4c576c6735ce510cecc70d2e6617)

Change-Id: Ia833145839578760ed9b49c626e8d4ab86ceacbe

Change-Id: Ieb6a4c75389dbafe832db0c3a2efdd133a062e73

Make it clearer that setSecurityManager method always throws
a SecurityException. It's easy to miss the current
"@throws SecurityManager always" line when you're reading the
docs.

Change-Id: Id5475bd952bd4e4a27a0058034c1b508b28a4f0f

Turns out there was one connect variant I hadn't properly addressed.

Bug: http://code.google.com/p/android/issues/detail?id=24318
Change-Id: I7af697bdf6ea17ad1f8ea904d7d8a051ffb970d2

* commit 'cf6f6b4c':

* commit 'e7daa0e9':

Newer compiler complains on return type mismatch. Change the error
return from NULL -> 0 to please the compiler.

Change-Id: Ib3e5f4a0e3e2507009cb309bb7965efbb968b7cb

Bug: 6009802

Update the TrustManagerImpl Api to allow clearing stored certificates.
This is needed so we can remove CAs when credential storage is updated.

Change-Id: I024f7e8b12b60ea0ee35d7f94280e0e3d6db039f

* commit '129c08d9':

Change-Id: If997d65f5b89d15b786cbad642a0ebc59e1851ae

* commit '81670623':

This implements the NativeCrypto piece necessary to do basic block
cipher operations. More work will need to be done to enable useful
modes.

This gives us the ability to replace BouncyCastle's ECB mode that it
bases the higher level CBC, CTR, etc modes on. However, calling through
JNI to OpenSSL for 16-byte blocks for AES ends up being the same speed
as the Java implementation.

Further enhancements to use large blocks during the JNI call should show
marked improvements in speed.

Change-Id: I594a6d13ce5101a1ef2877b84edaa5e5b65e1e71

Change-Id: I905b379ef5c5382dfe06a46817e92e3ebced89e8

When GoogleTV had to work around this for glibc, I realized that maybe I was
the problem, not the C library...

Bug: http://code.google.com/p/android/issues/detail?id=15722
Change-Id: Ia3fbf2f4578ad60d84e538819660c181248f1e1c

We had an ugly bug where HttpEngine was inspecting its response
cache's implementation type, and calling stats tracking methods
if that type permitted it.

This worked for all the libcore tests, but not in practice since
the public HttpResponseCache class WRAPS the libcore HttpResponseCache.
The new interface makes the new APIs called by HttpEngine on its
cache explicit, and permits the public android.net.HttpResponseCache
to implement the API.

Bug: http://code.google.com/p/android/issues/detail?id=25418
Change-Id: Ic9697af96635256e7ffc24f86628a85fbe2defd4

Previously if the response had a Content-Length or Transfer-Encoding
header we assumed it had a body. This was untrue of HEAD requests.
This is not a regression.

Bug: http://code.google.com/p/android/issues/detail?id=24672
Change-Id: I264db2014e681a6f8e93cd40b0f1640c95103d21

Bug: http://code.google.com/p/android/issues/detail?id=24327
Change-Id: Ib31b60fe01e3d0b5a1ed7125ddf047e09fd64056

Bug: http://code.google.com/p/android/issues/detail?id=24748
Change-Id: Iafab3a6c55007a47b68af5581148f5556f2fc808

am f04956f8: am cddd42f4: am 99918765: Fix for number pattern tests for devices not supporting "ar" locale

* commit 'f04956f8':

am 6f32bcec: am 6e678367: am 99918765: Fix for number pattern tests for devices not supporting "ar" locale

* commit '6f32bcec':
  Fix for number pattern tests for devices not supporting "ar" locale

* commit 'cddd42f4':

* commit '6e678367':
  Fix for number pattern tests for devices not supporting "ar" locale

* commit '99918765':
  Fix for number pattern tests for devices not supporting "ar" locale

* commit '99918765':
  Fix for number pattern tests for devices not supporting "ar" locale

Bug: http://code.google.com/p/android/issues/detail?id=25124
Change-Id: I31ed6818f6828cc0da53aa0c3ca867d88c5f66d3