GitLab

Central directories of zip files should be read lazily.
They are needed only to service calls to
BaseDexClassLoader#findResouce(s). Most android processes
don't use these methods since they use android resources.

bug:6797061

Change-Id: I1a5b5d03572601707e1fb1fd4424c1ae2fd2217d

Change-Id: I1e5f377e655f72b8d2e497e83cf4207f5b786e01

Bug: 7378963
Change-Id: I45fb1350baaf591e59b78bb3cf97091667cb079c

Bug: 7453821
Change-Id: I69408d0bb4063e34441ed1d7632fd1ccac39965b

Bug: 7355008
Change-Id: Ifdedca5cf5c96c86f94a0deead61e1798d1a495f

We can't rely on consuming all the listen(2) backlog. For the tests we've
seen fail because they sometimes connect rather than time out, switch to
an unroutable address.

(cherry-pick of babccbf9.)

Bug: 6971145
Change-Id: Ibfa412ff1ad7da7e63842d0162cc67a706e2b27e

Bug: 6951157
Change-Id: I4a23a90a4a54d71f9cc5f0c608a30000c8e1b033

Support_PortManager is a bad idea, badly implemented. Kill some more of it.
Only OldDatagramSocketTest still uses Support_PortManager, so hopefully we
can kill this completely soon.

Bug: 7347233
Change-Id: I9dec7b23d59edc8c7a5e391dd5ebccd2a92f82b5

Bug: 7337970
Change-Id: I7da3e22bc2c773258d30f0919d51d386f7b92e7d

Bug: 7337970
Change-Id: I17318afadd04d972f907fe968e050ece98116c6a

Includes cherry-pick of 847f22ad

Bug: 7192453
Change-Id: Ib5fa1e313d942d2c1034e8e7831af285ad24d71d

No clue what's wrong here, but we're seeing some flakiness in the
"connect" tests. Hopefully there will be a clue in this extra
output next time we see a failure.

(cherry-pick of 73587588.)

Bug: 6971145
Change-Id: I8236b2fb368581e2b6965898f109370e7e0af2e5

(cherry-pick of 0f129636.)

Bug: 6972612
Change-Id: Ibc919c7023898f9e521f222d69c7efb50a170923

Depending on the timing of the two threads, SocketTimeoutException or SSLProtocolException can be thrown.

Bug: 6964910
Change-Id: Ia921f0dc61d1f2eb39a8d41c4ff6d551c924977e

It appears that OpenSSL has a bug where any DSA key verification
operation that fails will push an error on the OpenSSL error stack. This
change will clear the error state whenever a key verification is
performed.

(cherry picked from commit 912d5102)

Bug: 7302619
Change-Id: I16d1afad20f2d3934a01963ab62eea45b6caa6a1

Based on https://github.com/square/okhttp/pull/30
(git fetch git://github.com/square/okhttp.git && git cherry-pick -x b8c51dc5bdc89d5487cbc110d6fe40389de16480)

Bug: http://code.google.com/p/android/issues/detail?id=37221
Bug: 7291446

(cherry picked from commit 4bedacd5)

Change-Id: I68c55e9259db91dbccca30824c00726094570304

This isn't right for all locales, but it's right for "tl" in "tl" itself,
which is how we render language names in Settings.

(Yes, I know that "tl" is for "Tagalog" and "fil" is for "Filipino", but
Google policy is to use the latter name. If our resource system weren't
limited to two-letter language codes, we'd be using the modern three-letter
"fil" instead of the obsolete "tl", but we can't fix that right now. Bukas.)

Bug: 7291355
Change-Id: I7d9021dc55d9f1d3f8a75b10c890a2d5f91982d6

Our Pipe originally used socket(2) to create AF_INET/AF_INET6
sockets. This was clearly a bad idea.

I rewrote it to use socketpair(2) and AF_UNIX, but this was
before the big "expose POSIX" rewrite, so it required a bunch
of hacks in the native code, so I went with pipe(2) instead.

The trouble with pipe(2) is that we end up using FileChannel
to implement the Pipe.SinkChannel and Pipe.SourceChannel, but
the kernel won't wake a read(2) on a pipe if another thread
calls close(2) on it, so we started failing interrupt tests.

This (final?) rewrite is hopefully the best of all worlds:
we don't have any INTERNET permission needs, the code is as
simple as the pipe(2) code, and interruption works.

Bug: 7084342
Bug: 2735373
Bug: http://code.google.com/p/android/issues/detail?id=9431

(cherry picked from commit 32180823)

Change-Id: Ib92cdf8c818f6bba1f00e191f1b624ce9e693754