OSDN Git Service
Brian Carlstrom [Fri, 2 Apr 2010 21:13:53 +0000 (14:13 -0700)]
New "from scratch" Junit tests for javax.net.ssl
These new tests focusi on areas where I am making OpenSSL
implementation chages, where I found differences from the RI when
working on that code, and places where we were missing functionality
compared to the spec. They all work on the RI, many fail on Dalvik
luni/src/test/java/javax/net/ssl/AllTests.java
luni/src/test/java/javax/net/ssl/SSLContextTest.java
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/SSLSocketFactoryTest.java
luni/src/test/java/javax/net/ssl/SSLSocketTest.java
Adding my javax.net.ssl.AllTests to the main list
luni/src/test/java/tests/AllTests.java
Adding Bouncy Castle provider /usr/share/java/bcprov.jar to host
classpath for testing with --java-home /usr/lib/jvm/java-6-openjdk
tools/runner/java/dalvik/runner/JavaVm.java
Fix usage comment to reboot to bootloader
run-core-tests
Change-Id: I7a7998fc7bedc7a00b1836517d043c6a27c8bd4d
Jesse Wilson [Thu, 1 Apr 2010 21:44:29 +0000 (14:44 -0700)]
Merge "New method-level granularity and output streaming for vogar." into dalvik-dev
Jesse Wilson [Thu, 1 Apr 2010 06:51:56 +0000 (23:51 -0700)]
New method-level granularity and output streaming for vogar.
Elliott Hughes [Thu, 1 Apr 2010 21:31:56 +0000 (14:31 -0700)]
Merge "Tidy up our getAvailableLocales methods to actually ask ICU4C." into dalvik-dev
Elliott Hughes [Thu, 1 Apr 2010 20:56:32 +0000 (13:56 -0700)]
Tidy up our getAvailableLocales methods to actually ask ICU4C.
These specialized methods are little used, and in several cases ICU itself
just returns the list of locales, but that's ICU's business, not ours. As
long as ICU is in charge of our locale-specific data, it should be responsible
for answering questions about what locale-specific data is available...
Change-Id: Idc8a66bbf7fcbc6b06e30929e6a7af3fe30ab7d1
Andy McFadden [Thu, 1 Apr 2010 20:06:00 +0000 (13:06 -0700)]
Merge "Mark external alloc functions as deprecated." into dalvik-dev
Elliott Hughes [Thu, 1 Apr 2010 18:15:14 +0000 (11:15 -0700)]
Merge "Add Java 6's java.net.IDN." into dalvik-dev
Elliott Hughes [Wed, 31 Mar 2010 22:45:15 +0000 (15:45 -0700)]
Add Java 6's java.net.IDN.
harmony's tests and my code, though ICU4C does all the hard work.
I've added a test of my own to demonstrate some weird RI behavior (that I've
emulated in our implementation).
Bug:
2497395
Change-Id: I8146f72a8a3204449ee3d0d9065dadc1c1c77fcc
Elliott Hughes [Thu, 1 Apr 2010 01:34:53 +0000 (18:34 -0700)]
Fix an off-by-one error in fd range checking.
Change-Id: I15f932bc246ebc7fec9471a873bef8bd58ef15fc
Carl Shapiro [Wed, 31 Mar 2010 23:27:24 +0000 (16:27 -0700)]
Merge "Add the ability to treat the zygote heap as a root collect just the forked application heap." into dalvik-dev
Elliott Hughes [Wed, 31 Mar 2010 22:52:09 +0000 (15:52 -0700)]
Merge "Add Java 6's ResourceBundle/Properties API." into dalvik-dev
Andy McFadden [Wed, 31 Mar 2010 22:50:17 +0000 (15:50 -0700)]
Mark external alloc functions as deprecated.
First step toward disabling them.
Bug
2537379.
Change-Id: I8664a5f1d7dcff11ddeb97b5e700dad76bffc723
Elliott Hughes [Wed, 31 Mar 2010 20:59:55 +0000 (13:59 -0700)]
Add Java 6's ResourceBundle/Properties API.
I've pretty much taken the upstream ResourceBundle implementations
as-is, putting back our string-to-locale conversion, removing a bit
of duplication and non-free, non-spec EBCDIC support, and hard-coding
the text of the MissingResourceExceptions (since harmony's changed
its message catalog from ours, so I had to touch those bits of the
code anyway).
(Why haven't I bothered to pay much attention to the resource bundle
implementations? Because I already rewrote our only code that was
using them to not use them, and third-party developers should be
using Android's resource system instead. There's very little chance
anyone needs Java resource bundles. I paid some attention to Properties,
because they're still somewhat useful.)
Also remove various unused messages, and update our tests. I've mostly
_not_ taken the upstream tests, because it would require a lot of work
that we'll be doing anyway when we switch to using their test suite
properly.
I ran the jtreg tests we're able to run, and the normal-case ones (plus
the stress test) seemed okay.
Bug:
2497395
Change-Id: I91606df0dc1a45e6974fbb27a0d334af87254f0b
Elliott Hughes [Wed, 31 Mar 2010 15:55:17 +0000 (08:55 -0700)]
Merge "Add Java 6's Calendar API changes." into dalvik-dev
Elliott Hughes [Wed, 31 Mar 2010 00:47:11 +0000 (17:47 -0700)]
Add Java 6's Calendar API changes.
Also correct two javadoc mistakes I added, and bring back the latest harmony
java6 CalendarTest.
Bug:
2497395
Change-Id: If481390948ca93d5f709a8ed4585991889de9f8b
Jesse Wilson [Wed, 31 Mar 2010 06:37:04 +0000 (23:37 -0700)]
Merge "Splitting TestRun into Action and Outcome." into dalvik-dev
Jesse Wilson [Tue, 30 Mar 2010 22:53:36 +0000 (15:53 -0700)]
Splitting TestRun into Action and Outcome.
Changing Vogar to prefer the word "action" over "test" to avoid
deemphisizing benchmarks and main classes.
Moving expectation management out of TestRun and into Expectation
and the ExpectationStore. The parse method moved, but it is unchanged.
Renamed ExpectedResult to Expectation.
Currently outcome has 2 names: an outcome name and an action name.
At the moment these are always equal. In a follow up to this change
I intend to allow a single action to have multiple outcomes, which
will allow JUnit tests to be tracked at the right granularity when
executed by vogar.
Carl Shapiro [Fri, 12 Mar 2010 04:39:47 +0000 (20:39 -0800)]
Add the ability to treat the zygote heap as a root collect just the
forked application heap.
Change-Id: I8807897ae426f8274018d950fec44a2182a90525
Elliott Hughes [Tue, 30 Mar 2010 22:19:23 +0000 (15:19 -0700)]
Merge "Trivial tidying up of networking code." into dalvik-dev
Elliott Hughes [Tue, 30 Mar 2010 21:21:19 +0000 (14:21 -0700)]
Trivial tidying up of networking code.
Removing cruft and renaming things.
Change-Id: I317726c40a1addf3a1ad9d905239438424082549
Jesse Wilson [Tue, 30 Mar 2010 20:05:32 +0000 (13:05 -0700)]
Merge "Fixing vogar script to use the latest name" into dalvik-dev
Jesse Wilson [Tue, 30 Mar 2010 20:05:09 +0000 (13:05 -0700)]
Fixing vogar script to use the latest name
Jesse Wilson [Tue, 30 Mar 2010 01:09:29 +0000 (18:09 -0700)]
Renaming dalvik.runner to Vogar!
Also moving commands and target classes to their own directories. This is all
in preparation to move the code out to code.google.com
Elliott Hughes [Tue, 30 Mar 2010 03:51:48 +0000 (20:51 -0700)]
Fix a bug, and protect against an unrelated class of bugs.
If the Java array allocation in InetAddress.cpp failed, we'd free NULL instead of the
previously-allocated structure. This is a new bug in froyo, but only happens in out of
memory situations, so doesn't seem worth fixing there.
Unrelatedly, let's disallow assignment and copying of all our RAII classes. This isn't
a mistake I've seen made, but it's easy to protect against, so we may as well do so
consistently.
Change-Id: I2433b31ff983d388788b09e59e08d661f1725ecd
Elliott Hughes [Tue, 30 Mar 2010 01:35:30 +0000 (18:35 -0700)]
Merge "Add Java 6's exponent separator to DecimalFormatSymbols." into dalvik-dev
Elliott Hughes [Tue, 30 Mar 2010 00:39:51 +0000 (17:39 -0700)]
Add Java 6's exponent separator to DecimalFormatSymbols.
Bug:
2497395
Change-Id: Ic552fa828649bae882e508a62a44073d1038b5c0
Jesse Wilson [Tue, 30 Mar 2010 00:21:31 +0000 (17:21 -0700)]
Merge "Tweak our test & runner so XPath tests can be executed on Hudson." into dalvik-dev
Jesse Wilson [Mon, 29 Mar 2010 23:13:48 +0000 (16:13 -0700)]
Tweak our test & runner so XPath tests can be executed on Hudson.
Elliott Hughes [Mon, 29 Mar 2010 22:20:01 +0000 (15:20 -0700)]
Fix build.
Rather than try to be clever with header files (which didn't work with glibc
for the sim build) let's make Math.copySign have the StrictMath behavior,
and have StrictMath call Math. (The other way round, though it might seem
more logical, wouldn't solve the problem. We already have numerous cases of
StrictMath calling Math anyway.)
Change-Id: Ifff065ddc8fbd5d5f8d4d5b67bc9ac07a719eb00
Jesse Wilson [Mon, 29 Mar 2010 20:52:59 +0000 (13:52 -0700)]
Merge "Merge commit '
8812fdd7' into manualmerge" into dalvik-dev
Jesse Wilson [Mon, 29 Mar 2010 20:52:17 +0000 (13:52 -0700)]
Merge commit '
8812fdd7' into manualmerge
Conflicts:
libcore/xml/src/test/java/tests/api/javax/xml/parsers/SAXParserFactoryTest.java
Elliott Hughes [Mon, 29 Mar 2010 20:43:29 +0000 (13:43 -0700)]
Merge "Fix StrictMath.copySign's behavior with NaN." into dalvik-dev
Elliott Hughes [Sat, 27 Mar 2010 06:23:38 +0000 (23:23 -0700)]
Fix StrictMath.copySign's behavior with NaN.
StrictMath's copySign method is defined to treat all NaNs as positive, regardless
of their sign bit. Rather than add yet more code on the Java side (which already
called down to native code three times), let's just make one call to native code.
This change fixes failures in already-committed StrictMath tests.
The only other potential instances of this class of error were in the
max and min methods, but they all handle NaNs before calling
(double|float)To(Raw)?LongBits, so the choice there really is arbitrary.
Change-Id: I439dbdff9068cb420b78a6330cde9d7d0d12c0ef
Elliott Hughes [Sat, 27 Mar 2010 00:46:16 +0000 (17:46 -0700)]
Fix all our java.text.Bidi failures.
Two are invalid tests where we behave the same as the RI. The third reflects
a difference between ICU4C and the RI, so I've added code to recognize that
case and paper over the crack. I've also tidied up a bit more.
With any luck, I'll never see these files again...
Change-Id: I278502fbdea039e9a2318df830dc0b55851e96c0
Elliott Hughes [Fri, 26 Mar 2010 23:16:30 +0000 (16:16 -0700)]
Merge "Java 6 changed CollationKey from final to abstract." into dalvik-dev
Jesse Wilson [Fri, 26 Mar 2010 23:08:43 +0000 (16:08 -0700)]
am
2be0c815: (-s ours) Suppressing additional tests that fail in Froyo.
Merge commit '
2be0c815dda289b09f3bb26f751c124f30eaf612' into dalvik-dev
* commit '
2be0c815dda289b09f3bb26f751c124f30eaf612':
Suppressing additional tests that fail in Froyo.
Elliott Hughes [Fri, 26 Mar 2010 21:46:11 +0000 (14:46 -0700)]
Java 6 changed CollationKey from final to abstract.
I've also taken the opportunity to tidy up our implementation a little,
though my hands are tied by (a) the fact that our concrete classes are
in a separate package from our abstract classes and (b) frameworks/base
actually pokes about with our icu4jni collation code (http://b/
2417080).
I've also tidied up a bunch of dead code. In particular, it's silly for
us to check parameters in Java that will be checked in native code (and
that one would assume will be valid most of the time anyway).
Bug:
1635883
Change-Id: I7db3c1ff1f0d23cb85604f9c8eb995e4488d7c0a
Jesse Wilson [Fri, 26 Mar 2010 22:21:11 +0000 (15:21 -0700)]
Merge "Fixing tests to handle changes in our behaviour since DOM 3." into froyo
Jesse Wilson [Fri, 26 Mar 2010 21:49:30 +0000 (14:49 -0700)]
Suppressing additional tests that fail in Froyo.
Our File class doesn't do SecurityManager checks at the same time as it
did previously.
We added a failing logging test but not the fix.
We moved the SQL drivers around, causing some SQL tests to fail.
A new logging test doesn't load its resources using the safe pattern.
Change-Id: I34ab81e4001969aa536f61a89c277ae2b879d49d
Jesse Wilson [Fri, 26 Mar 2010 21:03:36 +0000 (14:03 -0700)]
Fixing tests to handle changes in our behaviour since DOM 3.
Our exception priority has changed for DOM attributes. We
previously used to throw DOMExceptions with namespace error
codes and now throw DOMExceptions with character error codes
when the attribute name is malformed. This caused changes to
many tests.
Another notable behaviour change is that we now supply the
qname (like the RI) where previously we did not. It is optional,
but we now include it for RI-consistency.
Yet another behaviour change is that we don't look at System
properties when choosing a SAX implementation. This simplifies
our internals significantly. End users who want an alternative
SAX implementation should construct it manually.
Also adding @KnownFailure tags for new tests that we have
never yet passed.
Change-Id: I6f81bedd7c2a0867086dc507b3220c2b07c4d3d3
Elliott Hughes [Fri, 26 Mar 2010 18:30:00 +0000 (11:30 -0700)]
Start cleaning up the Charset implementation.
This was going to be https://issues.apache.org/jira/browse/HARMONY-6461,
but I couldn't resist cleaning up some of the surrounding code, and ended
up cleaning up some of our native code too. In the course of the afternoon
I spent on this, I lost my conviction that the upstream change makes
sense, so I reverted that, leaving this change just pure cleanup.
(Note that the cleanup work is incomplete. This is an improvement, but
there's plenty left to do. I just don't want to get too distracted until
all the Java 6 changes are done.)
Change-Id: I56841db5f6c038bbf7942e83a148dca546519269
Elliott Hughes [Fri, 26 Mar 2010 18:05:38 +0000 (11:05 -0700)]
Merge "Use more idiomatic naming in test code." into dalvik-dev
Elliott Hughes [Fri, 26 Mar 2010 17:51:48 +0000 (10:51 -0700)]
Use more idiomatic naming in test code.
(I introduced this in https://android-git.corp.google.com/g/46305.)
Change-Id: I100202fc18e246dd7b66870ce75543642505336d
Ben Cheng [Thu, 25 Mar 2010 23:55:50 +0000 (16:55 -0700)]
am
63868feb: Use correct resource flags for Dalvik ld/st instructions to enable code motion.
Merge commit '
63868feb2cc6a1fb1c76d7b54296ef4869f632be' into dalvik-dev
* commit '
63868feb2cc6a1fb1c76d7b54296ef4869f632be':
Use correct resource flags for Dalvik ld/st instructions to enable code motion.
Ben Cheng [Thu, 25 Mar 2010 23:16:02 +0000 (16:16 -0700)]
Use correct resource flags for Dalvik ld/st instructions to enable code motion.
Change-Id: I9b371af4150b6245c0ff59eea63d83406edbbcee
Jesse Wilson [Thu, 25 Mar 2010 21:40:39 +0000 (14:40 -0700)]
am
f2f7880a: Test and document our handling of nulls with getString().
Merge commit '
f2f7880a40aed1c8d9c27db49226e47396556aac' into dalvik-dev
* commit '
f2f7880a40aed1c8d9c27db49226e47396556aac':
Test and document our handling of nulls with getString().
Jesse Wilson [Thu, 25 Mar 2010 20:31:43 +0000 (13:31 -0700)]
Test and document our handling of nulls with getString().
Our behaviour is consistent with Crockford's.
The test confirms that the behaviour is consistent with the report
of that bug, which the submitter claims is not how it should behave.
http://code.google.com/p/android/issues/detail?id=7257
Change-Id: Ibace4bd995e3cbc8fb6c9dc509f8f4491865a647
Bill Buzbee [Thu, 25 Mar 2010 20:11:50 +0000 (13:11 -0700)]
am
d630700a: Merge "Jit: Fix for
2542488 JIT codegen bug with overlapping wide operands"
Merge commit '
d630700a9fbc6d317e0b4df73cbbc74a2c8fadaa' into dalvik-dev
* commit '
d630700a9fbc6d317e0b4df73cbbc74a2c8fadaa':
Jit: Fix for
2542488 JIT codegen bug with overlapping wide operands
Bill Buzbee [Thu, 25 Mar 2010 20:07:45 +0000 (13:07 -0700)]
Merge "Jit: Fix for
2542488 JIT codegen bug with overlapping wide operands"
Elliott Hughes [Thu, 25 Mar 2010 18:30:13 +0000 (11:30 -0700)]
Apply https://issues.apache.org/jira/browse/HARMONY-4307.
They didn't add a test for this; supposedly it fixed a flaky AWT test. But
the change looks plausible.
Change-Id: I358849a20d5e38d01d6c77a4c335002bb7bba095
Elliott Hughes [Thu, 25 Mar 2010 17:39:48 +0000 (10:39 -0700)]
Clean up a few toStrings.
I couldn't find any more that were using getClass() rather than
getClass().getName().
Change-Id: I30b375748d36bc36d1e6999349fc472496456746
Bill Buzbee [Thu, 25 Mar 2010 17:38:34 +0000 (10:38 -0700)]
Jit: Fix for
2542488 JIT codegen bug with overlapping wide operands
Change-Id: I7b922e223fe1f5242d1f3db1fa18f54aaed725af
Elliott Hughes [Thu, 25 Mar 2010 17:25:17 +0000 (10:25 -0700)]
am
ec7c8c98: Fix java.util.Random\'s constructors.
Merge commit '
ec7c8c98a4094580224eb9c400249c5c0984cf29' into dalvik-dev
* commit '
ec7c8c98a4094580224eb9c400249c5c0984cf29':
Fix java.util.Random's constructors.
Elliott Hughes [Thu, 25 Mar 2010 06:38:48 +0000 (23:38 -0700)]
Fix java.util.Random's constructors.
Subclasses rely on having their overridden setSeed called by Random's
constructors, and the RI actually documents this behavior. (The
documentation even changed between Java 5 and Java 6 to make it _more_
explicit.)
This patch keeps that part of I6239d93bb46876ef1c4a5e155a6dc1ac6fab4eae
that improved our randomness, but reverts the attempt to fix Random's
uncouth behavior.
Also a regression test so we don't try to fix Random again in future.
Bug:
2502231
Change-Id: Ieea1009145c74eac9475c0cd5066dabad20eb114
Kenny Root [Thu, 25 Mar 2010 02:40:12 +0000 (19:40 -0700)]
am
f87ab961: Merge "Don\'t leak memory on failed pipe or mutex init"
Merge commit '
f87ab9616697b8bae08c5e8007cbdd0039a1f8ce' into dalvik-dev
* commit '
f87ab9616697b8bae08c5e8007cbdd0039a1f8ce':
Don't leak memory on failed pipe or mutex init
Ben Cheng [Thu, 25 Mar 2010 02:40:10 +0000 (19:40 -0700)]
am
3d5d8736: Merge "Fix for the JIT blocking mode plus some code cleanup."
Merge commit '
3d5d87364d062734753bfd26336e96a7e8d03360' into dalvik-dev
* commit '
3d5d87364d062734753bfd26336e96a7e8d03360':
Fix for the JIT blocking mode plus some code cleanup.
Andy McFadden [Thu, 25 Mar 2010 02:40:04 +0000 (19:40 -0700)]
am
c66affa4: Switch to VMWAIT while outputting trace data.
Merge commit '
c66affa440de113df861928c528fcbbead2ead45' into dalvik-dev
* commit '
c66affa440de113df861928c528fcbbead2ead45':
Switch to VMWAIT while outputting trace data.
Kenny Root [Thu, 25 Mar 2010 00:07:15 +0000 (17:07 -0700)]
Merge "Don't leak memory on failed pipe or mutex init"
Kenny Root [Thu, 25 Mar 2010 00:01:22 +0000 (17:01 -0700)]
Don't leak memory on failed pipe or mutex init
SSL allocation might fail on pipe or mutex creation, so free allocated
memory when that happens.
Change-Id: Ibe3813cb652db1e880e3acf0dc1fa37fbe0ad455
Elliott Hughes [Wed, 24 Mar 2010 23:23:29 +0000 (16:23 -0700)]
Merge "Add Java 6's getAvailableLocales/getInstance to DateFormatSymbols and DecimalFormatSymbols." into dalvik-dev
Elliott Hughes [Wed, 24 Mar 2010 21:04:55 +0000 (14:04 -0700)]
Add Java 6's getAvailableLocales/getInstance to DateFormatSymbols and DecimalFormatSymbols.
This patch also improves the documentation of all getAvailableLocales methods
to clarify exactly what "available" means. Note that at the moment, many of
our implementations just return Locale.getAvailableLocales. It turns out that
ICU does the same in most cases, but I'll come back in a separate patch and
add code so we just pass these down to ICU, and wash our hands of the matter.
I've rewritten DateFormatSymbols.equals, which was very wrong. The time zone
names comparison had an invalid optimization, and its slow-path loop was wrong
too: it only ever tested values against themselves, which isn't likely to have
been the author's intention.
I've added toString overrides to DateFormatSymbols and DecimalFormatSymbols,
because I need them every time I work on these classes.
I've made the constants in DecimalFormatSymbols static final, and given them
idiomatic names. (I still think we might benefit from breaking these into
separate fields, as in the serialized form, but that's a separate issue.)
Finally, and unrelatedly, I've added a comment to BreakIteratorProvider that
I missed in my last change to that file.
Change-Id: I5d6cb30f9afdb502d38353d8a624dc2f0fef41ac
Ben Cheng [Wed, 24 Mar 2010 22:49:50 +0000 (15:49 -0700)]
Merge "Fix for the JIT blocking mode plus some code cleanup."
Ben Cheng [Wed, 24 Mar 2010 22:24:19 +0000 (15:24 -0700)]
Fix for the JIT blocking mode plus some code cleanup.
Bug:
2517606
Change-Id: I2b5aa92ceaf23d484329330ae20de5966704280b
Andy McFadden [Wed, 24 Mar 2010 20:35:19 +0000 (13:35 -0700)]
Show current values of mutexes in thread dump.
Experimental feature; may or may not be useful. Dump the current value
of half a dozen VM mutexes. This will tell us if they're unlocked,
locked, or contended-for at the point somebody does a thread dump. For
example, when the GC notices the HeapWorker watchdog has expired:
I/dalvikvm( 2828): DALVIK THREADS:
I/dalvikvm( 2828): (mutexes: tll=0 tsl=0 tscl=0 ghl=1 hwl=1 hwll=0)
In this case, gcHeapLock and heapWorkerLock are held, the rest are not.
If this turns out to be useful we can put more effort into the output
format. The pthread lib doesn't insert the thread ID unless you're
using recursive or errorcheck mutexes, which (for performance reasons)
we try not to use.
Change-Id: I272534d8c9db6340cfc23bfdddb47beefc2efbfb
Andy McFadden [Wed, 24 Mar 2010 20:03:08 +0000 (13:03 -0700)]
Switch to VMWAIT while outputting trace data.
One of the traces from a monkey run suggests we stalled out during the
open/write/close part of writing the trace data. To reduce the chances
of spin-on-suspending here we now switch to VMWAIT during the I/O.
For bug
2541030.
Change-Id: Ice832988213741976917fdd36afa12a694a42e81
Elliott Hughes [Wed, 24 Mar 2010 01:06:35 +0000 (18:06 -0700)]
Improve the ThreadGroup documentation.
(Because I was asked a question about ThreadGroup today, and the answer is
always "don't use ThreadGroup"...)
Change-Id: Ibaa2f1830e1fe435c9712f3b08506639b9521b9e
Elliott Hughes [Wed, 24 Mar 2010 00:41:53 +0000 (17:41 -0700)]
Merge "Remove explicit 8192 arguments to BufferedReader and friends." into dalvik-dev
Elliott Hughes [Tue, 23 Mar 2010 23:31:13 +0000 (16:31 -0700)]
am
b7b9be13: Merge "Improve java.net.InetAddress.getLocalHost documentation."
Merge commit '
b7b9be1343b59b429b6de9b28ffffbbbde8478c2' into dalvik-dev
* commit '
b7b9be1343b59b429b6de9b28ffffbbbde8478c2':
Improve java.net.InetAddress.getLocalHost documentation.
Elliott Hughes [Tue, 23 Mar 2010 23:26:01 +0000 (16:26 -0700)]
Remove explicit 8192 arguments to BufferedReader and friends.
These were clearly added just to shut up our own warning, and are now
unnecessary and misleading to future maintainers.
There's one barely-related change: InputStreamReader and OutputStreamWriter
are very similar, and this patch makes them more similar, and adds a few
missing modifiers from their fields.
Change-Id: I959011f914ff215e92bbfa41c1bac66465803685
Elliott Hughes [Tue, 23 Mar 2010 23:23:02 +0000 (16:23 -0700)]
Merge "Improve java.net.InetAddress.getLocalHost documentation."
Brian Carlstrom [Tue, 23 Mar 2010 23:09:29 +0000 (16:09 -0700)]
Merge "Fix --debug-port to --debug in vogar usage" into dalvik-dev
Brian Carlstrom [Tue, 23 Mar 2010 23:05:10 +0000 (16:05 -0700)]
Fix --debug-port to --debug in vogar usage
Change-Id: I012100578caa4f5768b7926b4fbe489c80f8d32b
Elliott Hughes [Tue, 23 Mar 2010 22:45:12 +0000 (15:45 -0700)]
Merge "Remove the "default buffer size" warnings." into dalvik-dev
Elliott Hughes [Tue, 23 Mar 2010 22:22:02 +0000 (15:22 -0700)]
Remove the "default buffer size" warnings.
Everyone seems to agree these have outlived their usefulness.
Change-Id: Ia2a53ffa1f9498333b4b2d51d5c1db488b89abaf
Bill Buzbee [Tue, 23 Mar 2010 21:55:08 +0000 (14:55 -0700)]
am
02388cf1: Jit: disable for Sapphire.
Merge commit '
02388cf18a87ba0243cddc6ccf3eaf8668118723' into dalvik-dev
* commit '
02388cf18a87ba0243cddc6ccf3eaf8668118723':
Jit: disable for Sapphire.
Bill Buzbee [Tue, 23 Mar 2010 21:34:11 +0000 (14:34 -0700)]
Jit: disable for Sapphire.
Change-Id: I3c48f239ef9a83fac1ef05d18ea1b00ab27879ff
Elliott Hughes [Tue, 23 Mar 2010 21:21:27 +0000 (14:21 -0700)]
Improve java.net.InetAddress.getLocalHost documentation.
Bug:
1518707
Change-Id: I741a7a28325320949e84e997e6a49d3356c9a308
Andy McFadden [Tue, 23 Mar 2010 20:01:33 +0000 (13:01 -0700)]
am
d4e0952b: Further refinements to "kill other thread".
Merge commit '
d4e0952b73f9054adcc927a1c12699ba63f1672a' into dalvik-dev
* commit '
d4e0952b73f9054adcc927a1c12699ba63f1672a':
Further refinements to "kill other thread".
Andy McFadden [Tue, 23 Mar 2010 19:34:43 +0000 (12:34 -0700)]
Further refinements to "kill other thread".
Switched from SIGSEGV to SIGSTKFLT. Otherwise it looks like we're
seg-faulting in code that was actually executing just fine.
Ignore SIGSEGV before returning. This avoids creating a second dump
when dvmAbort() is called. (If the caller's stack trace were at all
interesting, we wouldn't be here, and we can do without the log spam.)
Added some more details to the logging.
For bug
2517042.
Change-Id: Ifff7fd75d6dbf7e3663152787f8921dda215fe40
Elliott Hughes [Tue, 23 Mar 2010 00:56:03 +0000 (17:56 -0700)]
Merge "Revert the removal of java.nio.NIOAccess, which frameworks/base/ uses." into dalvik-dev
Elliott Hughes [Tue, 23 Mar 2010 00:50:33 +0000 (17:50 -0700)]
Revert the removal of java.nio.NIOAccess, which frameworks/base/ uses.
This was removed in https://android-git.corp.google.com/g/45796, but turns
out to be used. I think we can do better in the long term, but right now
this just removes some of the duplication from the original, and adds
comments explaining where it's used.
Bug:
2535509
Change-Id: Ie8e717537b8b8c70c8689f2dbb2c3a38501ecc57
Elliott Hughes [Tue, 23 Mar 2010 00:00:55 +0000 (17:00 -0700)]
Merge "Add the Java 6 java.text.spi and java.util.spi interfaces." into dalvik-dev
Elliott Hughes [Mon, 22 Mar 2010 23:36:58 +0000 (16:36 -0700)]
Add the Java 6 java.text.spi and java.util.spi interfaces.
We don't currently use these, and don't necessarily plan to. Full support
would have a run-time cost, and it's not obvious that it would be particularly
useful.
Code search can't actually find any users of this stuff in the wild outside
of the various VM implementations and their test suites.
Bug:
2497395
Change-Id: Ie25aef73ece6d1fd169fdcb7b2f847761d77914d
Bill Buzbee [Mon, 22 Mar 2010 22:43:25 +0000 (15:43 -0700)]
am
64ee717b: Jit: Re-enable for Sapphire
Merge commit '
64ee717be7cf4aad64371e73cfd180aa4d85dd07' into dalvik-dev
* commit '
64ee717be7cf4aad64371e73cfd180aa4d85dd07':
Jit: Re-enable for Sapphire
Bill Buzbee [Mon, 22 Mar 2010 20:59:45 +0000 (13:59 -0700)]
Jit: Re-enable for Sapphire
Change-Id: Icbfd8781756c2c7cde88140ed40766e29448b1ff
Elliott Hughes [Mon, 22 Mar 2010 19:26:38 +0000 (12:26 -0700)]
Remove all remaining "@since Android" tags.
I've fixed a few typos, and removed a few of the more egregiously nonsensical
or incorrect comments that were nearby.
Change-Id: I35851baebd532f949cc269f4738a26eeb9b6e697
Elliott Hughes [Mon, 22 Mar 2010 18:16:46 +0000 (11:16 -0700)]
Add Java 6's java.util.Arrays changes.
I've kept our binary search implementation and just generalized the
interface.
I've gone out of my way to preserve exception priority. I know we don't
agree that it's necessary, but it is important if we want to be able to
run other people's tests. If someone wants to write a new high-quality
test suite (which would be a great thing to have), we should remove the
hacks. (I've commented them.) Otherwise, I've gone out of my way to keep
the near-duplicates forced on us by Java's primitive type system as
identical as possible.
This passes all harmony and jtreg tests.
Change-Id: I91fbf707dac76124c6dbe59b0b30b7ded9a69529
Elliott Hughes [Fri, 19 Mar 2010 23:23:09 +0000 (16:23 -0700)]
Merge "Java 6 java.nio.Buffer changes." into dalvik-dev
Elliott Hughes [Fri, 19 Mar 2010 23:06:18 +0000 (16:06 -0700)]
Java 6 java.nio.Buffer changes.
We no longer need the duplicated lower-quality documentation on
each subclass, and the NIOAccess class turns out to be unused.
Bug:
2497395
Change-Id: Ib7ce8f51aac2b78bf56fff4526af1c2c8818e277
Andy McFadden [Fri, 19 Mar 2010 22:40:13 +0000 (15:40 -0700)]
am
e9f1038c: Merge "Use pthread_kill(3) instead of kill(2)."
Merge commit '
e9f1038c3e71dba123532c62741257867d68e08b' into dalvik-dev
* commit '
e9f1038c3e71dba123532c62741257867d68e08b':
Use pthread_kill(3) instead of kill(2).
Andy McFadden [Fri, 19 Mar 2010 22:35:40 +0000 (15:35 -0700)]
Merge "Use pthread_kill(3) instead of kill(2)."
Dan Bornstein [Fri, 19 Mar 2010 22:21:55 +0000 (15:21 -0700)]
Merge "Remove a superfluous declaration (left over from when DexClassLoader was duplicated from PathClassLoader)." into dalvik-dev
Andy McFadden [Fri, 19 Mar 2010 22:18:57 +0000 (15:18 -0700)]
Use pthread_kill(3) instead of kill(2).
Another swing at useful native traces. pthread_kill(3) uses tkill(2)
instead of kill(2), so we may have a better chance of the signal landing
on the right thread.
For bug
2517042.
Change-Id: I62871c1e21b11442ea52c861397674f1644f4da9
Elliott Hughes [Fri, 19 Mar 2010 21:49:47 +0000 (14:49 -0700)]
Track Java 6's BreakIterator API changes.
Bug:
2497395
Change-Id: Iaef75bbd946bebe6201b5f52564bcaf8a4d0e56e
Elliott Hughes [Fri, 19 Mar 2010 01:25:22 +0000 (18:25 -0700)]
Clean up the Java side of the ICU interface a bit.
My original intention was just to add the missing "final" on a few classes,
but our BreakIterator implementation struck me as excessively bloated and
confusing.
Change-Id: I2d2dccafe8ec91124f3c83909c9ec647cc2d51e2
Dan Bornstein [Fri, 19 Mar 2010 21:01:59 +0000 (14:01 -0700)]
Remove a superfluous declaration (left over from when DexClassLoader was
duplicated from PathClassLoader).
Change-Id: Iaf3e8a5b23df1a8bbf670bf6d6045fffef6c8fcb
Jesse Wilson [Fri, 19 Mar 2010 20:50:30 +0000 (13:50 -0700)]
am
7713b900: Merge "Fixing namespace+prefix mode in Expat and removing optional fields from callbacks."
Merge commit '
7713b900d12c7a46334db46b33fbe848167189dd' into dalvik-dev
* commit '
7713b900d12c7a46334db46b33fbe848167189dd':
Fixing namespace+prefix mode in Expat and removing optional fields from callbacks.
Jesse Wilson [Fri, 19 Mar 2010 20:39:28 +0000 (13:39 -0700)]
Merge "Fixing namespace+prefix mode in Expat and removing optional fields from callbacks."
Jesse Wilson [Fri, 19 Mar 2010 19:51:43 +0000 (12:51 -0700)]
Fixing namespace+prefix mode in Expat and removing optional fields from callbacks.
The first part is related to bug 6632:
http://code.google.com/p/android/issues/detail?id=6632
I added these optional fields back when I was originally updating
the XML parser for Froyo. I've decided to remove them to simplify
migrating between Android and the RI. It should also save some
object allocations.
Note that the RI v5 and the RI v6 behave differently for optional
values on attributes; this motivated me to add the otherwise
unfortunate assertOneOf() method to the testcase. (We behave more
like RI v6, which is to supply the values upon request)
Change-Id: Icfa5d29976a86bf194b3ed7c0d9e2275c3bff9dd
Barry Hayes [Wed, 17 Mar 2010 19:12:39 +0000 (12:12 -0700)]
Added flags to the vm:
-Xgc:[no]preverify
-Xgc:[no]postverify
to run verify routines over the heap pre- and post-gc.
Changed the Verify.h interface. It now publishes an entry point for
verifying a HeapBitmap, rather than the HeapBitmap callback.
Added a dvmHeapSuspendAndVerify to Heap.h for verification outside of
the GC.
Added callbacks before and after GC, under the locks, under flag control.
Processing of properties to produce flags is in a different project,
frameworks/base
Change-Id: I3f3896583fe9e7239bbe2f374d7ed4c5dd5d3e82