| bf9c47975631cda6ead2eda00d3db3c93444eeba |
|
14-Sep-2012 |
Elliott Hughes <enh@google.com> |
Manually inline StrictMath.min(JJ) and StrictMath.max(JJ).
Also add new intrinsic tests.
Bug: 7146208
Change-Id: Ic454c8c5b218c51ce1c21f0c11163455392b4e72
|
| 08343a4ef77616fc420e51d1095c0d6a266041ba |
|
16-Feb-2011 |
Elliott Hughes <enh@google.com> |
Remove our implementations of inline natives.
It turns out that our String.equals was getting used. Making dalvikvm's inline
natives be native methods ensures that doesn't happen.
I've also added a test to ensure we explicitly test both variants.
Change-Id: Ie2a4b5289092da5ffca4a4273692f1048ab3b841
|
| 750d24b13461035a89c9af8ec7555588da1cc3d0 |
|
27-Jan-2011 |
Elliott Hughes <enh@google.com> |
Better tests for Float.parseFloat/Double.parseDouble, plus some fixes.
I'm not happy with the magic 1024s, and I had hoped to have one hack to
cover both harmony bugs. But this passes all our existing tests, all
harmony's existing tests, and the new tests I wrote while looking at
the code. (The new tests all pass on the RI too.)
I've also made a few doc comments slightly less unclear.
Change-Id: Ib72fd20bb3b157c7db5a09913247eed3cc746a2e
|
| 2fc5dcd5614f910f25d794d272834752a72e63b1 |
|
09-Dec-2010 |
Elliott Hughes <enh@google.com> |
Apply various Intellij quick fixes to java.lang.
There's one real bug here: our Integer.toString was supposed to have a cache
for small negative values, but an accidentally-introduced temporary meant we
were never using the cached values. Other than that, this is just cleanup.
Change-Id: I457f9bd166c9a029ba8b439f3bbfa926f9b84cc9
|
| 706de1164836051ab31cc69eb77a6bba1a723896 |
|
27-Aug-2010 |
Elliott Hughes <enh@google.com> |
Minor tidy-up of Float.toString and Double.toString.
This patch looks larger than it is. I've moved duplicated constants out of
Math and StrictMath and into Float and Double where they belong. I've also
moved a table out of BigDecimal so I can reuse it in NumberConverter, which
I've renamed to RealToString.
The main active ingredient here is that there's no longer StringBuilder usage
to prepend '-'. This actually brings us down from 74ns to 62ns for an easy
case like 123.45f.
[cherry-pick of 910106e29fe98f14b2c36312a7498287273ba826 from dalvik-dev to gingerbread; plus build fix]
Bug: 2934304
Change-Id: Id79a6fb0d739e673b7428d240fc48bdae9b8bb43
|
| d16c314465f7f5905fe8852eaa09d4e7e14ec835 |
|
25-Aug-2010 |
Elliott Hughes <enh@google.com> |
Faster implementations of Math/StrictMath's copySign methods.
These are only faster than the old native implementations now we have
intrinsics for float/int and double/long conversion, but given those,
they're about 4x faster.
We were already using Java for getExponent, the other obvious candidate
for an all-Java implementation.
[cherry-pick of a61c63c54ba46bb8a1dcbd888e3b29d6cedfb1ff from dalvik-dev to gingerbread]
Bug: 2935622
|
| 910106e29fe98f14b2c36312a7498287273ba826 |
|
27-Aug-2010 |
Elliott Hughes <enh@google.com> |
Minor tidy-up of Float.toString and Double.toString.
This patch looks larger than it is. I've moved duplicated constants out of
Math and StrictMath and into Float and Double where they belong. I've also
moved a table out of BigDecimal so I can reuse it in NumberConverter, which
I've renamed to RealToString.
The main active ingredient here is that there's no longer StringBuilder usage
to prepend '-'. This actually brings us down from 74ns to 62ns for an easy
case like 123.45f.
Bug: 2934304
Change-Id: I65461753247eeb9ff890ffe66b0bc79d66d2ec9c
|
| a61c63c54ba46bb8a1dcbd888e3b29d6cedfb1ff |
|
25-Aug-2010 |
Elliott Hughes <enh@google.com> |
Faster implementations of Math/StrictMath's copySign methods.
These are only faster than the old native implementations now we have
intrinsics for float/int and double/long conversion, but given those,
they're about 4x faster.
We were already using Java for getExponent, the other obvious candidate
for an all-Java implementation.
Bug: 2935622
Change-Id: I77549c88b7448b6087a67b37f8c317579b4f92b8
|
| e22935d3c7040c22b48d53bd18878844f381287c |
|
13-Aug-2010 |
Elliott Hughes <enh@google.com> |
Remove most of our C-style casts.
After being burned by an incorrect C-style cast that cast away const, I've been
keen to remove them all and turn on -Wold-style-cast. This patch doesn't get us
that far, but it does kill the majority of our C-style casts. In turn, the
majority of the casts that it removes are the ones from our tables of native
methods to be registered.
The new NATIVE_METHOD macro also _enforces_ our convention of using the
"Class_nativeMethod" style of naming. Mostly this works out fine. In some
cases (most notably ExpatParser and ExpatAttributes) I've had to un-overload
a few functions, but I don't like overloading anyway, and in the particular
case of a native method, where the stack trace doesn't show a line number,
overloading makes it one step harder to work out which native method you're
actually in. So good riddance to that. The only unfortunate case is
Math.copySign, where there are two overloads corresponding to copysign(3)
and copysignf(3). I had to add an extra layer of indirection there. In my
defense, we've never shipped these functions before, they're unlikely to
become anyone's hotspot, and the right fix is to be doing such trivial work
on the Java side anyway, with intrinsics making the conversion between
float/double and int/long cheap.
This patch also replaces other C-style casts, primarily in
"OSNetworkSystem.cpp".
This patch also removes unnecessary uses of the "struct" keyword.
This patch also fixes a "may be used uninitialized" warning (now error) in
the sim build for "ICU.cpp".
The remaining C-style casts are in the hairy float-parsing code. That stuff --
and turning on -Wold-style-cast -- will have to wait for another day.
Change-Id: I9b3ee14aefd4676f980f6a7ca757595d78d80e6a
|
| 5d40b59c6bba79dc978f173ad64cdfbd8a937018 |
|
08-Jul-2010 |
Elliott Hughes <enh@google.com> |
Clean up File.createTempFile and make Math.random thread-safe.
File.createTempFile has long annoyed me: using a SecureRandom but then always
incrementing by one is silly. Using a new Random is more random, more scalable,
and slightly faster than reusing a SecureRandom according to my benchmark.
Relatedly, access to the Random used by Math.random wasn't synchronized, which
is a bug.
I've also fixed a typo in a comment in a test. Woo!
Change-Id: Ief72ec0e378ed3d5807c278c589f7b0e44c21d3e
|
| 1d7a5c66d9973f3383e7e4f9f8be5262161a1421 |
|
23-Jun-2010 |
Elliott Hughes <enh@google.com> |
Fix Math.atan2 documentation.
The StrictMath.atan2 documentation is already correct.
Bug: http://code.google.com/p/android/issues/detail?id=9260
Change-Id: Ibd86cf7fe73a120578840dce4126a902a827fb40
|
| 162a12c1442641a95fe95859fa4e561b22db049f |
|
25-May-2010 |
Elliott Hughes <enh@google.com> |
Remove @hide from Java 6 API.
I've left the two new spi packages @hidden, because I think we shouldn't
support them without convincing demand from developers (and I don't
believe there could be such a thing --- they just don't make sense, and
if we add anything, it should be the ability to provide extra ICU data).
Also fix a handful of javadoc syntax errors in Arrays.java and TreeMap.java.
Bug: 2497395
Change-Id: I4176b72daff0face4ed6c7ee1d1f4267d52006b4
|
| f33eae7e84eb6d3b0f4e86b59605bb3de73009f3 |
|
13-May-2010 |
Elliott Hughes <enh@google.com> |
Remove all trailing whitespace from the dalvik team-maintained parts of libcore.
Gentlemen, you may now set your editors to "strip trailing whitespace"...
Change-Id: I85b2f6c80e5fbef1af6cab11789790b078c11b1b
|
| eeefcae2980c8db05ec08303b5b112afce232d26 |
|
06-May-2010 |
Elliott Hughes <enh@google.com> |
Remove the "Answers" javadoc barbarism.
I didn't realize how little of it was left, or I'd have done this earlier.
Change-Id: Ic4c23c3779bbee0ed4a8117584c5332e04536b5d
|
| ace7e6ff6285c073a80aaad68321b17bdcd09a4b |
|
27-Mar-2010 |
Elliott Hughes <enh@google.com> |
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
|
| 5820c2d5ea8199005b9c1c9dbf34b3d5cdbb8c10 |
|
11-Mar-2010 |
Elliott Hughes <enh@google.com> |
Tidy up java.lang.Math and java.lang.StrictMath.
Removes duplication, fixes typos, and removes some unnecessary object creation.
Note that otherwise-duplicated code that calls a native method needs to remain
duplicated, because the native methods have different implementations. Other
than that, it's not clear to me that the remaining textual differences (mainly
scalb and its implementation details) are meaningful, but it's not clear that
they're not, so I'm leaving them for now. We can always come back later.
This new code still passes all the junit and jtreg tests.
Change-Id: Ica28a01fd1469a162e05deccdb9e6f919246d9f3
|
| 73342e0a7adba5a494309c1fcee0c64fee4f79f1 |
|
11-Mar-2010 |
Elliott Hughes <enh@google.com> |
Copy harmony's jdk6 Math and StrictMath.
There's some ugliness here I want to remove, but it'll be less confusing if
I commit the upstream code first without my changes...
...that said, I've reverted the upstream Math.pow change because it it's just
cruft; we already pass their tests and jtreg's more thorough pow tests (see
http://blogs.sun.com/darcy/entry/finding_a_bug_in_fdlibm). My guess is that
their real problem was that they were using the buggy fdlibm 5.2 until after
they made the Math.pow change. We've always used 5.3, so we were fine.
Change-Id: I5a6c080d9fd6b60dc7bf77ac10096a913766f512
|
| 4fefecee9d4a5d2a4510f516b4015607b19e8d09 |
|
13-Nov-2009 |
Jean-Baptiste Queru <jbq@google.com> |
eclair snapshot
|
| 5839b909d9528b7726e678a4b696ed37df15d897 |
|
10-Oct-2009 |
Jesse Wilson <jessewilson@google.com> |
Udating luni to Harmony r823222.
Highlights:
- InputStream.skip concurrency issue
- "better" messages in bound exceptions for streams and arrays
- prefer fewer writes to underlying streams (using byte[] buffers)
- Rename subclasses to not reuse names from their superclasses
- PlatformAddressFactory.allocMap bugfix
Plus some spelling fixes, style fixes, serial version UIDs and other
boilerplate improvements.
|
| a389b4a499f40379b0b204d7ba1c2057663d95c0 |
|
11-Aug-2009 |
Jesse Wilson <jessewilson@google.com> |
Update Luni to Harmony r802921.
Notable changes:
- replaced StringBuffer with StringBuilder in several places
- fixed a problem with BufferedInputStream's newline characters (EBCDIC)
- cleanup Timer's finalizer helper object
- new cache for the canonical path of a file
- fixed concurrency issue with ArrayList
- floating point parsing now trims length for very small numbers
- encoding specified "UTF-8" when converting some byte[]s to strings (JarURLConnection, Util, OSFileSystem)
- Harmony now implements floor and ceil in Java. We continue to use native code.
|
| f5597e626ecf7949d249dea08c1a2964d890ec11 |
|
25-Jul-2009 |
Jesse Wilson <jessewilson@google.com> |
Integrate luni module (but not tests) to Harmony r772995.
Notable changes
- Stripped "@since Android 1.0" from many files. Most files
are now 100% the same in Dalvik and Harmony.
- AbstractStringBuilder.reverse() supports surrogates
- AbstractStringBuilder shares less to waste less memory
- Bitset optimized
- BufferedInputStream changed to support unsynchronized close()
- BufferedOutputStream does flushInternal
- BufferedReader supports EBCDIC NEL
- Collections.synchronizedList().indexOf() does a copy for more concurrency
- Classes in nio module changed: DatagramChannelImpl, SocketChannelImpl
and ServerSocketChannelImpl (these depend on internal APIs changed in
this update)
- DataInputStream/DataOutputStream now use a small buffer to limit the
number of times the underlying stream is accessed
- Date now has a minutes offset, more efficient toString()
- ExposedByteArrayInputStream: new internal class
- DeleteOnExit moved to top-level class
- FileDescriptor.isValid() now non-native
- Float, Double lessThan optimized (fix for compare(-0.0F, 0.0F) still pending)
- FileURLConnection now guesses content types from streams
- HashMap iterator changes
- Hashtable iterator changes
- INetworkSystem
- removes bind2(), createMulticastSocket, sendStream(),
- renames createSocket to createStreamSocket
- JarURLConnection rewritten
- LinkedHashMap: new iterator
- Locale, Currency, TimeZone: now use ICU in Harmony, plain Java in Dalvik
- ObjectInputStream: Accessor objects in Harmony, direct native in Dalvik
- ProxyClassFile - many changes
- String - optimized ascii for toLowerCase, toUpperCase, compare
- Timer - rewritten
- TreeMap - rewritten
- URLClassLoader - new
- URLConnection - new guessContentTypeFromStream(), uses org.apache.harmony.awt.www.content
to lookup content type handlers
|
| adc854b798c1cfe3bfd4c27d68d5cee38ca617da |
|
04-Mar-2009 |
The Android Open Source Project <initial-contribution@android.com> |
auto import from //depot/cupcake/@135843
|
| 1c0fed63c71ddb230f3b304aac12caffbedf2f21 |
|
04-Mar-2009 |
The Android Open Source Project <initial-contribution@android.com> |
auto import from //depot/cupcake/@135843
|
| dd828f42a5c83b4270d4fbf6fce2da1878f1e84a |
|
18-Dec-2008 |
The Android Open Source Project <initial-contribution@android.com> |
Code drop from //branches/cupcake/...@124589
|
| fdb2704414a9ed92394ada0d1395e4db86889465 |
|
21-Oct-2008 |
The Android Open Source Project <initial-contribution@android.com> |
Initial Contribution
|