OSDN Git Service

android-x86/dalvik.git
14 years agoam fb8eb6b1: Merge "Print the self thread id and the owning thread id in the descript...
Carl Shapiro [Tue, 2 Mar 2010 23:18:35 +0000 (15:18 -0800)]
am fb8eb6b1: Merge "Print the self thread id and the owning thread id in the description of the IllegalMonitorExcpetion throw by unlockMonitor."

Merge commit 'fb8eb6b1534c142ccc91844e6507a3f67d16a457' into dalvik-dev

* commit 'fb8eb6b1534c142ccc91844e6507a3f67d16a457':
  Print the self thread id and the owning thread id in the description

14 years agoam 59c1f5cb: Use assertEquals rather than assertTrue/assertFalse.
Elliott Hughes [Tue, 2 Mar 2010 23:18:32 +0000 (15:18 -0800)]
am 59c1f5cb: Use assertEquals rather than assertTrue/assertFalse.

Merge commit '59c1f5cb8bea87144b473170adb0b0dae7267937' into dalvik-dev

* commit '59c1f5cb8bea87144b473170adb0b0dae7267937':
  Use assertEquals rather than assertTrue/assertFalse.

14 years agoMerge "Print the self thread id and the owning thread id in the description of the...
Carl Shapiro [Tue, 2 Mar 2010 23:12:53 +0000 (15:12 -0800)]
Merge "Print the self thread id and the owning thread id in the description of the IllegalMonitorExcpetion throw by unlockMonitor."

14 years agoUse assertEquals rather than assertTrue/assertFalse.
Elliott Hughes [Tue, 2 Mar 2010 23:04:09 +0000 (15:04 -0800)]
Use assertEquals rather than assertTrue/assertFalse.

14 years agoam 971b5067: Add our home-grown Normalizer tests.
Elliott Hughes [Tue, 2 Mar 2010 21:59:04 +0000 (13:59 -0800)]
am 971b5067: Add our home-grown Normalizer tests.

Merge commit '971b50675ea0c140abb326ffb75fe79442eb6eec' into dalvik-dev

* commit '971b50675ea0c140abb326ffb75fe79442eb6eec':
  Add our home-grown Normalizer tests.

14 years agovogar should chdir to android_root
Brian Carlstrom [Tue, 2 Mar 2010 21:51:32 +0000 (13:51 -0800)]
vogar should chdir to android_root

I often run vogar from various working directories but several
internal paths assume that it is run from the android_root.

14 years agoAdd our home-grown Normalizer tests.
Elliott Hughes [Tue, 2 Mar 2010 21:24:31 +0000 (13:24 -0800)]
Add our home-grown Normalizer tests.

Originally https://android-git.corp.google.com/g/42517.

14 years agoAdded assert to dvmInterpHandleFillArrayData.
Barry Hayes [Wed, 24 Feb 2010 17:19:07 +0000 (09:19 -0800)]
Added assert to dvmInterpHandleFillArrayData.
Arrays of objects should never make it to there.

14 years agoMove allocation of the uninitialized class object on to the object heap.
Barry Hayes [Mon, 1 Mar 2010 17:27:59 +0000 (09:27 -0800)]
Move allocation of the uninitialized class object on to the object heap.

14 years agoam 10ebc7d0: Merge "Implement (but @hide) java.text.Normalizer from Java 6."
Elliott Hughes [Tue, 2 Mar 2010 04:11:37 +0000 (20:11 -0800)]
am 10ebc7d0: Merge "Implement (but @hide) java.text.Normalizer from Java 6."

Merge commit '10ebc7d0b84dcb98e1a7eeac96ef06acdfc8d184' into dalvik-dev

* commit '10ebc7d0b84dcb98e1a7eeac96ef06acdfc8d184':
  Implement (but @hide) java.text.Normalizer from Java 6.

14 years agoMerge "Implement (but @hide) java.text.Normalizer from Java 6."
Elliott Hughes [Tue, 2 Mar 2010 04:07:07 +0000 (20:07 -0800)]
Merge "Implement (but @hide) java.text.Normalizer from Java 6."

14 years agoam acf4a11b: Merge "Implementing almost all of Document.normalizeDocument()."
Jesse Wilson [Tue, 2 Mar 2010 03:19:31 +0000 (19:19 -0800)]
am acf4a11b: Merge "Implementing almost all of Document.normalizeDocument()."

Merge commit 'acf4a11bf201ccdf4a4d2b429d33757556cdc0ed' into dalvik-dev

* commit 'acf4a11bf201ccdf4a4d2b429d33757556cdc0ed':
  Implementing almost all of Document.normalizeDocument().

14 years agoMerge "Implementing almost all of Document.normalizeDocument()."
Jesse Wilson [Tue, 2 Mar 2010 03:15:57 +0000 (19:15 -0800)]
Merge "Implementing almost all of Document.normalizeDocument()."

14 years agoImplementing almost all of Document.normalizeDocument().
Jesse Wilson [Tue, 2 Mar 2010 01:31:53 +0000 (17:31 -0800)]
Implementing almost all of Document.normalizeDocument().

This follows the rules specified by DOMConfiguration. In particular:
 - replacing CDATA nodes with text
 - splitting CDATA nodes
 - merging text nodes
 - stripping comments
 - detecting invalid characters

I haven't added the normalization code for XML validation or namespaces.

14 years agoImplement (but @hide) java.text.Normalizer from Java 6.
Elliott Hughes [Tue, 2 Mar 2010 00:28:05 +0000 (16:28 -0800)]
Implement (but @hide) java.text.Normalizer from Java 6.

Based on https://android-git.corp.google.com/g/42516.

Includes the harmony tests from their Java 6 branch.

Bug: 719001

14 years agoam 98376a27: Merge "Allow unpriviledged gdbserver to attach to debuggable processes."
David 'Digit' Turner [Mon, 1 Mar 2010 23:57:17 +0000 (15:57 -0800)]
am 98376a27: Merge "Allow unpriviledged gdbserver to attach to debuggable processes."

Merge commit '98376a27029e3ce4de36e5dc2379600be33ee4ae' into dalvik-dev

* commit '98376a27029e3ce4de36e5dc2379600be33ee4ae':
  Allow unpriviledged gdbserver to attach to debuggable processes.

14 years agoMerge "Allow unpriviledged gdbserver to attach to debuggable processes."
David 'Digit' Turner [Mon, 1 Mar 2010 23:45:58 +0000 (15:45 -0800)]
Merge "Allow unpriviledged gdbserver to attach to debuggable processes."

14 years agoAllow unpriviledged gdbserver to attach to debuggable processes.
David 'Digit' Turner [Tue, 23 Feb 2010 00:56:58 +0000 (16:56 -0800)]
Allow unpriviledged gdbserver to attach to debuggable processes.

This patch is used to allow an unprivileged gdbserver binary to
perform a ptrace attach to a debuggable process.

The kernel will only allow the attach to succeed under the following
conditions:

- gdbserver and the process are run under the exact same uids/gids
- the process is dumpable

The second condition requires this patch. The first one is ensured
by different changes to the platform (that don't directly touch Dalvik).

This is part of the work performed to allow native debugging of
applications on production devices.

14 years agoam def26c1a: Merge "Added a simple check for stray tracked refs."
Andy McFadden [Mon, 1 Mar 2010 21:43:35 +0000 (13:43 -0800)]
am def26c1a: Merge "Added a simple check for stray tracked refs."

Merge commit 'def26c1aff8c80079ebfa766773f5405d054d901' into dalvik-dev

* commit 'def26c1aff8c80079ebfa766773f5405d054d901':
  Added a simple check for stray tracked refs.

14 years agoMerge "Added a simple check for stray tracked refs."
Andy McFadden [Mon, 1 Mar 2010 21:39:38 +0000 (13:39 -0800)]
Merge "Added a simple check for stray tracked refs."

14 years agoam 448cc386: Merge "Added an x86-atom "TODO" list."
Andy McFadden [Mon, 1 Mar 2010 20:43:52 +0000 (12:43 -0800)]
am 448cc386: Merge "Added an x86-atom "TODO" list."

Merge commit '448cc386f87564ce251407f7907b7855d317edb7' into dalvik-dev

* commit '448cc386f87564ce251407f7907b7855d317edb7':
  Added an x86-atom "TODO" list.

14 years agoMerge "Added an x86-atom "TODO" list."
Andy McFadden [Mon, 1 Mar 2010 20:40:06 +0000 (12:40 -0800)]
Merge "Added an x86-atom "TODO" list."

14 years agoAdded an x86-atom "TODO" list.
Andy McFadden [Mon, 1 Mar 2010 20:16:12 +0000 (12:16 -0800)]
Added an x86-atom "TODO" list.

Besides the obvious things (unimplemented opcodes), this points out a
couple of changes that were made to the interpreter and require a
corresponding change for x86-atom.

14 years agoAdd a missing android-changed marker, fix a half-finished one.
Elliott Hughes [Mon, 1 Mar 2010 19:22:54 +0000 (11:22 -0800)]
Add a missing android-changed marker, fix a half-finished one.

14 years agoUse more efficient idioms for converting types.
Elliott Hughes [Mon, 1 Mar 2010 18:59:42 +0000 (10:59 -0800)]
Use more efficient idioms for converting types.

(Because we only include the Java 5 JDBC driver, there was less to fix than I
thought; this code is copy & pasted into the other drivers.)

14 years agoAdded a simple check for stray tracked refs.
Andy McFadden [Mon, 1 Mar 2010 18:52:00 +0000 (10:52 -0800)]
Added a simple check for stray tracked refs.

The idea is to make sure the main thread is getting through the VM init
process with an empty tracked references table.  If we find any, we
print a warning and dump the table, but continue on anyway.

My test was a trivial dvmMalloc(), which turned up a flaw in the ref
table dump code: it didn't deal with "raw" dvmMalloc sections.  So
that's fixed as well.

14 years agoam ad71aa20: Merge "New implementation for DOMConfiguration."
Jesse Wilson [Mon, 1 Mar 2010 18:26:05 +0000 (10:26 -0800)]
am ad71aa20: Merge "New implementation for DOMConfiguration."

Merge commit 'ad71aa2046502db784a8c861b30b11ddf88044f1' into dalvik-dev

* commit 'ad71aa2046502db784a8c861b30b11ddf88044f1':
  New implementation for DOMConfiguration.

14 years agoam 16525029: Merge "Fix a FormatFlagsConversionMismatchException bug I introduced...
Elliott Hughes [Mon, 1 Mar 2010 18:26:01 +0000 (10:26 -0800)]
am 16525029: Merge "Fix a FormatFlagsConversionMismatchException bug I introduced in Froyo."

Merge commit '165250293744f0aadcb0a0bfba1b53b8a4fb99b7' into dalvik-dev

* commit '165250293744f0aadcb0a0bfba1b53b8a4fb99b7':
  Fix a FormatFlagsConversionMismatchException bug I introduced in Froyo.

14 years agoMerge "New implementation for DOMConfiguration."
Jesse Wilson [Mon, 1 Mar 2010 18:21:05 +0000 (10:21 -0800)]
Merge "New implementation for DOMConfiguration."

14 years agoNew implementation for DOMConfiguration.
Jesse Wilson [Sat, 27 Feb 2010 08:00:54 +0000 (00:00 -0800)]
New implementation for DOMConfiguration.

This API is disgusting. Its not regular, not typesafe, sparsely
implemented and overly specific. I'm implementing the minimum I can
get away with - exactly the same route taken by the RI.

The Parameter stuff feels a little bit like overkill, but it allowed
me to group related chunks of code together and avoid long chains of
equalsIgnoreCase if statements.

This is necessary to implement Document.normalize(), one of the
last remaining APIs in DOMv3.

Also fixing our implementation of Node.normalize() and adding tests
to discover a bug in that code. Previously a document like this:
  "<foo>abc<br>def</foo>"
Would be normalized to this:
  "<foo>abcdef<br></foo>"
Which is a horrible bug! Yuck.

Implementation and tests for Document.normalize() are forthcoming.

14 years agoMerge "Fix a FormatFlagsConversionMismatchException bug I introduced in Froyo."
Elliott Hughes [Mon, 1 Mar 2010 18:19:46 +0000 (10:19 -0800)]
Merge "Fix a FormatFlagsConversionMismatchException bug I introduced in Froyo."

14 years agoam fcc61ec2: Merge "Don\'t set x86 arch variant."
Andy McFadden [Mon, 1 Mar 2010 03:36:45 +0000 (19:36 -0800)]
am fcc61ec2: Merge "Don\'t set x86 arch variant."

Merge commit 'fcc61ec20932de83d385131983b47f82a2008cd3' into dalvik-dev

* commit 'fcc61ec20932de83d385131983b47f82a2008cd3':
  Don't set x86 arch variant.

14 years agoam dd55f1bb: Merge "Adding tests to parse document attributes from the DOM."
Jesse Wilson [Mon, 1 Mar 2010 03:36:41 +0000 (19:36 -0800)]
am dd55f1bb: Merge "Adding tests to parse document attributes from the DOM."

Merge commit 'dd55f1bba1216e91206255f31de587062ed115b6' into dalvik-dev

* commit 'dd55f1bba1216e91206255f31de587062ed115b6':
  Adding tests to parse document attributes from the DOM.

14 years agoam f109fc4a: Merge "Update a few things in x86-atom."
Andy McFadden [Mon, 1 Mar 2010 03:36:37 +0000 (19:36 -0800)]
am f109fc4a: Merge "Update a few things in x86-atom."

Merge commit 'f109fc4aa8b26984e135cb8321bdde69b64c7878' into dalvik-dev

* commit 'f109fc4aa8b26984e135cb8321bdde69b64c7878':
  Update a few things in x86-atom.

14 years agoam 40094c16: Tweak the interpreter entries and 2nd level trace filter to capture...
Ben Cheng [Mon, 1 Mar 2010 03:36:32 +0000 (19:36 -0800)]
am 40094c16: Tweak the interpreter entries and 2nd level trace filter to capture more traces.

Merge commit '40094c16d9727cc1e047a7d4bddffe04dd566211' into dalvik-dev

* commit '40094c16d9727cc1e047a7d4bddffe04dd566211':
  Tweak the interpreter entries and 2nd level trace filter to capture more traces.

14 years agoresolved conflicts for merge of e84bea92 to dalvik-dev
Brian Carlstrom [Sun, 28 Feb 2010 21:51:14 +0000 (13:51 -0800)]
resolved conflicts for merge of e84bea92 to dalvik-dev

14 years agoMerge "Upgrade our sqlite JDBC driver to version 20100131." into dalvik-dev
Elliott Hughes [Sun, 28 Feb 2010 19:31:52 +0000 (11:31 -0800)]
Merge "Upgrade our sqlite JDBC driver to version 20100131." into dalvik-dev

14 years agoFix a FormatFlagsConversionMismatchException bug I introduced in Froyo.
Elliott Hughes [Sun, 28 Feb 2010 19:10:10 +0000 (11:10 -0800)]
Fix a FormatFlagsConversionMismatchException bug I introduced in Froyo.

We want to show the characters representing the flags, not the decimal
values of their ASCII representation. "Use overloading judiciously",
as the man says, or cause errors like this.

Seen in FRE56B (see https://android-git.corp.google.com/g/42528).

14 years agoam c85663a4: Merge "Fix a couple of typos in JIT function names."
Elliott Hughes [Sun, 28 Feb 2010 09:41:22 +0000 (01:41 -0800)]
am c85663a4: Merge "Fix a couple of typos in JIT function names."

Merge commit 'c85663a4856c476c4ead93bdf19e8a91b2d15e22' into dalvik-dev

* commit 'c85663a4856c476c4ead93bdf19e8a91b2d15e22':
  Fix a couple of typos in JIT function names.

14 years agoam ac9f2112: Merge "Jit: fix typo that caused incorrect def mask for blr\'s."
Bill Buzbee [Sun, 28 Feb 2010 09:41:17 +0000 (01:41 -0800)]
am ac9f2112: Merge "Jit: fix typo that caused incorrect def mask for blr\'s."

Merge commit 'ac9f2112303e1353b875999d6481e2047789af9e' into dalvik-dev

* commit 'ac9f2112303e1353b875999d6481e2047789af9e':
  Jit: fix typo that caused incorrect def mask for blr's.

14 years agoam f76b01c6: Merge "Document that Class.getSigners() always returns null."
Elliott Hughes [Sun, 28 Feb 2010 09:41:11 +0000 (01:41 -0800)]
am f76b01c6: Merge "Document that Class.getSigners() always returns null."

Merge commit 'f76b01c6377ad4653dcdc37b2ca35db785219f80' into dalvik-dev

* commit 'f76b01c6377ad4653dcdc37b2ca35db785219f80':
  Document that Class.getSigners() always returns null.

14 years agoam ca5ac879: Merge "Rearrange JDWP cleanup."
Andy McFadden [Sun, 28 Feb 2010 09:41:04 +0000 (01:41 -0800)]
am ca5ac879: Merge "Rearrange JDWP cleanup."

Merge commit 'ca5ac8792c0dc7cb33728ba3389fcc7d7b76772b' into dalvik-dev

* commit 'ca5ac8792c0dc7cb33728ba3389fcc7d7b76772b':
  Rearrange JDWP cleanup.

14 years agoam 4d5e5a0f: Merge "Keep the order from getaddrinfo() unchanged if preferIPv6Addresse...
Elliott Hughes [Sun, 28 Feb 2010 09:40:56 +0000 (01:40 -0800)]
am 4d5e5a0f: Merge "Keep the order from getaddrinfo() unchanged if preferIPv6Addresses() is true."

Merge commit '4d5e5a0f65e3cca6662fac8e832712fba61b9d2c' into dalvik-dev

* commit '4d5e5a0f65e3cca6662fac8e832712fba61b9d2c':
  Keep the order from getaddrinfo() unchanged if preferIPv6Addresses() is true.

14 years agoam cbcd79b3: Merge "Print "JIT" in the thread dump if the top frame is in JIT\'ed...
Ben Cheng [Sun, 28 Feb 2010 09:40:48 +0000 (01:40 -0800)]
am cbcd79b3: Merge "Print "JIT" in the thread dump if the top frame is in JIT\'ed code."

Merge commit 'cbcd79b31ddecad8ae38e5f45ac63e7225a877ff' into dalvik-dev

* commit 'cbcd79b31ddecad8ae38e5f45ac63e7225a877ff':
  Print "JIT" in the thread dump if the top frame is in JIT'ed code.

14 years agoam ae9cc0e6: Merge "Added printf attribute to exception function."
Andy McFadden [Sun, 28 Feb 2010 09:40:40 +0000 (01:40 -0800)]
am ae9cc0e6: Merge "Added printf attribute to exception function."

Merge commit 'ae9cc0e699cc680873b4fc72a020afd167722047' into dalvik-dev

* commit 'ae9cc0e699cc680873b4fc72a020afd167722047':
  Added printf attribute to exception function.

14 years agoam b4c05977: Optimize more easy multiplications by constants.
Elliott Hughes [Sun, 28 Feb 2010 09:40:32 +0000 (01:40 -0800)]
am b4c05977: Optimize more easy multiplications by constants.

Merge commit 'b4c05977c28c38d2f81b48d0cb15559dc3d05564' into dalvik-dev

* commit 'b4c05977c28c38d2f81b48d0cb15559dc3d05564':
  Optimize more easy multiplications by constants.

14 years agoam 88a0f970: Enhance the jit profiler to print more statistics and be more verbose.
Ben Cheng [Sun, 28 Feb 2010 09:40:23 +0000 (01:40 -0800)]
am 88a0f970: Enhance the jit profiler to print more statistics and be more verbose.

Merge commit '88a0f970e47dc0091d2c9965aa9bd06667e5f4b7' into dalvik-dev

* commit '88a0f970e47dc0091d2c9965aa9bd06667e5f4b7':
  Enhance the jit profiler to print more statistics and be more verbose.

14 years agoam 59a007be: am cf4e8b02: Merge "This is a contribution of x86-atom targeted assembly...
Andy McFadden [Sun, 28 Feb 2010 09:40:11 +0000 (01:40 -0800)]
am 59a007be: am cf4e8b02: Merge "This is a contribution of x86-atom targeted assembly for the fast byte-code interpreter engine. This is an initial contribution with minimal optimizations that target the Intel ATOM processor. We expect to continuously improve this cod

Merge commit '59a007be670601e40f036f63904b3f51c91a4d95' into dalvik-dev

* commit '59a007be670601e40f036f63904b3f51c91a4d95':
  This is a contribution of x86-atom targeted assembly for the fast byte-code interpreter engine. This is an initial contribution with minimal optimizations that target the Intel ATOM processor. We expect to continuously improve this code. It is expected that there will be a discussion on the potential merge of this code and similar efforts (i.e. the mterp/x86 directory first included with cupcake). While this code is intended to target ATOM and not a generic X-86 processor, we were able to show the following improvements over the c-portable interpreter using the simulator build:

14 years agoam 11834964: Clarify a few items in the dex spec.
Dan Bornstein [Sun, 28 Feb 2010 09:40:02 +0000 (01:40 -0800)]
am 11834964: Clarify a few items in the dex spec.

Merge commit '118349640ffa44bbc55ed355d0aaa2b993de5e5b' into dalvik-dev

* commit '118349640ffa44bbc55ed355d0aaa2b993de5e5b':
  Clarify a few items in the dex spec.

14 years agoam ba05d3b5: Merge "Add class init stats to alloc counters (API change)."
Andy McFadden [Sun, 28 Feb 2010 09:39:53 +0000 (01:39 -0800)]
am ba05d3b5: Merge "Add class init stats to alloc counters (API change)."

Merge commit 'ba05d3b507d5e1de794d00e7e634862c2a7cf8c2' into dalvik-dev

* commit 'ba05d3b507d5e1de794d00e7e634862c2a7cf8c2':
  Add class init stats to alloc counters (API change).

14 years agoam 4795e033: Merge "Implementing still more DOM API for text nodes. - Text.isElementC...
Jesse Wilson [Sun, 28 Feb 2010 09:39:43 +0000 (01:39 -0800)]
am 4795e033: Merge "Implementing still more DOM API for text nodes. - Text.isElementContentWhitespace() - Text.getWholeText() - Text.replaceWholeText()"

Merge commit '4795e033e5d93d72438d5004dd77ea82014f4657' into dalvik-dev

* commit '4795e033e5d93d72438d5004dd77ea82014f4657':
  Implementing still more DOM API for text nodes.

14 years agoam 60a65773: Minor fix for Adb.waitForNonEmptyDirectory so that waiting for /sdcard...
Brian Carlstrom [Sun, 28 Feb 2010 09:39:32 +0000 (01:39 -0800)]
am 60a65773: Minor fix for Adb.waitForNonEmptyDirectory so that waiting for /sdcard after reboot works again.

Merge commit '60a65773a7199f0a6b8f36bba00f11b84f8c0227' into dalvik-dev

* commit '60a65773a7199f0a6b8f36bba00f11b84f8c0227':
  Minor fix for Adb.waitForNonEmptyDirectory so that waiting for /sdcard after reboot works again.

14 years agoam 01961a9d: Merge "Fixed bug 2438465, prevented blocks with move-result from being...
Jeff Hao [Sun, 28 Feb 2010 09:39:21 +0000 (01:39 -0800)]
am 01961a9d: Merge "Fixed bug 2438465, prevented blocks with move-result from being combined."

Merge commit '01961a9d1f152c308b0c372214a1f204b3c0a5cf' into dalvik-dev

* commit '01961a9d1f152c308b0c372214a1f204b3c0a5cf':
  Fixed bug 2438465, prevented blocks with move-result from being combined.

14 years agoam d1ad6ae8: DalvikRunner --tee option to send output to a file or stdout at runtime
Brian Carlstrom [Sun, 28 Feb 2010 09:39:10 +0000 (01:39 -0800)]
am d1ad6ae8: DalvikRunner --tee option to send output to a file or stdout at runtime

Merge commit 'd1ad6ae8509d5890a7175a1dcd0a3f7fceb75d23' into dalvik-dev

* commit 'd1ad6ae8509d5890a7175a1dcd0a3f7fceb75d23':
  DalvikRunner --tee option to send output to a file or stdout at runtime

14 years agoam b0d01b01: Fixed bug 2438465, prevented blocks with move-result from being combined.
jeffhao [Sun, 28 Feb 2010 09:38:58 +0000 (01:38 -0800)]
am b0d01b01: Fixed bug 2438465, prevented blocks with move-result from being combined.

Merge commit 'b0d01b0178081c98b8cdb2fba2d84f275a0c595e' into dalvik-dev

* commit 'b0d01b0178081c98b8cdb2fba2d84f275a0c595e':
  Fixed bug 2438465, prevented blocks with move-result from being combined.

14 years agoam 05c9d91a: Add a test case for the jsr / invoke bug. (See the included info.txt...
Dan Bornstein [Sun, 28 Feb 2010 09:38:44 +0000 (01:38 -0800)]
am 05c9d91a: Add a test case for the jsr / invoke bug. (See the included info.txt for more details.)

Merge commit '05c9d91a7fd80fef3141b832fc8caeef44612c86' into dalvik-dev

* commit '05c9d91a7fd80fef3141b832fc8caeef44612c86':
  Add a test case for the jsr / invoke bug. (See the included

14 years agoUpgrade our sqlite JDBC driver to version 20100131.
Elliott Hughes [Sat, 27 Feb 2010 02:13:57 +0000 (18:13 -0800)]
Upgrade our sqlite JDBC driver to version 20100131.

This fixes a bunch of KnownFailures, and adds a bunch of new features. I've
updated the tests correspondingly (though doubtless we could add a lot more
tests now, if we wanted to).

I regenerated Constants.java manually, and explained how in its javadoc.

I've added the upstream VERSION file so it's easier to see what upstream
version we're at. (Constants now contains the version number too, but that's
less obvious.)

All our changes now have android-changed markers. I'll see about getting them
pushed upstream next week.

(This change best reviewed with "ignore whitespace". It looks like we made a
bunch of whitespace "corrections" when we imported this source, which just
makes it harder to stay in sync. I've taken the upstream copies of files that
only had whitespace differences.)

14 years agoBreak the sqlite JDBC driver out from our JDBC implementation.
Elliott Hughes [Fri, 26 Feb 2010 22:46:53 +0000 (14:46 -0800)]
Break the sqlite JDBC driver out from our JDBC implementation.

The JDBC driver is from a different source
(http://www.ch-werner.de/javasqlite/overview-summary.html) and is only
really needed for testing.

Bug 2468870 asks that we don't eagerly register the native methods for
these classes. That bug is fixed by this change.

Bug 2198667 asks that we stop shipping this JDBC driver as part of the
base system. That bug is not addressed by this change: the classes and
native code are now in their own, separate, .jar and .so files -- so
they'll be easier to remove in future -- but for now those files are
still in /system/framework and /system/lib respectively.

Bug: 2468870
Bug: 2198667

14 years agoMerge "Don't set x86 arch variant."
Andy McFadden [Fri, 26 Feb 2010 22:25:32 +0000 (14:25 -0800)]
Merge "Don't set x86 arch variant."

14 years agoDon't set x86 arch variant.
Andy McFadden [Fri, 26 Feb 2010 21:34:01 +0000 (13:34 -0800)]
Don't set x86 arch variant.

This was a workaround for a missing bit in the core build.  No longer
needed.

14 years agoMerge "Adding tests to parse document attributes from the DOM."
Jesse Wilson [Fri, 26 Feb 2010 21:16:39 +0000 (13:16 -0800)]
Merge "Adding tests to parse document attributes from the DOM."

14 years agoAdding tests to parse document attributes from the DOM.
Jesse Wilson [Fri, 26 Feb 2010 20:17:32 +0000 (12:17 -0800)]
Adding tests to parse document attributes from the DOM.

This adds implementation and tests for these methods:
getInputEncoding(), getDocumentURI().

And tests for these methods:
getXmlEncoding(), getXmlVersion(), getXmlStandalone().

14 years agoMerge "Update a few things in x86-atom."
Andy McFadden [Fri, 26 Feb 2010 18:16:38 +0000 (10:16 -0800)]
Merge "Update a few things in x86-atom."

14 years agoUpdate a few things in x86-atom.
Andy McFadden [Fri, 26 Feb 2010 15:35:20 +0000 (07:35 -0800)]
Update a few things in x86-atom.

This replaced unuses opcodes EC and EF with C stubs for the breakpoint
and execute-inline/range instructions, which were added after the last
bunch of stuff we rolled out to eclair open-source.  I also rebuilt the
"out" dir to pick up the various other changes.

14 years agoTweak the interpreter entries and 2nd level trace filter to capture more traces.
Ben Cheng [Thu, 25 Feb 2010 04:58:44 +0000 (20:58 -0800)]
Tweak the interpreter entries and 2nd level trace filter to capture more traces.

Real changes:
1) Add a new entry point from JIT to the interpreter to request hot traces w/o
   doing chaining.
2) Increase the granularity of the secondary profile filter to match 64-byte
   chunks using 64 entries.

The remaining are just cosmetic changes.

14 years agoMerge "First pass at reorganizing org.apache.harmony.xnet.provider.jsse native code...
Brian Carlstrom [Fri, 26 Feb 2010 00:15:50 +0000 (16:15 -0800)]
Merge "First pass at reorganizing org.apache.harmony.xnet.provider.jsse native code into a single file."

14 years agoMerge "Jit: fix typo that caused incorrect def mask for blr's." into dalvik-dev
Bill Buzbee [Fri, 26 Feb 2010 00:10:39 +0000 (16:10 -0800)]
Merge "Jit: fix typo that caused incorrect def mask for blr's." into dalvik-dev

14 years agoJit: fix typo that caused incorrect def mask for blr's.
Bill Buzbee [Thu, 25 Feb 2010 23:38:40 +0000 (15:38 -0800)]
Jit: fix typo that caused incorrect def mask for blr's.

14 years agoMerge "Fix a couple of typos in JIT function names."
Elliott Hughes [Fri, 26 Feb 2010 00:04:11 +0000 (16:04 -0800)]
Merge "Fix a couple of typos in JIT function names."

14 years agoMerge "Relocate function declaration out the GC and into the compiler." into dalvik-dev
Carl Shapiro [Thu, 25 Feb 2010 23:54:18 +0000 (15:54 -0800)]
Merge "Relocate function declaration out the GC and into the compiler." into dalvik-dev

14 years agoFirst pass at reorganizing org.apache.harmony.xnet.provider.jsse native code into...
Brian Carlstrom [Thu, 11 Feb 2010 06:56:47 +0000 (22:56 -0800)]
First pass at reorganizing org.apache.harmony.xnet.provider.jsse native code into a single file.

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.

14 years agoRelocate function declaration out the GC and into the compiler.
Carl Shapiro [Thu, 25 Feb 2010 23:47:08 +0000 (15:47 -0800)]
Relocate function declaration out the GC and into the compiler.

14 years agoMerge "Jit: fix typo that caused incorrect def mask for blr's."
Bill Buzbee [Thu, 25 Feb 2010 23:43:39 +0000 (15:43 -0800)]
Merge "Jit: fix typo that caused incorrect def mask for blr's."

14 years agoFix a couple of typos in JIT function names.
Elliott Hughes [Thu, 25 Feb 2010 23:41:42 +0000 (15:41 -0800)]
Fix a couple of typos in JIT function names.

(I saw these the other day, but preferred a separate patch.)

14 years agoJit: fix typo that caused incorrect def mask for blr's.
Bill Buzbee [Thu, 25 Feb 2010 23:38:40 +0000 (15:38 -0800)]
Jit: fix typo that caused incorrect def mask for blr's.

14 years agoRemove the sole use of WITH_HPROF_STACK_UNREACHABLE macro. This
Carl Shapiro [Thu, 25 Feb 2010 23:13:05 +0000 (15:13 -0800)]
Remove the sole use of WITH_HPROF_STACK_UNREACHABLE macro.  This
feature never seems to have existed.

14 years agoMerge "Eliminate the unused hprof "unreachable" feature. This code dates back to...
Carl Shapiro [Thu, 25 Feb 2010 23:04:55 +0000 (15:04 -0800)]
Merge "Eliminate the unused hprof "unreachable" feature.  This code dates back to the Perforce repository and has never been enabled.  Nobody seems to know what it was once, if ever, used for so it may as well go." into dalvik-dev

14 years agoMerge "Document that Class.getSigners() always returns null."
Elliott Hughes [Thu, 25 Feb 2010 22:59:44 +0000 (14:59 -0800)]
Merge "Document that Class.getSigners() always returns null."

14 years agoDocument that Class.getSigners() always returns null.
Elliott Hughes [Thu, 25 Feb 2010 22:42:17 +0000 (14:42 -0800)]
Document that Class.getSigners() always returns null.

Bug: 1594865
Bug: http://code.google.com/p/android/issues/detail?id=1766

14 years agoEliminate the unused hprof "unreachable" feature. This code dates
Carl Shapiro [Thu, 25 Feb 2010 03:47:40 +0000 (19:47 -0800)]
Eliminate the unused hprof "unreachable" feature.  This code dates
back to the Perforce repository and has never been enabled.  Nobody
seems to know what it was once, if ever, used for so it may as well
go.

14 years agoMerge "Rearrange JDWP cleanup."
Andy McFadden [Thu, 25 Feb 2010 22:12:32 +0000 (14:12 -0800)]
Merge "Rearrange JDWP cleanup."

14 years agoRearrange JDWP cleanup.
Andy McFadden [Thu, 25 Feb 2010 20:31:04 +0000 (12:31 -0800)]
Rearrange JDWP cleanup.

This moves the "get rid of the breakpoints" part before the "tell the
rest of the VM that there's no debugger attached" part.  Hitting a
breakpoint when there's no debugger attached is not expected.

For bug 2461667.

14 years agoMerge "Keep the order from getaddrinfo() unchanged if preferIPv6Addresses() is true."
Elliott Hughes [Thu, 25 Feb 2010 16:22:52 +0000 (08:22 -0800)]
Merge "Keep the order from getaddrinfo() unchanged if preferIPv6Addresses() is true."

14 years agoMerge "Fix a test by syncing with upstream." into dalvik-dev
Elliott Hughes [Thu, 25 Feb 2010 16:17:58 +0000 (08:17 -0800)]
Merge "Fix a test by syncing with upstream." into dalvik-dev

14 years agoPrint the self thread id and the owning thread id in the description
Carl Shapiro [Wed, 24 Feb 2010 10:30:55 +0000 (02:30 -0800)]
Print the self thread id and the owning thread id in the description
of the IllegalMonitorExcpetion throw by unlockMonitor.

14 years agoPrint the self thread id and the owning thread id in the description
Carl Shapiro [Wed, 24 Feb 2010 10:30:55 +0000 (02:30 -0800)]
Print the self thread id and the owning thread id in the description
of the IllegalMonitorExcpetion throw by unlockMonitor.

14 years agoMerge "Allocate an object and mark bitmap which span the entire virtual memory reserv...
Carl Shapiro [Thu, 25 Feb 2010 02:40:54 +0000 (18:40 -0800)]
Merge "Allocate an object and mark bitmap which span the entire virtual memory reservation of the heap.  Eliminates indirection during the marking phase." into dalvik-dev

14 years agoMerge "Print "JIT" in the thread dump if the top frame is in JIT'ed code."
Ben Cheng [Thu, 25 Feb 2010 02:34:25 +0000 (18:34 -0800)]
Merge "Print "JIT" in the thread dump if the top frame is in JIT'ed code."

14 years agoMerge "Added printf attribute to exception function."
Andy McFadden [Thu, 25 Feb 2010 02:28:14 +0000 (18:28 -0800)]
Merge "Added printf attribute to exception function."

14 years agoFix a test by syncing with upstream.
Elliott Hughes [Thu, 25 Feb 2010 02:16:28 +0000 (18:16 -0800)]
Fix a test by syncing with upstream.

This test has been failing for us because of a local modification.

14 years agoPrint "JIT" in the thread dump if the top frame is in JIT'ed code.
Ben Cheng [Thu, 25 Feb 2010 01:27:01 +0000 (17:27 -0800)]
Print "JIT" in the thread dump if the top frame is in JIT'ed code.

14 years agoIf we're in an Android build tree, we should use core.jar as our "SDK".
Elliott Hughes [Thu, 25 Feb 2010 00:58:08 +0000 (16:58 -0800)]
If we're in an Android build tree, we should use core.jar as our "SDK".

This fixes running junit tests for non-public classes, since bdc's --sdk
change went in. While --sdk is perfect for real people, it's probably not
what we want to use ourselves. Since this script is already hard-coded
for use in an Android build tree, this seems like a reasonable place to
apply the fix.

14 years agoOptimize more easy multiplications by constants.
Elliott Hughes [Thu, 25 Feb 2010 00:36:18 +0000 (16:36 -0800)]
Optimize more easy multiplications by constants.

Rather than make these changes in the libraries (*10 being a common case),
let's do them once and for all in the JIT.

The 2^n-1 case could be better if we generated RSB instructions, but the
current "fake" RSB is still better than a full multiply.

Thumb doesn't support reg/reg/reg/shift instructions, so we can't optimize
the "population count <= 2" cases (such as *10) there.

Tested on sholes, passion, and passion-running-sapphire (and visually
inspected to check we weren't trying to generate Thumb2 instructions there).
Also tested with the self-verifier.

14 years agoAllocate an object and mark bitmap which span the entire virtual
Carl Shapiro [Fri, 19 Feb 2010 08:46:33 +0000 (00:46 -0800)]
Allocate an object and mark bitmap which span the entire virtual
memory reservation of the heap.  Eliminates indirection during the
marking phase.

14 years agoEnhance the jit profiler to print more statistics and be more verbose.
Ben Cheng [Wed, 24 Feb 2010 23:00:40 +0000 (15:00 -0800)]
Enhance the jit profiler to print more statistics and be more verbose.

Specifically, the per-trace invocation percentage is printed now.

D/dalvikvm(  671): TRACEPROFILE 0x42b6eed8   15449863  2.13% [0(+2), 253] Ljava/nio/Buffer;position;()I
D/dalvikvm(  671): TRACEPROFILE 0x42b7b3ac   15410023  2.13% [0x4(+3), 450] Lorg/apache/xml/serializer/CharInfo;shouldMapTextChar;(I)Z
D/dalvikvm(  671): TRACEPROFILE 0x42b7b2cc   15346582  2.12% [0(+2), 449] Lorg/apache/xml/serializer/CharInfo;shouldMapTextChar;(I)Z
D/dalvikvm(  671): TRACEPROFILE 0x42b7b40c   15013650  2.07% [0xb4(+2), 1527] Lorg/apache/xml/serializer/ToStream;characters;([CII)V
D/dalvikvm(  671): TRACEPROFILE 0x42b7b4f0   14907957  2.06% [0xc7(+2), 1537] Lorg/apache/xml/serializer/ToStream;characters;([CII)V
D/dalvikvm(  671): TRACEPROFILE 0x42b7b308   14898588  2.06% [0xad(+3), 1525] Lorg/apache/xml/serializer/ToStream;characters;([CII)V
D/dalvikvm(  671): TRACEPROFILE 0x42b7b52c   14798762  2.04% [0xc4(+2), 1523] Lorg/apache/xml/serializer/ToStream;characters;([CII)V
D/dalvikvm(  671): TRACEPROFILE 0x42b7b570   14763770  2.04% [0xf3(+2), 1577] Lorg/apache/xml/serializer/ToStream;characters;([CII)V

And the top 10 traces will be recompiled in verbose mode to facilitate code
quality analysis.

14 years agoam cf4e8b02: Merge "This is a contribution of x86-atom targeted assembly for the...
Andy McFadden [Wed, 24 Feb 2010 19:18:33 +0000 (11:18 -0800)]
am cf4e8b02: Merge "This is a contribution of x86-atom targeted assembly for the fast byte-code interpreter engine. This is an initial contribution with minimal optimizations that target the Intel ATOM processor. We expect to continuously improve this code. It is expe

Merge commit 'cf4e8b02ecfb5ec17a27ed4188dd7a9db14ab7bf'

* commit 'cf4e8b02ecfb5ec17a27ed4188dd7a9db14ab7bf':
  This is a contribution of x86-atom targeted assembly for the fast byte-code interpreter engine. This is an initial contribution with minimal optimizations that target the Intel ATOM processor. We expect to continuously improve this code. It is expected that there will be a discussion on the potential merge of this code and similar efforts (i.e. the mterp/x86 directory first included with cupcake). While this code is intended to target ATOM and not a generic X-86 processor, we were able to show the following improvements over the c-portable interpreter using the simulator build:

14 years agoClarify a few items in the dex spec.
Dan Bornstein [Wed, 24 Feb 2010 18:54:08 +0000 (10:54 -0800)]
Clarify a few items in the dex spec.

In particular, do a bit of "may" vs. "must" maintenance, and be specific
about references as arguments to filled-new-array.

Change-Id: Iae4cb734e05cd29230e2c9a343108bbfb3c20193

14 years agoAdded printf attribute to exception function.
Andy McFadden [Wed, 24 Feb 2010 17:39:52 +0000 (09:39 -0800)]
Added printf attribute to exception function.

14 years agoMerge "Add class init stats to alloc counters (API change)."
Andy McFadden [Wed, 24 Feb 2010 17:39:09 +0000 (09:39 -0800)]
Merge "Add class init stats to alloc counters (API change)."

14 years agoRemove a bogus assertion.
Carl Shapiro [Wed, 24 Feb 2010 10:44:53 +0000 (02:44 -0800)]
Remove a bogus assertion.