OSDN Git Service

android-x86/art.git
9 years agoMerge "Enable multi-dex content per file in dexlist/dump."
Aart Bik [Fri, 10 Jul 2015 17:25:39 +0000 (17:25 +0000)]
Merge "Enable multi-dex content per file in dexlist/dump."

9 years agoMerge "Revert "Fuse long and FP compare & condition on ARM64 in Optimizing.""
Roland Levillain [Fri, 10 Jul 2015 17:19:29 +0000 (17:19 +0000)]
Merge "Revert "Fuse long and FP compare & condition on ARM64 in Optimizing.""

9 years agoRevert "Fuse long and FP compare & condition on ARM64 in Optimizing."
Roland Levillain [Fri, 10 Jul 2015 17:18:11 +0000 (17:18 +0000)]
Revert "Fuse long and FP compare & condition on ARM64 in Optimizing."

This reverts commit 5cfe61f27ed9203498169355bb95db756486d292.

Change-Id: I9879e76e7f8315cace05700e3b571a6a4749bf1a

9 years agoMerge "Fuse long and FP compare & condition on ARM64 in Optimizing."
Roland Levillain [Fri, 10 Jul 2015 15:10:46 +0000 (15:10 +0000)]
Merge "Fuse long and FP compare & condition on ARM64 in Optimizing."

9 years agoMerge "Fuse long and FP compare & condition on ARM in Optimizing."
Roland Levillain [Fri, 10 Jul 2015 10:51:14 +0000 (10:51 +0000)]
Merge "Fuse long and FP compare & condition on ARM in Optimizing."

9 years agoMerge "Initialize pDeoptimize quick entry pointer for Mips."
Roland Levillain [Fri, 10 Jul 2015 10:08:06 +0000 (10:08 +0000)]
Merge "Initialize pDeoptimize quick entry pointer for Mips."

9 years agoFuse long and FP compare & condition on ARM64 in Optimizing.
Roland Levillain [Fri, 10 Jul 2015 09:38:11 +0000 (10:38 +0100)]
Fuse long and FP compare & condition on ARM64 in Optimizing.

Bug: 21120453
Change-Id: I701e808600fb5ba9ff4d0f5e19e4ce22b1d34b29

9 years agoMerge "ART: Change merges with Undefined to Undefined"
Andreas Gampe [Fri, 10 Jul 2015 01:27:00 +0000 (01:27 +0000)]
Merge "ART: Change merges with Undefined to Undefined"

9 years agoMerge "Remove unused WorkStealing thread pool"
Mathieu Chartier [Fri, 10 Jul 2015 00:37:31 +0000 (00:37 +0000)]
Merge "Remove unused WorkStealing thread pool"

9 years agoRemove unused WorkStealing thread pool
Mathieu Chartier [Thu, 9 Jul 2015 23:58:42 +0000 (16:58 -0700)]
Remove unused WorkStealing thread pool

Bug: 10141986
Change-Id: Id2042320afc91b34b6203e82fb56424281c8d64b

9 years agoMerge "Clean up some includes"
Mathieu Chartier [Thu, 9 Jul 2015 23:43:48 +0000 (23:43 +0000)]
Merge "Clean up some includes"

9 years agoClean up some includes
Mathieu Chartier [Thu, 9 Jul 2015 22:15:09 +0000 (15:15 -0700)]
Clean up some includes

Change-Id: Ia03a3b54b235df38d5cfe096fef1aebe2b80eb29

9 years agoMerge "ART: Improve the symbolize script"
Andreas Gampe [Thu, 9 Jul 2015 21:50:13 +0000 (21:50 +0000)]
Merge "ART: Improve the symbolize script"

9 years agoART: Improve the symbolize script
Andreas Gampe [Thu, 9 Jul 2015 21:39:56 +0000 (14:39 -0700)]
ART: Improve the symbolize script

Allow a list of files to pull. This improves turnaround time when
tracking down perf hotspots.

Bug: 21760614

(cherry picked from commit 6f0a9c305b3e51b852fa9f188590076d6612864e)

Change-Id: I824e875053940d59eb00f9d796ef402632905c39

9 years agoART: Change merges with Undefined to Undefined
Andreas Gampe [Thu, 9 Jul 2015 18:30:14 +0000 (11:30 -0700)]
ART: Change merges with Undefined to Undefined

The result of a merge with an Undefined type should be Undefined.
Conflicts are allowed to be copied around, but Undefined registers
should not be touched at all, except to be written into.

Add a success test case (the register isn't used) and a fail test
case (the register is tried to be copied).

Bug: 22331663
Bug: 22371999
Change-Id: I9697ce31c1d2ab5aee0433dcf1253bcca79c2983

9 years agoEnable multi-dex content per file in dexlist/dump.
Aart Bik [Thu, 9 Jul 2015 18:45:28 +0000 (11:45 -0700)]
Enable multi-dex content per file in dexlist/dump.

Rationale:
Unlike the original Dalvik-based utilities, the Art-based
utilities are able to extract multi-dex content from one file.
So no need for any artificial restriction on number of dex files.

Bug: 22322814
Change-Id: I7c603d22c7e0caa5ae712b905d9bb83982f1b918

9 years agoMerge "ART: Fix test"
Andreas Gampe [Thu, 9 Jul 2015 18:29:36 +0000 (18:29 +0000)]
Merge "ART: Fix test"

9 years agoART: Fix test
Andreas Gampe [Thu, 9 Jul 2015 18:01:07 +0000 (11:01 -0700)]
ART: Fix test

Fixes a breakage from 38536287f61c9c0fc3bab8c1950cf8c74881482a. Ensure
that v4 is not undefined when an exception may be thrown (at the
new-instance).

Bug: 22331663
Bug: 22371999
Change-Id: Ice36ddda5ca92ea32fa88523dd616d4f2a61027e

9 years agoMerge "Fix makefile test against host os."
Calin Juravle [Thu, 9 Jul 2015 14:52:13 +0000 (14:52 +0000)]
Merge "Fix makefile test against host os."

9 years agoFix makefile test against host os.
Calin Juravle [Thu, 9 Jul 2015 14:48:03 +0000 (15:48 +0100)]
Fix makefile test against host os.

Change-Id: I7637845277012a189bf5f3598f9f0b1f599fbf75

9 years agoMerge "ART: Fix bug in GraphBuilder"
David Brazdil [Thu, 9 Jul 2015 14:42:05 +0000 (14:42 +0000)]
Merge "ART: Fix bug in GraphBuilder"

9 years agoART: Fix bug in GraphBuilder
David Brazdil [Thu, 9 Jul 2015 13:36:05 +0000 (14:36 +0100)]
ART: Fix bug in GraphBuilder

This fixes a bug where the GraphBuilder would split a throwing catch
block but would not update info about which blocks throw.

Change-Id: If5415f0c320aa488e06eb042e8fea6f03e30246a

9 years agoMerge "ART: Add regression test"
David Brazdil [Thu, 9 Jul 2015 10:37:09 +0000 (10:37 +0000)]
Merge "ART: Add regression test"

9 years agoART: Add regression test
David Brazdil [Wed, 8 Jul 2015 18:13:19 +0000 (19:13 +0100)]
ART: Add regression test

This adds a test for CL I624c0f91676d9baaada6f33be9d7091f68d57535.

Change-Id: Ib9ea1400082ad03c4355983d95268e084c0b6a6e

9 years agoMerge "Support compiling run-tests with jack"
Sebastien Hertz [Thu, 9 Jul 2015 09:11:35 +0000 (09:11 +0000)]
Merge "Support compiling run-tests with jack"

9 years agoMerge "ART: Allow to set and copy conflicts in the verifier"
Andreas Gampe [Thu, 9 Jul 2015 03:02:05 +0000 (03:02 +0000)]
Merge "ART: Allow to set and copy conflicts in the verifier"

9 years agoART: Allow to set and copy conflicts in the verifier
Andreas Gampe [Thu, 9 Jul 2015 00:22:57 +0000 (17:22 -0700)]
ART: Allow to set and copy conflicts in the verifier

As long as conflicts are not actively being used, it is OK to set
them in a register line or copy them around.

Bug: 22331663
Change-Id: I61999e2d9c92f9bdedcb0a5dea54df3c5bb130ca

9 years agoMerge "Fix potential bugs in allocation tracker"
Mathieu Chartier [Wed, 8 Jul 2015 21:46:48 +0000 (21:46 +0000)]
Merge "Fix potential bugs in allocation tracker"

9 years agoMerge "Have test-art-host-dexdump specifically depend on the host binaries."
Ying Wang [Wed, 8 Jul 2015 21:46:14 +0000 (21:46 +0000)]
Merge "Have test-art-host-dexdump specifically depend on the host binaries."

9 years agoHave test-art-host-dexdump specifically depend on the host binaries.
Ying Wang [Wed, 8 Jul 2015 21:09:10 +0000 (14:09 -0700)]
Have test-art-host-dexdump specifically depend on the host binaries.

dexdump2 and dexlist2 are phony targets representing both the host and
target binaries.
Also set environmental variable ANDROID_HOST_OUT=$(realpath $(HOST_OUT))
before calling art/test/dexdump/run-all-tests.
art/test/dexdump/run-all-tests references ANDROID_HOST_OUT, which may be
not set up by the user if he didn't run build/envsetup.sh.

Bug: 22343844
Change-Id: I38b5f3efccd5817d7d9ff0e91b5c6d070b546d77

9 years agoMerge "Revert "Make Art dexdump/list the real utilities.""
Aart Bik [Wed, 8 Jul 2015 21:21:15 +0000 (21:21 +0000)]
Merge "Revert "Make Art dexdump/list the real utilities.""

9 years agoMerge "Use trap + async timeout for tests"
Mathieu Chartier [Wed, 8 Jul 2015 21:21:55 +0000 (21:21 +0000)]
Merge "Use trap + async timeout for tests"

9 years agoRevert "Make Art dexdump/list the real utilities."
Aart Bik [Wed, 8 Jul 2015 21:20:13 +0000 (21:20 +0000)]
Revert "Make Art dexdump/list the real utilities."

This reverts commit d10ee1ce1ccd92c003ec49c15edf339e74c13894.

REASON:
win_sdk is not happy with my migration :-(

Change-Id: I2ed196da3f5883112883525f883d25b36e48352a

9 years agoUse trap + async timeout for tests
Mathieu Chartier [Wed, 8 Jul 2015 21:03:02 +0000 (14:03 -0700)]
Use trap + async timeout for tests

Allows tests to be killed with ctrl-c.

Bug: 19414810
Change-Id: I09fbee370e71386ed48b814407868c2fe7a0b058

9 years agoMerge "ART: Avoid soft- after hard-fail in verifier"
Andreas Gampe [Wed, 8 Jul 2015 20:56:40 +0000 (20:56 +0000)]
Merge "ART: Avoid soft- after hard-fail in verifier"

9 years agoMerge "Make Art dexdump/list the real utilities."
Aart Bik [Wed, 8 Jul 2015 20:10:39 +0000 (20:10 +0000)]
Merge "Make Art dexdump/list the real utilities."

9 years agoFix potential bugs in allocation tracker
Man Cao [Tue, 7 Jul 2015 01:53:15 +0000 (18:53 -0700)]
Fix potential bugs in allocation tracker

Add a missing visit allocation records call in concurrent copying collecter.
Handle null class objects if we support class unloading, and issues
with disallow and allow new allocation records.

Bug: 20037135
Change-Id: I59b7321c281e0d79a620501b2f43e36d2a576203

9 years agoMake Art dexdump/list the real utilities.
Aart Bik [Wed, 8 Jul 2015 18:30:04 +0000 (11:30 -0700)]
Make Art dexdump/list the real utilities.

Note: depends on pending CL that removes
      old dexdump/list and libdex.

Bug: 22322814
Change-Id: I99259553f0f4655a633989feb8d67aa11de8b29d

9 years agoMerge "Address small comments in HGraphBuilder"
David Brazdil [Wed, 8 Jul 2015 18:02:56 +0000 (18:02 +0000)]
Merge "Address small comments in HGraphBuilder"

9 years agoAddress small comments in HGraphBuilder
David Brazdil [Wed, 8 Jul 2015 15:44:00 +0000 (16:44 +0100)]
Address small comments in HGraphBuilder

Change-Id: Ife0a19f6e07146c8e03922f1330527d092f5e236

9 years agoMerge "Revert "Revert "ART: Ignore try blocks with no throwing instructions"""
David Brazdil [Wed, 8 Jul 2015 18:00:22 +0000 (18:00 +0000)]
Merge "Revert "Revert "ART: Ignore try blocks with no throwing instructions"""

9 years agoMerge "Migrated dexlist from Dalvik (libdex) into Art (libart)"
Aart Bik [Wed, 8 Jul 2015 17:47:23 +0000 (17:47 +0000)]
Merge "Migrated dexlist from Dalvik (libdex) into Art (libart)"

9 years agoRevert "Revert "ART: Ignore try blocks with no throwing instructions""
David Brazdil [Wed, 8 Jul 2015 17:26:51 +0000 (17:26 +0000)]
Revert "Revert "ART: Ignore try blocks with no throwing instructions""

The original CL broke libcore tests because monitor-exit instructions
did not have any side-effects and got removed by DCE once not labelled
throwing any more.

This reverts commit efe374d7c25c1d48945a9198d96469de99e0c1bd.

Change-Id: I624c0f91676d9baaada6f33be9d7091f68d57535

9 years agoMerge "Revert "ART: Ignore try blocks with no throwing instructions""
David Brazdil [Wed, 8 Jul 2015 16:58:19 +0000 (16:58 +0000)]
Merge "Revert "ART: Ignore try blocks with no throwing instructions""

9 years agoRevert "ART: Ignore try blocks with no throwing instructions"
David Brazdil [Wed, 8 Jul 2015 16:58:07 +0000 (16:58 +0000)]
Revert "ART: Ignore try blocks with no throwing instructions"

Turns out monitor-exit *can* throw... Need to investigate

This reverts commit 8f8ee680bec71a28d9d7b7538e8c7ca100a18184.

Change-Id: I8b42690918833c917b6a7fc3ceea932b7c1a6f15

9 years agoMerge "Revert "Use the object class as top in reference type propagation""
Calin Juravle [Wed, 8 Jul 2015 15:57:32 +0000 (15:57 +0000)]
Merge "Revert "Use the object class as top in reference type propagation""

9 years agoRevert "Use the object class as top in reference type propagation"
Calin Juravle [Wed, 8 Jul 2015 15:57:18 +0000 (15:57 +0000)]
Revert "Use the object class as top in reference type propagation"

failing on the build bot on some targets but not locally. needs more investigation.

This reverts commit 20e6071362b84a9782b633a893c29ebde458205e.

Change-Id: I6965483f569fb862f9bdb66d459b747ded54de71

9 years agoMerge "ART: Release inputs in Long.reverse intrinsic in x86"
Andreas Gampe [Wed, 8 Jul 2015 15:50:00 +0000 (15:50 +0000)]
Merge "ART: Release inputs in Long.reverse intrinsic in x86"

9 years agoART: Release inputs in Long.reverse intrinsic in x86
Andreas Gampe [Tue, 7 Jul 2015 20:25:58 +0000 (13:25 -0700)]
ART: Release inputs in Long.reverse intrinsic in x86

In the worst case we are using two temps each for input and output.
Then we do not have a temp left over for the swap operations. The
input is dead, however, after the first swap. So try to release it
(a no-op if it isn't actually a temp).

Bug: 22324327
Change-Id: I1fc50159afdad14160e34abeaf4670958171d6b2

9 years agoMerge "ART: Ignore try blocks with no throwing instructions"
David Brazdil [Wed, 8 Jul 2015 15:23:53 +0000 (15:23 +0000)]
Merge "ART: Ignore try blocks with no throwing instructions"

9 years agoART: Ignore try blocks with no throwing instructions
David Brazdil [Wed, 8 Jul 2015 11:11:42 +0000 (12:11 +0100)]
ART: Ignore try blocks with no throwing instructions

In order to avoid complex removal of redundant exceptional edges in
the SSA builder, this patch modified the graph builder to consider
blocks without throwing instructions as not in a try block, even if
covered by a TryItem.

In some corner cases, this may generate more TryBoundaries than
necessary, but those can be removed once the SSA form is built.

Change-Id: I158c4542b2c1964a8dd532f82e921b9cb1997e1e

9 years agoSupport compiling run-tests with jack
Sebastien Hertz [Tue, 24 Feb 2015 16:39:50 +0000 (17:39 +0100)]
Support compiling run-tests with jack

This CL adds support to compile run-test source files with jack. When
a test needs to rely on class files, we use jill to convert them to a
jack library.

We need to pass the full classpath to jack containing at least core
classes (like java.lang.Object). This means the Android tree must
have been compiled with jack first so we find all the necessary
classes.jack files.

Some tests still rely on dex files generated with the old toolchain.
We keep building them this way for the moment and will update them
later, when they get ready for Jack.

Also updates a few tests dealing with garbage collection to avoid a
situation where a reference can be retained by a local DEX register.

Bug: 19467889
Change-Id: I9eedd0705d1186d28a2acd37ea42a1762cd0ace2

9 years agoMerge "Use the object class as top in reference type propagation"
Calin Juravle [Wed, 8 Jul 2015 15:00:00 +0000 (15:00 +0000)]
Merge "Use the object class as top in reference type propagation"

9 years agoMerge "Fix oatdump after ArtMethod change."
Nicolas Geoffray [Wed, 8 Jul 2015 14:59:35 +0000 (14:59 +0000)]
Merge "Fix oatdump after ArtMethod change."

9 years agoFix oatdump after ArtMethod change.
Nicolas Geoffray [Wed, 8 Jul 2015 14:55:28 +0000 (15:55 +0100)]
Fix oatdump after ArtMethod change.

ArtMethod is now word aligned, and not necessarily 8 byte aligned.

Change-Id: I8bd6b2f156d7f2ab191d6f2675b013713bbf258d

9 years agoFuse long and FP compare & condition on ARM in Optimizing.
Roland Levillain [Mon, 6 Jul 2015 17:11:54 +0000 (18:11 +0100)]
Fuse long and FP compare & condition on ARM in Optimizing.

Also:
- Stylistic changes in corresponding parts on the x86 and
  x86-64 code generators.
- Update and improve the documentation of
  art::arm::Condition.

Bug: 21120453
Change-Id: If144772046e7d21362c3c2086246cb7d011d49ce

9 years agoMerge "Revert "Revert "Remove interpreter entrypoint in ArtMethod."""
Nicolas Geoffray [Wed, 8 Jul 2015 13:52:41 +0000 (13:52 +0000)]
Merge "Revert "Revert "Remove interpreter entrypoint in ArtMethod."""

9 years agoRevert "Revert "Remove interpreter entrypoint in ArtMethod.""
Nicolas Geoffray [Wed, 8 Jul 2015 10:11:59 +0000 (10:11 +0000)]
Revert "Revert "Remove interpreter entrypoint in ArtMethod.""

The start of the interned strings in the image was not aligned
properly, now that ArtMethods just need to be word aligned.

This reverts commit 7070ccd8b6439477eafeea7ed3736645d78e003f.

bug:22242193

Change-Id: I580c23310c33c239fe0e5d15c72f23a936f58ed1

9 years agoUse the object class as top in reference type propagation
Calin Juravle [Wed, 1 Jul 2015 17:41:04 +0000 (18:41 +0100)]
Use the object class as top in reference type propagation

This properly types all instructions, making it safe to query the type
at any time.

Change-Id: I3ee2f0f79253cdf45b10ddab37ecb473345ca53a

9 years agoMerge "Accept 0xFFFFFFFF as dex debug info offset."
David Srbecky [Wed, 8 Jul 2015 11:08:32 +0000 (11:08 +0000)]
Merge "Accept 0xFFFFFFFF as dex debug info offset."

9 years agoMerge "Fuse long and FP compare & condition on x86/x86-64 in Optimizing."
Roland Levillain [Wed, 8 Jul 2015 10:36:47 +0000 (10:36 +0000)]
Merge "Fuse long and FP compare & condition on x86/x86-64 in Optimizing."

9 years agoMerge "Revert "Remove interpreter entrypoint in ArtMethod.""
Nicolas Geoffray [Wed, 8 Jul 2015 09:42:05 +0000 (09:42 +0000)]
Merge "Revert "Remove interpreter entrypoint in ArtMethod.""

9 years agoRevert "Remove interpreter entrypoint in ArtMethod."
Nicolas Geoffray [Wed, 8 Jul 2015 09:41:54 +0000 (09:41 +0000)]
Revert "Remove interpreter entrypoint in ArtMethod."

Build failures on bots. Investigating.

This reverts commit fa2c054b28d4b540c1b3651401a7a091282a015f.

Change-Id: Id65b2009aa66cb291fb8c39758a58e0b0d22616c

9 years agoMerge "Remove interpreter entrypoint in ArtMethod."
Nicolas Geoffray [Wed, 8 Jul 2015 08:56:49 +0000 (08:56 +0000)]
Merge "Remove interpreter entrypoint in ArtMethod."

9 years agoRemove interpreter entrypoint in ArtMethod.
Nicolas Geoffray [Wed, 1 Jul 2015 13:32:54 +0000 (14:32 +0100)]
Remove interpreter entrypoint in ArtMethod.

Saves 4/8 bytes for each ArtMethod.

Change-Id: I110ecdddf8516b0759a31fa157609643e6d60b15

9 years agoMigrated dexlist from Dalvik (libdex) into Art (libart)
Aart Bik [Wed, 8 Jul 2015 00:09:41 +0000 (17:09 -0700)]
Migrated dexlist from Dalvik (libdex) into Art (libart)

Rationale:
The new dexlist (temporarily called dexlist2 until we are
satisfied with the migration) is a re-implementation of the
original dexlist utility that was based on Dalvik functions
in libdex into a new dexlist that is now based on Art functions
in libart instead.

Bug: 22322814
Change-Id: I0f034d66c1a1f3c5450423f57fcdc23e2f5a1e29

9 years agoInitialize pDeoptimize quick entry pointer for Mips.
Douglas Leung [Wed, 8 Jul 2015 00:27:34 +0000 (17:27 -0700)]
Initialize pDeoptimize quick entry pointer for Mips.

This fixes art test 449-checker-bce.

Change-Id: I9a87c430168ea35984289b0b4a417aa0c659b6d5

9 years agoAccept 0xFFFFFFFF as dex debug info offset.
David Srbecky [Tue, 7 Jul 2015 18:13:29 +0000 (19:13 +0100)]
Accept 0xFFFFFFFF as dex debug info offset.

Bug: 22023362
Change-Id: I03e5c74f5ece4dbe1fab37d8b5f857582b04f57c

9 years agoMerge "Build rule for dexdump regression test on host."
Aart Bik [Tue, 7 Jul 2015 17:52:34 +0000 (17:52 +0000)]
Merge "Build rule for dexdump regression test on host."

9 years agoMerge "ART: Fix opsize in LoadArgDirect"
Andreas Gampe [Tue, 7 Jul 2015 16:46:06 +0000 (16:46 +0000)]
Merge "ART: Fix opsize in LoadArgDirect"

9 years agoART: Fix opsize in LoadArgDirect
Andreas Gampe [Mon, 6 Jul 2015 21:00:39 +0000 (14:00 -0700)]
ART: Fix opsize in LoadArgDirect

If the destination register is a reference, use kReference for the
op size.

Bug: 22244733
Change-Id: Idf52f2ee4c65b5dc41cb66257d95281dc6f32255

9 years agoFuse long and FP compare & condition on x86/x86-64 in Optimizing.
Mark Mendell [Fri, 10 Apr 2015 17:18:51 +0000 (13:18 -0400)]
Fuse long and FP compare & condition on x86/x86-64 in Optimizing.

This is a preliminary implementation of fusing long/float/double
compares with conditions to avoid materializing the result from the
compare and condition.

The information from a HCompare is transferred to the HCondition if it
is legal.  There must be only a single use of the HCompare, the HCompare
and HCondition must be in the same block, the HCondition must not need
materialization.

Added GetOppositeCondition() to HCondition to return the flipped
condition.

Bug: 21120453
Change-Id: I1f1db206e6dc336270cd71070ed3232dedc754d6
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
9 years agoMerge "Improve code generation for ARM64 VisitArrayGet/Set."
Nicolas Geoffray [Tue, 7 Jul 2015 10:48:12 +0000 (10:48 +0000)]
Merge "Improve code generation for ARM64 VisitArrayGet/Set."

9 years agoImprove code generation for ARM64 VisitArrayGet/Set.
Alexandre Rames [Tue, 7 Jul 2015 10:34:16 +0000 (11:34 +0100)]
Improve code generation for ARM64 VisitArrayGet/Set.

We prefer the code sequence
   add temp, obj, #offset
   ldr out, [temp, index LSL #shift_amount]
to
   add temp, obj, index LSL #shift_amount
   ldr out, [temp, #offset]

Change-Id: I98f51a1b5a5ecd84c677d6dbd4c4bfc0f157f5e2

9 years agoMerge "Make dexdump2 compile for 64bit devices."
Nicolas Geoffray [Tue, 7 Jul 2015 09:13:48 +0000 (09:13 +0000)]
Merge "Make dexdump2 compile for 64bit devices."

9 years agoMake dexdump2 compile for 64bit devices.
Nicolas Geoffray [Tue, 7 Jul 2015 09:12:10 +0000 (10:12 +0100)]
Make dexdump2 compile for 64bit devices.

Change-Id: I4c990272b20062632d236671415cfb3dbb26c867

9 years agoMerge "Added next condition check to branch instruction"
Roland Levillain [Tue, 7 Jul 2015 09:09:06 +0000 (09:09 +0000)]
Merge "Added next condition check to branch instruction"

9 years agoMerge "Make oatdump work with .art files and code generated by Optimizing."
Roland Levillain [Tue, 7 Jul 2015 08:52:10 +0000 (08:52 +0000)]
Merge "Make oatdump work with .art files and code generated by Optimizing."

9 years agoBuild rule for dexdump regression test on host.
Aart Bik [Tue, 7 Jul 2015 00:45:30 +0000 (17:45 -0700)]
Build rule for dexdump regression test on host.

Bug: 17442393

Change-Id: I7795ba014e57080572769286cb3f8e3d4931e706

9 years agoART: Avoid soft- after hard-fail in verifier
Andreas Gampe [Tue, 7 Jul 2015 00:20:18 +0000 (17:20 -0700)]
ART: Avoid soft- after hard-fail in verifier

Follow-up to 414000ec4d728b5c85f8c6dee4f867fecde59b01.

Bug: 21886894
Bug: 22245124
Change-Id: I3728187bd8526cdf84f95b123e6fad5d6c5a8b3c

9 years agoMerge "Disable sanitization of dex2oat on device."
Evgenii Stepanov [Mon, 6 Jul 2015 23:20:48 +0000 (23:20 +0000)]
Merge "Disable sanitization of dex2oat on device."

9 years agoDisable sanitization of dex2oat on device.
Evgenii Stepanov [Mon, 6 Jul 2015 21:01:25 +0000 (14:01 -0700)]
Disable sanitization of dex2oat on device.

Bug: 22233158
Change-Id: I54990bb167a65275ed28e0cbbab56b936eb70ca4

9 years agoMerge "List dependence of dexdump gtest on dexdump binary."
Aart Bik [Mon, 6 Jul 2015 22:24:11 +0000 (22:24 +0000)]
Merge "List dependence of dexdump gtest on dexdump binary."

9 years agoList dependence of dexdump gtest on dexdump binary.
Aart Bik [Mon, 6 Jul 2015 21:27:27 +0000 (14:27 -0700)]
List dependence of dexdump gtest on dexdump binary.

Change-Id: I1293009a370eff808b5e31cc89110eb9fe3ce32f

9 years agoMerge "Migrated dexdump from Dalvik (libdex) into Art (libart)"
Aart Bik [Mon, 6 Jul 2015 20:25:55 +0000 (20:25 +0000)]
Merge "Migrated dexdump from Dalvik (libdex) into Art (libart)"

9 years agoMigrated dexdump from Dalvik (libdex) into Art (libart)
Aart Bik [Wed, 1 Jul 2015 21:52:26 +0000 (14:52 -0700)]
Migrated dexdump from Dalvik (libdex) into Art (libart)

Rationale:
The new dexdump (temporarily called dexdump2 until we are
satisfied with the migration) is a re-implementation of the
original dexdump utility that was based on Dalvik functions
in libdex into a new dexdump that is now based on Art functions
in libart instead.

The output is identical to the original for *correct* DEX files.
Output in error messages and the usage() may differ, however,
since the new utility relies on Art parsing and verification.

NOTE 1:
ODEX files are no longer supported.

NOTE 2:
Where possible, I kept the file as close to the original
as possible, including some archaic C idioms on memory
allocation; those can be improved over time.

NOTE 3:
I used the standard Android.mk format for the new dexdump,
but this probably needs to be Art-i-fied.

NOTE 4:
Some minor issues that need resolution are marked with a TODO.

Bug: 17442393
Change-Id: I753743f64afcf4b84b8d33efbd1cfcb7908f0c3e

9 years agoMerge "Revert "Revert "Fix LSRA bug with explicit register temporaries"""
Nicolas Geoffray [Mon, 6 Jul 2015 15:28:24 +0000 (15:28 +0000)]
Merge "Revert "Revert "Fix LSRA bug with explicit register temporaries"""

9 years agoMerge "ART: Check switch and array data better in the verifier"
Andreas Gampe [Mon, 6 Jul 2015 15:25:27 +0000 (15:25 +0000)]
Merge "ART: Check switch and array data better in the verifier"

9 years agoART: Check switch and array data better in the verifier
Andreas Gampe [Thu, 2 Jul 2015 05:05:59 +0000 (22:05 -0700)]
ART: Check switch and array data better in the verifier

Incorrect padding (or just malformed code) can lead to pseudo
instructions for data payload not being correctly discoverable in
a linear scan. This is problematic for the optimizing compiler,
which uses such a scan to build its graph.

Use the opcode flag in the insn_flags_ to ensure that a pseudo
instruction referenced from a base instruction was actually
correctly discovered. This does not ensure the letter of the dex
specification for now, which calls for nops for padding. However,
it is unclear what the difference between unreachable code and
padding is.

Bug: 22237618
Bug: https://code.google.com/p/android/issues/detail?id=178522
Change-Id: I823b3f541f326e6e77b15f3e13a7608a39d4b1fb

9 years agoMerge "ART: Unlink old file in OS::CreateEmptyFile"
Andreas Gampe [Mon, 6 Jul 2015 15:21:45 +0000 (15:21 +0000)]
Merge "ART: Unlink old file in OS::CreateEmptyFile"

9 years agoMerge "Revert "ART: Unlink target oat file before compiling""
Andreas Gampe [Mon, 6 Jul 2015 15:21:35 +0000 (15:21 +0000)]
Merge "Revert "ART: Unlink target oat file before compiling""

9 years agoRevert "Revert "Fix LSRA bug with explicit register temporaries""
Nicolas Geoffray [Mon, 6 Jul 2015 15:12:53 +0000 (15:12 +0000)]
Revert "Revert "Fix LSRA bug with explicit register temporaries""

This reverts commit a5fc140ff315dda9bc0a8e59963ed547676cd941.

Change-Id: Ic322484176e55d0c7cd7250d629b9e5046006a4f

9 years agoMerge "Revert "Fix LSRA bug with explicit register temporaries""
Calin Juravle [Mon, 6 Jul 2015 15:10:41 +0000 (15:10 +0000)]
Merge "Revert "Fix LSRA bug with explicit register temporaries""

9 years agoRevert "Fix LSRA bug with explicit register temporaries"
Calin Juravle [Mon, 6 Jul 2015 15:09:54 +0000 (15:09 +0000)]
Revert "Fix LSRA bug with explicit register temporaries"

register_allocator_test32 fails.

This reverts commit 283b8541546e7673d33d104241623d07c91cf500.

Change-Id: I2a46f3c68de3e8273e402102065c13797045c481

9 years agoMerge "Fix LSRA bug with explicit register temporaries"
Nicolas Geoffray [Mon, 6 Jul 2015 13:51:54 +0000 (13:51 +0000)]
Merge "Fix LSRA bug with explicit register temporaries"

9 years agoFix LSRA bug with explicit register temporaries
Mark Mendell [Fri, 3 Jul 2015 12:26:41 +0000 (08:26 -0400)]
Fix LSRA bug with explicit register temporaries

A temporary with an explicit RegisterLocation, such as ESI on x86 didn't
have the register marked as allocated.  This caused it to not be
saved/restored in the prologue/epilogue, causing problems in the caller
routine, which expected it to be saved.  Found while implementing
https://android-review.googlesource.com/#/c/157522/.

Change-Id: I22ca2b24c2d21b1c6ab6cfb7dec26cb38034a891
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
9 years agoMerge "ART: Fix graph for switch leaving a try block"
David Brazdil [Mon, 6 Jul 2015 09:29:27 +0000 (09:29 +0000)]
Merge "ART: Fix graph for switch leaving a try block"

9 years agoART: Fix graph for switch leaving a try block
David Brazdil [Fri, 3 Jul 2015 09:34:57 +0000 (10:34 +0100)]
ART: Fix graph for switch leaving a try block

Blocks that GraphBuilder creates for switch-case logic are given
a dex_pc of the branch targets they serve, while in fact they should
be considered part of the switch instruction itself and get its pc.
This caused the try/catch algorithm to either miss try boundaries or
create bogus edges.

This patch fixed the dex_pc of the switch-case blocks and modifies
the try/catch logic to iterate over all blocks as opposed to just
branch targets since multiple blocks can now cover the same dex_pc.

Change-Id: I30fe4f8db0647b869979197a3bc847cf212a7315

9 years agoMerge "Supress OsTest#test_xattr on the bots."
Narayan Kamath [Mon, 6 Jul 2015 08:49:50 +0000 (08:49 +0000)]
Merge "Supress OsTest#test_xattr on the bots."

9 years agoSupress OsTest#test_xattr on the bots.
Narayan Kamath [Fri, 3 Jul 2015 12:20:51 +0000 (13:20 +0100)]
Supress OsTest#test_xattr on the bots.

bug: 22258911
Change-Id: Iba57eb8660e0f46e2b152ca05f49f64a233e0f44