OSDN Git Service
Wojciech Staszkiewicz [Wed, 28 Sep 2016 20:36:59 +0000 (20:36 +0000)]
Merge "Gather divergences in single directory"
am:
1c3ca75e63
Change-Id: I1fb9c199ae801b6a3f5336b2ba7bcf9f6d82ab62
Treehugger Robot [Wed, 28 Sep 2016 20:22:28 +0000 (20:22 +0000)]
Merge "Gather divergences in single directory"
Wojciech Staszkiewicz [Tue, 27 Sep 2016 20:09:34 +0000 (13:09 -0700)]
Gather divergences in single directory
Gather all divergences in a single directory to make investigating
them easier.
Test: ./run_jfuzz_test_nightly.py
Change-Id: I0fcdfb2f99e33d8bce444e5d5bf5092f282f03b1
Igor Murashkin [Wed, 28 Sep 2016 18:17:50 +0000 (18:17 +0000)]
Merge "build: Fix golem build to exclude libartd.so from public.libraries.txt"
am:
6bee259767
Change-Id: I66761db635d71fa5b0d808e3b402a935f2da903a
Treehugger Robot [Wed, 28 Sep 2016 18:11:47 +0000 (18:11 +0000)]
Merge "build: Fix golem build to exclude libartd.so from public.libraries.txt"
Andreas Gampe [Wed, 28 Sep 2016 18:08:49 +0000 (18:08 +0000)]
Merge "ART: Dump referenced type in IRT overflows"
am:
d30e51ac25
Change-Id: Icae4a3a3bc24a24ba148ead726aa3823e106dd34
Treehugger Robot [Wed, 28 Sep 2016 18:01:40 +0000 (18:01 +0000)]
Merge "ART: Dump referenced type in IRT overflows"
Andreas Gampe [Wed, 28 Sep 2016 17:44:17 +0000 (17:44 +0000)]
Merge "ART: Factor out IMT from ArtMethod"
am:
7f7fb9c0dc
Change-Id: I578904513b435ad50eecc5b7ca676a6b52c45b86
Treehugger Robot [Wed, 28 Sep 2016 17:41:06 +0000 (17:41 +0000)]
Merge "ART: Factor out IMT from ArtMethod"
Andreas Gampe [Wed, 28 Sep 2016 01:45:02 +0000 (18:45 -0700)]
ART: Dump referenced type in IRT overflows
When the runtime dumps a reference table, e.g., when aborting for
an overflow, dump the type of stored referents for reference types
to aid in debugging leaks.
Bug:
31600693
Test: m test-art-host-gtest-reference_table_test
Change-Id: Ia892dc84ca8827dd93a8b75d6f571c392f94859c
(cherry picked from commit
280f32b095f55f24dc557f9a9067d223901214ce)
Igor Murashkin [Wed, 28 Sep 2016 16:04:07 +0000 (09:04 -0700)]
build: Fix golem build to exclude libartd.so from public.libraries.txt
Golem doesn't build libartd.so, so at runtime the JNI loading would fail
when trying to initialize the android library namespaces.
Test: build-art-target-golem
Bug:
31401609
Change-Id: I2ee988e84d0a87b3cd1338d29fc28a874905cccc
Andreas Gampe [Mon, 26 Sep 2016 19:04:26 +0000 (12:04 -0700)]
ART: Factor out IMT from ArtMethod
Move IMT Conflict Table to its own file. Move IMT index computation to
ImTable.
In preparation for code using this independent of specific ArtMethods.
Bug:
31594153
Test: m test-art-host
Change-Id: Ifd6e304bb86724c5284984c4655c43d3af357f84
Neil Fuller [Wed, 28 Sep 2016 12:33:01 +0000 (12:33 +0000)]
Merge "Add support for Executable.getParameters() metadata"
am:
1b6cf7fbca
Change-Id: Ifff7a9d2906cbfb5678357a111297b3f6d08b5dd
Neil Fuller [Wed, 28 Sep 2016 12:29:53 +0000 (12:29 +0000)]
Merge "Add support for Executable.getParameters() metadata"
Neil Fuller [Fri, 9 Sep 2016 13:24:51 +0000 (14:24 +0100)]
Add support for Executable.getParameters() metadata
java.lang.reflect.Executable native code to create Parameter
objects and supporting code for obtaining the system annotations
that hold the parameter metadata.
Bug:
30391692
Test: test-art-host and CtsLibcoreTestCases
Change-Id: I23d7e36014716967ce189fba5955cc5e064fe8d0
Nicolas Geoffray [Wed, 28 Sep 2016 08:27:42 +0000 (08:27 +0000)]
Merge "benchmark: Add jni benchmarks for use in golem"
am:
fd8339e58e
Change-Id: I0dee551a90754c31ebdaece9c19b27e317b0ad3b
Nicolas Geoffray [Wed, 28 Sep 2016 08:24:12 +0000 (08:24 +0000)]
Merge "benchmark: Add jni benchmarks for use in golem"
Wojciech Staszkiewicz [Wed, 28 Sep 2016 02:22:33 +0000 (02:22 +0000)]
Merge "Handle tests which fail with no methods compiled"
am:
161243e480
Change-Id: I574fbd4a6089fc873bfaee22ab8ceb336568b900
Wojciech Staszkiewicz [Wed, 28 Sep 2016 02:22:19 +0000 (02:22 +0000)]
Merge "Store bisection logs in test directory"
am:
36cbbde646
Change-Id: I17b58431a9850cc4546e7ee480ae577b224013b5
Treehugger Robot [Wed, 28 Sep 2016 02:19:49 +0000 (02:19 +0000)]
Merge "Handle tests which fail with no methods compiled"
Treehugger Robot [Wed, 28 Sep 2016 02:14:34 +0000 (02:14 +0000)]
Merge "Store bisection logs in test directory"
Hiroshi Yamauchi [Tue, 27 Sep 2016 23:59:40 +0000 (23:59 +0000)]
Merge "Fix a CC deadlock in 129-ThreadGetId."
am:
b737745c79
Change-Id: Ie8024f2735625f760e3d5ab063de97b86e91a4d6
Treehugger Robot [Tue, 27 Sep 2016 23:53:08 +0000 (23:53 +0000)]
Merge "Fix a CC deadlock in 129-ThreadGetId."
Wojciech Staszkiewicz [Tue, 27 Sep 2016 20:21:24 +0000 (20:21 +0000)]
Merge "Add a true divergence mode to run_jfuzz_test.py"
am:
05ce52b406
Change-Id: Ic9db912f397700212df5dae4ac9b9b946c4e4d9b
Treehugger Robot [Tue, 27 Sep 2016 20:12:21 +0000 (20:12 +0000)]
Merge "Add a true divergence mode to run_jfuzz_test.py"
Mathieu Chartier [Tue, 27 Sep 2016 18:26:16 +0000 (18:26 +0000)]
Merge "Add basic support for object pointer poisoning"
am:
83b13b32f3
Change-Id: I4d44d8e16303089d89351ac24a0600c80fa495d6
Mathieu Chartier [Tue, 27 Sep 2016 18:17:00 +0000 (18:17 +0000)]
Merge "Add basic support for object pointer poisoning"
Hiroshi Yamauchi [Mon, 26 Sep 2016 23:56:24 +0000 (16:56 -0700)]
Fix a CC deadlock in 129-ThreadGetId.
GC should consider a thread as "waiting for GC thread flip" or
"transitioning to runnable" only if its suspend count is 1.
See
31683379#7 for the deadlock scenario.
Bug:
31683379
Bug:
12687968
Test: test-art with CC and gcstress, N9 libartd boot, Ritz EAAC.
Change-Id: Icd2548bd6d9a4f8d7b54ed20150a4801af9e26a3
Igor Murashkin [Thu, 22 Sep 2016 22:46:29 +0000 (15:46 -0700)]
benchmark: Add jni benchmarks for use in golem
Also adds libartbenchmark.so as a build dependency for golem targets
Test: make build-art-host-golem build-art-target-golem
Bug:
31401609
Change-Id: I3f6b11f93684cc8eb3534bb02bf2d3eb05e39dcf
Andreas Gampe [Tue, 27 Sep 2016 15:18:05 +0000 (15:18 +0000)]
Merge "ART: Fix dependencies"
am:
38e1882858
Change-Id: If040142b21b56fe3e172e76a8d0efe9afd5afff6
Andreas Gampe [Tue, 27 Sep 2016 15:10:57 +0000 (15:10 +0000)]
Merge "ART: Fix dependencies"
Nicolas Geoffray [Tue, 27 Sep 2016 09:36:15 +0000 (09:36 +0000)]
Merge "Cleanup String.<init> handling."
am:
adc4e8297c
Change-Id: I96e1fd961cb7e93358ec53dfc8c8e03e3598f43e
Nicolas Geoffray [Tue, 27 Sep 2016 09:31:33 +0000 (09:31 +0000)]
Merge "Cleanup String.<init> handling."
Nicolas Geoffray [Tue, 27 Sep 2016 09:31:14 +0000 (09:31 +0000)]
Merge "Fix no-prebuild failure."
am:
fc916a6c9a
Change-Id: I034c417313e32207b71dcb219f6499bca7233cfb
Nicolas Geoffray [Mon, 26 Sep 2016 16:56:07 +0000 (17:56 +0100)]
Cleanup String.<init> handling.
Move everything to one place (currently well_known_classes.cc, but
no strong preference) and define a macro to easily handle the list
of affected methods.
test: m test-art-host
test: m test-art-target
Change-Id: Ib8372d130d5458516a1f1ae31014afc76037fc34
Nicolas Geoffray [Tue, 27 Sep 2016 09:25:58 +0000 (09:25 +0000)]
Merge "Fix no-prebuild failure."
Nicolas Geoffray [Tue, 27 Sep 2016 09:24:36 +0000 (10:24 +0100)]
Fix no-prebuild failure.
Printed executable is now dalvikvm* instead of art.
test:run-test --host 118
Change-Id: Id2c3a7f6ba90ee3cb4e57a49e366c46377fc4c45
Richard Uhler [Tue, 27 Sep 2016 07:25:35 +0000 (07:25 +0000)]
Merge "ahat: Show GC Root Paths."
am:
52591a6682
Change-Id: I672498e2f21801c26c1052386c1caa2b25ab68ee
Treehugger Robot [Tue, 27 Sep 2016 07:17:52 +0000 (07:17 +0000)]
Merge "ahat: Show GC Root Paths."
Andreas Gampe [Tue, 27 Sep 2016 03:17:42 +0000 (20:17 -0700)]
ART: Fix dependencies
Add libbase to libart-simulator.
Bug:
31338270
Test: m test-art-host
Change-Id: I956465f8d380e856cddf7f53c5aae358847b52ed
Andreas Gampe [Tue, 27 Sep 2016 02:55:48 +0000 (02:55 +0000)]
Merge "ART: Fix dependencies"
am:
2ac59375f8
Change-Id: I5fb1ac633b743e2a18ae3d93c1dcf22f32423366
Andreas Gampe [Tue, 27 Sep 2016 02:48:37 +0000 (02:48 +0000)]
Merge "ART: Fix dependencies"
Andreas Gampe [Tue, 27 Sep 2016 01:09:47 +0000 (18:09 -0700)]
ART: Fix dependencies
Add libbase to libart-compiler.
Bug:
31338270
Test: m test-art-host
Change-Id: I6398baaaee15063f4c275f40109d943c04af066f
Wojciech Staszkiewicz [Tue, 27 Sep 2016 00:51:52 +0000 (17:51 -0700)]
Store bisection logs in test directory
Store bisection search logs in test directory for easier access.
Test: ./test/run-test --never-clean --no-prebuild --bisection-search
037-inherit
Change-Id: I67e472b933516f0e486f47fd2c7a4ae06212db44
Wojciech Staszkiewicz [Tue, 27 Sep 2016 00:39:26 +0000 (17:39 -0700)]
Handle tests which fail with no methods compiled
Check if test passes with no methods compiled and only then bisect
it.
Test: ./bisection_search.py -cp classes.dex --expected-output out --class
Test
Change-Id: If63af7304ae124a1920ec2a3f57d33a6b1f71bb9
Andreas Gampe [Mon, 26 Sep 2016 23:38:36 +0000 (23:38 +0000)]
Merge "ART: Fix dependencies"
am:
fdaa803ba3
Change-Id: Ib7bbf19b3ed243120c8bd5e45e3bb3a23b3ce330
Mathieu Chartier [Mon, 26 Sep 2016 18:39:52 +0000 (11:39 -0700)]
Add basic support for object pointer poisoning
ObjPtr is a pointer that checks for heap corruption and is meant
to replace mirror::Object* in places where a mirror::Object* is a
local variable. Whenever there is a possible suspend point, the
current thread's object pointers are all invalidated. This is done
by storing a cookie in the object pointer associated with what thread
created it.
Added test case in object_test.
Example failure:
object_test F 25379 25379 object_pointer.h:70] Check failed:
IsValid() Invalid cookie, expected 0 but got 2
Bug:
31113334
Test: test-art-host-gtest-object_test
Change-Id: I9fa80ccaf2f0448621942935af702a243a3e1ee6
Andreas Gampe [Mon, 26 Sep 2016 23:26:51 +0000 (23:26 +0000)]
Merge "ART: Fix dependencies"
Andreas Gampe [Mon, 26 Sep 2016 23:07:32 +0000 (23:07 +0000)]
Merge "ART: Revert base/logging conditional hacks"
am:
42c49d236c
Change-Id: I0cb701ab52fb3047360bc96097e47b0806f90f64
Treehugger Robot [Mon, 26 Sep 2016 23:00:02 +0000 (23:00 +0000)]
Merge "ART: Revert base/logging conditional hacks"
Andreas Gampe [Mon, 26 Sep 2016 22:52:31 +0000 (22:52 +0000)]
Merge "ART: Fix nits"
am:
cda79e87e0
Change-Id: Ie552080275f58b1354f47fefbc3d5a06e6eb65f5
Treehugger Robot [Mon, 26 Sep 2016 22:48:55 +0000 (22:48 +0000)]
Merge "ART: Fix nits"
Wojciech Staszkiewicz [Mon, 26 Sep 2016 21:52:05 +0000 (21:52 +0000)]
Merge "Implement nightly jfuzz test runner"
am:
25243492bf
Change-Id: If37fa2111a9709e7c21f308539d7a7faa4c901ed
Treehugger Robot [Mon, 26 Sep 2016 21:42:08 +0000 (21:42 +0000)]
Merge "Implement nightly jfuzz test runner"
Andreas Gampe [Mon, 26 Sep 2016 21:18:25 +0000 (14:18 -0700)]
ART: Fix dependencies
Add libbase as a dependency to libartbenchmark.
Bug:
31338270
Test: m test-art-host
Change-Id: I23c9f4219710fd8b5691837c5bc197b993aafe7b
Andreas Gampe [Mon, 26 Sep 2016 21:07:57 +0000 (14:07 -0700)]
ART: Fix nits
Follow-up to commit
3fec9ac0d5af1358d216eb2fdc2000ec0205f3f0.
Bug:
31338270
Test: m test-art-host
Change-Id: Ia50f450bef144167b946224c26242c529302996e
Wojciech Staszkiewicz [Sat, 24 Sep 2016 01:02:55 +0000 (18:02 -0700)]
Add a true divergence mode to run_jfuzz_test.py
Adds a --true_divergence switch which prevents tool from bisecting
TIMEOUT vs ERROR and TIMEOUT vs SUCCESS return code divergences.
Test: ./run_jfuzz_test.py
Change-Id: I53dd1c2b6934aafcc178fcc5dffd0b10d7a7fc5f
Igor Murashkin [Mon, 26 Sep 2016 19:56:36 +0000 (19:56 +0000)]
Merge "jni: Punt to generic JNI for mips"
am:
20426c3403
Change-Id: I078b4bdafdd14b9e2876db26c2a5a2d7d2ea550f
Andreas Gampe [Mon, 26 Sep 2016 19:56:19 +0000 (19:56 +0000)]
Merge "ART: Use libbase logging"
am:
a4850b0fa1
Change-Id: Id19ac0187a9213fe6b2bd2135f994e60e772dc90
Andreas Gampe [Mon, 26 Sep 2016 19:51:53 +0000 (12:51 -0700)]
ART: Revert base/logging conditional hacks
This partially reverts commit
bda1d606f2d31086874b68edd9254e3817d8049c.
ART was switched to libbase logging.
Bug:
31338270
Test: m test-art-host
Change-Id: I1a2f90d1ddb67d45ebe28d970b3ee7fd2d16a730
Treehugger Robot [Mon, 26 Sep 2016 19:53:37 +0000 (19:53 +0000)]
Merge "jni: Punt to generic JNI for mips"
Wojciech Staszkiewicz [Sat, 24 Sep 2016 00:41:27 +0000 (17:41 -0700)]
Implement nightly jfuzz test runner
New script runs multiple instances of run_jfuzz_test.py and has
predefined arguments suitable for nightly runs.
Test: ./run_jfuzz_test_nightly.py
Change-Id: I8c0060d72c69bbdf88571a9272f2d3fef74697eb
Treehugger Robot [Mon, 26 Sep 2016 19:45:22 +0000 (19:45 +0000)]
Merge "ART: Use libbase logging"
Andreas Gampe [Tue, 13 Sep 2016 17:47:28 +0000 (10:47 -0700)]
ART: Use libbase logging
Move most of our logging infrastructure over to system/core/base.
Retain VLOG.
Using unified Android infrastructure has two main advantages. First,
it reduces the complexity/maintenance burden in ART. Second, it
allows to detach logging for the cases where we do not want or need
a runtime, e.g., dexdump, the disassembler, etc. As a part of the
latter, libbase is also supported for all hosts (including Windows).
From a developer viewpoint, there are minor behavior changes for the
LOG statements (see above), but otherwise usage is the same. Explicit
severity enum items are in the android::base namespace now.
Bug:
31338270
Test: m test-art-host
Change-Id: I5abcb2f45f5b03d49951874c48544f72a283a91b
Igor Murashkin [Mon, 26 Sep 2016 17:10:17 +0000 (10:10 -0700)]
jni: Punt to generic JNI for mips
Preopting dex the bootclasspath causes dex2oat to crash for MIPS builds.
Disable JNI compiler for MIPS until it supports @CriticalNative.
Test: lunch aosp_mips-eng && make && make test-art-host
Bug:
31743474
Change-Id: I802f1e61863c4bb54a1bd884a4c9ce9a75bab62b
David Sehr [Mon, 26 Sep 2016 14:42:26 +0000 (14:42 +0000)]
Merge "Fix CTS crash from DexFile refactoring"
am:
0cfe19af3b
Change-Id: Ifd5c3e906c3175ba25dc179d6378be0f1aa80c84
David Sehr [Mon, 26 Sep 2016 14:34:49 +0000 (14:34 +0000)]
Merge "Fix CTS crash from DexFile refactoring"
Nicolas Geoffray [Mon, 26 Sep 2016 12:45:49 +0000 (12:45 +0000)]
Merge "Compensate in compiler for verifier shortcomings."
am:
3cc69a8944
Change-Id: I7b485879dbaee0feea345d2834fef348405eec83
Treehugger Robot [Mon, 26 Sep 2016 12:39:16 +0000 (12:39 +0000)]
Merge "Compensate in compiler for verifier shortcomings."
Nicolas Geoffray [Fri, 23 Sep 2016 14:40:41 +0000 (15:40 +0100)]
Compensate in compiler for verifier shortcomings.
The verifier does not differentiate zero and null, so a move-object
of zero can be used as a non-object later on.
Change the compiler to ignore the object conversion when the input
is zero or a phi (which might just hold zeros). The type propagation
will then do proper inferencing of the types.
Also remove some stalled comments in ssa_builder.cc.
bug:
31313170
test: dex2oat
b31313170.apk
test: run-test 800
test: m test-art-host-run-test
Change-Id: I579d667415a7decf8ff2c2238dae4c13eec5d0e0
Richard Uhler [Tue, 20 Sep 2016 09:41:47 +0000 (10:41 +0100)]
ahat: Show GC Root Paths.
The Dominator Path in the objects view is replaced by an augmented
Sample Path from GC Root, which includes non-dominator objects
along a sample path and field names.
Also, use blanks instead of "0" in heap tables when the size is 0.
This cleans up the pages a little, and conveniently lets us
distinguish between dominator and non-dominator objects in the Sample
Path from GC Root.
Test: m ahat-test, with new InstanceUtils.gcRootPath test added.
Bug:
27299030
Change-Id: I53d75f9dcb3157c2b5b3afc74958711536cd67b6
Sebastien Hertz [Mon, 26 Sep 2016 07:13:56 +0000 (07:13 +0000)]
Merge "Add DCE check to test 458-checker-instruct-simplification"
am:
a1d66b9050
Change-Id: Ib44b71f95409e604f7c73ed68c1d8df0b39b853f
Sebastien Hertz [Mon, 26 Sep 2016 07:09:43 +0000 (07:09 +0000)]
Merge "Add DCE check to test 458-checker-instruct-simplification"
Dimitry Ivanov [Sat, 24 Sep 2016 01:36:43 +0000 (01:36 +0000)]
Merge "Use liblog for liblog (instead of libcutils)"
am:
ad1f0ad512
Change-Id: Ifdfc800cfba522e9bf9465682f435935702f8f06
Treehugger Robot [Sat, 24 Sep 2016 01:31:25 +0000 (01:31 +0000)]
Merge "Use liblog for liblog (instead of libcutils)"
Mathieu Chartier [Sat, 24 Sep 2016 00:44:09 +0000 (00:44 +0000)]
Merge "Smarter image layout"
am:
48ec36f8a0
Change-Id: I1c5001b9f1d891f67fa36fced91b049936ab64d7
Treehugger Robot [Sat, 24 Sep 2016 00:37:51 +0000 (00:37 +0000)]
Merge "Smarter image layout"
Dimitry Ivanov [Fri, 23 Sep 2016 23:52:05 +0000 (16:52 -0700)]
Use liblog for liblog (instead of libcutils)
Because one wouldn't use libcutils if there is a liblog.
Bug: http://b/
31289077
Bug: http://b/
27171986
Test: remove liblog from libcutils and make
Change-Id: I5e7eb3a67a483835933e1b5587d43b697e39ceac
Mathieu Chartier [Tue, 20 Sep 2016 22:33:31 +0000 (15:33 -0700)]
Smarter image layout
Put strings in the dex file that resolves them.
Depth first traversal with overrides for class and dex cache. The
work list keeps track of what oat_index with each pushed item. This
means the static fields of a class will usually be in the same image.
Added layout test to image_test to make sure things are somewhat
reasonably attributed.
Bug:
28640955
Test: test-art-host
(cherry picked from commit
4e9c4e746617bad6a012d799d2f5cf9e01d24ea2)
Change-Id: I67a536c33aeed603b252d8e0f75622c9efbf2559
David Sehr [Thu, 22 Sep 2016 21:05:37 +0000 (14:05 -0700)]
Fix CTS crash from DexFile refactoring
Avoid path to nullptr access introduced by the previous refactoring.
Bug:
22322814
Change-Id: Id8874b26c072a11b0494e8126f8b0602a7c5b9e8
Test: m test-art-host, cts-tradefed run cts --skip-preconditions --skip-device-info --module vm-tests-tf
Robert Sesek [Fri, 23 Sep 2016 17:09:56 +0000 (17:09 +0000)]
Merge "Add CanWriteToDalvikCache() check before writing a boot marker or image cache."
am:
c16672a026
Change-Id: I1e7e21dcac13cfdc42d7efdeae44a14971099dc7
Treehugger Robot [Fri, 23 Sep 2016 17:05:40 +0000 (17:05 +0000)]
Merge "Add CanWriteToDalvikCache() check before writing a boot marker or image cache."
Aart Bik [Fri, 23 Sep 2016 15:22:21 +0000 (15:22 +0000)]
Merge "Minor jfuzz improvements."
am:
304533ed84
Change-Id: I951fc02aad45d18502139610345e8f16b9c2ebf7
Aart Bik [Fri, 23 Sep 2016 15:18:21 +0000 (15:18 +0000)]
Merge "Minor jfuzz improvements."
Sebastien Hertz [Fri, 23 Sep 2016 14:51:42 +0000 (16:51 +0200)]
Add DCE check to test 458-checker-instruct-simplification
Follow-up to
9837caff9cbfb61e3dabc6fbecb398d4d659c017 by checking
that dead code is actually removed.
Test: make -j32 test-art-host
Change-Id: Ic7848be893c6fe5e65f5bcdd525597072bf67857
Nicolas Geoffray [Fri, 23 Sep 2016 13:21:44 +0000 (13:21 +0000)]
Merge "Fix mips build after sharpening change."
am:
ae3016324f
Change-Id: I70fcc86a9c2bbd1ecc8e4ac7cb0f5429305b88a7
Treehugger Robot [Fri, 23 Sep 2016 13:17:32 +0000 (13:17 +0000)]
Merge "Fix mips build after sharpening change."
Nicolas Geoffray [Fri, 23 Sep 2016 11:37:48 +0000 (12:37 +0100)]
Fix mips build after sharpening change.
test: m build-art-target
Change-Id: Ie1866a944be351878c7f860c640923bc0029f697
Nicolas Geoffray [Fri, 23 Sep 2016 09:54:02 +0000 (09:54 +0000)]
Merge "Clean-up sharpening and compiler driver."
am:
cbb651fdf2
Change-Id: I5f0a556f72934a00dc24a0bdfb634d5baa0362e5
Treehugger Robot [Fri, 23 Sep 2016 09:49:47 +0000 (09:49 +0000)]
Merge "Clean-up sharpening and compiler driver."
Nicolas Geoffray [Fri, 23 Sep 2016 08:23:26 +0000 (08:23 +0000)]
Merge "Use android_root instead of hard coded /system in test script."
am:
411caada9b
Change-Id: I75157ade60a8f1088707a15b275cf3be8ab13025
Nicolas Geoffray [Fri, 23 Sep 2016 08:19:08 +0000 (08:19 +0000)]
Merge "Use android_root instead of hard coded /system in test script."
Nicolas Geoffray [Thu, 22 Sep 2016 12:17:41 +0000 (13:17 +0100)]
Clean-up sharpening and compiler driver.
Remove dependency on compiler driver for sharpening
and dex2dex (the methods called on the compiler driver were
doing unnecessary work), and remove the now unused methods
in compiler driver.
Also remove test that is now invalid, as sharpening always
succeeds.
test: m test-art-host m test-art-target
Change-Id: I54e91c6839bd5b0b86182f2f43ba5d2c112ef908
Wojciech Staszkiewicz [Fri, 23 Sep 2016 06:40:50 +0000 (06:40 +0000)]
Merge "Prepare run_jfuzz_test.py to report bugs"
am:
ca8bad9136
Change-Id: Iaa785870e47a5a9498def3e2a5a5271442456cd6
Treehugger Robot [Fri, 23 Sep 2016 06:36:10 +0000 (06:36 +0000)]
Merge "Prepare run_jfuzz_test.py to report bugs"
Alex Light [Fri, 23 Sep 2016 00:48:44 +0000 (00:48 +0000)]
Merge "Disable test 902-hello-transformation on target"
am:
35505bfe62
Change-Id: Ib06711ac0e6d7dd4a8d1d961ffbea6efd7726e65
Treehugger Robot [Fri, 23 Sep 2016 00:43:24 +0000 (00:43 +0000)]
Merge "Disable test 902-hello-transformation on target"
Aart Bik [Thu, 22 Sep 2016 20:50:11 +0000 (13:50 -0700)]
Minor jfuzz improvements.
Rationale:
Added simple class hierarchy to jfuzz output (still rather
rudimentary, but at least covers all method invocation kinds).
Also cleaned up the array dimensions logic.
Finally, increased granularity of time-based seeding.
Test: ./tools/jfuzz/run_jfuzz_test.py
Change-Id: I6582a75e5be593ba56146e1cd56f801e85f051c5
Alex Light [Thu, 22 Sep 2016 21:54:33 +0000 (14:54 -0700)]
Disable test 902-hello-transformation on target
This test doesn't work on device at the moment due to restrictions on
the android linker.
Bug:
31681198
Test: mma -j40 test-art-target-run-test-902-hello-transformation
Change-Id: I0ce72aa938a5c274083bcb0c60be5370a2bb02c3
Nicolas Geoffray [Thu, 22 Sep 2016 21:55:58 +0000 (22:55 +0100)]
Use android_root instead of hard coded /system in test script.
The buildbot uses its own ANDROID_ROOT.
Spotted while debugging 902 test failures on target.
Change-Id: I9b2047ce510e93f205771f7384f644450b0325e6