OSDN Git Service
Andreas Gampe [Tue, 6 Oct 2015 01:21:29 +0000 (01:21 +0000)]
am
bcb71a2c: Merge "MIPS64r6 Assembler Tests"
* commit '
bcb71a2ce5bcb516f76fc9fe838b61b0c48e1210':
MIPS64r6 Assembler Tests
Andreas Gampe [Tue, 6 Oct 2015 01:15:07 +0000 (01:15 +0000)]
Merge "MIPS64r6 Assembler Tests"
Mathieu Chartier [Tue, 6 Oct 2015 01:14:38 +0000 (01:14 +0000)]
am
0edc0bdb: Merge "Disable class unloading when tracing is active"
* commit '
0edc0bdb1dc8018ec0000702e6de60e99537b519':
Disable class unloading when tracing is active
Chris Larsen [Thu, 17 Sep 2015 20:34:00 +0000 (13:34 -0700)]
MIPS64r6 Assembler Tests
Assembler tests for:
- SQRT.fmt - ABS.fmt - ROUND.L.fmt - ROUND.W.fmt
- CEIL.L.fmt - CEIL.W.fmt - FLOOR.L.fmt - FLOOR.W.fmt
- SEL.fmt - RINT.fmt - CLASS.fmt - MIN.fmt
- MAX.fmt - cvt.d.l - BITSWAP - DBITSWAP
- DSBH - DSHD - WSBH - ROTR
- SELEQZ - SELNEZ - CLZ - CLO
- DCLZ - DCLO - SC - SCD
- LL - LLD
These are the assembler instructions which were added to support
intrinsic functions on MIPS64. Tests for additional assembler
instructions will follow.
Support added to the testing infrastructure for:
- Assembler instructions which use three registers; previously
instructions were limited to one, or two, registers.
- Immediate values which have their sizes specified by the number of
bits required to store them rather than the number of bytes, in both
signed and unsigned versions.
Change-Id: I38c07dcbf2539825b25bed13aac05a26fa594b0b
Mathieu Chartier [Tue, 6 Oct 2015 01:10:45 +0000 (01:10 +0000)]
Merge "Disable class unloading when tracing is active"
Mathieu Chartier [Mon, 5 Oct 2015 23:41:10 +0000 (16:41 -0700)]
Disable class unloading when tracing is active
Tracing keeps a lot of pointers to ArtMethods that could belong to
classes that we are going to unload. A quick fix for this is to
disable class unloading when tracing is active.
Bug:
24414774
Bug:
22720414
Change-Id: Ia5619cbd7c9fd558eaa2a5000871d287213c4a76
Mathieu Chartier [Mon, 5 Oct 2015 21:35:20 +0000 (21:35 +0000)]
am
6639985f: Merge "Fix test 087 for interpreter"
* commit '
6639985f39a1983494fc7de2e4639fdc88e81228':
Fix test 087 for interpreter
Mathieu Chartier [Mon, 5 Oct 2015 21:29:40 +0000 (21:29 +0000)]
Merge "Fix test 087 for interpreter"
Mathieu Chartier [Mon, 5 Oct 2015 21:26:32 +0000 (14:26 -0700)]
Fix test 087 for interpreter
Dex register was keeping BrokenDexLoader live. Fixes interpreter
tests.
Change-Id: I9d2d0907ad5d38dc40434dbde88a5e0cefbb5c80
Mathieu Chartier [Mon, 5 Oct 2015 20:03:33 +0000 (20:03 +0000)]
am
02706fcb: Merge "Delete DexFiles in closeDexFiles"
* commit '
02706fcb35325ba11825d575eb7278e3b9d6dfac':
Delete DexFiles in closeDexFiles
Mathieu Chartier [Mon, 5 Oct 2015 19:57:21 +0000 (19:57 +0000)]
Merge "Delete DexFiles in closeDexFiles"
Mathieu Chartier [Mon, 5 Oct 2015 19:52:57 +0000 (19:52 +0000)]
am
3a9122a5: Merge "Add exclusion between instrumentation and GC"
* commit '
3a9122a5f9f057e9acb20083cd7e076f3bbf8577':
Add exclusion between instrumentation and GC
Mathieu Chartier [Mon, 5 Oct 2015 19:48:24 +0000 (19:48 +0000)]
Merge "Add exclusion between instrumentation and GC"
Mathieu Chartier [Fri, 25 Sep 2015 23:48:57 +0000 (16:48 -0700)]
Delete DexFiles in closeDexFiles
TODO: Also unmap oat code.
Bug:
22720414
Change-Id: I4d003d9b7f8c22890e7b40121a73081524e71f42
Mathieu Chartier [Fri, 2 Oct 2015 22:53:37 +0000 (15:53 -0700)]
Add exclusion between instrumentation and GC
Instrumentation calls VisitClasses while exclusive holding the
mutator lock. This can cause deadlocks since VisitClasses needs to
decode JNI weak globals. If the system weak slow path flag is set,
then we wait holding locks while exclusive holding the mutator lock.
This causes a deadlock since the GC cannot acquire the mutator lock
to sweep system weaks.
This fixes a deadlock seen in one of the tracing tests.
Change-Id: I580152118e068a70f309dcc19df4144afec835dd
Andreas Gampe [Mon, 5 Oct 2015 18:34:19 +0000 (18:34 +0000)]
am
002117f9: Merge "MIPS64: Code cleanup."
* commit '
002117f95896ffa5db74bee808ae61e876b6e8b0':
MIPS64: Code cleanup.
Andreas Gampe [Mon, 5 Oct 2015 18:30:05 +0000 (18:30 +0000)]
Merge "MIPS64: Code cleanup."
Hiroshi Yamauchi [Mon, 5 Oct 2015 17:40:59 +0000 (17:40 +0000)]
am
4b6d025b: Merge "Make it possible to change the read barrier type as build config."
* commit '
4b6d025b914e82652ca72e2917b70b2231ead13f':
Make it possible to change the read barrier type as build config.
Hiroshi Yamauchi [Mon, 5 Oct 2015 17:34:29 +0000 (17:34 +0000)]
Merge "Make it possible to change the read barrier type as build config."
Andreas Gampe [Mon, 5 Oct 2015 16:58:14 +0000 (16:58 +0000)]
am
0a051ef5: Merge "ART: Add more deopt tests"
* commit '
0a051ef572327013fa30e68f3c1f9a7591316970':
ART: Add more deopt tests
Andreas Gampe [Mon, 5 Oct 2015 16:51:14 +0000 (16:51 +0000)]
Merge "ART: Add more deopt tests"
Andreas Gampe [Mon, 5 Oct 2015 16:29:34 +0000 (16:29 +0000)]
am
adb6fbac: Merge "ART: Add CheckJNI lock checking"
* commit '
adb6fbac5223873d99bdf19797cc070e3ca43758':
ART: Add CheckJNI lock checking
Andreas Gampe [Mon, 5 Oct 2015 15:59:06 +0000 (15:59 +0000)]
Merge "ART: Add CheckJNI lock checking"
Andreas Gampe [Mon, 28 Sep 2015 20:16:33 +0000 (13:16 -0700)]
ART: Add CheckJNI lock checking
JNI MonitorEnter and MonitorExit have similar rules to structured
locking. Count locks in CheckJNI mode.
Bug:
23502994
Change-Id: Ie3f53d3aa669a6bd0c7153c50c168116b43764d9
Andreas Gampe [Thu, 1 Oct 2015 00:13:59 +0000 (17:13 -0700)]
ART: Add more deopt tests
Extend libarttest with the ability to turn off asserts for method
state. Use this to put asserts into test functions, but turn them
off if the test wouldn't react as we want, e.g., when we're in
interpreter mode.
Extend run-test 449 by adding asserts for the method state for the
expected deopt cases. This tests both standard behavior, as well
as single-frame deoptimization.
Bug:
21611912
Change-Id: I27fa5e3e44d8c5eab57962d144c6aec96587bf45
Nicolas Geoffray [Mon, 5 Oct 2015 12:37:20 +0000 (12:37 +0000)]
am
1aba494d: Merge "Block callee save fp registers in debuggable."
* commit '
1aba494df441c31836e5b7d4d8e408420a93cb8d':
Block callee save fp registers in debuggable.
Nicolas Geoffray [Mon, 5 Oct 2015 12:32:25 +0000 (12:32 +0000)]
Merge "Block callee save fp registers in debuggable."
Nicolas Geoffray [Mon, 5 Oct 2015 10:15:37 +0000 (11:15 +0100)]
Block callee save fp registers in debuggable.
This is a simple but conservative implementation. We could
extend it by using the registers but still saving them before
a call and at method entry.
bug:
21057237
Change-Id: Ia2e9e0e2efae0b01625e0f4165d0535c4bf9ba62
Nicolas Geoffray [Mon, 5 Oct 2015 09:17:01 +0000 (09:17 +0000)]
am
d0f5b02e: Merge "Fix typo from previous commit."
* commit '
d0f5b02e197f6540cd285e505d6cf0e6bef0bb82':
Fix typo from previous commit.
Nicolas Geoffray [Mon, 5 Oct 2015 09:12:14 +0000 (09:12 +0000)]
Merge "Fix typo from previous commit."
Nicolas Geoffray [Mon, 5 Oct 2015 08:51:36 +0000 (09:51 +0100)]
Fix typo from previous commit.
core_dex2oat_dependency has a '|' dependency, so we should do :=
instead of += to force recompilation of the image.
Change-Id: Ic59b86666fa69038a67459e7be11c57c20bf815b
Nicolas Geoffray [Mon, 5 Oct 2015 07:43:39 +0000 (07:43 +0000)]
am
c8fb997c: Merge "Revert "Don\'t use floating point callee saves in debuggable.""
* commit '
c8fb997c694e3587f4b618b2a1727206b8d4a327':
Revert "Don't use floating point callee saves in debuggable."
Nicolas Geoffray [Mon, 5 Oct 2015 07:43:38 +0000 (07:43 +0000)]
am
8a50a4aa: Merge "Change IllegalAccess/IncompatibleClassChange order."
* commit '
8a50a4aa4636bdc3e0443ea94f5042db18c03ed2':
Change IllegalAccess/IncompatibleClassChange order.
Nicolas Geoffray [Mon, 5 Oct 2015 07:41:02 +0000 (07:41 +0000)]
Merge "Revert "Don't use floating point callee saves in debuggable.""
Nicolas Geoffray [Mon, 5 Oct 2015 07:40:35 +0000 (07:40 +0000)]
Revert "Don't use floating point callee saves in debuggable."
bug:
24602865
bug:
24605078
This reverts commit
88a95ba893fcda974d492917dd77a9b11693dbf2.
Change-Id: Iba97eeab5c2ba725f66cc138f740dac337344828
Nicolas Geoffray [Mon, 5 Oct 2015 07:38:17 +0000 (07:38 +0000)]
Merge "Change IllegalAccess/IncompatibleClassChange order."
Calin Juravle [Sat, 3 Oct 2015 02:49:12 +0000 (02:49 +0000)]
am
d4ca43bb: Merge "Add aarch64 and mips64 support for testing 64 bit field set."
* commit '
d4ca43bbdca1632eca6891758ad103305f2e7d32':
Add aarch64 and mips64 support for testing 64 bit field set.
Calin Juravle [Sat, 3 Oct 2015 02:45:06 +0000 (02:45 +0000)]
Merge "Add aarch64 and mips64 support for testing 64 bit field set."
Calin Juravle [Fri, 2 Oct 2015 22:56:06 +0000 (23:56 +0100)]
Add aarch64 and mips64 support for testing 64 bit field set.
Change-Id: Ibc4a9aecca19a7c8bee5404730d64fbeda9bb2e2
Alex Light [Fri, 2 Oct 2015 21:46:12 +0000 (21:46 +0000)]
am
089fabf4: Merge "Move getPublicFieldRecursive to native JNI"
* commit '
089fabf4d39e59016d92b20e930b86398892534b':
Move getPublicFieldRecursive to native JNI
Alex Light [Fri, 2 Oct 2015 21:40:34 +0000 (21:40 +0000)]
Merge "Move getPublicFieldRecursive to native JNI"
Mathieu Chartier [Fri, 2 Oct 2015 21:01:41 +0000 (21:01 +0000)]
am
a3ac4d4c: Merge "Perform write barrier on declaring class when adding instrumentation"
* commit '
a3ac4d4c39ad3b14392d8fc564cef71655250d1e':
Perform write barrier on declaring class when adding instrumentation
Mathieu Chartier [Fri, 2 Oct 2015 20:53:48 +0000 (20:53 +0000)]
Merge "Perform write barrier on declaring class when adding instrumentation"
Mathieu Chartier [Fri, 2 Oct 2015 20:08:39 +0000 (13:08 -0700)]
Perform write barrier on declaring class when adding instrumentation
Since the instrumentation is marked from the declaring class we need
to mark the card so that mod-union tables and card rescanning know
about the update.
This fixes some moving GC crashes on eng builds.
Bug:
24133607
(cherry picked from commit
b5bba5966045258fd8b755fb65e381cd7fb9c311)
Change-Id: Ibe8430e32e161d33ce0a6782e8dc252e5c86c8da
Hiroshi Yamauchi [Fri, 2 Oct 2015 20:41:34 +0000 (13:41 -0700)]
Make it possible to change the read barrier type as build config.
Bug:
12687968
Change-Id: Ib43dab4d4e68a5a962347fcc2163c3100cb5b760
Richard Uhler [Fri, 2 Oct 2015 19:59:36 +0000 (19:59 +0000)]
am
a5b63f2c: Merge "Generate an hprof file to test ahat."
* commit '
a5b63f2c46b1a5bf0452032dd51198adfb3aa309':
Generate an hprof file to test ahat.
Richard Uhler [Fri, 2 Oct 2015 19:52:39 +0000 (19:52 +0000)]
Merge "Generate an hprof file to test ahat."
Mathieu Chartier [Fri, 2 Oct 2015 16:45:51 +0000 (16:45 +0000)]
am
42888b28: Merge "Change DecodeWeakGlobal to DecodeJObject for class unloading"
* commit '
42888b28d2a0f112187c906296d4b28674c8bee2':
Change DecodeWeakGlobal to DecodeJObject for class unloading
Mathieu Chartier [Fri, 2 Oct 2015 16:40:06 +0000 (16:40 +0000)]
Merge "Change DecodeWeakGlobal to DecodeJObject for class unloading"
Nicolas Geoffray [Fri, 2 Oct 2015 16:14:53 +0000 (17:14 +0100)]
Change IllegalAccess/IncompatibleClassChange order.
This is to have the same order as ClassLinker::ResolveMethod. In
a JIT environment, hitting the dex cache would not make it to the
expected IllegalAccessError and we would throw the wrong exception.
Change-Id: Ifc70f1a93260a24e952fa7a659be1be5bdc666de
Vladimir Marko [Fri, 2 Oct 2015 16:09:57 +0000 (16:09 +0000)]
am
edbff6f5: Merge "ART: Use an iterative way in DoDFSPreOrderSSARename()"
* commit '
edbff6f51abc2bfb43be78b9a1beaa382c498a1f':
ART: Use an iterative way in DoDFSPreOrderSSARename()
Vladimir Marko [Fri, 2 Oct 2015 16:03:24 +0000 (16:03 +0000)]
Merge "ART: Use an iterative way in DoDFSPreOrderSSARename()"
Nicolas Geoffray [Fri, 2 Oct 2015 15:18:37 +0000 (15:18 +0000)]
am
acd70887: Merge "KNOWN_BROKEN/FAILING does not work for gtests."
* commit '
acd7088735d8299c90c1e480be146c71adcd8d08':
KNOWN_BROKEN/FAILING does not work for gtests.
Nicolas Geoffray [Fri, 2 Oct 2015 15:15:47 +0000 (15:15 +0000)]
Merge "KNOWN_BROKEN/FAILING does not work for gtests."
Nicolas Geoffray [Fri, 2 Oct 2015 13:41:18 +0000 (14:41 +0100)]
KNOWN_BROKEN/FAILING does not work for gtests.
Instead disable the test in the gtest file.
Change-Id: Iede5de7ca572c9454437981a5fa63cda444d0497
Nicolas Geoffray [Fri, 2 Oct 2015 11:42:41 +0000 (11:42 +0000)]
am
cc245eab: Merge "Fix lint error."
* commit '
cc245eab2148547f9df37ddc5f767656dc033b81':
Fix lint error.
Nicolas Geoffray [Fri, 2 Oct 2015 11:39:37 +0000 (11:39 +0000)]
Merge "Fix lint error."
Nicolas Geoffray [Fri, 2 Oct 2015 11:38:40 +0000 (12:38 +0100)]
Fix lint error.
Change-Id: Icb10d23bdac6881b48d2025cac76dc130aefbf73
Nicolas Geoffray [Fri, 2 Oct 2015 10:40:45 +0000 (10:40 +0000)]
am
00f24284: Merge "Test fails on mips."
* commit '
00f24284fd4bd12a7fc1865da4984dd6a7c71a94':
Test fails on mips.
Nicolas Geoffray [Fri, 2 Oct 2015 10:40:43 +0000 (10:40 +0000)]
am
d0abb395: Merge "Do more inlining when JITting."
* commit '
d0abb39540aa85306647801ef75fcc5c35fab69a':
Do more inlining when JITting.
Nicolas Geoffray [Fri, 2 Oct 2015 10:36:22 +0000 (10:36 +0000)]
Merge "Test fails on mips."
Nicolas Geoffray [Fri, 2 Oct 2015 10:34:17 +0000 (10:34 +0000)]
Merge "Do more inlining when JITting."
Nicolas Geoffray [Fri, 2 Oct 2015 09:39:55 +0000 (10:39 +0100)]
Test fails on mips.
bug:
24596015
Change-Id: Iad6ab18b287b398af82b6a0574cb74204619755d
Nicolas Geoffray [Fri, 2 Oct 2015 09:00:46 +0000 (09:00 +0000)]
am
230033ba: Merge "Don\'t force the boot image to using quick."
* commit '
230033ba185ef83f7be84a4839d4d2e195ab50f4':
Don't force the boot image to using quick.
Nicolas Geoffray [Thu, 1 Oct 2015 13:34:20 +0000 (14:34 +0100)]
Do more inlining when JITting.
We now check the verification status of the class.
This triggered a bug when an inlined method was not compiled
(typically in a JIT configuration), and the path for deopting in
StackVisitor was using the wrong ArtMethod in order to know the
compiler that compiled the current frame.
Change-Id: I81d3ca0cf5cd3864b83b63dd954c58e1f2adaad4
Nicolas Geoffray [Fri, 2 Oct 2015 08:01:09 +0000 (08:01 +0000)]
Merge "Don't force the boot image to using quick."
Calin Juravle [Fri, 2 Oct 2015 06:45:04 +0000 (06:45 +0000)]
am
1604cb97: Merge "Fix x86_64 gtest."
* commit '
1604cb973e96ede4af275e676f0f35af46641f03':
Fix x86_64 gtest.
Calin Juravle [Fri, 2 Oct 2015 06:40:57 +0000 (06:40 +0000)]
Merge "Fix x86_64 gtest."
Calin Juravle [Fri, 2 Oct 2015 06:27:51 +0000 (07:27 +0100)]
Fix x86_64 gtest.
Calling convention for setting a 64bit field changed.
The CL fixes the test for x86_64 and disable it for other arches.
Change-Id: Iea3df0965f65b10c59ed975f91ca8069fde6e13c
Calin Juravle [Fri, 2 Oct 2015 01:56:39 +0000 (01:56 +0000)]
am
77189602: Merge "Enable compilation of members with access check failures."
* commit '
77189602d7bd291a7de87967b87b9b5a9b2eb68a':
Enable compilation of members with access check failures.
Calin Juravle [Fri, 2 Oct 2015 01:56:37 +0000 (01:56 +0000)]
am
487e7b3a: Merge "Revert "Revert "Support unresolved fields in optimizing""
* commit '
487e7b3aca83d6e7d52f86b73eaa359f23993fcb':
Revert "Revert "Support unresolved fields in optimizing"
Calin Juravle [Fri, 2 Oct 2015 01:53:08 +0000 (01:53 +0000)]
Merge "Enable compilation of members with access check failures."
Calin Juravle [Fri, 2 Oct 2015 01:52:46 +0000 (01:52 +0000)]
Merge "Revert "Revert "Support unresolved fields in optimizing""
Mathieu Chartier [Fri, 2 Oct 2015 01:29:16 +0000 (01:29 +0000)]
am
2a09e92a: Merge "Make SetStateUnsafe run checkpoints before going to suspended"
* commit '
2a09e92a87ff788479ba481b3b216609424aabdc':
Make SetStateUnsafe run checkpoints before going to suspended
Calin Juravle [Thu, 17 Sep 2015 13:15:12 +0000 (14:15 +0100)]
Enable compilation of members with access check failures.
Change-Id: I0de73099b53612072c3e6f1235c22f96339fe440
Calin Juravle [Tue, 29 Sep 2015 03:52:17 +0000 (04:52 +0100)]
Revert "Revert "Support unresolved fields in optimizing"
The CL also changes the calling convetion for 64bit static field set
to use kArg2 instead of kArg1. This allows optimizing to keep
the asumptions:
- arm pairs are always of form (even_reg, odd_reg)
- ecx_edx is not used as a register on x86.
This reverts commit
e6f49b47b6a4dc9c7684e4483757872cfc7ff1a1.
Change-Id: I93159917565824084abc96775f31be1a4249f2f3
Mathieu Chartier [Fri, 2 Oct 2015 01:23:04 +0000 (01:23 +0000)]
Merge "Make SetStateUnsafe run checkpoints before going to suspended"
Mathieu Chartier [Thu, 1 Oct 2015 22:58:41 +0000 (15:58 -0700)]
Make SetStateUnsafe run checkpoints before going to suspended
This prevents cases where we can get a checkpoint request in runnable
and then unsafe set to suspended. This was bad behavior since the
checkpoint may not be run in a timely manner.
Bug:
24191051
Change-Id: I73605c1a1154f67699f3902141f3232ab064793f
Hiroshi Yamauchi [Fri, 2 Oct 2015 01:05:15 +0000 (01:05 +0000)]
am
c448024b: Merge "Improvements for table lookup read barriers."
* commit '
c448024b005bb2debb851f4de1b7fb5e544c3a80':
Improvements for table lookup read barriers.
Hiroshi Yamauchi [Fri, 2 Oct 2015 01:01:25 +0000 (01:01 +0000)]
Merge "Improvements for table lookup read barriers."
Hiroshi Yamauchi [Thu, 1 Oct 2015 03:12:11 +0000 (20:12 -0700)]
Improvements for table lookup read barriers.
- Implement fast paths for the GC root read barrier routines.
- Avoid unnecessary CAS operations.
Bug:
12687968
Change-Id: Iceef44e253062af5bf2295a521a9c64403deafe1
Mathieu Chartier [Fri, 2 Oct 2015 00:08:44 +0000 (17:08 -0700)]
Change DecodeWeakGlobal to DecodeJObject for class unloading
DecodeWeakGlobal returns the sentinel object for cleared JNI weak
globals. This was causing a memory leak since it wouldn't delete
class tables and linear allocs due to never returning null. The bug
was found by yamauchi.
Bug:
22720414
Change-Id: Iff4681495232b9a9756dbdb51d10ea72691a85dd
Alex Light [Thu, 1 Oct 2015 20:35:09 +0000 (13:35 -0700)]
Move getPublicFieldRecursive to native JNI
Benchmark: libcore/benchmarks/src/benchmarks/regression/ReflectionBenchmark.java
Previous benchmarks:
Class_getField 814.26 ns; σ=6.44 ns @ 3 trials
GetInterfaceStaticField 1552.28 ns; σ=38.22 ns @ 10 trials
GetSuperClassField 939.85 ns; σ=2.94 ns @ 3 trials
New benchmarks:
Class_getField 735.97 ns; σ=0.98 ns @ 3 trials
GetInterfaceStaticField 1178.98 ns; σ=6.14 ns @ 3 trials
GetSuperClassField 853.76 ns; σ=16.86 ns @ 10 trials
Bug:
24209213
Change-Id: I08073cb36b007ac33af010dab917a6f1e51107c0
Mathieu Chartier [Thu, 1 Oct 2015 22:16:26 +0000 (22:16 +0000)]
am
25217af2: Merge "Add more error printing to TransitionFromSuspendedToRunnable"
* commit '
25217af2a7cae96b32ba566aaf697288f3374c99':
Add more error printing to TransitionFromSuspendedToRunnable
Andreas Gampe [Thu, 1 Oct 2015 22:11:39 +0000 (22:11 +0000)]
am
506b08cb: Merge "Revert "Revert "ART: Remove unnecessary SHARED_REQUIRES in ArtMethod"""
* commit '
506b08cbd5e44bedc32f31ea74ddcc921c98ec99':
Revert "Revert "ART: Remove unnecessary SHARED_REQUIRES in ArtMethod""
Mathieu Chartier [Thu, 1 Oct 2015 22:08:56 +0000 (22:08 +0000)]
Merge "Add more error printing to TransitionFromSuspendedToRunnable"
Andreas Gampe [Thu, 1 Oct 2015 22:05:49 +0000 (22:05 +0000)]
Merge "Revert "Revert "ART: Remove unnecessary SHARED_REQUIRES in ArtMethod"""
Mathieu Chartier [Thu, 1 Oct 2015 21:46:29 +0000 (14:46 -0700)]
Add more error printing to TransitionFromSuspendedToRunnable
Bug:
24191051
Change-Id: I5c66ba5c76bf86e8d603110989b9da883ca4ca65
Andreas Gampe [Thu, 1 Oct 2015 21:05:46 +0000 (21:05 +0000)]
am
9b29f3ef: Merge "MIPS64: Implement intrinsic java.lang.Thread.currentThread()"
* commit '
9b29f3ef7df18c84d922634709e38a1913709ad7':
MIPS64: Implement intrinsic java.lang.Thread.currentThread()
Andreas Gampe [Thu, 1 Oct 2015 20:29:43 +0000 (20:29 +0000)]
Merge "MIPS64: Implement intrinsic java.lang.Thread.currentThread()"
Chris Larsen [Thu, 1 Oct 2015 18:35:18 +0000 (11:35 -0700)]
MIPS64: Code cleanup.
1. Add comment to explain logic for GenMinMax in intrinsics code.
2. Declare enumerated type for class.s/class.d mask values.
3. Change intrinsics code to use the enumerated values described in
item 2.
4. Change "CLASS_MASK" to "kFPLeaveUnchanged" to match ART coding
standards.
Change-Id: Ib1fe4b01515595b46e5f101e0082bb9bbcf0c688
Nicolas Geoffray [Thu, 1 Oct 2015 18:05:43 +0000 (18:05 +0000)]
am
8446ed41: Merge "Fix another poisoning problem."
* commit '
8446ed412d3a4ca2c70737d6f40c4129b0ee8ab1':
Fix another poisoning problem.
Nicolas Geoffray [Thu, 1 Oct 2015 17:41:27 +0000 (17:41 +0000)]
Merge "Fix another poisoning problem."
Nicolas Geoffray [Thu, 1 Oct 2015 14:50:27 +0000 (15:50 +0100)]
Fix another poisoning problem.
We were using the wrong temp.
Change-Id: Id79d5079cc85f61eb1a45d741a67f24d33e8fa03
Richard Uhler [Thu, 10 Sep 2015 23:45:54 +0000 (16:45 -0700)]
Generate an hprof file to test ahat.
This change sets up the infrastructure to write test cases for ahat
that make use of an hprof file automatically generated from a sample
program.
Change-Id: Id11f656afb69c96a26655cc4caeb745ad844f431
David Brazdil [Thu, 1 Oct 2015 14:22:19 +0000 (14:22 +0000)]
am
b3577f0a: Merge "ART: Fix a static_cast int32_t -> uint64_t bug."
* commit '
b3577f0ab8f1765d3554f575d99adbc65bfaf1b6':
ART: Fix a static_cast int32_t -> uint64_t bug.
David Brazdil [Thu, 1 Oct 2015 14:16:40 +0000 (14:16 +0000)]
Merge "ART: Fix a static_cast int32_t -> uint64_t bug."
David Brazdil [Thu, 1 Oct 2015 13:32:56 +0000 (14:32 +0100)]
ART: Fix a static_cast int32_t -> uint64_t bug.
HConstant::GetValueAsUint64 is used by SsaChecker to verify that
equivalent phis are created only for untyped constants. The test
would fail because a static_cast would sign extend the value of the
IntConstant.
Bug:
24561315
Change-Id: I818ce6a2080994a7c4395d084c1df7fd615a246d
Nicolas Geoffray [Thu, 1 Oct 2015 10:52:12 +0000 (10:52 +0000)]
am
9664c089: Merge "CUSTOM_TARGET_LINKER is set by the buildbot already."
* commit '
9664c08964c7db92151335023c6ea595f75f4033':
CUSTOM_TARGET_LINKER is set by the buildbot already.
Nicolas Geoffray [Thu, 1 Oct 2015 10:47:57 +0000 (10:47 +0000)]
Merge "CUSTOM_TARGET_LINKER is set by the buildbot already."