OSDN Git Service
Mingyao Yang [Wed, 18 Nov 2015 23:33:40 +0000 (23:33 +0000)]
Merge "Fix an assert in lse."
Mingyao Yang [Wed, 18 Nov 2015 23:20:37 +0000 (15:20 -0800)]
Fix an assert in lse.
Wrong assert. Just deleted it.
Change-Id: I6f051609d87b2c6da081ee63f783ce99462f8a09
Mingyao Yang [Wed, 18 Nov 2015 21:29:47 +0000 (21:29 +0000)]
Merge "Revert "Revert "Enable store elimination for singleton objects."""
Mingyao Yang [Mon, 2 Nov 2015 18:56:59 +0000 (10:56 -0800)]
Revert "Revert "Enable store elimination for singleton objects.""
This reverts commit
55d02cf056f993aeafebd54e7b7c68c7a48507c9, and
makes the following change:
Currently we leverage loop side effects to decide whether heap values are
killed by the loop. Stores need to be kept if heap values may be killed
by loops and the corresponding loads cannot be eliminated. Similar thing
need to be done for each predecessor when we merge predecessor heap values.
To do that, the HInstanceFieldSet instruction itself is put in the heap
value array instead of the value of the store instruction. The store
instruction may be added to possibly_removed_stores_ first, but can later
be removed from possibly_removed_stores_ when it's found out that the store
needs to be kept due to merging/loop side effects.
Change-Id: I4f7bb1960f7b47240873e00ff1adac46fc102a02
Mathieu Chartier [Wed, 18 Nov 2015 20:06:43 +0000 (20:06 +0000)]
Merge "Remove incorrect iterator increment"
Mathieu Chartier [Wed, 18 Nov 2015 18:24:43 +0000 (10:24 -0800)]
Remove incorrect iterator increment
Erase already goes to the next element.
Bug:
25738102
Change-Id: Iea31a7fb27ce9f810acf019a7b59dda7704b8f26
Vladimir Marko [Wed, 18 Nov 2015 17:09:04 +0000 (17:09 +0000)]
Merge "Opt compiler: Arm64 packed-switch jump tables."
Zheng Xu [Wed, 18 Nov 2015 09:46:25 +0000 (17:46 +0800)]
Opt compiler: Arm64 packed-switch jump tables.
In this patch, we set a rough threshold and only generate jump table
with limited number of HIRs in the graph. This is because current VIXL
can only handle Adr with label in the range of +/-1Mb.
Change-Id: I42bff2095ec26caeacc5efc90afebe34e229b518
Nicolas Geoffray [Wed, 18 Nov 2015 10:56:51 +0000 (10:56 +0000)]
Merge "Fix bogus DCHECK and rename DidForkFromZygote."
Nicolas Geoffray [Tue, 17 Nov 2015 13:50:21 +0000 (13:50 +0000)]
Fix bogus DCHECK and rename DidForkFromZygote.
The method can also be called for non-zygote forked processes.
(cherry picked from commit
9d157e48e0abc48f08fd98e4c5e918d029c1b84c)
Change-Id: Id4f57b77c1a7f802f80e066afcf30afecc198bfc
Nicolas Geoffray [Wed, 18 Nov 2015 10:47:50 +0000 (10:47 +0000)]
Merge "Remove DEBUG_JIT from Zygote flags."
Calin Juravle [Wed, 18 Nov 2015 10:41:41 +0000 (10:41 +0000)]
Merge "Implement common super type in reference type propagation."
Sebastien Hertz [Wed, 18 Nov 2015 10:29:22 +0000 (10:29 +0000)]
Merge "Support deoptimization only to set vreg"
Sebastien Hertz [Wed, 18 Nov 2015 10:17:46 +0000 (10:17 +0000)]
Merge "Fix StackVisitor::GetVReg for reference in shadow frame"
Nicolas Geoffray [Thu, 5 Nov 2015 11:32:24 +0000 (11:32 +0000)]
Remove DEBUG_JIT from Zygote flags.
The flag is being obsolete by the move to JIT.
(cherry picked from commit
0f042e04efba887557e40f981bd9c41b121c4652)
Change-Id: I32c3183e791690964c00fab02a6ee7bbacf6b665
Hiroshi Yamauchi [Tue, 17 Nov 2015 21:53:34 +0000 (21:53 +0000)]
Merge "RosAlloc fast path inline in assembly for x86 32bit."
Mathieu Chartier [Tue, 17 Nov 2015 20:10:24 +0000 (20:10 +0000)]
Merge "Refactor some patching logic"
Calin Juravle [Wed, 11 Nov 2015 16:58:31 +0000 (16:58 +0000)]
Implement common super type in reference type propagation.
Currently only if both types are classes.
Change-Id: I06e98211ead56875a42bd17f099e319b107a50d4
Vladimir Marko [Tue, 17 Nov 2015 17:40:57 +0000 (17:40 +0000)]
Merge "Minor fixes and cleaning of arm64 static and direct calls code."
Vladimir Marko [Tue, 17 Nov 2015 17:40:34 +0000 (17:40 +0000)]
Merge "ART: add kryo to a53 #835769 & #843419 erratum exception list"
David Brazdil [Tue, 17 Nov 2015 16:46:49 +0000 (16:46 +0000)]
Merge "ART: Refactor GenerateTestAndBranch"
Richard Uhler [Tue, 17 Nov 2015 16:21:12 +0000 (16:21 +0000)]
Merge "Change roots view to "rooted"."
David Brazdil [Thu, 12 Nov 2015 18:37:00 +0000 (18:37 +0000)]
ART: Refactor GenerateTestAndBranch
Each code generator implements a method for generating condition
evaluation and branching to arbitrary labels. This patch refactors
it for better clarity but also to generate fewer jumps when the true
branch is the fallthrough successor.
This is preliminary work for implementing HSelect.
Change-Id: Iaa545a5ecbacb761c5aa241fa69140cf6eb5952f
Mingyao Yang [Fri, 31 Jul 2015 23:40:24 +0000 (16:40 -0700)]
Support deoptimization only to set vreg
This CL removes the old API that was used to update vreg. We now rely
exclusively on deoptimization to change the value of a vreg (mainly
from the debugger). This allows to have only one mechanism working
with both Quick and Optimizing compilers. It also remove run-test
455-set-vreg which was the only user of the old API.
This CL also renames StackVisitor::SetVReg[Pair]FromDebugger to
StackVisitor::SetVReg[Pair] to reflect that change.
Bug:
25428216
Change-Id: Ib09510185280dbd0f6d02c52549ae9671c187e32
Sebastien Hertz [Tue, 17 Nov 2015 09:35:39 +0000 (10:35 +0100)]
Fix StackVisitor::GetVReg for reference in shadow frame
Also fixes 457-regs run-test that used to pass an incorrect
vreg kind.
Bug:
25428216
Change-Id: Ic31b51ca0f72350db6325f23087b06e93055d750
Mathieu Chartier [Fri, 13 Nov 2015 18:42:08 +0000 (10:42 -0800)]
Refactor some patching logic
Reduce duplication since app images in-place relocation will require
this code also.
Bug:
22858531
Change-Id: Ibb901b67267e27ef3bc2a0baff77189d4dcd018a
Alex Light [Mon, 16 Nov 2015 18:10:37 +0000 (18:10 +0000)]
Merge "Add ability to skip run-tests with ART_TEST_RUN_TEST_SKIP."
Calin Juravle [Mon, 16 Nov 2015 17:00:23 +0000 (17:00 +0000)]
Merge "Ignore NullConstants when computing the type for Phis."
Aart Bik [Mon, 16 Nov 2015 16:41:21 +0000 (16:41 +0000)]
Merge "Implemented control of "export only" feature in dexdump."
Roland Levillain [Mon, 16 Nov 2015 16:09:48 +0000 (16:09 +0000)]
Merge "Mark kQuickReadBarrierForRootSlow as direct entry point on MIPS."
Roland Levillain [Mon, 16 Nov 2015 15:59:08 +0000 (15:59 +0000)]
Mark kQuickReadBarrierForRootSlow as direct entry point on MIPS.
Change-Id: I35caa32c9773646692706642188ab522237a68ac
Roland Levillain [Mon, 16 Nov 2015 15:34:28 +0000 (15:34 +0000)]
Merge "x86/x86-64 read barrier support for concurrent GC in Optimizing."
Calin Juravle [Mon, 16 Nov 2015 11:04:13 +0000 (11:04 +0000)]
Merge "Move the trace file out of the profiles folder."
Roland Levillain [Fri, 13 Nov 2015 10:07:31 +0000 (10:07 +0000)]
x86/x86-64 read barrier support for concurrent GC in Optimizing.
This first implementation uses slow paths to instrument heap
reference loads and GC root loads for the concurrent copying
collector, respectively calling the artReadBarrierSlow and
artReadBarrierForRootSlow (new) runtime entry points.
Notes:
- This implementation does not instrument HInvokeVirtual
nor HInvokeInterface instructions (for class reference
loads), as the corresponding read barriers are not stricly
required with the current concurrent copying collector.
- Intrinsics which may eventually call (on slow path) are
disabled when read barriers are enabled, as the current
slow path infrastructure does not support this case.
- When read barriers are enabled, the code generated for a
HArraySet instruction always go into the array set slow
path for object arrays (delegating the operation to the
runtime), as we are lacking a mechanism to keep a
temporary register live accross a runtime call (needed for
the instrumentation of type checking code, which requires
two successive read barriers).
Bug:
12687968
Change-Id: I14cd6107233c326389120336f93955b28ffbb329
Hiroshi Yamauchi [Sat, 14 Nov 2015 02:15:09 +0000 (02:15 +0000)]
Merge "Fix class unloading with the CC collector."
Alex Light [Fri, 13 Nov 2015 18:51:10 +0000 (10:51 -0800)]
Add ability to skip run-tests with ART_TEST_RUN_TEST_SKIP.
One can skip running specific run-tests under any configuration by
setting the ART_TEST_RUN_TEST_SKIP variable to a list of the run-test
names to skip.
$ # None of the variants of 022-interfaces will be run.
$ ART_TEST_RUN_TEST_SKIP=022-interface mm test-art-host-run-test
Change-Id: I33f24dd7f1eb9f4d6dc407a4abc3c5919437d4fa
Hiroshi Yamauchi [Thu, 12 Nov 2015 07:45:34 +0000 (23:45 -0800)]
Fix class unloading with the CC collector.
Avoid unnecessarily decoding dex cache and class loader weak roots,
which would trigger read barriers.
Re-enable 141-class-unload with the CC collector.
Bug:
12687968
Bug:
24468364
Change-Id: Ib4c19f25000873cab0e06047040442d135285745
Alex Light [Fri, 13 Nov 2015 19:07:37 +0000 (19:07 +0000)]
Merge "Disable test 964 with gcstress since it often times out."
Alex Light [Fri, 13 Nov 2015 18:39:34 +0000 (10:39 -0800)]
Disable test 964 with gcstress since it often times out.
Change-Id: Ice9b25d373b849d4415f9189fe94d8ccd6d05cbe
Ajay Dudani [Thu, 12 Nov 2015 22:28:58 +0000 (14:28 -0800)]
ART: add kryo to a53 #835769 & #843419 erratum exception list
Kryo is custom-designed 64-bit ARM CPU and does not need this A53
erratum handling.
https://www.qualcomm.com/news/snapdragon/2015/09/02/snapdragon-820-and-kryo-cpu-heterogeneous-computing-and-role-custom
Change-Id: Id8f6c616ab2ed19056fd1e3829b85b94a6e6eb84
Mathieu Chartier [Fri, 13 Nov 2015 18:33:54 +0000 (18:33 +0000)]
Merge "Revert "Revert "Add missing null check to String::ToCharArray"""
Aart Bik [Fri, 13 Nov 2015 18:06:01 +0000 (10:06 -0800)]
Implemented control of "export only" feature in dexdump.
Rationale:
Export only is now under explicit control of a new -e flag, and
can be used to control both -l xml and -l plain. The result is
that, by default, xml output now shows all items (old behavior
can be forced with -e -l xml). This solution feels a bit cleaner
than adding a flag to override export only when -l xml is given
(old dexdump forced this tight relation).
Bug:
25673756 (feature request).
Change-Id: I7cfd48bf91195d2a81d134936719d267fda399f5
Mathieu Chartier [Fri, 13 Nov 2015 16:36:59 +0000 (08:36 -0800)]
Revert "Revert "Add missing null check to String::ToCharArray""
Fix flaky test.
Bug:
25641543
This reverts commit
b52a39c76e0cb0638c92cdcf6b5c5aa551420e1a.
Change-Id: I6b0d3087dba3680252e65af554530e2769cde55c
David Srbecky [Fri, 13 Nov 2015 15:53:30 +0000 (15:53 +0000)]
Merge "Stream DWARF debug section directly to disk."
Nicolas Geoffray [Fri, 13 Nov 2015 08:16:55 +0000 (08:16 +0000)]
Merge "Revert "Add missing null check to String::ToCharArray""
Nicolas Geoffray [Fri, 13 Nov 2015 08:16:23 +0000 (08:16 +0000)]
Revert "Add missing null check to String::ToCharArray"
Does not work with the interpreter.
Bug:
25641543
This reverts commit
952d608062eec2d7f47f9b45dba935ad8b4d23e5.
Change-Id: Ic112fa69c7ddd119cbccc5b65007b5ee4dfccd09
Mathieu Chartier [Fri, 13 Nov 2015 01:32:44 +0000 (01:32 +0000)]
Merge "Add missing null check to String::ToCharArray"
Mathieu Chartier [Fri, 13 Nov 2015 00:53:41 +0000 (16:53 -0800)]
Add missing null check to String::ToCharArray
Added test.
Bug:
25641543
Change-Id: Ic9a21ce8bc530dbedf14334ad47f5faa90ae4ddc
Alex Light [Fri, 13 Nov 2015 00:50:11 +0000 (00:50 +0000)]
Merge "Fix bug with test 964."
Alex Light [Thu, 12 Nov 2015 23:38:24 +0000 (15:38 -0800)]
Fix bug with test 964.
An indentation error meant that some test cases would not be
generated. Also fixed run-test script since fixing this causes the
output to be larger than the default ulimit.
Change-Id: I454f89078f8307f400b66627fdab13feaf492512
Mathieu Chartier [Thu, 12 Nov 2015 19:19:36 +0000 (19:19 +0000)]
Merge "Add class loader argument to openDexFileNative"
Mathieu Chartier [Thu, 12 Nov 2015 18:00:58 +0000 (10:00 -0800)]
Add class loader argument to openDexFileNative
Bug:
22858531
Change-Id: I315919d91822db0c73cf16b21d660d5870d5746f
Alexandre Rames [Thu, 12 Nov 2015 14:44:19 +0000 (14:44 +0000)]
Minor fixes and cleaning of arm64 static and direct calls code.
Fixes:
The proper way to avoid the MacroAssembler to generate code before or
after an instruction is to block the pools (usually via
`vixl::BlockPoolsScope`). Here we can use
`vixl::SingleEmissionCheckScope`, that checks we generate only one
instruction and also blocks the pools.
In practice the current code would have worked fine because VIXL would
not have generated anything after `Bl()` or `Ldr()`, but that was not
guaranteed.
Cleaning:
- `XRegisterFrom()` returns an X register. Calling `.X()` is not
required.
- Since we are sure (after the previous fixes) that nothing will be
emitted around the instructions we care about, update the code to
bind labels before the instructions for simplicity.
Change-Id: I42d49976721e380e66bcd7a5b345f1777009434a
Vladimir Marko [Thu, 12 Nov 2015 15:20:08 +0000 (15:20 +0000)]
Merge "Optimizing/X86: PC-relative dex cache array addressing."
Vladimir Marko [Thu, 12 Nov 2015 11:56:22 +0000 (11:56 +0000)]
Merge "Optimizing: Clean up constant location handling."
Tamas Berghammer [Thu, 12 Nov 2015 11:06:38 +0000 (11:06 +0000)]
Merge "Revert "Mark __jit_debug_register_code as a unique symbol""
Tamas Berghammer [Thu, 12 Nov 2015 11:05:50 +0000 (11:05 +0000)]
Revert "Mark __jit_debug_register_code as a unique symbol"
This reverts commit
e7132a9a5d98103fa52a750287476b063fa88863.
We are still using ld on some targets instead of gold when compiling for ararch64. Revert until we figure out why.
Change-Id: If09fac8d72485dc8adbc81e44a2af84ed91dcd91
Sebastien Hertz [Thu, 12 Nov 2015 11:00:38 +0000 (11:00 +0000)]
Merge "JDWP: default implementation of Method.IsObsolete"
Vladimir Marko [Mon, 2 Nov 2015 14:36:43 +0000 (14:36 +0000)]
Optimizing/X86: PC-relative dex cache array addressing.
Add PC-relative dex cache array addressing for X86 and use
it for better invoke-static/-direct dispatch. Also delay
the initialization to the PC-relative base until needed.
Change-Id: Ib8634d5edce4920cd70172fd13211809cf6948d1
Roland Levillain [Thu, 12 Nov 2015 10:43:30 +0000 (10:43 +0000)]
Merge "MIPS64: fix calling conventions in LoadClass and field accesses"
Tamas Berghammer [Thu, 12 Nov 2015 10:40:23 +0000 (10:40 +0000)]
Merge "Mark __jit_debug_register_code as a unique symbol"
Vladimir Marko [Thu, 12 Nov 2015 10:16:29 +0000 (10:16 +0000)]
Merge "ART: Add comments to Scoped-/ArenaAllocator."
lwang85 [Tue, 10 Nov 2015 09:26:31 +0000 (17:26 +0800)]
RosAlloc fast path inline in assembly for x86 32bit.
Port patch https://android-review.googlesource.com/#/c/180014/ for
x86 32bit.
Measurement:
AndroidGCTest(ms, the lower the better): 73372->68165, 7.10% improvement.
Signed-off-by: lwang85 <li.w.wang@intel.com>
Change-Id: I9ac3fc56f98574c41fc070dbf12ce7474d6a3610
Mathieu Chartier [Thu, 12 Nov 2015 03:23:54 +0000 (03:23 +0000)]
Merge "Image space cleanup for app images"
Mathieu Chartier [Tue, 10 Nov 2015 22:13:23 +0000 (14:13 -0800)]
Image space cleanup for app images
Removed Heap::GetImageSpace, added Heap::GetBootImageSpace.
Generalized some logic in the class linker for image spaces.
Bug:
22858531
Change-Id: Ib3e12bb061a247e232d3dc93b0d6b35eb3a34d25
Hans Boehm [Thu, 12 Nov 2015 02:51:20 +0000 (02:51 +0000)]
Merge "Get class loader to remember interrupts"
Hans Boehm [Thu, 12 Nov 2015 00:13:57 +0000 (16:13 -0800)]
Get class loader to remember interrupts
Bug:
25606036
Monitor::Wait should not reset the interrupt flag unless it actually
throws the exception.
Change-Id: Id6a35c888160043831f30055f9bd39e7f0440439
Alexey Frunze [Wed, 11 Nov 2015 21:31:12 +0000 (13:31 -0800)]
MIPS64: fix calling conventions in LoadClass and field accesses
This fixes failures in test 529-checker-unresolved
Change-Id: I6170c22059e9711b2fcc965b00d6e34edd839539
Alex Light [Wed, 11 Nov 2015 23:36:15 +0000 (23:36 +0000)]
Merge "Make test 968's name shorter"
Alex Light [Wed, 11 Nov 2015 23:32:24 +0000 (15:32 -0800)]
Make test 968's name shorter
It was causing intermittent CI breaks due to the long name of the
generated oat file.
Change-Id: I71b4b3aa1a13df2696b276998cafe40006d54b66
Aart Bik [Wed, 11 Nov 2015 22:38:10 +0000 (22:38 +0000)]
Merge "Revert "lambda: unit test for lambda expressions at Java Language source level.""
Aart Bik [Wed, 11 Nov 2015 22:36:19 +0000 (22:36 +0000)]
Revert "lambda: unit test for lambda expressions at Java Language source level."
REASON: Buildbots are having jack issues
This reverts commit
152ee552652691ecaec2db420d468cb178647952.
Change-Id: I650fe1a94f1b97c14a689599703498506dd70b23
David Srbecky [Wed, 11 Nov 2015 21:36:22 +0000 (21:36 +0000)]
Merge "Move .debug_line before .debug_info."
David Srbecky [Wed, 11 Nov 2015 20:19:38 +0000 (20:19 +0000)]
Stream DWARF debug section directly to disk.
Change the structure so that the debug sections are written to
disk as we go. There are still some temporary buffers, however,
we no longer hold all of the data in memory before writing it.
We can not avoid buffering of some things (e.g. .debug_str).
Change-Id: Id4940cf10ae5b6f3cac7fb8d20197f0304079b92
David Srbecky [Wed, 11 Nov 2015 20:38:16 +0000 (20:38 +0000)]
Move .debug_line before .debug_info.
Swap the order since .debug_info depends on .debug_line, but
not vice versa. This is for disk streaming convenience.
Change-Id: Iabe18fea97aa5527aedfc55c9bdf66e5587568bc
Vladimir Marko [Wed, 11 Nov 2015 18:45:50 +0000 (18:45 +0000)]
ART: Add comments to Scoped-/ArenaAllocator.
Change-Id: Iba7350114fb998ac9ee1466ef12f4a5024a18e8e
Vladimir Marko [Wed, 11 Nov 2015 18:44:25 +0000 (18:44 +0000)]
Merge "ART: Fix arena allocation for valgrind."
Aart Bik [Wed, 11 Nov 2015 18:30:36 +0000 (18:30 +0000)]
Merge "lambda: unit test for lambda expressions at Java Language source level."
Aart Bik [Wed, 11 Nov 2015 00:49:45 +0000 (16:49 -0800)]
lambda: unit test for lambda expressions at Java Language source level.
Rationale: having a checkers-flavored unit test will enable testing
both if the correct HIR is generated using the checker
as well as the correctness of execution (interpreter for now,
but native code later).
Change-Id: I588e9955efaaf53e5c5e3398399e009041cd65f0
Calin Juravle [Wed, 4 Nov 2015 16:16:47 +0000 (16:16 +0000)]
Move the trace file out of the profiles folder.
/dalvik-cache/profiles/ is going away. Create the trace file under
/data/misc/trace.
This enable tracing without forcing SElinux into permissive mode.
Bug:
24698874
Bug:
25612377
Change-Id: I78d09ef629bebe635c3baabaf88f01996d3333e5
Alex Light [Wed, 11 Nov 2015 18:07:14 +0000 (18:07 +0000)]
Merge "Correct exception behavior for default methods"
Vladimir Marko [Tue, 10 Nov 2015 20:54:22 +0000 (20:54 +0000)]
ART: Fix arena allocation for valgrind.
Move the zero-initialization check after marking the newly
allocated chunk as defined and check only the allocated
space without the red zone. Also mark unallocated space as
inaccessible instead of just undefined.
Change-Id: I74fc65f5b53acb74cec4e5a0146f41dacf4a1470
Richard Uhler [Wed, 11 Nov 2015 17:13:23 +0000 (09:13 -0800)]
Change roots view to "rooted".
Instead of just showing roots in the roots view, show all objects
whose immediate dominator is the SENTINEL_ROOT.
Bug:
24613815
Change-Id: I96429d75395edfe109222e88d31cdc0bd87a7767
Calin Juravle [Wed, 11 Nov 2015 14:59:46 +0000 (14:59 +0000)]
Ignore NullConstants when computing the type for Phis.
NullConstants are always typed as Object and they widen the
type of Phis unnecessarily.
Change-Id: I2b29e67617a22120642cb23d2bc6c1ff75b4e7c2
Vladimir Marko [Wed, 11 Nov 2015 12:37:46 +0000 (12:37 +0000)]
Optimizing: Clean up constant location handling.
Locations builder should use ConstantLocation() when the
code generator relies on a location to be constant. Code
generator should interrogate locations, not inputs, about
being const.
Change-Id: Ic35bb84aa9f83e0977b151a0430aca6c88f19cf0
Tamas Berghammer [Wed, 2 Sep 2015 10:24:33 +0000 (11:24 +0100)]
Mark __jit_debug_register_code as a unique symbol
This symbol is used to notify the debugger about JIT library load/unload
events. Because of ICF, several functions with the same body put to the
same adress by the linker what causes the debugger to get notifictaion
about JIT load/unload events in several case where it isn't really happened.
This change disable ICF for the __jit_debug_register_code symbol to prevent
this behavior with keeping the symbol unique in the object file.
This is the recommit of
5caf652d7bfb43bdd2f3addfe8021a5fb8e3f796 after
fixing the compilation issue on sdk_phone_armv7-sdk (mac) target.
Change-Id: I027d2c7964098d47ed350ee6b7c74d82c88a5d1f
Calin Juravle [Wed, 11 Nov 2015 13:55:39 +0000 (13:55 +0000)]
Merge "Revert "Revert "Run type propagation after inliner only when needed."""
Calin Juravle [Mon, 26 Oct 2015 14:05:01 +0000 (14:05 +0000)]
Revert "Revert "Run type propagation after inliner only when needed.""
This reverts commit
271743601650308c7ac5c7a3ec35025d8130a298.
Change-Id: I173e27a0a4d7d54f90ca459eb48d280d1d40ab70
Roland Levillain [Wed, 11 Nov 2015 12:40:26 +0000 (12:40 +0000)]
Merge "Ignore failures of WeakHashMapTest#test_keySet in libcore tests."
Roland Levillain [Wed, 11 Nov 2015 12:16:52 +0000 (12:16 +0000)]
Ignore failures of WeakHashMapTest#test_keySet in libcore tests.
The outcome of this test depends on a weak reference being
collected before a timeout, which sometimes fails. This
test used to pass, but recent changes have made this
non-determinism more frequent. Until we rewrite this
properly (if ever), just ignore these failures.
Bug:
25437292
Change-Id: I1f041f6dfc1318026bce966575477486207f5351
Vladimir Marko [Wed, 11 Nov 2015 11:27:03 +0000 (11:27 +0000)]
Merge "Optimizing/ARM: Improve shifts of long values by a constant."
Vladimir Marko [Tue, 10 Nov 2015 19:31:26 +0000 (19:31 +0000)]
Optimizing/ARM: Improve shifts of long values by a constant.
Change-Id: Id66ef8cdb9e64306f2be547370b90cc100a3e086
Alex Light [Fri, 9 Oct 2015 22:59:48 +0000 (15:59 -0700)]
Correct exception behavior for default methods
Default methods are defined to throw an IncompatibleClassChangeError
(ICCE) when they are called and there is no "best" implementation.
Previously we would simply throw an ICCE during class loading as soon
as we noticed that this would happen if called. This makes us wait
until we actually attempt to execute the method. Furthermore, this
allows us to use other, non-conflicting, methods on the object as
normal.
Furthermore, this makes us correctly throw AbstractMethodErrors in
cases where all default implementations of a method are overridden by
abstract declarations.
Adds 3 tests for this new behavior.
Bug:
24618811
Change-Id: Id891958a81f9b3862b2ce5919636aabef7d3422e
Mathieu Chartier [Tue, 10 Nov 2015 22:09:40 +0000 (22:09 +0000)]
Merge "Don't use fixed addresses in mem map test"
Mathieu Chartier [Tue, 10 Nov 2015 18:32:52 +0000 (10:32 -0800)]
Don't use fixed addresses in mem map test
We obtain a valid address instead of assuming there won't be a map
there.
These checks used to be OK when we didn't start the runtime.
Change-Id: Ie636ed3773d2c0c62d87e57623ee4ce380c64db0
Andreas Gampe [Tue, 10 Nov 2015 21:55:08 +0000 (21:55 +0000)]
Merge "ART: Fix leak in Heap"
Andreas Gampe [Tue, 10 Nov 2015 21:34:17 +0000 (13:34 -0800)]
ART: Fix leak in Heap
Delete a mutex in the destructor.
Change-Id: I7fb8f3f9e7a30baf352f384f38520eb0dd4906eb
Vladimir Marko [Tue, 10 Nov 2015 20:04:08 +0000 (20:04 +0000)]
Merge "Optimizing: Clean up Equal/NotEqual constant folding for nulls."
Hiroshi Yamauchi [Tue, 10 Nov 2015 19:55:05 +0000 (19:55 +0000)]
Merge "Rosalloc fast path in assembly for x86_64."
Vladimir Marko [Tue, 10 Nov 2015 19:53:01 +0000 (19:53 +0000)]
Optimizing: Clean up Equal/NotEqual constant folding for nulls.
Change-Id: I17766395092ec61df61ef0b9ae4c37fd38164a3b