OSDN Git Service
Bill Buzbee [Mon, 16 Jun 2014 20:16:00 +0000 (20:16 +0000)]
Merge "ART: Instanceof returns true if ref == 0 on x86-64"
Hiroshi Yamauchi [Mon, 16 Jun 2014 19:56:59 +0000 (19:56 +0000)]
Merge "Add read barriers for the class and the intern tables."
Bill Buzbee [Mon, 16 Jun 2014 19:52:09 +0000 (19:52 +0000)]
Merge "Use GetLow() to access low register of a pair"
Mark Mendell [Mon, 16 Jun 2014 17:03:42 +0000 (13:03 -0400)]
Use GetLow() to access low register of a pair
A couple of spots were missed.
Change-Id: I14bd1d1bec4b497b742ef9edaf762e0c8e4ebf79
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
buzbee [Mon, 16 Jun 2014 17:47:23 +0000 (17:47 +0000)]
Merge "Arm64 hard-float"
buzbee [Thu, 12 Jun 2014 21:56:32 +0000 (14:56 -0700)]
Arm64 hard-float
Basic enabling of hard-float for Arm64. In future CLs we'll
consolidate the various targets - there is a lot of overlap.
Compilation remains turned off in this CL, but I expect
to enable a subset shortly. With compilation fully enabled
(including the EXPERIMENTAL opcodes with the exception of
REM and THROW), we get the following run-test results:
003-omnibus-opcode failures:
Classes.checkCast
Classes.arrayInstance
UnresTest2
Haven't gone deep, but these appear to be related to throw/catch and/or
stacktrace.
For REM, the generated code looks reasonable to me - my guess is that
we've got something wrong on the transition to the runtime. Haven't
looked deeper yet, though.
The bulk of the other failure also appear to be related to transitioning
to the runtime system, or handling try/catch.
run-test status:
Status with optimizations disabled, REM_FLOAT/DOUBLE and THROW disabled:
succeeded tests: 94
failed tests: 22
failed: 003-omnibus-opcodes
failed: 004-annotations
failed: 009-instanceof2
failed: 024-illegal-access
failed: 025-access-controller
failed: 031-class-attributes
failed: 044-proxy
failed: 045-reflect-array
failed: 046-reflect
failed: 058-enum-order
failed: 062-character-encodings
failed: 063-process-manager
failed: 064-field-access
failed: 068-classloader
failed: 071-dexfile
failed: 083-compiler-regressions
failed: 084-class-init
failed: 086-null-super
failed: 087-gc-after-link
failed: 100-reflect2
failed: 107-int-math2
failed: 201-built-in-exception-detail-messages
Change-Id: Ib66209285cad8998d77a14781de300af02a96b15
Alexei Zavjalov [Mon, 9 Jun 2014 16:27:46 +0000 (23:27 +0700)]
ART: Instanceof returns true if ref == 0 on x86-64
On x86-64 platform rl_result and kArg0 is not the same, so we have to
handle this in the proper way.
Change-Id: Ia651f5dd8dd2cf584f510739ed58b46197dfee94
Signed-off-by: Alexei Zavjalov <alexei.zavjalov@intel.com>
Hiroshi Yamauchi [Fri, 13 Jun 2014 23:44:55 +0000 (16:44 -0700)]
Add read barriers for the class and the intern tables.
Add read barriers for the strong roots in the intern table and the
(strong) roots in the class table to make possible concurrent scanning
of them.
Bug:
12687968
Change-Id: If6edc33a37e65a8494e66dc3b144138b1530367f
buzbee [Sat, 14 Jun 2014 20:10:01 +0000 (20:10 +0000)]
Merge "Quick compiler:interpret on type mismatch"
buzbee [Sat, 14 Jun 2014 19:33:09 +0000 (12:33 -0700)]
Quick compiler:interpret on type mismatch
The Dex specification is a bit loose - particularly in regards
to types. The Quick compiler, though, makes some assumptions
about types. In particular, it doesn't expect to encounter the
use of a long or double virtual register pair that was defined
by two independent 32-bit operations.
dx does not create such patterns (at least in recent memory).
However, at least one such case exists in the wild. The next
version of the Dex specification will add more type constraints
and formally disallow such cases. Meanwhile, existing code will
be handled by identifying these cases an reverting to interpretation
for the offending method.
Fix for internal b/
15616104
Change-Id: Ibe9c423be9a952ff58cf8d985aa164885b8dd2ae
Andreas Gampe [Fri, 13 Jun 2014 22:50:49 +0000 (22:50 +0000)]
Merge "ART: Hide unreachable basic blocks in the compiler"
Andreas Gampe [Fri, 13 Jun 2014 20:44:40 +0000 (13:44 -0700)]
ART: Hide unreachable basic blocks in the compiler
Unreachable blocks are not handled uniformly in the optimization
passes. Uniformly hide them to avoid initialization errors.
Bug:
15573463
Change-Id: Ia9e89fa357d5672a6cd8389f28a06ff618fe60ee
Dave Allison [Fri, 13 Jun 2014 18:51:05 +0000 (18:51 +0000)]
Merge "Don't use signal handlers in interpreter only mode"
Dave Allison [Fri, 13 Jun 2014 16:09:41 +0000 (09:09 -0700)]
Don't use signal handlers in interpreter only mode
Bug:
15543631
Change-Id: I8f251b5a66487dce3ac3127bdf3b5bc824d0e4bc
Ian Rogers [Sat, 14 Jun 2014 06:51:15 +0000 (06:51 +0000)]
Merge "x86_64: Enable compilation"
Dmitry Petrochenko [Fri, 6 Jun 2014 08:18:14 +0000 (15:18 +0700)]
x86_64: Enable compilation
This patch enables compilation of all methods for x86_64
except image.
Change-Id: Ie210809f2595cc25da688a4ad0363c258bcf8233
Signed-off-by: Dmitry Petrochenko <dmitry.petrochenko@intel.com>
Brian Carlstrom [Sat, 14 Jun 2014 05:40:21 +0000 (05:40 +0000)]
Merge "Add extra information to InitImageMethodVisitor CHECK"
Brian Carlstrom [Sat, 14 Jun 2014 01:23:20 +0000 (18:23 -0700)]
Add extra information to InitImageMethodVisitor CHECK
Bug:
15616246
(cherry picked from commit
21f1740e845505bf437dcf128912ce218ff92af9)
Change-Id: Ib698229cd56151c682d44e8dd59d8df18fb5a393
Hiroshi Yamauchi [Fri, 13 Jun 2014 23:51:57 +0000 (23:51 +0000)]
Merge "Avoid a non-root monitor vector in Dbg::GetOwnedMonitors()."
Hiroshi Yamauchi [Fri, 13 Jun 2014 22:08:05 +0000 (15:08 -0700)]
Avoid a non-root monitor vector in Dbg::GetOwnedMonitors().
Following up on CL 96571, remove the need to use stack-scoped vectors,
one of which is a non-root vector that holds Object*, which could
cause a GC bug.
Bug:
12687968
Change-Id: I8582fe6bfe71bae7018dbf844fe75ebe989b6535
Ian Rogers [Fri, 13 Jun 2014 22:05:35 +0000 (22:05 +0000)]
Merge "x86_64: Disable all optimizations and fix bugs"
Chao-ying Fu [Wed, 11 Jun 2014 01:11:11 +0000 (18:11 -0700)]
x86_64: Disable all optimizations and fix bugs
This disables all optimizations and ensures that art tests still pass.
Change-Id: I43217378d6889bb04f4d064f8d53cb3ff4c20aa0
Signed-off-by: Chao-ying Fu <chao-ying.fu@intel.com>
Signed-off-by: Serguei Katkov <serguei.i.katkov@intel.com>
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
Brian Carlstrom [Fri, 13 Jun 2014 21:18:04 +0000 (21:18 +0000)]
Merge "Optionally add debug symbols to ELF files made by quick compiler."
Alex Light [Fri, 6 Jun 2014 22:45:32 +0000 (15:45 -0700)]
Optionally add debug symbols to ELF files made by quick compiler.
Added debug symbols to ELF files created by dex2oat using
the quick compiler. Adds two flags --include-debug-symbols and
--no-include-debug-symbols for dex2oat that control the inclusion of these
debug symbols. Debug info is added by default if kIsDebugBuild is true.
Fixed bug where Intel DWARF information would not correctly deal with
deduplicated code the binary.
Changed the portable compiler code path in dex2oat.cc so that symbols
will not be stripped when run with --include-debug-symbols.
Change-Id: Ia2eb2f654dedf0e5e8606f7744e05b8d14155fb1
Andreas Gampe [Fri, 13 Jun 2014 18:44:18 +0000 (18:44 +0000)]
Merge "ART: Clean up DexFile verifier API"
Andreas Gampe [Fri, 13 Jun 2014 18:35:12 +0000 (11:35 -0700)]
ART: Clean up DexFile verifier API
Use an out arg instead of widening a return type.
Change-Id: I70744eec408b0c6f3487f168faf510869277daff
Ian Rogers [Fri, 13 Jun 2014 18:34:21 +0000 (18:34 +0000)]
Merge "Don't verify runtime only opcodes in the data-flow phase of verification."
Ian Rogers [Fri, 13 Jun 2014 17:31:28 +0000 (10:31 -0700)]
Don't verify runtime only opcodes in the data-flow phase of verification.
Bug:
15570344
Change-Id: I0304e8742a1d0318783ba72862e684ab91f63d0e
Andreas Gampe [Fri, 13 Jun 2014 18:01:51 +0000 (18:01 +0000)]
Merge "ART: Check the number of invocation args in verifier"
Andreas Gampe [Fri, 13 Jun 2014 02:35:05 +0000 (19:35 -0700)]
ART: Check the number of invocation args in verifier
Check the number of invocation arguments against the method's
prototype signature. It could happen that the method wasn't
available, yet, in which case we would let a clearly wrong
instruction through.
This generalizes https://android-review.googlesource.com/#/c/97779/
Bug:
15570483
Change-Id: Ie81aff3c1166a2b2bf1385414dff2e22fbb40ef2
Andreas Gampe [Thu, 12 Jun 2014 23:44:57 +0000 (23:44 +0000)]
Merge "ART: Invoke_virtual needs at least receiver parameter"
Andreas Gampe [Thu, 12 Jun 2014 22:44:40 +0000 (15:44 -0700)]
ART: Invoke_virtual needs at least receiver parameter
Makes the method verifier reject invoke_virtual instructions that
claim they do not need any parameters.
Bug:
15570946
Change-Id: Ia76fddb7a08551d6c08654950e7085f09eacf430
Bill Buzbee [Thu, 12 Jun 2014 21:01:46 +0000 (21:01 +0000)]
Merge "AArch64: Enable MOVE_*, some CONST_*, CMP_*."
Zheng Xu [Thu, 12 Jun 2014 02:22:33 +0000 (10:22 +0800)]
AArch64: Enable MOVE_*, some CONST_*, CMP_*.
With the fixes of GenArithImmOpLong, GenShiftOpLong, OpRegImm,
OpRegRegImm, OpRegRegImm64, EncodeLogicalImmediate and fmov.
Change-Id: I8cae4f921d5150a6b8e4803ca4dee553928d1a58
Andreas Gampe [Sat, 14 Jun 2014 02:34:53 +0000 (02:34 +0000)]
Merge "ART: Dex file verifier can't blindly use GetDescriptor"
Andreas Gampe [Fri, 13 Jun 2014 22:31:40 +0000 (15:31 -0700)]
ART: Dex file verifier can't blindly use GetDescriptor
The dex file verifier cannot call GetDescriptor without checking
the index first.
Bug:
15615908
Change-Id: Iec5e20b77aac3e6c25fb94b1b630d0846c0b8f33
Andreas Gampe [Sat, 14 Jun 2014 02:34:44 +0000 (02:34 +0000)]
Merge "ART: Ignore timing issues in debug builds"
Andreas Gampe [Fri, 13 Jun 2014 22:46:46 +0000 (15:46 -0700)]
ART: Ignore timing issues in debug builds
When in debug builds, do not warn about slow verification and
compilation.
Change-Id: Ib982e1c7cf40cf624688bbf6c41f0a7cc53bbc67
Ian Rogers [Sat, 14 Jun 2014 01:03:01 +0000 (01:03 +0000)]
Merge "Tidy x86 assembler."
Ian Rogers [Fri, 13 Jun 2014 23:38:24 +0000 (16:38 -0700)]
Tidy x86 assembler.
Use helper functions to compute when the kind has a SIB, a ModRM and RegReg
form.
Change-Id: I86a5cb944eec62451c63281265e6974cd7a08e07
Andreas Gampe [Fri, 13 Jun 2014 17:25:16 +0000 (17:25 +0000)]
Merge "ART: Add initialize_type stubs for ARM64; indexof test"
Andreas Gampe [Mon, 9 Jun 2014 21:55:53 +0000 (14:55 -0700)]
ART: Add initialize_type stubs for ARM64; indexof test
Add down-call stubs for ARM64. Add a test for indexof in stub_test.
Change-Id: I7bc045e29e4ca11ded5b318c47544edc1266afdb
Bill Buzbee [Thu, 12 Jun 2014 21:53:06 +0000 (21:53 +0000)]
Merge "AArch64: Add support for inlined methods"
Serban Constantinescu [Wed, 11 Jun 2014 15:43:35 +0000 (16:43 +0100)]
AArch64: Add support for inlined methods
This patch adds support for Arm64 inlined methods.
Change-Id: Ic6aeed6d2d32f65cd1e63cf482f83cdcf958798a
Andreas Gampe [Thu, 12 Jun 2014 21:47:57 +0000 (21:47 +0000)]
Merge "ART: Dex file verifier needs to check proto idx of method id."
Andreas Gampe [Thu, 12 Jun 2014 04:46:05 +0000 (21:46 -0700)]
ART: Dex file verifier needs to check proto idx of method id.
Bug:
15570490
Change-Id: I4e90aea6e77711ee37d14d96a1fc9a2fbb980d94
Bill Buzbee [Thu, 12 Jun 2014 17:39:09 +0000 (17:39 +0000)]
Merge "AArch64: fix MarkGCCard, enabling more MIR opcodes."
Nick Kralevich [Thu, 12 Jun 2014 23:05:01 +0000 (23:05 +0000)]
Merge "consolidate profiles and profile-cache"
Andreas Gampe [Thu, 12 Jun 2014 22:51:22 +0000 (22:51 +0000)]
Merge "ART: Do not allow quickened dex codes in compiler mode."
Andreas Gampe [Thu, 12 Jun 2014 21:35:51 +0000 (14:35 -0700)]
ART: Do not allow quickened dex codes in compiler mode.
The method verifier will reject quickened dex opcodes when the
runtime is in compiler mode, as we do not expect them here.
Bug:
15570344
Change-Id: I072567835139860213f9882cac995370b4548a79
Nick Kralevich [Thu, 12 Jun 2014 01:29:07 +0000 (18:29 -0700)]
consolidate profiles and profile-cache
Ownership problems originally prevented using the same directory
for /data/dalvik-cache/profiles and /data/dalvik-cache/profile-cache.
Now that ownership is assigned via init.rc, we can consolidate these
two directories.
Old profile files are stored in /data/dalvik-cache/profiles/ with
the "@old" extension. We use "@" instead of "." to avoid conflicts
should someone create a package ending in ".old".
Change-Id: Ic1b44009faa30d704855e97631006c4b990a4ad3
Ian Rogers [Wed, 11 Jun 2014 23:45:26 +0000 (23:45 +0000)]
Merge "Fix typo comment by removing."
Ian Rogers [Wed, 11 Jun 2014 23:44:24 +0000 (16:44 -0700)]
Fix typo comment by removing.
Change-Id: I09764f194317634164be9431328013205fba2910
Ian Rogers [Wed, 11 Jun 2014 23:39:15 +0000 (23:39 +0000)]
Merge "ART: Add mul_long bytecode for X86-64"
Alexei Zavjalov [Wed, 11 Jun 2014 11:26:40 +0000 (18:26 +0700)]
ART: Add mul_long bytecode for X86-64
This patch implements the inlined version of the mul-long
bytecode for X86-64 platform.
Change-Id: I828e5861e17024d33f09223e0245380aef63a622
Signed-off-by: Chao-ying Fu <chao-ying.fu@intel.com>
Signed-off-by: Alexei Zavjalov <alexei.zavjalov@intel.com>
Bill Buzbee [Thu, 12 Jun 2014 17:39:32 +0000 (17:39 +0000)]
Merge "ART: arm64 explicit stack overflow checks"
Stuart Monteith [Fri, 6 Jun 2014 16:05:08 +0000 (17:05 +0100)]
ART: arm64 explicit stack overflow checks
Implement only the explicit checks for the quick backend for arm64.
Implicit checks require fault handlers, which are currently unimplemented.
CMN + CMP have extended versions implemented for comparisons against the
stack pointer. More extended opcode implementations will need to follow.
Change-Id: I8db297aec73df818b20fe410297800c886701c76
Nicolas Geoffray [Thu, 12 Jun 2014 09:04:03 +0000 (09:04 +0000)]
Merge "Final CL to enable register allocation on x86."
Nicolas Geoffray [Wed, 4 Jun 2014 10:12:39 +0000 (11:12 +0100)]
Final CL to enable register allocation on x86.
This CL implements:
1) Resolution after allocation: connecting the locations
allocated to an interval within a block and between blocks.
2) Handling of fixed registers: some instructions require
inputs/output to be at a specific location, and the allocator
needs to deal with them in a special way.
3) ParallelMoveResolver::EmitNativeCode for x86.
Change-Id: I0da6bd7eb66877987148b87c3be6a983b4e3f858
Andreas Gampe [Thu, 12 Jun 2014 01:36:15 +0000 (01:36 +0000)]
Merge "ART: Check indices in dex file verifier"
Ian Rogers [Wed, 11 Jun 2014 23:17:59 +0000 (23:17 +0000)]
Merge "Fix mem_map_test for Mips."
Douglas Leung [Wed, 11 Jun 2014 18:47:09 +0000 (11:47 -0700)]
Fix mem_map_test for Mips.
Mips' kuseg window is from 0x0 to 0x80000000.
Change-Id: I3e90b04030672c003cdb4f7e9cda414b673b2d04
Signed-off-by: Douglas Leung <douglas@mips.com>
Hiroshi Yamauchi [Wed, 11 Jun 2014 23:04:03 +0000 (23:04 +0000)]
Merge "Remove Object* weak roots from the debugger."
Hiroshi Yamauchi [Mon, 9 Jun 2014 19:11:20 +0000 (12:11 -0700)]
Remove Object* weak roots from the debugger.
The weak roots were converted to JNI weak refs.
Since the weak roots are now normal JNI weak refs, we eliminate the
need to insert read barriers for those weak roots in the debugger and
the need for the GC to have a separate step to update them as part of
the system weak sweeping.
Bug:
12687968
Change-Id: If16396d4713457b8af4f1ea6a0c6ec6799cb615e
Brian Carlstrom [Wed, 11 Jun 2014 23:01:40 +0000 (23:01 +0000)]
Merge changes I70a3b6fe,Ib12f3474
* changes:
Fix build by converting from bionic to elfutils version of elf.h
Revert "Revert "Rewrite ElfWriterQuick to make it more modular.""
Brian Carlstrom [Wed, 11 Jun 2014 21:22:02 +0000 (14:22 -0700)]
Fix build by converting from bionic to elfutils version of elf.h
Change-Id: I70a3b6fedae6d290a4f01d4ae8fdd726b8aaa90f
Brian Carlstrom [Wed, 11 Jun 2014 21:54:46 +0000 (14:54 -0700)]
Revert "Revert "Rewrite ElfWriterQuick to make it more modular.""
This reverts commit
35f72251e722cad03a08e8ceacbd5c244eab5c29.
Brian Carlstrom [Wed, 11 Jun 2014 21:35:05 +0000 (21:35 +0000)]
Merge "Revert "Rewrite ElfWriterQuick to make it more modular.""
Brian Carlstrom [Wed, 11 Jun 2014 21:24:53 +0000 (21:24 +0000)]
Revert "Rewrite ElfWriterQuick to make it more modular."
This reverts commit
ad13cb2db8c4a64298ed85c82ae7366ee73f7640.
Change-Id: Iee1d1fd08a008d7d382c181cb352ef46a6eaf5d3
Andreas Gampe [Wed, 11 Jun 2014 21:56:39 +0000 (21:56 +0000)]
Merge "x86_64: Fix issue in JNI compiler"
avignate [Wed, 4 Jun 2014 10:59:44 +0000 (17:59 +0700)]
x86_64: Fix issue in JNI compiler
This patch fixed 64 bit conversion issue in Immediate.
The issue is inside type conversion of Immediate:
explicit Immediate(int64_t value) : value_(value) {}.
In case of the following example we'll have unexpected value in Immediate:
size_t t = 1;
Immediate(-t) will contain value
4294967295 because by conversion
rules -t is first transformed to unsigned and then transformed
to 64bit (size64_t). The issue can be fixed by using long value
as a parameter of Immediate constructor.
Added tests for BuildFrame, RemoveFrame, IncreaseFrameSize and
DecreaseFrameSize to assembler_x86_64_test.
Change-Id: I0652bac83e4266fd4153bc6a4e9d3aae7cc4cb6f
Signed-off-by: avignate <aleksey.v.ignatenko@intel.com>
Signed-off-by: Dmitry Petrochenko <dmitry.petrochenko@intel.com>
Ian Rogers [Wed, 11 Jun 2014 21:47:08 +0000 (21:47 +0000)]
Merge "ART: Update the DecodedInstruction for the Fused extended bytecodes."
Brian Carlstrom [Wed, 11 Jun 2014 20:33:09 +0000 (20:33 +0000)]
Merge "Rewrite ElfWriterQuick to make it more modular."
Ian Rogers [Wed, 11 Jun 2014 20:14:09 +0000 (20:14 +0000)]
Merge "Fix compilation errors when setting kTracing = true."
Douglas Leung [Wed, 11 Jun 2014 18:41:11 +0000 (11:41 -0700)]
Fix compilation errors when setting kTracing = true.
Change-Id: I946c98021166bd099983a8f50a18b57d8eac6af6
Signed-off-by: Douglas Leung <douglas@mips.com>
Ian Rogers [Wed, 11 Jun 2014 20:10:02 +0000 (20:10 +0000)]
Merge "Fix host Mac build."
Ian Rogers [Wed, 11 Jun 2014 20:00:44 +0000 (13:00 -0700)]
Fix host Mac build.
Ptr - ptr is an int not size_t on the Mac.
Fix use of assembly literal macro.
Remove forgotten about WITH_HOST_DALVIK.
Change-Id: I64db1927b2c6a145b6a96ffcc33d704914bba1bb
Ian Rogers [Wed, 11 Jun 2014 17:29:23 +0000 (17:29 +0000)]
Merge "Remove deprecated WITH_HOST_DALVIK."
Ian Rogers [Fri, 6 Jun 2014 03:48:42 +0000 (20:48 -0700)]
Remove deprecated WITH_HOST_DALVIK.
Bug:
13751317
Fix the Mac build:
- disable x86 selector removal that causes OS/X 10.9 kernel panics,
- madvise don't need does zero memory on the Mac, factor into MemMap
routine,
- switch to the elf.h in elfutils to avoid Linux kernel dependencies,
- we can't rely on exclusive_owner_ being available from other pthread
libraries so maintain our own when futexes aren't available (we
can't rely on the OS/X 10.8 hack any more),
- fix symbol naming in assembly code,
- work around C library differences,
- disable backtrace in DumpNativeStack to avoid a broken libbacktrace
dependency,
- disable main thread signal handling logic,
- align the stack in stub_test,
- use $(HOST_SHLIB_SUFFIX) rather than .so in host make file variables.
Not all host tests are passing on the Mac with this change. dex2oat
works as does running HelloWorld.
Change-Id: I5a232aedfb2028524d49daa6397a8e60f3ee40d3
Ian Rogers [Tue, 10 Jun 2014 23:56:41 +0000 (23:56 +0000)]
Merge "Workaround frame size issues."
Ian Rogers [Tue, 10 Jun 2014 23:31:03 +0000 (16:31 -0700)]
Workaround frame size issues.
x86 and x86-64 are exceeding the frame size for the switch interpreter.
The SOMETIMES_INLINE hack doesn't work with GCC as inline and the noinline
attribute are mutually exclusive. As a temporary solution move the effected
code to the the interpreter_common.cc file.
Bug:
14882674
Change-Id: Id5383ef5436046b36565cd1d76de8e3d59f42cff
Ian Rogers [Tue, 10 Jun 2014 23:34:42 +0000 (23:34 +0000)]
Merge "X86_64: Proper IMT fix"
Brian Carlstrom [Tue, 10 Jun 2014 23:21:40 +0000 (23:21 +0000)]
Merge "Fix for test-art-host"
Brian Carlstrom [Tue, 10 Jun 2014 22:42:52 +0000 (15:42 -0700)]
Fix for test-art-host
Courtesy of dallison@
Change-Id: I160a5c84c176f22168bf236899349b6087494807
Vladimir Marko [Thu, 12 Jun 2014 18:12:51 +0000 (18:12 +0000)]
Merge "Rewrite use/def masks to support 128 bits."
Vladimir Marko [Fri, 6 Jun 2014 13:50:36 +0000 (14:50 +0100)]
Rewrite use/def masks to support 128 bits.
Reduce LIR memory usage by holding masks by pointers in the
LIR rather than directly and using pre-defined const masks
for the common cases, allocating very few on the arena.
Change-Id: I0f6d27ef6867acd157184c8c74f9612cebfe6c16
Vladimir Marko [Thu, 12 Jun 2014 14:27:31 +0000 (14:27 +0000)]
Merge "Add D/CHECK_CONSTEXPR macros for use in constexpr context."
Vladimir Marko [Wed, 11 Jun 2014 11:12:07 +0000 (12:12 +0100)]
Add D/CHECK_CONSTEXPR macros for use in constexpr context.
Make most RegStorage methods constexpr, make StorageSize()
const.
Change-Id: Ie63faa3b081094b3dd30352c6240fdf10ef8b87e
Vladimir Marko [Thu, 12 Jun 2014 17:15:23 +0000 (17:15 +0000)]
Merge "Fix reflection access check for attached native threads."
Vladimir Marko [Thu, 12 Jun 2014 14:22:31 +0000 (15:22 +0100)]
Fix reflection access check for attached native threads.
Bug:
15539150
Change-Id: Ie580cad1c0169d550162bcc878b07643259323c4
Vladimir Marko [Thu, 12 Jun 2014 16:08:46 +0000 (16:08 +0000)]
Merge "Clean up ArmMirToLir::LoadDispBody()/StoreDispBody()."
Sebastien Hertz [Wed, 11 Jun 2014 12:32:31 +0000 (12:32 +0000)]
Merge "Fix exception reporting from interpreter"
Sebastien Hertz [Fri, 23 May 2014 06:59:42 +0000 (08:59 +0200)]
Fix exception reporting from interpreter
To comply with JDWP exception report rules, we must report an exception at the
location of the throw (or the first instruction encountered after a native
call). To do this, we use the CatchLocationFinder visitor to look for a catch
handler until we reach a native frame or the top frame.
Because interpreter handles pending exception on a method-by-method basis, we
need a flag to remember we already reported the exception and avoid reporting
it multiple times when unwinding methods. The drawback is we need to maintain
the state of this flag. We clear it when the exception is cleared. In the case
we temporarily clear the exception (when finding a catch handler for instance),
we restore the flag to its previous value at the same time we restore the
pending exception.
Bump oat version to force recompilation because we modify Thread offsets.
Bug:
14402770
Change-Id: Ic059c58f80b2023b118038301f8f0a24f1e18241
Andreas Gampe [Wed, 11 Jun 2014 05:15:19 +0000 (05:15 +0000)]
Merge "ART: Fix MIPS build"
Andreas Gampe [Wed, 11 Jun 2014 04:57:00 +0000 (21:57 -0700)]
ART: Fix MIPS build
Wrong name for an assembly file. Also, registers need $ prefix
in our assembler.
Change-Id: Ic2d5961ec7d0c8e65ca5b27d681487da847c9065
Andreas Gampe [Wed, 11 Jun 2014 03:13:42 +0000 (03:13 +0000)]
Merge "ART: Move __memcmp16 from Bionic to ART"
Andreas Gampe [Tue, 10 Jun 2014 23:10:56 +0000 (16:10 -0700)]
ART: Move __memcmp16 from Bionic to ART
Handle __memcmp16 / MemCmp16 in ART. Import assembly implementations
for arm and mips from Bionic. Use a generic C version for all other
platforms.
Removes the memcmp16 quick entrypoint, as it is never used. Bump
the oat version and update thread.cc and checks to reflect the
structural change.
Change-Id: I54a5a1da2a0a43ef271c8aeda0bf2276b8b11ac6
Vladimir Marko [Mon, 16 Jun 2014 10:32:40 +0000 (10:32 +0000)]
Merge "ART: PassMEDataHolder should have a data field"
Jean Christophe Beyler [Mon, 9 Jun 2014 17:18:26 +0000 (10:18 -0700)]
ART: PassMEDataHolder should have a data field
For passes that would like to pass data around while working, let them actually
allocate it if need be in the Start, use it during the Worker, and
potentially free it during the End.
Change-Id: I03238ef17158f7e2615d6defe5d7f15432c8a511
Signed-off-by: Jean Christophe Beyler <jean.christophe.beyler@intel.com>
Calin Juravle [Mon, 16 Jun 2014 09:39:33 +0000 (09:39 +0000)]
Merge "Relax limits for profiler options"