| f517e283d477dd2ae229ee3f054120c6953895db |
|
28-Apr-2016 |
Andreas Gampe <agampe@google.com> |
ART: Disambiguate access-checks mode from lock-counting
Lock-counting (when structural locking verification failed) is a
special sub-mode of access-checks and must be disambiguated, because
we currently use access-checks mode class-wide when at least one
method soft-fails, but do not stop the compiler/JIT to compile
the "working" methods. So we may end up in the access-checks
interpreter for a working method through deopt without knowing
which locks are already held.
Bug: 28351535
Change-Id: I083032f064d88df8f8f0611ad8b57d1b39cd09fb
|
| f3f9cf6b65c4bcf9ea44253188d8d910b7cf7e64 |
|
14-Apr-2016 |
Nicolas Geoffray <ngeoffray@google.com> |
Add weight to compiled/interpreter transitions.
Also:
- Cleanup logging.
- Check ArtMethod status before adding compilation requests.
- Don't request osr compilation if we know AddSamples does not come
from a back edge.
Bug: 27865109
(cherry picked from commit 71cd50fb67fa48667b0ab59aa436a582c04ba43d)
Change-Id: Icbe89fe6cc495b113616391a8f257758d34b4b60
|
| bb11c8b1219f5b4b3154c2c83fca19ec8add6646 |
|
12-Apr-2016 |
Nicolas Geoffray <ngeoffray@google.com> |
Remove the JIT from the instrumentation framework.
This was slowing down the interpreter for no reason.
Also, call AddSamples for invoke-static and invoke-direct.
bug:27865109
(cherry picked from commit 274fe4adcb0610a9920be7814d9beb9cac6417ce)
Change-Id: I3519456ac8e0c7211cbe3f12e88d134beee87479
|
| 1d011d9306fd4ff57d72411775d415a86f5ed398 |
|
04-Apr-2016 |
Bill Buzbee <buzbee@google.com> |
Revert "Revert "Revert "Revert "ART: Improve JitProfile perf in arm/arm64 mterp""""
Bug: 28081559
This reverts commit 961ea9fe42edcc2c57469bf451d1ca421da5cd59.
Change-Id: I98a5bb8112646706ae7bd73bf6393cb956466be3
|
| d77798356724ebab2230f4623d4f187952c0fa35 |
|
05-Apr-2016 |
Vladimir Marko <vmarko@google.com> |
Revert "Remove many `instrumentation` variables from goto interpreter."
This reverts commit 4cb22b4f6d0afbdf424369df99c6eb11999f2473.
Change-Id: I1d84965c2cde2ffc2ebdb46fb5251691ea12569b
|
| 961ea9fe42edcc2c57469bf451d1ca421da5cd59 |
|
01-Apr-2016 |
Hiroshi Yamauchi <yamauchi@google.com> |
Revert "Revert "Revert "ART: Improve JitProfile perf in arm/arm64 mterp"""
This reverts commit 4a8ac9cee4312ac910fabf31c64d28d4c8362836.
570-checker-osr intermittently failing.
Bug: 27939339
|
| 4cb22b4f6d0afbdf424369df99c6eb11999f2473 |
|
31-Mar-2016 |
Vladimir Marko <vmarko@google.com> |
Remove many `instrumentation` variables from goto interpreter.
Clean up after
https://android-review.googlesource.com/210440 .
Change-Id: I7e7b1863a5627365de8d9a74f559ffe56cb4fdfa
|
| d6190dcdfb1f247ff0bf8268dacf413c93b58cf5 |
|
31-Mar-2016 |
Calin Juravle <calin@google.com> |
Revert "Revert "Revert "ART: Improve JitProfile perf in arm/arm64 mterp"""
This reverts commit 4a8ac9cee4312ac910fabf31c64d28d4c8362836.
Bug: 27939339
|
| 4a8ac9cee4312ac910fabf31c64d28d4c8362836 |
|
25-Mar-2016 |
Bill Buzbee <buzbee@google.com> |
Revert "Revert "ART: Improve JitProfile perf in arm/arm64 mterp""
Ready for review.
This reverts commit 6aef867f4d1a98a12bcdd65e9bf2ff894f0f2d7e.
Change-Id: I5d53ed2bedc7e429ce7d3cdf80b6696a9628740e
|
| 6aef867f4d1a98a12bcdd65e9bf2ff894f0f2d7e |
|
25-Mar-2016 |
Calin Juravle <calin@google.com> |
Revert "ART: Improve JitProfile perf in arm/arm64 mterp"
This reverts commit c1d6b341eed646e5adafc6c4fd4e3748f0292368.
|
| c1d6b341eed646e5adafc6c4fd4e3748f0292368 |
|
02-Mar-2016 |
buzbee <buzbee@google.com> |
ART: Improve JitProfile perf in arm/arm64 mterp
ART currently requires two profiling-related things from the
interpreters: hotness updates and OSR switch checks. The hotness
updates previously used the existing instrumentation framework - which
is flexible, but quite heavyweight. For most things, the
instrumentation framework overhead is acceptable, but because we do a
hotness update on every backwards branch the overhead is unacceptable.
Prior to this CL, branch profiling dominates interpreter cost.
Here, we bypass the instrumentation framework for hotness updates
and deliver a significant performance improvement. Running
interpreter-only (dalvikvm -Xint) on a Nexus 6, we see the logic
subtest of Caffeinemark improving from 2600 to 9200, and the
overall score going from 1979 to over 3000. Compared to the
C++ switch interpreter, we see a 6x improvement on the branchy logic
subtest and a 2.6x improvement overall.
Compared with the previous mterp which did not have support for
jit profiling, we see a few (1% to 5%) performance loss on the
standard command-line benchmarks. I consider this acceptable
(we could create an alternate non-profiling mterp which would
have no penalty, but I don't consider this overhead big enough to
justify that).
Change-Id: I50b5b8c5ed8ebda3c8b4e65d27ba7393c3feae04
|
| f1dcaccfac4a9e022ea49752a774552b2f791489 |
|
24-Feb-2016 |
buzbee <buzbee@google.com> |
ART: Profile all branches for on-stack replacement
Change the switch, goto and mterp interpreters to profile
not-taken as well as taken branches. This allows for on-stack
replacement when the cfg has been rearranged such that the loop
header was originally the fallthrough of a Dalvik byte-code branch.
Note that this increases the already-heavy cost of branch profiling.
Measuring on a Nexus 6 using a very branchy benchmark (logic subtest
from Caffeinemark), we see:
No profiling Taken only Taken & not-taken
mterp 9728 3434 2384
C++ goto 3914 2422 2037
C++ switch 2986 2411 2112
As measured, the cost of branch profiling is dominating execution
time. This will be addressed in follow-up CLs.
Change-Id: Ibc858f317398dd991ed8e4f3c3d72bd4c9a60594
|
| b331febbab8e916680faba722cc84b66b84218a3 |
|
05-Feb-2016 |
Nicolas Geoffray <ngeoffray@google.com> |
Revert "Revert "Implement on-stack replacement for arm/arm64/x86/x86_64.""
This reverts commit bd89a5c556324062b7d841843b039392e84cfaf4.
Change-Id: I08d190431520baa7fcec8fbdb444519f25ac8d44
|
| bd89a5c556324062b7d841843b039392e84cfaf4 |
|
05-Feb-2016 |
David Brazdil <dbrazdil@google.com> |
Revert "Implement on-stack replacement for arm/arm64/x86/x86_64."
DCHECK whether loop headers are covered fails.
This reverts commit 891bc286963892ed96134ca1adb7822737af9710.
Change-Id: I0f9a90630b014b16d20ba1dfba31ce63e6648021
|
| 891bc286963892ed96134ca1adb7822737af9710 |
|
29-Jan-2016 |
Nicolas Geoffray <ngeoffray@google.com> |
Implement on-stack replacement for arm/arm64/x86/x86_64.
High-level overview:
- osr_method_threshold is used to know when to compile a method
in osr mode (-> treat all loops as irreducible).
- branch instructions in the compiler query whether they can
jump to an osr method.
- An osr entry point is found through the stack maps: if a stack
map is duplicated in the CodeInfo, it is an osr entry point.
Change-Id: Ifb39338cd281e2c7eccce67f4e18d46428be71e4
|
| 734f3aa9f758236b6f1ace6347895af90b0d813d |
|
28-Jan-2016 |
buzbee <buzbee@google.com> |
Fix "Never Interpret" option for all interpreters
CL 196596 added support for an option to bypass interpretation.
However, it only covers 2 our of 3 interpreters (missing mterp).
This change moves the control up a level to the common interpreter
entry where it will take effect before we select which interpreter
to use.
Also, it corrects a somewhat academic bug in that the existing
code that assumes that (dex_pc == 0) means that we just entered a
method. If a method's dex code internally branched to address 0,
we could issue bogus method entry events. By moving this test up
a level, we should avoid this situation. Note, though, that dx
would never generate this pattern, and it's hard to imagine even
hand-generated dex code that would trigger a deoptimization in this
situation.
Change-Id: I6684bbf63570e02f5b01ce423c656889a890de7d
|
| 05d241565f36df825cf56a4f1b61bfb7e4dcb056 |
|
06-Jan-2016 |
Siva Chandra <sivachandra@google.com> |
Add option to never interpret.
Change-Id: Ib6d6170fa60c77c2e6a8844964f14fa0de4c9e7b
|
| 81f0f953c4bb159997046c962d44cb1898b1778d |
|
20-Jan-2016 |
Nicolas Geoffray <ngeoffray@google.com> |
Move backward branch instrumentation to all branch.
This is in preparation for osr, and later dead branch profiling.
Change-Id: I4597c0784c894c310f295dd534967d67dcefa419
|
| eb7c144a6aff7da673ba53d501c46f00311d4d7f |
|
31-Aug-2015 |
Alex Light <allight@google.com> |
Add initial default method support to Art
This commit starts the process of adding default methods and their
associated pieces to ART.
This adds full support for calling default methods using
invoke-interface and invoke-virtual on objects implementing the
interfaces. Verifier is changed to allow this when the runtime is
started with -Xexperimental:default-methods.
This also adds support for defining and calling static methods on
interface classes with invoke-static.
Directly calling overridden default methods using invoke-super is not
yet supported.
This adds 5 new run-tests for this functionality.
Bug: 24618811
Change-Id: I35ca800d99d3329348b277789b70ceeeba6e7f03
|
| 03ec930faded5bbfa096533ce60b6893847922db |
|
28-Aug-2015 |
Andreas Gampe <agampe@google.com> |
ART: Lock counting in the interpreter
To support structured locking when balanced-locking verification
failed, count lock and unlock operations in the verifier.
Bug: 23502994
Change-Id: I2bb915da6e3d43c49723a943b42d4d5a7c939aa1
|
| 6918bf13eb855b3aa8ccdddda2d27ae8c60cec56 |
|
28-Sep-2015 |
Igor Murashkin <iam@google.com> |
Revert "Revert "lambda: Experimental support for capture-variable and liberate-variable""
This reverts commit 7bbb80ab52c203e44d2ded2c947b3b03b4b31ec4.
Change-Id: If806ce5c6c5e96fdb2c3761dee096f74e7e5b001
|
| 7bbb80ab52c203e44d2ded2c947b3b03b4b31ec4 |
|
27-Sep-2015 |
Nicolas Geoffray <ngeoffray@google.com> |
Revert "lambda: Experimental support for capture-variable and liberate-variable"
Test fails.
This reverts commit b72123440d8541362ebdb131436f9dbdda5fd329.
Change-Id: Ic9ed92f8c826d8465eb36b746dc44af05caf041c
|
| b72123440d8541362ebdb131436f9dbdda5fd329 |
|
25-Sep-2015 |
Igor Murashkin <iam@google.com> |
lambda: Experimental support for capture-variable and liberate-variable
Supports capturing/liberating any primitive variables.
No support for capturing objects/lambdas yet since they would both
need GC changes to track roots through closures.
Change-Id: Ibfb68bfe4c579dbf93823aac4c0e6ac8f6360c5d
|
| 05792b98980741111b4d0a24d68cff2a8e070a3a |
|
03-Aug-2015 |
Vladimir Marko <vmarko@google.com> |
ART: Move DexCache arrays to native.
This CL has a companion CL in libcore/
https://android-review.googlesource.com/162985
Change-Id: Icbc9e20ad1b565e603195b12714762bb446515fa
|
| 90443477f9a0061581c420775ce3b7eeae7468bc |
|
17-Jul-2015 |
Mathieu Chartier <mathieuc@google.com> |
Move to newer clang annotations
Also enable -Wthread-safety-negative.
Changes:
Switch to capabilities and negative capabilities.
Future work:
Use capabilities to implement uninterruptible annotations to work
with AssertNoThreadSuspension.
Bug: 20072211
Change-Id: I42fcbe0300d98a831c89d1eff3ecd5a7e99ebf33
|
| 2ee54e249ad21c74f29a161e248bebe7d22fddf1 |
|
18-Jun-2015 |
Igor Murashkin <iam@google.com> |
runtime: Partially implement box-lambda and unbox-lambda experimental opcodes
These opcodes are not yet fully specified, and *will* change before they become shippable.
Do not write production code against experimental opcodes.
--
Implement partial interpreter support for new dex instructions box/unbox-lambda.
* box-lambda will take a closure and convert it into an Object
* unbox-lambda will take an Object and convert it to a closure
(Currently does not implement object identity or variable capture).
All new opcodes are disabled by default, use runtime option -Xexperimental-lambdas to enable them.
Change-Id: I3c15ccf8a26ccecd1d35808a8c1b4149220f6019
|
| 158f35c98e2ec0d40d2c032b8cdce5fb60944a7f |
|
11-Jun-2015 |
Igor Murashkin <iam@google.com> |
interpreter: Add experimental lambda opcodes for invoke/create-lambda
These opcodes are not yet fully specified, and *will* change before they become shippable.
Do not write production code against experimental opcodes.
--
Implement partial interpreter support for new dex instructions invoke/create-lambda, and a
new opcode format 25x.
* Does not verify, in fact verification will soft fail when we see those opcodes.
* Compilers will punt to interpreter since they don't support new opcodes.
* As there is no way to capture/liberate variables yet, the "closure" is just
an ArtMethod for the time being.
All new opcodes are disabled by default, use runtime option -Xexperimental-lambdas to enable them.
For example:
dalvikvm ... -Xexperimental-lambdas ...
dex2oat --runtime-arg -Xexperimental-lambdas ...
Change-Id: I6c996ca32a9b54ec45ec21d7a959b84dfb8a24eb
|
| 3d21bdf8894e780d349c481e5c9e29fe1556051c |
|
22-Apr-2015 |
Mathieu Chartier <mathieuc@google.com> |
Move mirror::ArtMethod to native
Optimizing + quick tests are passing, devices boot.
TODO: Test and fix bugs in mips64.
Saves 16 bytes per most ArtMethod, 7.5MB reduction in system PSS.
Some of the savings are from removal of virtual methods and direct
methods object arrays.
Bug: 19264997
(cherry picked from commit e401d146407d61eeb99f8d6176b2ac13c4df1e33)
Change-Id: I622469a0cfa0e7082a2119f3d6a9491eb61e3f3d
Fix some ArtMethod related bugs
Added root visiting for runtime methods, not currently required
since the GcRoots in these methods are null.
Added missing GetInterfaceMethodIfProxy in GetMethodLine, fixes
--trace run-tests 005, 044.
Fixed optimizing compiler bug where we used a normal stack location
instead of double on ARM64, this fixes the debuggable tests.
TODO: Fix JDWP tests.
Bug: 19264997
Change-Id: I7c55f69c61d1b45351fd0dc7185ffe5efad82bd3
ART: Fix casts for 64-bit pointers on 32-bit compiler.
Bug: 19264997
Change-Id: Ief45cdd4bae5a43fc8bfdfa7cf744e2c57529457
Fix JDWP tests after ArtMethod change
Fixes Throwable::GetStackDepth for exception event detection after
internal stack trace representation change.
Adds missing ArtMethod::GetInterfaceMethodIfProxy call in case of
proxy method.
Bug: 19264997
Change-Id: I363e293796848c3ec491c963813f62d868da44d2
Fix accidental IMT and root marking regression
Was always using the conflict trampoline. Also included fix for
regression in GC time caused by extra roots. Most of the regression
was IMT.
Fixed bug in DumpGcPerformanceInfo where we would get SIGABRT due to
detached thread.
EvaluateAndApplyChanges:
From ~2500 -> ~1980
GC time: 8.2s -> 7.2s due to 1s less of MarkConcurrentRoots
Bug: 19264997
Change-Id: I4333e80a8268c2ed1284f87f25b9f113d4f2c7e0
Fix bogus image test assert
Previously we were comparing the size of the non moving space to
size of the image file.
Now we properly compare the size of the image space against the size
of the image file.
Bug: 19264997
Change-Id: I7359f1f73ae3df60c5147245935a24431c04808a
[MIPS64] Fix art_quick_invoke_stub argument offsets.
ArtMethod reference's size got bigger, so we need to move other args
and leave enough space for ArtMethod* and 'this' pointer.
This fixes mips64 boot.
Bug: 19264997
Change-Id: I47198d5f39a4caab30b3b77479d5eedaad5006ab
|
| e401d146407d61eeb99f8d6176b2ac13c4df1e33 |
|
22-Apr-2015 |
Mathieu Chartier <mathieuc@google.com> |
Move mirror::ArtMethod to native
Optimizing + quick tests are passing, devices boot.
TODO: Test and fix bugs in mips64.
Saves 16 bytes per most ArtMethod, 7.5MB reduction in system PSS.
Some of the savings are from removal of virtual methods and direct
methods object arrays.
Bug: 19264997
Change-Id: I622469a0cfa0e7082a2119f3d6a9491eb61e3f3d
|
| c3cde2c00bcbe97c9b0c119919500fcbbe824baa |
|
10-Apr-2015 |
Sebastien Hertz <shertz@google.com> |
JDWP: properly combine location events
This CL properly groups JDWP events at the same location: Breakpoint,
Single-step, Method Entry and Method Exit. This is necessary if the
debugger is not the only instrumentation listener. This matches the
behavior of Dalvik, especially for methods with a single return
instruction.
The interpreter was tuned so the instrumentation callbacks were
called to satisfy the debugger with the idea the debugger was the
only instrumentation listener. This is not true when method tracing
is enabled at the same time.
When tracing is enabled, there is always a listener for MethodEntry
and MethodExit events (art::Trace class). However, if the debugger
is only listening to DexPcMoved event (to manage JDWP Breakpoint
event), it will not be notified of this event.
We now properly call all the instrumentation callbacks in the
interpreter and move the logic specific to debugging into the class
DebugInstrumentationListener. This allows to properly group JDWP
location events together depending on the sequence of instrumentation
callbacks.
We add Thread::tls_32bit_sized_values::debug_method_entry_ flag to
remember we just entered a method. It replaces the local variable
notified_method_entry_event in the interpreter and simplifies the
code.
Bump oat version to force recompilation because the layout of the
Thread class is modified.
Bug: 19829329
Bug: 20205350
(cherry picked from commit 9d6bf69ad3012a9d843268fdd5325b6719b6d5f2)
Change-Id: I204af9112e37d2eebc86661fb7c961a41c74e598
|
| 9d6bf69ad3012a9d843268fdd5325b6719b6d5f2 |
|
10-Apr-2015 |
Sebastien Hertz <shertz@google.com> |
JDWP: properly combine location events
This CL properly groups JDWP events at the same location: Breakpoint,
Single-step, Method Entry and Method Exit. This is necessary if the
debugger is not the only instrumentation listener. This matches the
behavior of Dalvik, especially for methods with a single return
instruction.
The interpreter was tuned so the instrumentation callbacks were
called to satisfy the debugger with the idea the debugger was the
only instrumentation listener. This is not true when method tracing
is enabled at the same time.
When tracing is enabled, there is always a listener for MethodEntry
and MethodExit events (art::Trace class). However, if the debugger
is only listening to DexPcMoved event (to manage JDWP Breakpoint
event), it will not be notified of this event.
We now properly call all the instrumentation callbacks in the
interpreter and move the logic specific to debugging into the class
DebugInstrumentationListener. This allows to properly group JDWP
location events together depending on the sequence of instrumentation
callbacks.
We add Thread::tls_32bit_sized_values::debug_method_entry_ flag to
remember we just entered a method. It replaces the local variable
notified_method_entry_event in the interpreter and simplifies the
code.
Bump oat version to force recompilation because the layout of the
Thread class is modified.
Bug: 19829329
Change-Id: I204af9112e37d2eebc86661fb7c961a41c74e598
|
| 848f70a3d73833fc1bf3032a9ff6812e429661d9 |
|
15-Jan-2014 |
Jeff Hao <jeffhao@google.com> |
Replace String CharArray with internal uint16_t array.
Summary of high level changes:
- Adds compiler inliner support to identify string init methods
- Adds compiler support (quick & optimizing) with new invoke code path
that calls method off the thread pointer
- Adds thread entrypoints for all string init methods
- Adds map to verifier to log when receiver of string init has been
copied to other registers. used by compiler and interpreter
Change-Id: I797b992a8feb566f9ad73060011ab6f51eb7ce01
|
| 2cebb24bfc3247d3e9be138a3350106737455918 |
|
22-Apr-2015 |
Mathieu Chartier <mathieuc@google.com> |
Replace NULL with nullptr
Also fixed some lines that were too long, and a few other minor
details.
Change-Id: I6efba5fb6e03eb5d0a300fddb2a75bf8e2f175cb
|
| 45b1597c152af90f6d5792d02b64fd4e7c81ac9d |
|
03-Apr-2015 |
Sebastien Hertz <shertz@google.com> |
Use va_list argument to abort transaction
Creates AbortTransactionV taking a va_list argument and renames
AbortTransaction to AbortTransactionF which calls AbortTransactionV.
This fixes the compiler_driver_test under valgrind.
Change-Id: Ia1c57330091c055ae9e46585a944ce0b78864920
|
| d7cbf8a6629942e7bd315ffae7e1c77b082f3e11 |
|
19-Mar-2015 |
Mathieu Chartier <mathieuc@google.com> |
Change RETURN_VOID_BARRIER to RETURN_VOID_NO_BARRIER
We want to default to having a barrier for the case where we don't
dex to dex.
Bug: 19762303
Change-Id: I60348d89eaf0b9e1e480298afcecbb5f52e8661b
|
| e84e4f77388ec9c1d63fb4b21605eedca2a2e932 |
|
18-Mar-2015 |
Colin Cross <ccross@android.com> |
Remove the Android.mk GCC-only source files hackery
The complexity in Android.mk to deal with GCC-only source files is
unnecessary, use #if !defined(__clang__) around the contents of
interpreter/interpreter_goto_table_impl.cc, the same way
interpreter/interpreter.cc does around references to it.
Bug: 17716550
Change-Id: I775c23b6790d38b0d73a92529c696a31e6a4ae83
|
| 0aa50ce2fb75bfc2e815a0c33adf9b049561923b |
|
10-Mar-2015 |
Nicolas Geoffray <ngeoffray@google.com> |
Remove ThrowLocation.
Note that this is a cleanup change, and has no functionality change.
The ThrowLocation had no use anymore.
Change-Id: I3d2126af1dc673cec3a0453ff3d56a172663a5f6
|
| 14691c5e786e8c2c5734f687e4c96217340771be |
|
05-Mar-2015 |
Nicolas Geoffray <ngeoffray@google.com> |
Compute the right catch location for the debugger.
Also remove tls ThrowLocation, it is not needed anymore.
Change-Id: I78fddf09ce968ca475e39c17fa76d699c589c8d9
|
| e5f13e57ff8fa36342beb33830b3ec5942a61cca |
|
24-Feb-2015 |
Mathieu Chartier <mathieuc@google.com> |
Revert "Revert "Add JIT""
Added missing EntryPointToCodePointer.
This reverts commit a5ca888d715cd0c6c421313211caa1928be3e399.
Change-Id: Ia74df0ef3a7babbdcb0466fd24da28e304e3f5af
|
| a5ca888d715cd0c6c421313211caa1928be3e399 |
|
24-Feb-2015 |
Nicolas Geoffray <ngeoffray@google.com> |
Revert "Add JIT"
Sorry, run-test crashes on target:
0-05 12:15:51.633 I/DEBUG (27995): Abort message: 'art/runtime/mirror/art_method.cc:349] Check failed: PcIsWithinQuickCode(reinterpret_cast<uintptr_t>(code), pc) java.lang.Throwable java.lang.Throwable.fillInStackTrace() pc=71e3366b code=0x71e3362d size=ad000000'
10-05 12:15:51.633 I/DEBUG (27995): r0 00000000 r1 0000542b r2 00000006 r3 00000000
10-05 12:15:51.633 I/DEBUG (27995): r4 00000006 r5 b6f9addc r6 00000002 r7 0000010c
10-05 12:15:51.633 I/DEBUG (27995): r8 b63fe1e8 r9 be8e1418 sl b6427400 fp b63fcce0
10-05 12:15:51.633 I/DEBUG (27995): ip 0000542b sp be8e1358 lr b6e9a27b pc b6e9c280 cpsr 40070010
10-05 12:15:51.633 I/DEBUG (27995):
Bug: 17950037
This reverts commit 2535abe7d1fcdd0e6aca782b1f1932a703ed50a4.
Change-Id: I6f88849bc6f2befed0c0aaa0b7b2a08c967a83c3
|
| 2535abe7d1fcdd0e6aca782b1f1932a703ed50a4 |
|
17-Feb-2015 |
Mathieu Chartier <mathieuc@google.com> |
Add JIT
Currently disabled by default unless -Xjit is passed in.
The proposed JIT is a method JIT which works by utilizing interpreter
instrumentation to request compilation of hot methods async during
runtime.
JIT options:
-Xjit / -Xnojit
-Xjitcodecachesize:N
-Xjitthreshold:integervalue
The JIT has a shared copy of a compiler driver which is accessed
by worker threads to compile individual methods.
Added JIT code cache and data cache, currently sized at 2 MB
capacity by default. Most apps will only fill a small fraction of
this cache however.
Added support to the compiler for compiling interpreter quickened
byte codes.
Added test target ART_TEST_JIT=TRUE and --jit for run-test.
TODO:
Clean up code cache.
Delete compiled methods after they are added to code cache.
Add more optimizations related to runtime checks e.g. direct pointers
for invokes.
Add method recompilation.
Move instrumentation to DexFile to improve performance and reduce
memory usage.
Bug: 17950037
Change-Id: Ifa5b2684a2d5059ec5a5210733900aafa3c51bca
|
| 90a33595bc637f5768a7726a186bdfe25efcd0d6 |
|
16-Jan-2015 |
Sebastien Hertz <shertz@google.com> |
Fix exception handling during deoptimization
When interpreting a deoptimized shadow frame, we may start with a
pending exception thrown by a previous deoptimized shadow frame (from
a previous invoke). Therefore, we need to handle it before executing
any instruction, otherwise we execute incorrect code.
Because we need the DEX pc of the throwing instruction to find a
matching catch handler, we initialize deoptimized shadow frames with
the current DEX pc at the time the stack is deoptimized.
When we are about to interpret a deoptimized shadow frame, we need to
update the shadow frame with the DEX pc of the next instruction to
interpret. There are three cases:
- if there is no pending exception, this is the instruction following
the current one.
- if there is a pending exception and we found a matching catch
handler, this is the first instruction of this handler.
- if there is a pending exception but there is no matching catch
handler, we do not execute the deoptimized shadow frame and continue
to its caller.
The verifier now fails when a method starts with a move-exception
instruction. Indeed we cannot start executing a method with a pending
exception.
Bug: 19057915
Bug: 19041195
Bug: 18607595
(cherry picked from commit 270a0e16c3b8e5b95cbfdbd8996ac137c7c6322b)
Change-Id: Ib4fdd0ad704b4f2369d229737c9cc77f559cea55
|
| 270a0e16c3b8e5b95cbfdbd8996ac137c7c6322b |
|
16-Jan-2015 |
Sebastien Hertz <shertz@google.com> |
Fix exception handling during deoptimization
When interpreting a deoptimized shadow frame, we may start with a
pending exception thrown by a previous deoptimized shadow frame (from
a previous invoke). Therefore, we need to handle it before executing
any instruction, otherwise we execute incorrect code.
Because we need the DEX pc of the throwing instruction to find a
matching catch handler, we initialize deoptimized shadow frames with
the current DEX pc at the time the stack is deoptimized.
When we are about to interpret a deoptimized shadow frame, we need to
update the shadow frame with the DEX pc of the next instruction to
interpret. There are three cases:
- if there is no pending exception, this is the instruction following
the current one.
- if there is a pending exception and we found a matching catch
handler, this is the first instruction of this handler.
- if there is a pending exception but there is no matching catch
handler, we do not execute the deoptimized shadow frame and continue
to its caller.
The verifier now fails when a method starts with a move-exception
instruction. Indeed we cannot start executing a method with a pending
exception.
Bug: 19057915
Bug: 19041195
Bug: 18607595
Change-Id: I355ac81e6ac098edc7e3cc8c13dbfa24a2969ab2
|
| 1cc7dbabd03e0a6c09d68161417a21bd6f9df371 |
|
18-Dec-2014 |
Andreas Gampe <agampe@google.com> |
ART: Reorder entrypoint argument order
Shuffle the ArtMethod* referrer backwards for easier removal.
Clean up ARM & MIPS assembly code.
Change some macros to make future changes easier.
Change-Id: Ie2862b68bd6e519438e83eecd9e1611df51d7945
|
| ffc605cd817e79d6c7602a87543bb31f24d3a99f |
|
10-Dec-2014 |
Mathieu Chartier <mathieuc@google.com> |
Add missing iget quick for bool, byte, char, short
Bug: 17791557
Bug: 17671806
Change-Id: Ifac4fbfba6c3a3f97131e85914b24756fb7f9722
|
| e94652f1e321b2c8b71acbe5b07d2ebf69fbdb99 |
|
02-Dec-2014 |
Ian Rogers <irogers@google.com> |
Remove MethodHelper from the interpreter.
Use ShadowFrame to get the executing method to avoid a handle for the current
method.
Various associated bits of header file clean-up and removal of an unnecessary
use of MethodHelper in CompilerDriver.
Change-Id: I3b6f4413701e8fc6b0c58b0041a0dd15472bedaa
|
| f72a11dd4b0dd86bc4b1baa37bfa47fc8d5572b5 |
|
30-Oct-2014 |
Ian Rogers <irogers@google.com> |
Add math routines with defined wrapping behavior for the interpreter.
Add a RSUB_INT_LIT16 instruction alias to make instruction opcode switch
statements easier to read.
Change-Id: I3bac07c9518665abf0b05b5c3105a90be22f780a
|
| ded66a01f81812e0129d17c3d08d5eda18433062 |
|
29-Oct-2014 |
Ian Rogers <irogers@google.com> |
Move MethodHelper::GetReturnType to mirror::ArtMethod.
Also, fix missing handle in HasSameSignatureWithDifferentClassLoaders.
Change-Id: I9e1ffd09be950ecc8346fc3c485760d82d9ecab3
|
| 6786a58659420c0631a1ebe1f2fde434b8bcd1e4 |
|
28-Oct-2014 |
Ian Rogers <irogers@google.com> |
Remove ResolveString from MethodHelper.
Change-Id: Ice0fff0680f876285539f78cd79d520d424e2f5e
|
| 647b1a86f518d8db0331b3d52a96392b7a62504b |
|
10-Oct-2014 |
Ian Rogers <irogers@google.com> |
Fix 2 new sets of clang compiler warnings.
Fix issues that are flagged by -Wfloat-equal and -Wmissing-noreturn.
In the case of -Wfloat-equal the current cases in regular code are deliberate,
so the change is to silence the warning. For gtest code the appropriate fix is
to switch from EXPECT_EQ to EXPECT_(FLOAT|DOUBLE)_EQ.
The -Wmissing-noreturn warning isn't enabled due to a missing noreturn in
gtest. This issue has been reported to gtest.
Change-Id: Id84c70c21c542716c9ee0c41492e8ff8788c4ef8
|
| 832336b3c9eb892045a8de1bb12c9361112ca3c5 |
|
09-Oct-2014 |
Ian Rogers <irogers@google.com> |
Don't copy fill array data to quick literal pool.
Currently quick copies the fill array data from the dex file to the literal
pool. It then has to go through hoops to pass this PC relative address down
to out-of-line code. Instead, pass the offset of the table to the out-of-line
code and use the CodeItem data associated with the ArtMethod. This reduces
the size of oat code while greatly simplifying it.
Unify the FillArrayData implementation in quick, portable and the interpreters.
Change-Id: I9c6971cf46285fbf197856627368c0185fdc98ca
|
| 7b078e8c04f3e1451dbdd18543c8b9692b5b067e |
|
10-Sep-2014 |
Ian Rogers <irogers@google.com> |
Compile time performance improvements focusing on interpret-only.
Reduce virtual method dispatch in the method verifier and make more code
inline-able.
Add a StringPiece with const char* equality operator to avoid redundant
StringPieces and strlens.
Remove back link from register line to verifier and pass as argument to reduce
size of RegisterLine.
Remove instruction length from instruction flags and compute from the
instruction, again to reduce size.
Add suspend checks to resolve and verify to allow for more easy monitor
inflation and reduce contention on Locks::thread_list_suspend_thread_lock_.
Change ThrowEarlierClassFailure to throw pre-allocated exception.
Avoid calls to Thread::Current() by passing self.
Template specialize IsValidClassName.
Make ANR reporting with SIGQUIT run using checkpoints rather than suspending
all threads. This makes the stack/lock analysis less lock error prone.
Extra Barrier assertions and condition variable time out is now returned as a
boolean both from Barrier and ConditionVariable::Wait.
2 threaded host x86-64 interpret-only numbers from 341 samples:
Before change: Avg 176.137ms 99% CI 3.468ms to 1060.770ms
After change: Avg 139.163% 99% CI 3.027ms to 838.257ms
Reduction in average compile time after change is 20.9%.
Slow-down without change is 26.5%.
Bug: 17471626 - Fix bug where RegTypeCache::JavaLangObject/String/Class/Throwable
could return unresolved type when class loading is disabled.
Bug: 17398101
Change-Id: Id59ce3cc520701c6ecf612f7152498107bc40684
|
| 37f05ef45e0393de812d51261dc293240c17294d |
|
17-Jul-2014 |
Fred Shih <ffred@google.com> |
Reduced memory usage of primitive fields smaller than 4-bytes
Reduced memory used by byte and boolean fields from 4 bytes down to a
single byte and shorts and chars down to two bytes. Fields are now
arranged as Reference followed by decreasing component sizes, with
fields shuffled forward as needed.
Bug: 8135266
Change-Id: I65eaf31ed27e5bd5ba0c7d4606454b720b074752
|
| cb6b0f31ede2275e79e6199ec391147585a37a2a |
|
12-Aug-2014 |
Ian Rogers <irogers@google.com> |
Avoid use of std::string where we have const char*.
Removing the ClassHelper caused std::string creation for all calls to
Class::GetDescriptor and a significant performance regression. Make the
std::string an out argument so the caller can maintain it and its life time
while allowing GetDescriptor to return the common const char* case.
Don't generate GC maps when compilation is disabled.
Remove other uses of std::string that are occuring on critical paths.
Use the cheaper SkipClass in CompileMethod in CompilerDriver.
Specialize the utf8 as utf16 comparison code for the common shorter byte
encoding.
Force a bit of inlining, remove some UNLIKELYs (they are prone to pessimizing
code), add some LIKELYs.
x86-64 host 1-thread interpret-only of 57 apks:
Before: 29.539s
After: 23.467s
Regular compile:
Before: 1m35.347s
After: 1m20.056s
Bug: 16853450
Change-Id: Ic705ea24784bee24ab80084d06174cbf87d557ad
Conflicts:
runtime/utils.cc
|
| 1ff3c98775a4577cf053dba9a0c2d5c21c07b298 |
|
12-Aug-2014 |
Ian Rogers <irogers@google.com> |
Avoid use of std::string where we have const char*.
Removing the ClassHelper caused std::string creation for all calls to
Class::GetDescriptor and a significant performance regression. Make the
std::string an out argument so the caller can maintain it and its life time
while allowing GetDescriptor to return the common const char* case.
Don't generate GC maps when compilation is disabled.
Remove other uses of std::string that are occuring on critical paths.
Use the cheaper SkipClass in CompileMethod in CompilerDriver.
Specialize the utf8 as utf16 comparison code for the common shorter byte
encoding.
Force a bit of inlining, remove some UNLIKELYs (they are prone to pessimizing
code), add some LIKELYs.
x86-64 host 1-thread interpret-only of 57 apks:
Before: 29.539s
After: 23.467s
Regular compile:
Before: 1m35.347s
After: 1m20.056s
Bug: 16853450
Change-Id: Ic705ea24784bee24ab80084d06174cbf87d557ad
|
| 4e99b3d8955131f3fc71aa113f0fa71f0092cb6f |
|
24-Jun-2014 |
Sebastien Hertz <shertz@google.com> |
Add missing class initialization during compilation and tests
Adds missing class initialization during compilation and tests, especially
java.lang.Class. Otherwise, we'd be able to execute code while the referring
class is not initialized or initializing.
Also adds mirror::Class::AssertInitializedOrInitializingInThread method to
check class initialization when entering the interpreter: the called method's
declaring class must either be initialized or be initializing by the current
thread (other threads must be waiting for the class initialization to complete
holding its lock). Note we only do this check in debug build.
Bump oat version to force compilation.
Bug: 15899971
Change-Id: I9a4edd3739a3ca4cf1c4929dcbb44cdf7a1ca1fe
|
| 9f1020305292a21fd14a402b189c765a125226ab |
|
23-May-2014 |
Sebastien Hertz <shertz@google.com> |
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
|
| bfd9a4378eacaf2dc2bbe05ad48c5164fc93c9fe |
|
22-May-2014 |
Mathieu Chartier <mathieuc@google.com> |
Change MethodHelper to use a Handle.
Added ConstHandle to help prevent errors where you modify the value
stored in the handle of the caller. Also fixed compaction bugs
related to not knowing MethodHelper::GetReturnType can resolve types.
This bug was present in interpreter RETURN_OBJECT.
Bug: 13077697
Change-Id: I71f964d4d810ab4debda1a09bc968af8f3c874a3
|
| 3035961cb41865b80b927546be0c708b6389cec6 |
|
22-May-2014 |
Hans Boehm <hboehm@google.com> |
Clean up and augment Atomic class. Replace QuasiAtomic MemBars.
Add a number of missing C++11 operations to Atomic class.
Invoke the 64 bit routines in QuasiAtomic when necessary.
Replace QuasiAtomic membars with fences that correspond to C++11 fences.
QuasiAtomic was moved to the top of the file. Only fence implementations
actually changed.
This replaces some buggy uses of MembarStoreStore, as reported
in b/14685856 .
Avoid some redundant fences for long volatile operations.
Incompletely converts low-level memory access operations to Atomic.
Change-Id: Iea828431a0cea46540eb74fcaa02071cab6fdcda
|
| ffddfdf6fec0b9d98a692e27242eecb15af5ead2 |
|
03-Jun-2014 |
Tim Murray <timmurray@google.com> |
DO NOT MERGE
Merge ART from AOSP to lmp-preview-dev.
Change-Id: I0f578733a4b8756fd780d4a052ad69b746f687a9
|
| da843e11bd5f2177a776c37de033228a7ea230af |
|
28-May-2014 |
Sebastien Hertz <shertz@google.com> |
Another workaround for stack overflow issue with clang
Bug: 14882674
Change-Id: I465fda533714021b39b847afe89e454758646b47
|
| 5c0049067a953d7a2383d3064a84a32d1efa3a24 |
|
21-May-2014 |
Sebastien Hertz <shertz@google.com> |
Secure move-exception in intepreter
Copies exception into the shadow frame before clearing it from its thread so
it's always reachable.
Change-Id: Ifdc68280424f5099aacf0724da94889881a99551
|
| f832284dd847ff077577bb5712225430bbbb3b67 |
|
16-May-2014 |
Mathieu Chartier <mathieuc@google.com> |
Delete ClassHelper and fix compaction bug in GetDirectInterface
Cleanup helps to prevent compaction bugs. Fixed a fairly serious
compaction error caused by calling ClassHelper::GetDirectInterface
without handling the case where it causes thread suspension due to
ResolveType.
Bug: 8981901
Change-Id: I82b3bb6dd48d21eb6ece7aae0733c4a23c2bc408
|
| e713d9338ad122d6b8c7997387d0c9fc464eea3e |
|
15-May-2014 |
Sebastien Hertz <shertz@google.com> |
Fix JDWP location event report on return instructions
After change https://android-review.googlesource.com/94334, the debugger may
listen to kDexPcChange event without listening for kMethodExit event. In this
case, we need to report kDexPcChange event on RETURN instructions.
Bug: 11874828
Change-Id: Ic61f4eec71e7ece494bee628d7f3358a616d31d3
|
| 8379b2256be5d2be4ad083a76e9f8ec403c4d405 |
|
24-Feb-2014 |
Sebastien Hertz <shertz@google.com> |
Combine JDWP location events
The runtime now sends location events BREAKPOINT, SINGLE_STEP, METHOD_ENTRY,
METHOD_EXIT and METHOD_EXIT_WITH_RETURN_VALUE in the same JDWP event packet
when they relate to the same location.
We update the Dbg::UpdateDebugger method to take initial event flags and
returned value. It allows to call this method from DebugInstrumentationListener
so we can treat method entry/exit events with breakpoint and single-step.
In the interpreter, we ensure we do not call Instrumentation::DexPcMovedEvent
when Instrumentation::MethodEnterEvent has just been called or when we're about
to call Instrumentation::MethodExitEvent. This prevents from sending duplicated
events.
I measured the average performance impact on some benchmarks with a Nexus 4
without a debugger attached:
* 1%-2% for the computed-goto-based interpreter (default interpreter)
* 5%-10% for the switch-based interpreter.
This is mostly due to the test of the boolean flag for the method entry event.
Bug: https://code.google.com/p/android/issues/detail?id=68427
Bug: 11874828
Change-Id: Ic4ff61375ff6b4ed5825adeac09f61f97b4be619
|
| 2fa98e26ddd71ca147869b84fc117751ae6ca92a |
|
07-May-2014 |
Ian Rogers <irogers@google.com> |
Fix typo.
Change-Id: I733f83bf8577d3fb41534faaf241f1f120870bec
|
| b2c7ead6bb5c98282cdfbc89db8984a004bea030 |
|
29-Apr-2014 |
Mathieu Chartier <mathieuc@google.com> |
Don't allow allocating finalizable objects during transactions.
It doesn't make sense to allocate finalizable objects during a
transcation since they will never get finalized without a started
runtime.
Before StatusInitialized in core.host.oatdump.txt: 3564
After StatusInitialized in core.host.oatdump.txt: 3564
Bug: 14078487
Change-Id: I7070536f7bb87bfc691d4268bd39a3eca492f48e
|
| d2fe10a3a34af171bf1631219cd2d6ff6b7778b5 |
|
15-Jan-2014 |
Sebastien Hertz <shertz@google.com> |
Remove blacklist
Removes the class initialization blacklist and use transaction to detect and
revert class initialization attempting to invoke native method. This only
concerns class initialization happening at compilation time when generating an
image (like boot.art for the system).
In transactional mode, we log every object's field assignment and array update.
Therefore we're able to abort a transaction to restore values of fields and
array as they were before the transaction starts. We also log changes to the
intern string table so we can restore its state prior to transaction start.
Since transactional mode only happens at compilation time, we don't need to log
all these changes at runtime. In order to reduce the overhead of testing if
transactional mode is on/off, we templatize interfaces of mirror::Object and
mirror::Array, respectively responsible for setting a field and setting an
array element.
For various reasons, we skip some specific fields from transaction:
- Object's class and array's length must remain unchanged so garbage collector
can compute object's size.
- Immutable fields only set during class loading: list of fields, method,
dex caches, vtables, ... as all classes have been loaded and verified before a
transaction occurs.
- Object's monitor for performance reason.
Before generating the image, we browse the heap to collect objects that need to
be written into it. Since the heap may still holds references to unreachable
objects due to aborted transactions, we trigger one collection at the end of
the class preinitialization phase.
Since the transaction is held by the runtime and all compilation threads share
the same runtime, we need to ensure only one compilation thread has exclusive
access to the runtime. To workaround this issue, we force class initialization
phase to run with only one thread. Note this is only done when generating image
so application compilation is not impacted. This issue will be addressed in a
separate CL.
Bug: 9676614
Change-Id: I221910a9183a5ba6c2b99a277f5a5a68bc69b5f9
|
| ef7d42fca18c16fbaf103822ad16f23246e2905d |
|
06-Jan-2014 |
Ian Rogers <irogers@google.com> |
Object model changes to support 64bit.
Modify mirror objects so that references between them use an ObjectReference
value type rather than an Object* so that functionality to compress larger
references can be captured in the ObjectRefererence implementation.
ObjectReferences are 32bit and all other aspects of object layout remain as
they are currently.
Expand fields in objects holding pointers so they can hold 64bit pointers. Its
expected the size of these will come down by improving where we hold compiler
meta-data.
Stub out x86_64 architecture specific runtime implementation.
Modify OutputStream so that reads and writes are of unsigned quantities.
Make the use of portable or quick code more explicit.
Templatize AtomicInteger to support more than just int32_t as a type.
Add missing, and fix issues relating to, missing annotalysis information on the
mutator lock.
Refactor and share implementations for array copy between System and uses
elsewhere in the runtime.
Fix numerous 64bit build issues.
Change-Id: I1a5694c251a42c9eff71084dfdd4b51fff716822
|
| abff6439db28fbbed95490bfff7e24d1fdf5b771 |
|
27-Jan-2014 |
Sebastien Hertz <shertz@google.com> |
Refactor array access for the interpreter.
Adds GetWithoutChecks and SetWithoutChecks methods in PrimitiveArray and use
them in the interpreter. Updates Get and Set methods to rely on them and adds
some DCHECK to control exception flow.
Renames IsValidIndex into CheckIsValidIndex to reflect it can throw an
exception. It's also more consistent with ObjectArray::CheckIsAssignable.
Make ThrowArrayIndexOutOfBoundsException private in Array since it's only used
by Array::CheckIsValidIndex.
Updates DoFilledNewArray to use SetWithoutChecks rather than Set.
Change-Id: I2fd314d77a67cf969843d499b86d04ca7b7a43e6
|
| b122a4bbed34ab22b4c1541ee25e5cf22f12a926 |
|
20-Nov-2013 |
Ian Rogers <irogers@google.com> |
Tidy up memory barriers.
Change-Id: I937ea93e6df1835ecfe2d4bb7d84c24fe7fc097b
|
| cbb2d20bea2861f244da2e2318d8c088300a3710 |
|
15-Nov-2013 |
Mathieu Chartier <mathieuc@google.com> |
Refactor allocation entrypoints.
Adds support for switching entrypoints during runtime. Enables
addition of new allocators with out requiring significant copy
paste. Slight speedup on ritzperf probably due to more inlining.
TODO: Ensuring that the entire allocation path is inlined so
that the switch statement in the allocation code is optimized
out.
Rosalloc measurements:
4583
4453
4439
4434
4751
After change:
4184
4287
4131
4335
4097
Change-Id: I1352a3cbcdf6dae93921582726324d91312df5c9
|
| ee1997a3b83334985e757f369c09e111b121661b |
|
19-Sep-2013 |
Sebastien Hertz <shertz@google.com> |
Improve interpreter handler table management.
We still have two handlers table:
- the "main" table which holds execution handlers of each instruction,
- the "alternative" table which holds handlers supporting instrumentation
before jumping to the corresponding instruction handler from the "main" table.
Instrumentation holds the index of the handler table the interpreter must use.
This index is represented by the InterpreterHandlerTable enum and is stored in
the Instrumentation::interpreter_handler_table_ field.
Interpreter's current handler table update happens:
- on backward branch
- after invoke
- when throwing exception.
In the case of the backward branch and exception, we only update the table if
any thread's flags is set. This allows to only do one test for handling thread
suspension and handler table update.
This CL also removes the local variable "instrumentation". Every handler which
needs it will get it from Runtime::Current()->GetInstrumentation().
Change-Id: Id886ea7ebf3dac1285f0ca701c098aee7ebaab8d
|
| e861ebd5d9490cc86200f3859f3d36fadad4588c |
|
10-Oct-2013 |
Mathieu Chartier <mathieuc@google.com> |
Fix interpreter bugs.
These showed up in compaction work.
Change-Id: Iac8eb0a1395c25aabba9f2e0ff6b01fc6180bdca
|
| 253ea073ec03a0be1e22df98957257594e316a39 |
|
02-Oct-2013 |
Hiroshi Yamauchi <yamauchi@google.com> |
Fix an Art debug build boot failure.
(The original change is 3b4c18933c24b8a33f38573c2ebcdb9aa16efeb5.)
Bug: 11003273
Bug: 9986565
Change-Id: Iaaf6395f171eb6e6b8c99386c20c4970c53ee00d
|
| c67148594b1580c278ae71e3ce5c6fd59bfa6bd3 |
|
30-Sep-2013 |
Sebastien Hertz <shertz@google.com> |
Enable thread analysis on template functions.
All template functions using thread analysis must be explicitly instantiated
with the thread analysis attributes to enable thread analysis. We use macros
to do this since there are many variants of a same function depending the
number of template arguments.
Also add documentation of these functions.
Change-Id: I3c79acc2f0a6a8dfb5c42924439145292dd68812
|
| c61124bdeaae94f977ffc36ac69535e792c226f2 |
|
10-Sep-2013 |
Sebastien Hertz <shertz@google.com> |
Cleanup invoke in interpreter.
Some cleanup in invocation stuff:
- Get the number of invoke arguments from instruction (vA) rather than get it
from its code item. This benefits to native invoke since we no longer need to
parse the method's shorty. Also pass the low 16 bits of instructions to avoid
fetching it twice when reading vA.
- Remove "is_static" tests by taking advantage of invoke type template argument
rather than testing method's access flags.
- Ensure Instruction::GetArgs is inlined.
- Check exception when initializing method's class when transitioning from
interpreter to compiled code (artInterpreterToCompiledCodeBridge).
- Move UnstartedRuntimeInvoke function to interpreter_common.cc and make it
static as it's only used by DoInvoke and DoInvokeVirtualQuick functions.
- Avoid duplicating code in ShadowFrame::Create.
Performance remains the same according to benchmarks. Hopefully, this should be
addressed in next CLs, especially by improving new shadow frame initialization.
Bug: 10668955
Change-Id: I514b8f098d0ef3e35921ceb770383aac1a9c7902
|
| a3faaf4bece7f42529c013fe87bd41de59798656 |
|
04-Sep-2013 |
Jeff Hao <jeffhao@google.com> |
Fix handling of unresolved references in verifier.
The verifier should not treat use of unresolved references as a reason to reject
the entire class. Instead, the verifier treats the instruction as a throw. If
that class is run, the interpreter with extra checks will throw an exception.
Bug: 10457426
Change-Id: I3799da843a7ffb3519bbf6dc13a6276519d9cb95
|
| 450dcb56ecbf6f729401e753f0a27e4170177ddd |
|
21-Sep-2013 |
Ian Rogers <irogers@google.com> |
Improve float to integral conversion.
Change-Id: I1597083cb2c04084ce825fe2e3c753fde8309cd8
|
| 3b588e09eac6fb2aff64595e2232e479703850fc |
|
11-Sep-2013 |
Sebastien Hertz <shertz@google.com> |
Optimize instruction data fetch in interpreter.
The computed goto implementation prevents the compiler from detecting we are
loading the first 16 bits of instruction twice: first one to get the opcode and
second one to fetch first instruction's operand(s) like vA and vB.
We now load the 16 bits into a local variable and decode opcode and operands
from this variable. And do the same in the switch-based implementation for
consistency.
The performance improvement is 5% in average on benchmark applications suite.
Also remove unused "Thread* self" parameter from DoIGetQuick and DoIPutQuick.
Bug: 10703860
Change-Id: I83026ed6e78f642ac3dcdc6edbb6056fe012005f
|
| 947ff080753c786a74f1cd7aeb09f717bb7074bd |
|
17-Sep-2013 |
Sebastien Hertz <shertz@google.com> |
Load shadow frame's this object only upon instrumentation.
We used to load the shadow frame's this object when entering the interpreter
and push it into thread's SIRT to make it visible to garbage collector. But
it is only used by instrumentation listeners.
We now move this load at each place an instrumentation listener is called. This
avoids loading it when instrumentation is disabled. This also allows to remove
the SIRT reference and the push/pop management it implies when entering/exiting
the interpreter.
The performance improvement is 6% in average on our benchmarks suite.
This CL also makes method Instrumentation::ExceptionCaughtEvent const so we can
use a "const instrumentation::Instrumentation*" object in interpreter.
Change-Id: I2caccba9a906f244c8057b24031250f9824cc711
|
| cdf2d4cbbb1d0cf3e9f80d3dca609a3435c74427 |
|
13-Sep-2013 |
Sebastien Hertz <shertz@google.com> |
Update interpreter handler table after invoke.
And introduce UPDATE_HANDLER_TABLE macro to reduce copy & paste.
Change-Id: I9ea7565df0db2b44581d0e6bcefb5f104e76ed01
|
| 043036f67fa37b3cfa3b05c41e5a824058730378 |
|
09-Sep-2013 |
Sebastien Hertz <shertz@google.com> |
Add missing memory barrier.
When DEX-to-DEX compiler is not run onto a method (because it's not preverified
at the time of compilation), we do not replace RETURN where a barrier is needed
into RETURN_VOID_BARRIER.
This CL fixes this by placing a barrier on RETURN instruction only when the
checks are enabled (non-preverified method).
Change-Id: I4eb4cf79bb4a74684579c578318e27f62f4d9e8a
|
| 1eda2268e84d384256814cb6c2ba2440a848f9ed |
|
09-Sep-2013 |
Sebastien Hertz <shertz@google.com> |
Move thread suspend check at safepoints.
Move CheckSuspend on backward branch, return and exception handling.
Bug: 10603072
Change-Id: Ic6c2c5066f133a345323d46edca7afde350849d8
|
| 8ece050d85fc244c72610244e440b0e00aa618fa |
|
07-Aug-2013 |
Sebastien Hertz <shertz@google.com> |
Add an interpreter using computed goto table.
This CL adds a new implementation of the interpreter using computed goto table.
In order to keep the switch-based implementation, it reorders things as the
following:
- Keep interpreter entrypoints into interpreter.h/.cc files.
- Move common interpreter parts to interpreter_common.h/.cc files.
- Separate both implementations to their own modules.
The interpreter implementation can be selected by changing the value of the
kInterpreterImplKind global variable (see interpreter.cc file). The default one
remains the switch-based implementation.
Also updates the exception handling (FindNextInstructionFollowingException) and
SPARSE_SWITCH switch handling (DoSparseSwitch) routines to share code between both
implementations.
Finally, adds a PACKED_SWITCH handling routine (DoPackedSwitch) so we are
consistent with SPARSE_SWITCH handling.
The computed goto implementation use two handlers table: one for normal
instruction handling and one for instrumentation handling. The current handlers
table to be used is updated on backward branch depending on whether there is
listener to DEX pc change.
Bug: 10602809
Change-Id: Ibb53bcc68be75c473fe5440835e78fc9a74381b3
|