OSDN Git Service
Alex Light [Tue, 20 Dec 2016 18:41:37 +0000 (18:41 +0000)]
Cleanup jvmti tests
am:
ca5b9dccce
Change-Id: I6acbcddf3eb32749d5c7e15bbe7a27d301f830ee
Alex Light [Tue, 20 Dec 2016 18:41:25 +0000 (18:41 +0000)]
Revert "Revert "Basic obsolete methods support""
am:
ce77fc0e7f
Change-Id: I086f8b0b3c23ec64ffb577e9946c477dc5d82794
Nicolas Geoffray [Tue, 20 Dec 2016 18:41:10 +0000 (18:41 +0000)]
Merge "ARM: VIXL32: Use 16-bit encoding for B(cond,...) where possible."
am:
1ffb6fc921
Change-Id: I2c171a03e02067998e8cebc8aedeb6265d6db89d
Nicolas Geoffray [Tue, 20 Dec 2016 13:52:48 +0000 (13:52 +0000)]
Merge "ARM: VIXL32: Use 16-bit encoding for B(cond,...) where possible."
Alexey Frunze [Tue, 20 Dec 2016 12:07:45 +0000 (12:07 +0000)]
Merge "MIPS64: Improve string and class loads."
am:
06ce6d4359
Change-Id: I6b35d073cb4088a65dcfe31f5403b1680aa9c53c
Treehugger Robot [Tue, 20 Dec 2016 12:00:20 +0000 (12:00 +0000)]
Merge "MIPS64: Improve string and class loads."
Nicolas Geoffray [Tue, 20 Dec 2016 09:10:52 +0000 (09:10 +0000)]
Merge "Sharpen HLoadClass from inliner."
am:
eaaaaef76b
Change-Id: I951c3f54a90a20d39b3e92c53649de64019be5e0
Nicolas Geoffray [Tue, 20 Dec 2016 09:10:35 +0000 (09:10 +0000)]
Merge "Make 141-class-unload more robust under JIT."
am:
ee3a0bde75
Change-Id: I4b31b791c4af840123fd1ae98d92de72bb4d1fe3
Nicolas Geoffray [Tue, 20 Dec 2016 09:06:17 +0000 (09:06 +0000)]
Merge "Sharpen HLoadClass from inliner."
Nicolas Geoffray [Tue, 20 Dec 2016 09:05:29 +0000 (09:05 +0000)]
Merge "Make 141-class-unload more robust under JIT."
Mingyao Yang [Tue, 20 Dec 2016 00:10:33 +0000 (00:10 +0000)]
Merge "CHA guard optimization (elimination/hoisting)."
am:
fb5a6bb830
Change-Id: Ic40219b4344e34190c875ecd9e9f86fc484cb1a2
Alex Light [Wed, 14 Dec 2016 23:29:42 +0000 (15:29 -0800)]
Cleanup jvmti tests
Deduplicate the 'run' scripts into the test/etc/run-test-jar file.
Test: mma -j40 test-art-host
Change-Id: I8cc87bacedb393bfa11cdf2345bfd7a4341caa37
Mingyao Yang [Tue, 20 Dec 2016 00:02:11 +0000 (00:02 +0000)]
Merge "CHA guard optimization (elimination/hoisting)."
Alexey Frunze [Wed, 14 Dec 2016 01:43:11 +0000 (17:43 -0800)]
MIPS64: Improve string and class loads.
This adds most kinds of string/class loads.
JIT string/class loads are TBD separately.
This also fixes Mips64Assembler::LoadLabelAddress()
(adding a constant to a 64-bit address must be done
using daddiu, not addiu).
Test: test-art-host-gtest
Test: booted MIPS64 (with 2nd arch MIPS32R6) in QEMU
Test: "make -j1 ART_TEST_DEFAULT_COMPILER=false ART_TEST_OPTIMIZING=true
ART_TEST_INTERPRETER=false ART_TEST_JIT=false
ART_TEST_PIC_TEST=true test-art-target-run-test64"
Change-Id: I1f94ece4cd202382c11167e1ed958e9d08d92822
Mingyao Yang [Thu, 17 Nov 2016 17:04:53 +0000 (09:04 -0800)]
CHA guard optimization (elimination/hoisting).
Test: manual by checking the dump-cfg output.
Change-Id: I254e168b9a85d2d3d23e02eea7e129c1bc9ab920
Alex Light [Thu, 15 Dec 2016 01:05:52 +0000 (01:05 +0000)]
Revert "Revert "Basic obsolete methods support""
This reverts commit
b81a9840b44480bfeacd74b8d9f51e06f295411d.
There were 2 issues with the original CL's test 916-obsolete-jit that
caused it to sporadically fail.
First, when checking if we had jitted the function under test in
916-obsolete-jit we failed to check to see if the function 'doCall',
which is used to work-around bugs in our deoptimization & compilation
systems, had also been jitted.
In the case where the 'sayHi' function had been jitted but the
'doCall' function had not we would (correctly) fail to redefine the
'Transform' class since we would not be able to deoptimize the 'sayHi'
function since it is under a quick_to_interpreter_bridge (runtime)
frame.
Secondly, the function Main.isInterpretedFunction was incorrect and
would always return false, regardless of the actual state of the
function. This would cause the test to fail as the
quick_to_interpreter_bridge frame prevented deoptimization of the
obsoleted function. Usually the warm-up period was enough to make sure
the methods were jitted anyway but this was not guaranteed.
Both of these problems become more likely to occur on systems with
more cpu contention such as the buildbots.
Test: stress --cpu 60 &; while ./test/run-test --host --jit 916; do ; done
Reason for revert: Fixed bug in test that was causing failures.
Original Tests:
Test: ./test/run-test --host 914
Test: ./test/run-test --host 915
Test: ./test/run-test --host 916
Test: mma -j40 test-art-host
Test: ART_TEST_JIT=true \
ART_TEST_INTERPRETER=true mma -j40 test-art-host
Test: ./art/tools/run-jdwp-tests.sh --mode=host --variant=X64
Test: ./art/tools/run-jdwp-tests.sh --mode=host --variant=X64 --no-jit
Bug:
32369913
Bug:
33630159
Change-Id: If1a92e47b90965a7dc21c5826185debe62bd1554
Aart Bik [Mon, 19 Dec 2016 21:46:48 +0000 (21:46 +0000)]
Merge "Improved induction var and range analysis around types."
am:
d54f43ca39
Change-Id: I580b92820d94de55bf8478d41ba8a2d5e3ab3832
Aart Bik [Mon, 19 Dec 2016 21:38:57 +0000 (21:38 +0000)]
Merge "Improved induction var and range analysis around types."
Nicolas Geoffray [Mon, 19 Dec 2016 21:20:15 +0000 (21:20 +0000)]
Merge "Revert "Revert "Revert "Use interpret-only instead of verify-at-runtime when testing JIT""""
am:
481593d874
Change-Id: Ie2517bba2c533d5a53113c0da326ca4c24000ff2
Nicolas Geoffray [Mon, 19 Dec 2016 21:12:49 +0000 (21:12 +0000)]
Merge "Revert "Revert "Revert "Use interpret-only instead of verify-at-runtime when testing JIT""""
Nicolas Geoffray [Mon, 19 Dec 2016 21:11:54 +0000 (21:11 +0000)]
Revert "Revert "Revert "Use interpret-only instead of verify-at-runtime when testing JIT"""
Reverting due to gtest crashes.
This reverts commit
3a3e4c10e4d08c9fa1fbeba6d74933964a3bb47f.
Change-Id: I7e1adf0d886a1cdaf3152469b3c11a33131ee498
Vladimir Marko [Mon, 19 Dec 2016 20:53:14 +0000 (20:53 +0000)]
Merge "Revert "Fix 616-cha assumptions.""
am:
761460ee54
Change-Id: I780638160fd9bb88c219f68052a5c31c03782ab4
Vladimir Marko [Mon, 19 Dec 2016 20:45:48 +0000 (20:45 +0000)]
Merge "Revert "Fix 616-cha assumptions.""
Vladimir Marko [Mon, 19 Dec 2016 20:44:29 +0000 (20:44 +0000)]
Revert "Fix 616-cha assumptions."
The change that this depends on has been reverted:
https://android-review.googlesource.com/316224
This reverts commit
5c77917ca94b33576b4a6bb850bdb578df378292.
Change-Id: I590dabcc7d12e661536b4234e45b7e763c34c35d
Vladimir Marko [Mon, 19 Dec 2016 19:50:11 +0000 (19:50 +0000)]
Merge "Revert "Fix JIT crash due to unverified dead code""
am:
263cc99a5f
Change-Id: I7fb96da9847c216a5a8e06b6f8470fc09f01cd0e
Vladimir Marko [Mon, 19 Dec 2016 19:45:32 +0000 (19:45 +0000)]
Merge "Revert "Fix JIT crash due to unverified dead code""
Vladimir Marko [Mon, 19 Dec 2016 19:31:26 +0000 (19:31 +0000)]
Revert "Fix JIT crash due to unverified dead code"
Broke method_verifier_test and image_test.
Bug:
31000839
This reverts commit
132eb43396bdb0b9bdacf069289f019d85d358fc.
Change-Id: I85923e6bdb3e76f54978658aad816d4c79fc339e
Calin Juravle [Mon, 19 Dec 2016 19:00:07 +0000 (19:00 +0000)]
Merge "Fix 616-cha assumptions."
am:
2d76677d7a
Change-Id: I60d4a592defe603c298ff96b2a3f7c0282243b83
Aart Bik [Fri, 16 Dec 2016 21:57:52 +0000 (13:57 -0800)]
Improved induction var and range analysis around types.
Rationale:
Lots of code should not depend on int only. This CL generalizes
the kinds of types that can be optimized after analysis. As part
of the CL, however, a minor cleanup regarding type safety of the
stored induction var analysis results is required. This further
improved our int benchmark, and brings the long benchmark up-to-par.
Test: m test-art-host-run-test
Change-Id: I5dfb623dabf9113de90c2f6da99328dda8f8b60b
Treehugger Robot [Mon, 19 Dec 2016 18:56:50 +0000 (18:56 +0000)]
Merge "Fix 616-cha assumptions."
Hiroshi Yamauchi [Mon, 19 Dec 2016 18:50:08 +0000 (18:50 +0000)]
Merge "Temporarily disable 153-reference-stress."
am:
9b8e6de85f
Change-Id: Id3a5b74c23eafd3f2a19fbf69cacca74cd921d26
Hiroshi Yamauchi [Mon, 19 Dec 2016 18:43:26 +0000 (18:43 +0000)]
Merge "Temporarily disable 153-reference-stress."
Vladimir Marko [Mon, 19 Dec 2016 18:39:37 +0000 (18:39 +0000)]
Merge "Fix arena allocation counting."
am:
a7ad00aad1
Change-Id: I90d2755e6e12917afc94ef58e46adb0d7fb73b50
Treehugger Robot [Mon, 19 Dec 2016 18:34:16 +0000 (18:34 +0000)]
Merge "Fix arena allocation counting."
Calin Juravle [Mon, 19 Dec 2016 17:18:31 +0000 (17:18 +0000)]
Merge "Fix JIT crash due to unverified dead code"
am:
b44553224c
Change-Id: I59641114a6415f01abab3b50d11cc63986f7c7e6
Calin Juravle [Mon, 19 Dec 2016 15:38:07 +0000 (15:38 +0000)]
Fix 616-cha assumptions.
616-cha relies that some classes are not loaded at startup. Disable the
app image to prevent that.
Test: ART_TEST_JIT=true m test-art-host-run-test-616-cha
Change-Id: I432c300a3cf3de7c0f2f9ab02d7576e92edac844
Calin Juravle [Mon, 19 Dec 2016 17:13:46 +0000 (17:13 +0000)]
Merge "Fix JIT crash due to unverified dead code"
Nicolas Geoffray [Fri, 16 Dec 2016 16:09:08 +0000 (16:09 +0000)]
Sharpen HLoadClass from inliner.
Also cleanup HLoadClass constructor.
Test: ART_TEST_JIT=true m test-art-host-run-test
Change-Id: I8f803b05fb8a7267d1421ca9c032e624f27efed3
Vladimir Marko [Mon, 19 Dec 2016 16:04:33 +0000 (16:04 +0000)]
Fix arena allocation counting.
Add the "CHA" allocation kind description. Move the
statistics to the heap to avoid excessively large frames
triggering errors with -Werror, -Wframe-larger-than=.
Test: m test-art-host
Test: m test-art-host with arena allocation counting.
Change-Id: I7947a3a17b4c1ed773742516b1f7ab9a58c92150
Calin Juravle [Fri, 28 Oct 2016 19:09:56 +0000 (20:09 +0100)]
Fix JIT crash due to unverified dead code
The JIT compiler assumes that it only gets completely verified code. To
work around potential unverified dead code it uses kAccDontBotherCompile
flag set during runtime verification. However, if a class is verified
during a prior dex2oat the flag is not persisted and JIT happily things
that everything is ok.
The simplest fix is to mark classes with potential unverified dex code
as verify at runtime. We only do this for apps and assume that
everything in the boot image is well formed.
Test: m test-art-host-706-jit-skip-compilation
Bug:
31000839
Change-Id: Ib73de1888581bb7202474cfd7aca70af4cc2cc00
Calin Juravle [Mon, 19 Dec 2016 13:49:18 +0000 (13:49 +0000)]
Merge "Revert "Revert "Use interpret-only instead of verify-at-runtime when testing JIT"""
am:
71464a10a4
Change-Id: I10552f5fba712022ce6a21e5f7b3fbcfe7f71ee0
Treehugger Robot [Mon, 19 Dec 2016 13:45:14 +0000 (13:45 +0000)]
Merge "Revert "Revert "Use interpret-only instead of verify-at-runtime when testing JIT"""
Nicolas Geoffray [Mon, 19 Dec 2016 13:01:46 +0000 (13:01 +0000)]
Merge "Don't unquicken RETURN_VOID_NO_BARRIER with vdex."
am:
c0e74baddb
Change-Id: I977bb2194de2ea070427663d72923646fc4bcea0
Nicolas Geoffray [Mon, 19 Dec 2016 12:58:45 +0000 (12:58 +0000)]
Merge "Don't unquicken RETURN_VOID_NO_BARRIER with vdex."
Nicolas Geoffray [Fri, 16 Dec 2016 11:22:05 +0000 (11:22 +0000)]
Make 141-class-unload more robust under JIT.
Make sure jit is stopped before counting maps.
bug:
29712509
test: 141-class-unload
Change-Id: I52c0fed74556fa228cd73c9c0bc6a1d019df0a1c
Calin Juravle [Mon, 19 Dec 2016 11:26:41 +0000 (11:26 +0000)]
Merge "Add Thread entry to signal if the thread can call into java"
am:
d96b96dd97
Change-Id: I70dd4e0fe4727daf13381a621042e9a55d80c862
Calin Juravle [Mon, 19 Dec 2016 11:26:40 +0000 (11:26 +0000)]
Add Thread entry to signal if the thread can call into java
am:
ccd56958eb
Change-Id: I6b5f99beea90fe4cdd1245a36e984adecc278530
Calin Juravle [Mon, 19 Dec 2016 11:22:55 +0000 (11:22 +0000)]
Merge "Add Thread entry to signal if the thread can call into java"
Nicolas Geoffray [Mon, 19 Dec 2016 09:35:20 +0000 (09:35 +0000)]
Merge "ARM: Fix vixl related branch issue in JNI."
am:
685e4a3252
Change-Id: Idb507ed12f2af87b2f00210b3c1becfd9603fe19
Nicolas Geoffray [Mon, 19 Dec 2016 09:35:03 +0000 (09:35 +0000)]
Merge "ARM: VIXL32: Implement some dispatch optimizations."
am:
76873db28c
Change-Id: I8d2ba47fd208c0da2a5ca086a7c72bcacf3733e3
Nicolas Geoffray [Mon, 19 Dec 2016 09:31:18 +0000 (09:31 +0000)]
Merge "ARM: Fix vixl related branch issue in JNI."
Artem Serov [Thu, 15 Dec 2016 17:56:27 +0000 (17:56 +0000)]
ARM: Fix vixl related branch issue in JNI.
Fix VIXL asserts in EncodeLabelFor in JNI macro assembler when
label is out of range of 16-bit branch (now a veneer is correctly
used).
bug:
32545704
Test: ART_USE_VIXL_ARM_BACKEND=true m test-art-host
Test: ART_USE_VIXL_ARM_BACKEND=true m test-art-target
Change-Id: Ie6401394cf364daeaaf107b42275997d2edf5b6d
Nicolas Geoffray [Mon, 19 Dec 2016 09:29:54 +0000 (09:29 +0000)]
Merge "ARM: VIXL32: Implement some dispatch optimizations."
Nicolas Geoffray [Fri, 16 Dec 2016 16:23:16 +0000 (16:23 +0000)]
Don't unquicken RETURN_VOID_NO_BARRIER with vdex.
The quickening of the RETURN_VOID instruction is based on
local information (no final fields) that doesn't get affected
by a boot image update.
Test: test-art-host, verifier_deps_test
bug:
30937355
Change-Id: I12b22d7fcda6dc681a32ff752c3871f6e84f19a1
Hiroshi Yamauchi [Fri, 16 Dec 2016 21:09:19 +0000 (13:09 -0800)]
Temporarily disable 153-reference-stress.
This is intermittently failing. See
33389022 #13.
Note this could fail either in CMS or CC, gcstress or not.
Bug:
33389022
Test: test-art-host.
Change-Id: I257c2c771fcab9405c1a15fa6e93364908fd1294
Calin Juravle [Thu, 15 Dec 2016 17:57:38 +0000 (17:57 +0000)]
Add Thread entry to signal if the thread can call into java
Compiler threads (AOT or JIT) should not call into Java as they have no
peers (which may lead to crashes, e.g. b/
33067273)
Bug:
32602185
Bug:
33067273
Test: m test-art-host-run-test; m test-art-host-gtest
Change-Id: I97dda7a5444643db3c5d5318339a65a602f709e8
Alex Light [Fri, 16 Dec 2016 19:00:16 +0000 (19:00 +0000)]
Merge "Allow redefined methods/classes to access fields"
am:
5b32b91d73
Change-Id: I75fff4a7353051adb0843bac579edf81247c1fb1
Treehugger Robot [Fri, 16 Dec 2016 18:51:17 +0000 (18:51 +0000)]
Merge "Allow redefined methods/classes to access fields"
Mingyao Yang [Fri, 16 Dec 2016 18:20:43 +0000 (18:20 +0000)]
Merge "Add HVariableInputSizeInstruction."
am:
89b31a3369
Change-Id: I4e3778a9ecf905e6a968e3fc6109209ad56087bb
Mingyao Yang [Fri, 16 Dec 2016 18:14:49 +0000 (18:14 +0000)]
Merge "Add HVariableInputSizeInstruction."
Alex Light [Thu, 15 Dec 2016 19:34:13 +0000 (11:34 -0800)]
Allow redefined methods/classes to access fields
Previously we were not updating the ArtFields of redefined classes.
This meant that the fields of redefined classes could not be accessed.
Bug:
32369913
Test: ./test/run-test --host 917
Test: mma -j40 test-art-host
Change-Id: If039e883e9a57970a0cffbc7dcbbaaf2d44490aa
Nicolas Geoffray [Fri, 16 Dec 2016 16:32:07 +0000 (16:32 +0000)]
Merge "Clear exception if FindClass fails."
am:
20b66e727a
Change-Id: I2454d59bdfaa455462ac5035b7cd0118233a2b7e
Nicolas Geoffray [Fri, 16 Dec 2016 16:20:15 +0000 (16:20 +0000)]
Merge "Clear exception if FindClass fails."
Nicolas Geoffray [Fri, 16 Dec 2016 16:14:39 +0000 (16:14 +0000)]
Clear exception if FindClass fails.
Test: verifier_deps_test.
bug:
30937355
Change-Id: I9916ec00010de012939462d4f30147e8b64fe279
Nicolas Geoffray [Fri, 16 Dec 2016 15:25:03 +0000 (15:25 +0000)]
Merge "Do not resolve types with vdex."
am:
0f69f7425c
Change-Id: Id2eda0a19fd4d09092801b6ffde76b1239782987
Nicolas Geoffray [Fri, 16 Dec 2016 15:11:15 +0000 (15:11 +0000)]
Merge "Do not resolve types with vdex."
Artem Serov [Mon, 12 Dec 2016 15:51:15 +0000 (15:51 +0000)]
ARM: VIXL32: Use 16-bit encoding for B(cond,...) where possible.
If conditional branch's target is known to be not far from branch
(within the range of 254 bytes), 16-bit encoding can be used.
However, we can't assume this by default for branches to a basic
block or a slow path; if we do and fail the range requirement,
veneer pool will be emitted.
Test: ART_USE_VIXL_ARM_BACKEND=true m test-art-host
Test: ART_USE_VIXL_ARM_BACKEND=true m test-art-target
Change-Id: I2fbe6d1a43bc2d1b54472c2c3fe05a575e5634f2
Artem Serov [Fri, 2 Dec 2016 19:19:58 +0000 (19:19 +0000)]
ARM: VIXL32: Implement some dispatch optimizations.
Implement dispatch optimizations for LoadClass, LoadString and
HInvokeStaticOrDirect. Still need to implement JitTables.
Test: ART_USE_VIXL_ARM_BACKEND=true m test-art-host
Test: ART_USE_VIXL_ARM_BACKEND=true m test-art-target
Change-Id: I6e24bc11f3915aa92eede687a864b7ea3a6ab165
Nicolas Geoffray [Thu, 15 Dec 2016 13:20:02 +0000 (13:20 +0000)]
Do not resolve types with vdex.
We don't need to, and it's time expensive.
test: test-art-host, 2x faster vdex compilation time.
bug:
30937355
Change-Id: Ic04ebfee9172ed676fbb38e6901615ed3473352a
Nicolas Geoffray [Fri, 16 Dec 2016 13:07:24 +0000 (13:07 +0000)]
Merge "ARM: VIXL32: Do a code cleanup."
am:
7e9a562e42
Change-Id: I244776733d4b2944d0988120f04f7c6e8e0fe4f7
Nicolas Geoffray [Fri, 16 Dec 2016 13:00:02 +0000 (13:00 +0000)]
Merge "ARM: VIXL32: Do a code cleanup."
Nicolas Geoffray [Fri, 16 Dec 2016 09:44:48 +0000 (09:44 +0000)]
Merge "Remove dex cache string from mirror::Class."
am:
913e6e6f36
Change-Id: Ie5c693e2e6e7daa549abbf1152749af597d0faef
Nicolas Geoffray [Fri, 16 Dec 2016 09:32:22 +0000 (09:32 +0000)]
Merge "Remove dex cache string from mirror::Class."
Nicolas Geoffray [Fri, 16 Dec 2016 09:05:12 +0000 (09:05 +0000)]
Merge "Flush JIT data cache when committing code."
am:
4c1667fd52
Change-Id: If410759aad4f2aa3c442de7b36b48342d3f25347
Nicolas Geoffray [Fri, 16 Dec 2016 08:57:36 +0000 (08:57 +0000)]
Merge "Flush JIT data cache when committing code."
Mathieu Chartier [Fri, 16 Dec 2016 00:30:39 +0000 (00:30 +0000)]
Merge "Fill in empty iftable early"
am:
a3dccf967a
Change-Id: Iff2e45b20e8dba1c4cf41a5458a9d70d03c89189
Mathieu Chartier [Fri, 16 Dec 2016 00:25:32 +0000 (00:25 +0000)]
Merge "Fill in empty iftable early"
Mathieu Chartier [Thu, 15 Dec 2016 22:21:04 +0000 (14:21 -0800)]
Fill in empty iftable early
Previously it would not get filled in and remain null for some
erroneous classes. This caused crashes for things like
GetIfTableCount.
Added regression test (sample failure):
class_linker_test F 12-15 14:20:09 28771 28771 class-inl.h:532] Check failed: ret != nullptr java.lang.Class<Erroneous>
Bug:
33650297
Bug:
33642884
Test: test-art-host
Change-Id: I506210bad17d6d2c58ca7870165f8a4a514f64fe
Aart Bik [Thu, 15 Dec 2016 21:59:30 +0000 (21:59 +0000)]
Merge "Fixed signal 11 bug by not testing resolved method first."
am:
2c43590dc2
Change-Id: I1def3c48276633519d8d6871f41a58fe654f6109
Treehugger Robot [Thu, 15 Dec 2016 21:50:00 +0000 (21:50 +0000)]
Merge "Fixed signal 11 bug by not testing resolved method first."
Aart Bik [Thu, 15 Dec 2016 21:44:58 +0000 (21:44 +0000)]
Merge "Fine tuning the induction analysis."
am:
e9f3269b54
Change-Id: Iaf633fbaa6626b60960f1c28d8f08295883c5272
Treehugger Robot [Thu, 15 Dec 2016 21:33:54 +0000 (21:33 +0000)]
Merge "Fine tuning the induction analysis."
Aart Bik [Thu, 15 Dec 2016 20:10:54 +0000 (20:10 +0000)]
Merge "Modify reporting when only true divergences are requested."
am:
45d75c6194
Change-Id: I14d99574fa0575dda01136ed7674ccc2f553ae88
Treehugger Robot [Thu, 15 Dec 2016 20:06:02 +0000 (20:06 +0000)]
Merge "Modify reporting when only true divergences are requested."
Mingyao Yang [Thu, 15 Dec 2016 20:02:53 +0000 (12:02 -0800)]
Add HVariableInputSizeInstruction.
Make HPhi and HInvoke subclasses of the new instruction.
Test: m test-art-host-run-test
Change-Id: I303c725876f1f4407b98702d92370be25193fc53
Andreas Gampe [Thu, 15 Dec 2016 19:40:22 +0000 (19:40 +0000)]
Merge "ART: Move to libbase StringPrintf"
am:
995a7db371
Change-Id: I1e08e8c07c20dbf69de2576d3aa1ffddcca9be82
Aart Bik [Tue, 13 Dec 2016 19:22:31 +0000 (11:22 -0800)]
Fine tuning the induction analysis.
Rationale:
Based on some self-imposed "blind" testing, improved
the induction variable analysis for typical cases
that provide a bit more elaborate HIR.
Test: test-art-host
Change-Id: I6e6bbf99928c29973178fa48f3942b14bf069944
Treehugger Robot [Thu, 15 Dec 2016 19:28:10 +0000 (19:28 +0000)]
Merge "ART: Move to libbase StringPrintf"
Aart Bik [Thu, 15 Dec 2016 17:36:31 +0000 (09:36 -0800)]
Fixed signal 11 bug by not testing resolved method first.
Test: test-art-host
Bug:
33656359
Change-Id: Idb1afccf811a2fbf7500fc1d953e118981ad36d4
Aart Bik [Thu, 15 Dec 2016 01:16:10 +0000 (17:16 -0800)]
Modify reporting when only true divergences are requested.
Rationale:
Old tool would still report divergences on time out
(and just skip extra processing). It is more desirable
to treat it as a regular time out, however.
Test: fuzzing itself
BUG:
31267855
Change-Id: I32e9758c464be7a22752a85c25dbb8cb6b889399
Nicolas Geoffray [Thu, 15 Dec 2016 17:58:48 +0000 (17:58 +0000)]
Flush JIT data cache when committing code.
Otherwise, executing compiled code may see old data from
the data cache.
Test: ART_TEST_JIT=true test-art-target on arm/arm64
Test: run-libcore-tests.sh (especially jsr166) on arm/arm64
Change-Id: Id037c68897aa0e9ccacd4c5121f3743fb722c6f3
Andreas Gampe [Wed, 14 Dec 2016 18:11:49 +0000 (10:11 -0800)]
ART: Move to libbase StringPrintf
Remove ART's StringPrintf implementation. Fix up clients. Add
missing includes where necessary.
Test: m test-art-host
Change-Id: I564038d5868595ac3bb88d641af1000cea940e5a
Nicolas Geoffray [Wed, 14 Dec 2016 14:59:04 +0000 (14:59 +0000)]
Remove dex cache string from mirror::Class.
The compiled code does not need it anymore.
Test: test-art-host test-art-target
Change-Id: I71c08ce12d450ab2b2cd018330d42a8b8c9e3da2
Vladimir Marko [Thu, 15 Dec 2016 11:36:53 +0000 (11:36 +0000)]
Merge "Add ClassLoader to app-image roots."
am:
9538f9e2a5
Change-Id: Ia692c0418b2e5277a3fac4718d525c24e55bacaf
Vladimir Marko [Thu, 15 Dec 2016 11:21:21 +0000 (11:21 +0000)]
Merge "Add ClassLoader to app-image roots."
Andreas Gampe [Thu, 15 Dec 2016 03:54:22 +0000 (03:54 +0000)]
Merge "ART: Blacklist tests that require JIT"
am:
f9b3457e81
Change-Id: I3dd28554f0898ebf20a505914783c3dc1cd75bcf
Andreas Gampe [Thu, 15 Dec 2016 03:46:26 +0000 (03:46 +0000)]
Merge "ART: Blacklist tests that require JIT"
Alex Light [Thu, 15 Dec 2016 01:14:42 +0000 (01:14 +0000)]
Merge "Revert "Basic obsolete methods support""
am:
b7cc593a34
Change-Id: I5417873f5c3c8cad926a074e6cbb097b2e474247
Alex Light [Thu, 15 Dec 2016 01:04:03 +0000 (01:04 +0000)]
Merge "Revert "Basic obsolete methods support""
Alex Light [Thu, 15 Dec 2016 00:59:05 +0000 (00:59 +0000)]
Revert "Basic obsolete methods support"
This reverts commit
d8936da27b792d1ca02e59c92456a1a53c7b9905.
Reason for revert: Some sort of race in JIT
Change-Id: Ibb4e520bb0721d6d7aa2c841a52eb5baff07740e