summaryrefslogtreecommitdiff
path: root/compiler/optimizing/optimizing_compiler.cc
AgeCommit message (Collapse)Author
2021-03-23Remove Vdex::GetQuickenedInfoOf and all its users.Nicolas Geoffray
Test: test.py Bug: 170086509 Change-Id: I1e1a4abf71245c0fd37f951c9af85f62feba18ca
2021-03-05Add code size to CodeInfo.David Srbecky
This is in preparation of removing it from OatQuickMethodHeader. Bug: 123510633 Test: m test-art-host-gtest Test: ./art/test.py -b -r --host Change-Id: I5c5adb4c040e329b81c1393aa1b80ee017729c8a
2021-01-28Add comment to art_jni_trampoline.Yabin Cui
Bug: 175226454 Test: N/A Change-Id: Ia83c3afa69d7e1efb9685cfaa5401242712fa02a
2021-01-25Revert^4 "Partial Load Store Elimination"Alex Light
This reverts commit 791df7a161ecfa28eb69862a4bc285282463b960. This unreverts commit fc1ce4e8be0d977e3d41699f5ec746d68f63c024. This unreverts commit b8686ce4c93eba7192ed7ef89e7ffd9f3aa6cd07. We incorrectly failed to include PredicatedInstanceFieldGet in a few conditions, including a DCHECK. This caused tests to fail under the read-barrier-table-lookup configuration. Reason for revert: Fixed 2 incorrect checks Bug: 67037140 Test: ./art/test/testrunner/run_build_test_target.py -j70 art-gtest-read-barrier-table-lookup Change-Id: I32b01b29fb32077fb5074e7c77a0226bd1fcaab4
2021-01-24Revert "Revert^2 "Partial Load Store Elimination""Nicolas Geoffray
This reverts commit fc1ce4e8be0d977e3d41699f5ec746d68f63c024. Bug: 67037140 Reason for revert: Fails read-barrier-table-lookup tests. Change-Id: I373867c728789bc14a4370b93a045481167d5f76
2021-01-22Revert^2 "Partial Load Store Elimination"Alex Light
This reverts commit 47ac53100303e7e864b7f6d65f17b23088ccf1d6. There was a bug in LSE where we would incorrectly record the shadow$_monitor_ field as not having a default initial value. This caused partial LSE to be unable to compile the Object.identityHashCode function, causing crashes. This issue was fixed in a parent CL. Also updated all Offsets in LSE_test to be outside of the object header regardless of configuration. Test: ./test.py --host Bug: 67037140 Reason for revert: Fixed issue with shadow$_monitor_ field and offsets Change-Id: I4fb2afff4d410da818db38ed833927dfc0f6be33
2021-01-22Revert "Partial Load Store Elimination"Nicolas Geoffray
This reverts commit b8686ce4c93eba7192ed7ef89e7ffd9f3aa6cd07. Bug: 67037140 Reason for revert: Fails a few tests. Change-Id: Icf0635bffbfbba93bf0a5b854a9582c418198136
2021-01-21Partial Load Store EliminationAlex Light
Add partial load-store elimination to the LSE pass. Partial LSE will move object allocations which only escape along certain execution paths closer to the escape point and allow more values to be eliminated. It does this by creating new predicated load and store instructions that are used when an object has only escaped some of the time. In cases where the object has not escaped a default value will be used. Test: ./test.py --host Test: ./test.py --target Bug: 67037140 Change-Id: Idde67eb59ec90de79747cde17b552eec05b58497
2021-01-06Optimizing: Add debugging output for HInstruction.Vladimir Marko
Allow printing individual instruction and its arguments with the HGraphVisualizer. Arguments are dumped "recursively" (but implemented with a queue instead of actual recursion). For example, printing the Return instruction from the method Main.testLoop17 in 530-checker-lse yields v28 Return [i27] dex_pc:23 loop:none i27 Add [i24,i26] dex_pc:22 loop:none i24 Phi [i5,i15] dex_pc:n/a reg:0 is_catch_phi:false loop:none i5 IntConstant dex_pc:0 1 loop:none i15 IntConstant dex_pc:5 2 loop:none i26 InstanceFieldGet [l6] dex_pc:20 field_name:TestClass.i field_type:Int32 loop:none l6 NullCheck [l1] dex_pc:1 env:[[i5,_,_,l1,i2]] loop:none l1 ParameterValue dex_pc:n/a loop:none Test: Manual; modify LSE to print the instruction above. Change-Id: Iaf41ba62cd6a5a36236ad0abca082ebffcf6a20e
2020-09-07Do not build intrinsic graph for signature polymorphic methods.Vladimir Marko
The InstructionBuilder cannot build HInvokePolymorphic for an intrinsic graph. It would be rather useless to allow this as signature polymorphic method intrinsics always need the slow-path for unhandled cases, and we would therefore reject the compiled code anyway. Instead, prevent the construction of the intrinsic graph for signature polymorphic methods. Test: testrunner.py --host --optimizing -t 712-varhandle-invocations Bug: 65872996 Change-Id: Id82f1f282383dbd45d6db2bd2a96d838b685d7ed
2020-09-07Move code item to the data pointer and remove code_item_offset.Nicolas Geoffray
This saves 4 bytes on 32bit and 8 bytes on 64bit on ArtMethod. Also update nterp to directly fetch the code item from the data pointer. Test: test.py Bug: 112676029 Change-Id: Ic01f43c7ccf2cbce1ec517478e81362232d36371
2020-08-21Improved LSE: Replacing loads with Phis.Vladimir Marko
Create "Phi placeholders" for tracking heap values that can merge from different values and try to match existing Phis or create new Phis to replace loads. For Phi placeholders from loop headers we do not know whether they are fed by unknown values through back-edges when processing the loop header, so we delay processing loads that depend on them until we walked the entire graph. We then try to match them with existing instructions (when the location is unchanged in the loop) or Phis or create new Phis if needed. If we find a loop Phi placeholder fed with unknown value from a back-edge, we mark the Phi placeholder unreplaceable and reprocess loads and stores to propagate the unknown value. This can sometimes allow other loads to be replaced. At the end we re-calculate the heap values to find stores that can be eliminated because they write over the same value. Golem results: art-opt-cc arm arm64 x86 x86-64 CaffeineFloat +6.7% +3.0% +5.9% +3.8% KotlinMicroWhen +33.7% +4.8% +1.8% +0.6% art-opt (more noisy than art-opt-cc) CaffeineFloat +4.1% +4.4% +7.8% +10.5% KotlinMicroWhen +33.6% +2.0% +1.8% +1.8% The MoveLiteralColumn benchmark seems to gain significantly (up to 22% on art-opt-cc but under 10% on art-opt) but it is very noisy and the results are therefore unreliable. Insignificant code size changes for aosp_blueline-userdebug: - before: arm boot*.oat: 15303468 arm64 boot*.oat: 18184736 services.odex: 25195944 grep -c pAllocObject boot.arm64.oatdump.txt: 27213 grep -c pAllocArray boot.arm64.oatdump.txt: 3620 - after: arm boot*.oat: 15299524 (-4KiB, -0.03%) arm64 boot*.oat: 18176528 (-8KiB, -0.05%) services.odex: 25191832 (-4KiB, -0.02%) grep -c pAllocObject boot.arm64.oatdump.txt: 27206 (-7) grep -c pAllocArray boot.arm64.oatdump.txt: 3615 (-5) Test: New tests in 530-checker-lse. Test: m test-art-host-gtest Test: testrunner.py --host --optimizing Test: blueline-userdebug boots. Bug: 77906240 Change-Id: Ia9fe0cd3530f9d3941650dfefc00a7f7fd821994
2020-08-10ARM: Allow FP args in core regs for @CriticalNative.Vladimir Marko
If a float or double argument needs to be passed in core register to a @CriticalNative method due to soft-float native ABI, insert a fake call to Float.floatToRawIntBits() or Double.doubleToRawLongBits() to satisfy type checks in the compiler. We cannot do that for intrinsics that expect those inputs in actual FP registers, so we still prevent such intrinsics from using `kCallCriticalNative`. This should be irrelevant if an actual intrinsic implementation is emitted. There are currently two unimplemented intrinsics that are affected by the carve-out, namely MathRoundDouble and FP16ToHalf, and four intrinsics implemented only when ARMv8A is supported, namely MathRint, MathRoundFloat, MathCeil and MathFloor. Test: testrunner.py --target --32 -t 178-app-image-native-method Bug: 112189621 Change-Id: Id14ef4f49f8a0e6489f97dc9588c0e6a5c122632
2020-07-28Dump ISA into .cfgEvgeny Astigeevich
The optimizing compiler dumps ISA features into .cfg which can be used by the checker. A new tool perf2cfg needs to know ISA a CFG corresponds to. The tool uses this to check that input perf data has the same ISA. This CL implements in the compiler dumping ISA in addition to ISA features: begin_compilation name "isa:some_isa isa_features:feature1,-feature2" method "isa:some_isa isa_features:feature1,-feature2" date 1580721972 end_compilation Bug: 147876827 Test: ./art/tools/checker/run_unit_tests.py Test: test.py --target --optimizing --jit --interpreter Test: test.py --host --optimizing --jit --interpreter Change-Id: I189eae0f4de61e6a49c01d925e7136b5f7027c91
2020-07-22Dump instruction set features in .cfgFabio Rinaldi
This commit adds a compilation block at the beginning of the .cfg dumped by the optimizing compiler when --dump-cfg is enabled. The compilation block appears in the following form: begin_compilation name "isa_features:feature1,-feature2" method "isa_features:feature1,-feature2" date 1580721972 end_compilation This compilation block dump is passed to checker script (see https://android-review.googlesource.com/c/platform/art/+/1290997) for checking if a certain instruction set feature was used at compile time. Author: Fabio Rinaldi Committer: Artem Serov Bug: 147876827 Test: ./art/tools/checker/run_unit_tests.py Test: test.py --target --optimizing Change-Id: If4309af4bab892f715aad1d3bd338f8ee11e497c
2020-06-18Introduce an enum for the compilation kind.Nicolas Geoffray
Test: test.py Change-Id: I5329e50a6b4521933b6b171c8c0fbc618c3f67cd
2020-06-17ART: Simplify HRem to reuse existing HDivEvgeny Astigeevich
A pattern seen in libcore and SPECjvm2008 workloads is a pair of HRem/HDiv having the same dividend and divisor. The code generator processes them separately and generates duplicated instructions calculating HDiv. This CL adds detection of such a pattern to the instruction simplifier. This optimization affects HInductionVarAnalysis and HLoopOptimization preventing some loop optimizations. To avoid this the instruction simplifier has the loop_friendly mode which means not to optimize HRems if they are in a loop. A microbenchmark run on Pixel 3 shows the following improvements: | little cores | big cores arm32 Int32 | +21% | +40% arm32 Int64 | +46% | +44% arm64 Int32 | +27% | +14% arm64 Int64 | +33% | +27% Test: 411-checker-instruct-simplifier-hrem Test: test.py --host --optimizing --jit --gtest --interpreter Test: test.py --target --optimizing --jit --interpreter Test: run-gtests.sh Change-Id: I376a1bd299d7fe10acad46771236edd5f85dfe56
2020-06-09ART: Introduce predicated vector instructions.Artem Serov
This CL introduces a minimal changes to the IR to support autovectorization with use of predicated execution of SIMD instructions (e.g. Arm SVE). Test: test-art-target, test-art-host. Change-Id: Ibb7c5520fec6b858fb29f0dde19ec65501831a3a
2020-06-08Run LSA as a part of the LSE pass.Vladimir Marko
Make LSA a helper class, not an optimization pass. Move all its allocations to ScopedArenaAllocator to reduce the peak memory usage a little bit. Test: m test-art-host-gtest Test: testrunner.py --host --optimizing Change-Id: I7fc634abe732d22c99005921ffecac5207bcf05f
2020-05-20Add compiler type to CompilerOptions.Vladimir Marko
Let CompilerOptions hold the information whether it is AOT or JIT compilation, or Zygote JIT for shared code. Test: m test-art-host-gtest Test: testrunner.py --host --optimizing --jit Test: aosp_taimen-userdebug boots. Change-Id: Id9200572406f8e43d99b8b61ef0e3edf43b52fff
2020-05-13Move HandleCache to HGraph.Vladimir Marko
This avoids passing the `VariableSizedHandleScope*` argument around and eliminates HGraph::inexact_object_rti_ and its initialization. The latter shall allow running Optimizing gtests that do not require type information without creating a Runtime in future. (To be implemented in a separate CL.) Test: m test-art-host-gtest Test: testrunner.py --host --optmizing Test: aosp_taimen-userdebug boots. Change-Id: I36fe9bc556c6d610d644c8c14cc74c9985a14d64
2020-05-13Move implementations from class_root.h to -inl.h .Vladimir Marko
Make it possible to include the definition of enum ClassRoot without pulling in a lot of other headers. Test: m test-art-host-gtest Test: testrunner.py --host --optimizing Test: aosp_taimen-userdebug boots. Change-Id: Ic90fdd70bfe0c5428a5c9a0d7901ea7e15b03488
2020-04-01Add explicit compiler option to enable test-specific features.David Srbecky
We can no longer rely on checking the "core.art" image name, since we plan to remove it and use the shipped boot image. This option enables test-specific features, such as $noinline$. Test: ./art/test.py -r --optimizing --64 Bug: 147817558 Change-Id: Iadac6e1b0f46e83efd0551fb8462a6b268ad33d8
2020-03-25Add more logging and sanity checks for JIT mini-debug-info.David Srbecky
Used when diagnosing b/151137723. Keep it around. Bug: 151137723 Test: test.py -r --jit Change-Id: I10cc613c7396607e221fdc1f5972d26c1ac03fa8
2020-03-12Refactor RemoveNativeDebugInfoForJit.David Srbecky
This is partial revert of CL/1099280 (Remove global maps). It somewhat resurrects the lazy method removal. The original goal was to only remove methods from the GC, and do all of them in bulk for simplicity and efficiency. However, this is proving infeasible since we have several corner cases which remove methods outside the GC code path. The behaviour for the GC code path is preserved by this CL. Instead of passing method array, the methods are individually marked for removal and then repacking is immediately forced. The only difference is that coroner cases are done lazily. Test: ./art/test.py -b -r --host --jit --64 Change-Id: I42729545d6b51df788d92f9cf149a6e065b90c68
2020-02-22Remove JIT native debug info on failed commits.David Srbecky
This could happen if single-implementation assumptions are invalid, and would result in temporary double entries in JIT debug info. Test: ./art/test.py -b -r --host --gcstress --jit --32 -t 068 Change-Id: I765c9310b6cc4fa440e4121780f8b9053388bd3d
2020-02-13Remove MIPS support from Optimizing.Vladimir Marko
Test: aosp_taimen-userdebug boots. Test: m test-art-host-gtest Test: testrunner.py --host --optimizing Bug: 147346243 Change-Id: I97fdc15e568ae3fe390efb1da690343025f84944
2020-02-04Only flush visualizer output at the end of the compilation.Nicolas Geoffray
Otherwise multiple threads compiling will mix their output. Test: run-tests Bug: 147094528 Change-Id: I905c8874cbe2c02c2d79577024d4606a3784b67a
2019-12-03JIT baseline: trigger optimized compilation on hotness threshold.Nicolas Geoffray
- Add a new hotness count in the ProfilingInfo to not conflict with interpreter hotness which may use it for OSR. - Add a baseline flag in the OatQuickMethodHeader to identify baseline compiled methods. - Add a -Xusetieredjit flag to experiment and test. Bug: 119800099 Test: test.py with Xusetieredjit to true Change-Id: I8512853f869f1312e3edc60bf64413dee9143c52
2019-11-26Honor the --baseline flag for the JIT.Nicolas Geoffray
Test: art -Xcompiler-option --baseline uses the baseline compiler. Change-Id: I2a5a3ed0038ca965ab1f0cebffb7900e90bae43a
2019-10-14Revert "Make compiler/optimizing/ symbols hidden."Vladimir Marko
This reverts commit e2727154f25e0db9a5bb92af494d8e47b181dfcf. Reason for revert: Breaks ASAN tests (ODR violation). Bug: 142365358 Change-Id: I38103d74a1297256c81d90872b6902ff1e9ef7a4
2019-10-14Make compiler/optimizing/ symbols hidden.Vladimir Marko
Make symbols in compiler/optimizing hidden by a namespace attribute. The unit intrinsic_objects.{h,cc} is excluded as it is needed by dex2oat. As the symbols are no longer exported, gtests are now linked with the static version of the libartd-compiler library. libart-compiler.so size: - before: arm: 2396152 arm64: 3345280 - after: arm: 2016176 (-371KiB, -15.9%) arm64: 2874480 (-460KiB, -14.1%) Test: m test-art-host-gtest Test: testrunner.py --host --optimizing --jit Bug: 142365358 Change-Id: I1fb04a33351f53f00b389a1642e81a68e40912a8
2019-10-10Revert^2 "JIT: Separate code allocation and initialization."David Srbecky
This reverts commit 63b0c26aae3e7237166dd781eb7a15fbc7c091c2. Test: ./art/test.py -b -r --host --all-gc -t 708 Reason for revert: Reland after bug fix. Change-Id: Ic13e2799bf4bdd8ca468f72cc0f3b72f224f2b08
2019-10-09Revert "JIT: Separate code allocation and initialization."Nicolas Geoffray
This reverts commit e1b36f09cb478a39ba443f6acb11cd1901c01c1d. Reason for revert: Crashes in art::jit::JitMemoryRegion::FreeCode https://logs.chromium.org/logs/art/buildbucket/cr-buildbucket.appspot.com/8900060997388741808/+/steps/test_libcore/0/stdout Change-Id: I40259baec9acbb7889732548f8b31e0a48651cd9
2019-10-09JIT: Separate code allocation and initialization.David Srbecky
Allocate(reserve) both code and data first, and then initialise(commit) both of them. This is useful since we know the address sooner for the purpose of debug info generation. Test: ./art/test.py -b -r --jit --host --64 Change-Id: I4971a8801004efbc6c2b27884834dda775b72664
2019-10-08JIT mini-debug-info: Generate the debug info sooner.David Srbecky
We need to add the debug info before the entry-point is set, otherwise there is window of time when we are missing it. Test: test.py -b -r --optimizing --jit --host --64 Change-Id: I3bbad0c96d68e9603bac131faee498b8b9f62699
2019-09-01Fix JIT data dual mapping for apps.David Srbecky
We don't use it now, but it would be nice to make it functional. Mark the read-only memory as const, and fix the compile errors. Test: test.py -b --host --jit Bug: 119800099 Change-Id: Ic1c45072f3c97f560e843f95fb87b95f754c6e03
2019-08-22JIT mini-debug-info: Allocate entries in the JIT data space.David Srbecky
Test: test.py -b --host --jit Test: device boots Bug: 119800099 Change-Id: I7efa1e6e6660239cbd6438b829e08dd9cd079343
2019-08-16JIT mini-debug-info: Remove global maps.David Srbecky
Keep the extra bookkeeping information in JITCodeEntry. Also do the compression eagerly during GC rather then lazily. Test: test.py -b --host --jit Bug: 119800099 Change-Id: Ie6cc682033a32c01d4c2cac242d8a4201116f940
2019-08-07Don't compile OSR methods that have phi equivalents at loop entry.Nicolas Geoffray
We currently don't handle this in the stack map, where we only encode one stack slot for a dex register. Bug: 136698025 Test: 721-osr Change-Id: Ib395ed1165387ad5446a463c307cc0a45e365885
2019-06-28Revert "Revert "Make the JIT zygote memory shared.""Nicolas Geoffray
This reverts commit 2fef66b294417d447630f9d98de68227eef476d3. Bug: 119800099 Bug: 136110523 Reason for revert: Fixed webview_zygote case. Change-Id: Iaae8c999463d77b7b1e62b55458493bdbc97a104
2019-06-26Revert "Make the JIT zygote memory shared."Nicolas Geoffray
This reverts commit 05f87217ddc9b4b9186710c0135b918f456c5aef. Bug: 119800099 Bug: 136110523 Reason for revert: testWebview flaking Change-Id: I96afa6bc9c56c4aaf5ed72ae370f6f69c096c559
2019-06-26Make the JIT zygote memory shared.Nicolas Geoffray
Test: boots Bug: 119800099 Change-Id: I75ff8a58eea4de5cb833139641b4e15b8394d9b1
2019-06-24Move code writing to data portion of the JIT cache in JitMemoryRegion.Nicolas Geoffray
So there is only one method that needs to write to the data portion of the region. Test: test.py --jit Change-Id: I5b7dc442526da54f83a208cb70c8cf86fad6ebcd
2019-06-10Pass the memory region to allocate into to the compiler / allocation.Nicolas Geoffray
Test: test.py Bug: 119800099 Change-Id: Ie3cba5abe3dd4f8756af5ecfd6c26320de314fe8
2019-04-29Fix DWARF line-number generation for JITed methods.David Srbecky
Don't strip, repack or compress debug-info if explicitly requested by the developer (using the -g compiler flag). If enabled, the DWARF debug info has about 1:1 size overhead relative to JIT code + data. Bug: 131422204 Test: Check that gdb shows line numbers for JITed method. Test: Hard-code always-enable generation and run maps. Change-Id: If06de8ae2317af4d57d84e8a8bfae86a597dd4e4
2019-03-27Use generic name for JNI trampolines.David Srbecky
JNI trampolines tend to be deduplicated which seems to lead to confusion. Bug: 32949969 Test: test.py -b --host -r -t 137-cfi Change-Id: Ic39998448a05f98fc1e068c4f384f35085121911
2019-03-26optimizing_compiler: Avoid extra std::string copyGreg Kaiser
The function OptimizationPassByName() takes a const std::string reference, so it's inefficient to pass it the result of std::string::c_str(). Test: TreeHugger Change-Id: I84fc717285b0278264cba3c9504290876025a01d
2019-03-06Revert "Revert^2 "Add code size to CodeInfo""Nicolas Geoffray
This reverts commit e1412dacbf1d2a809bd1fca658cc8cb8f61f8ee6. Bug: 123510633 Bug: 127305289 Reason for revert: b/127305289 Change-Id: I54557b05a44777f1fa2c15bde4fa648980f42eed
2019-02-20ART: Reduce dependencies on CompilerDriver.Vladimir Marko
Preparation for moving CompilerDriver and other stuff from libart-compiler.so to dex2oat. Test: m test-art-host-gtest Test: testrunner.py --host --optimizing Change-Id: Ic221ebca4b8c79dd1549316921ace655f2e3f0fe
generated by cgit v1.2.3 (git 2.25.1) at 2025-10-04 06:09:38 +0000