)]}'
{
  "log": [
    {
      "commit": "dba61481035b7944173181ec9ee02aea41dd0e29",
      "tree": "8629775e657cc8968e2f8fa2af6256c7a58c7dd4",
      "parents": [
        "cb2ce333ee2d1053919fe0acc390369bea3ad77e"
      ],
      "author": {
        "name": "Alex Light",
        "email": "allight@google.com",
        "time": "Wed Dec 21 08:20:29 2016 -0800"
      },
      "committer": {
        "name": "Alex Light",
        "email": "allight@google.com",
        "time": "Wed Dec 21 11:43:14 2016 -0800"
      },
      "message": "Revert \"Revert \"Revert \"Revert \"Basic obsolete methods support\"\"\"\"\n\nA GetDeclaringClass()-\u003eGetDexCache() got inserted during the\nmerge/review process meaning that we would try to access incorrect\ndex-cache in obsolete methods in some situations.\n\nAlso when using tracing we would loop forever (or at least until an\nOOM error) in test 916 due to tracing forcing InterpretOnly mode\nmeaning methods would never be jitted.\n\nBug: 32369913\nBug: 33630159\n\nTest: ART_TEST_TRACE\u003dtrue \\\n      ART_TEST_JIT\u003dtrue   \\\n      ART_TEST_INTERPRETER\u003dtrue mma -j40 test-art-host\n\nThis reverts commit f6abcda293b115a9d7d8a26376ea2dcf2d1dc510.\n\nChange-Id: I0773bfcba52e3cd51a83be815c6a50c189558f48\n"
    },
    {
      "commit": "f6abcda293b115a9d7d8a26376ea2dcf2d1dc510",
      "tree": "a30ce596848d14c95a73449dbf2bfccf26a510b9",
      "parents": [
        "ce77fc0e7f60a15354bb20c356537cbf8b53b722"
      ],
      "author": {
        "name": "Nicolas Geoffray",
        "email": "ngeoffray@google.com",
        "time": "Wed Dec 21 09:26:18 2016 +0000"
      },
      "committer": {
        "name": "Nicolas Geoffray",
        "email": "ngeoffray@google.com",
        "time": "Wed Dec 21 09:26:18 2016 +0000"
      },
      "message": "Revert \"Revert \"Revert \"Basic obsolete methods support\"\"\"\n\nFails in tracing mode\n\nBug: 32369913\nBug: 33630159\n\nThis reverts commit ce77fc0e7f60a15354bb20c356537cbf8b53b722.\n\nChange-Id: I1bdcf6ad467f2e31f9c5d0c3c987b90a4f5efc69\n"
    },
    {
      "commit": "ce77fc0e7f60a15354bb20c356537cbf8b53b722",
      "tree": "b4301967c423ab190a4274409d99a3d6e23f0e8a",
      "parents": [
        "d54f43ca39dfa92f08c2d760123f185f0f65fb86"
      ],
      "author": {
        "name": "Alex Light",
        "email": "allight@google.com",
        "time": "Thu Dec 15 01:05:52 2016 +0000"
      },
      "committer": {
        "name": "Alex Light",
        "email": "allight@google.com",
        "time": "Mon Dec 19 14:26:18 2016 -0800"
      },
      "message": "Revert \"Revert \"Basic obsolete methods support\"\"\n\nThis reverts commit b81a9840b44480bfeacd74b8d9f51e06f295411d.\n\nThere were 2 issues with the original CL\u0027s test 916-obsolete-jit that\ncaused it to sporadically fail.\n\nFirst, when checking if we had jitted the function under test in\n916-obsolete-jit we failed to check to see if the function \u0027doCall\u0027,\nwhich is used to work-around bugs in our deoptimization \u0026 compilation\nsystems, had also been jitted.\n\nIn the case where the \u0027sayHi\u0027 function had been jitted but the\n\u0027doCall\u0027 function had not we would (correctly) fail to redefine the\n\u0027Transform\u0027 class since we would not be able to deoptimize the \u0027sayHi\u0027\nfunction since it is under a quick_to_interpreter_bridge (runtime)\nframe.\n\nSecondly, the function Main.isInterpretedFunction was incorrect and\nwould always return false, regardless of the actual state of the\nfunction. This would cause the test to fail as the\nquick_to_interpreter_bridge frame prevented deoptimization of the\nobsoleted function. Usually the warm-up period was enough to make sure\nthe methods were jitted anyway but this was not guaranteed.\n\nBoth of these problems become more likely to occur on systems with\nmore cpu contention such as the buildbots.\n\nTest: stress --cpu 60 \u0026; while ./test/run-test --host --jit 916; do ; done\n\nReason for revert: Fixed bug in test that was causing failures.\n\nOriginal Tests:\n\nTest: ./test/run-test --host 914\nTest: ./test/run-test --host 915\nTest: ./test/run-test --host 916\nTest: mma -j40 test-art-host\nTest: ART_TEST_JIT\u003dtrue \\\n      ART_TEST_INTERPRETER\u003dtrue mma -j40 test-art-host\nTest: ./art/tools/run-jdwp-tests.sh --mode\u003dhost --variant\u003dX64\nTest: ./art/tools/run-jdwp-tests.sh --mode\u003dhost --variant\u003dX64 --no-jit\n\nBug: 32369913\nBug: 33630159\n\nChange-Id: If1a92e47b90965a7dc21c5826185debe62bd1554\n"
    },
    {
      "commit": "b81a9840b44480bfeacd74b8d9f51e06f295411d",
      "tree": "d79a3567b272b6eb4fb94d66a537cf92bef455db",
      "parents": [
        "d8936da27b792d1ca02e59c92456a1a53c7b9905"
      ],
      "author": {
        "name": "Alex Light",
        "email": "allight@google.com",
        "time": "Thu Dec 15 00:59:05 2016 +0000"
      },
      "committer": {
        "name": "Alex Light",
        "email": "allight@google.com",
        "time": "Thu Dec 15 00:59:05 2016 +0000"
      },
      "message": "Revert \"Basic obsolete methods support\"\n\nThis reverts commit d8936da27b792d1ca02e59c92456a1a53c7b9905.\n\nReason for revert: Some sort of race in JIT\n\nChange-Id: Ibb4e520bb0721d6d7aa2c841a52eb5baff07740e\n"
    },
    {
      "commit": "d8936da27b792d1ca02e59c92456a1a53c7b9905",
      "tree": "6a593dd426e7f4a15c651dd3b872f22fb64f09ce",
      "parents": [
        "7c73bf3518eea7544406eadfed6946411c54a267"
      ],
      "author": {
        "name": "Alex Light",
        "email": "allight@google.com",
        "time": "Mon Nov 28 16:24:32 2016 -0800"
      },
      "committer": {
        "name": "Alex Light",
        "email": "allight@google.com",
        "time": "Wed Dec 14 13:01:27 2016 -0800"
      },
      "message": "Basic obsolete methods support\n\nAdd support for executing obsolete methods following redefinitions.\nThis support includes methods that have been jitted. This does not add\nany additional validity checks to our redefinition functions.\n\nNote using work-arounds to ensure 916 works pending some fixes to the\ncompiler, deoptimizer, or both.\n\nTest: ./test/run-test --host 914\nTest: ./test/run-test --host 915\nTest: ./test/run-test --host 916\nTest: mma -j40 test-art-host\nTest: ART_TEST_JIT\u003dtrue \\\n      ART_TEST_INTERPRETER\u003dtrue mma -j40 test-art-host\nTest: ./art/tools/run-jdwp-tests.sh --mode\u003dhost --variant\u003dX64\nTest: ./art/tools/run-jdwp-tests.sh --mode\u003dhost --variant\u003dX64 --no-jit\n\nBug: 32369913\nBug: 33630159\n\nChange-Id: I78ef95f484146f1fb93c37fc50f56575bdab2432\n"
    },
    {
      "commit": "97d7e1cd7f733cb33a0e238bec6d7ed525638cd1",
      "tree": "d2f02420f82d76d10dc0d80f431c9f4aba03ac82",
      "parents": [
        "eebe142e935e9444c7c947d02246e7fbe88b2cc4"
      ],
      "author": {
        "name": "Vladimir Marko",
        "email": "vmarko@google.com",
        "time": "Tue Oct 04 14:44:28 2016 +0100"
      },
      "committer": {
        "name": "Vladimir Marko",
        "email": "vmarko@google.com",
        "time": "Wed Oct 05 13:21:44 2016 +0100"
      },
      "message": "Remove #include \"oat_file.h\" from class_linker.h .\n\nRefactor the OatClass and OatMethod related functions from\nClassLinker to OatFile and ArtMethod, respectively. Refactor\nthe remaining ClassLinker dependencies on OatFile to break\nthe #include dependency and reduce incremental build times.\n\nTest: m test-art-host\nChange-Id: Iebc5b9f81b48fbcf79821cc827a5d7c4a0261bf6\n"
    },
    {
      "commit": "0795f23920ee9aabf28e45c63cd592dcccf00216",
      "tree": "ff3f880c5e84f3316532b47d0e9a7729ade848ac",
      "parents": [
        "d1224dce59eb0019507e41da5e10f12dda66bee4"
      ],
      "author": {
        "name": "Mathieu Chartier",
        "email": "mathieuc@google.com",
        "time": "Tue Sep 27 18:43:30 2016 -0700"
      },
      "committer": {
        "name": "Mathieu Chartier",
        "email": "mathieuc@google.com",
        "time": "Thu Sep 29 17:31:09 2016 -0700"
      },
      "message": "Clean up ScopedThreadStateChange to use ObjPtr\n\nAlso fixed inclusion of -inl.h files in .h files by adding\nscoped_object_access-inl.h and scoped_fast_natvie_object_access-inl.h\n\nChanged AddLocalReference / Decode to use ObjPtr.\n\nChanged libartbenchmark to be debug to avoid linkage errors.\n\nBug: 31113334\n\nTest: test-art-host\n\nChange-Id: I4d2e160483a29d21e1e0e440585ed328b9811483\n"
    },
    {
      "commit": "f711f2cf3c28dfe865e36f17419a16f06a0ebb5a",
      "tree": "3657d68b2a2fb67f4919cb97ca279cf9310224e9",
      "parents": [
        "bbfa42aa4c4b5cbc2ed2d8bdea6ff76da615a8aa"
      ],
      "author": {
        "name": "Mingyao Yang",
        "email": "mingyao@google.com",
        "time": "Mon May 23 12:29:39 2016 -0700"
      },
      "committer": {
        "name": "Mingyao Yang",
        "email": "mingyao@google.com",
        "time": "Thu May 26 13:24:54 2016 -0700"
      },
      "message": "Partial fragment deoptimization\n\nWe used to do either single frame deoptimization, or full fragment\ndeoptimization which deoptimizes all the frames in a fragment.\nThis change allows some methods to be not deoptimizeable, likely due\nto some kind of optimization. So we need another deoptimization mode\nthat unwinds partial fragment. Deoptimizations are now generalized into\neither full or partial fragment. A full fragment deoptimization will\ndeopt all frames in the fragment, and then returns from the invoke stub\nto enter interpreter. A partial fragment deoptimization will deopt a\nsingle frame, or all frames up to the method that\u0027s not deoptimizeable,\nand then jumps to the interpreter bridge.\n\nCurrently code not deoptimizeable is the code in boot image since the\ncode may not be compiled with debuggable flag.\n\nBug: 28769520\nChange-Id: I875c694791cc8ebd5121abcd92ce7b0db95aca38\n"
    },
    {
      "commit": "0dfc9bc0cca3ae15dc75bbea3ca9b6d7508ce893",
      "tree": "abdc76069da7142d899eb776a0d82bdbbe316c94",
      "parents": [
        "1aba494df441c31836e5b7d4d8e408420a93cb8d"
      ],
      "author": {
        "name": "Andreas Gampe",
        "email": "agampe@google.com",
        "time": "Wed Sep 30 17:13:59 2015 -0700"
      },
      "committer": {
        "name": "Andreas Gampe",
        "email": "agampe@google.com",
        "time": "Mon Oct 05 08:57:04 2015 -0700"
      },
      "message": "ART: Add more deopt tests\n\nExtend libarttest with the ability to turn off asserts for method\nstate. Use this to put asserts into test functions, but turn them\noff if the test wouldn\u0027t react as we want, e.g., when we\u0027re in\ninterpreter mode.\n\nExtend run-test 449 by adding asserts for the method state for the\nexpected deopt cases. This tests both standard behavior, as well\nas single-frame deoptimization.\n\nBug: 21611912\nChange-Id: I27fa5e3e44d8c5eab57962d144c6aec96587bf45\n"
    },
    {
      "commit": "89df7bfc41a4de9685f84e7db07f77db3fd485fc",
      "tree": "f1a266870d0020b9b814022698569affc90e3844",
      "parents": [
        "931e26843bbb688eacfa67b40414c6b8f221a56a"
      ],
      "author": {
        "name": "Andreas Gampe",
        "email": "agampe@google.com",
        "time": "Wed Sep 30 13:13:21 2015 -0700"
      },
      "committer": {
        "name": "Andreas Gampe",
        "email": "agampe@google.com",
        "time": "Wed Sep 30 13:13:21 2015 -0700"
      },
      "message": "ART: Refactor libarttest C++ code\n\nRefactor some libarttest JNI functions into a common directory, so\nthey can be easily used by multiple tests without code duplication.\nThis is preparation for better deoptimization state testing.\n\nChange-Id: I43f7340b699b26930aabfd665c707e0485b17fc8\n"
    },
    {
      "commit": "a727e372d8f6929cd30b983f6969c7a50fc83bb6",
      "tree": "f71b33bf46053e0ccbd191ac74a850e87145eb98",
      "parents": [
        "706ea6a0294ef8bc3def050142c548c082fa8e30"
      ],
      "author": {
        "name": "Andreas Gampe",
        "email": "agampe@google.com",
        "time": "Tue Aug 25 09:22:37 2015 -0700"
      },
      "committer": {
        "name": "Andreas Gampe",
        "email": "agampe@google.com",
        "time": "Thu Aug 27 10:53:12 2015 -0700"
      },
      "message": "ART: Balanced locking\n\nChange the verifier to check for balanced locking. When balanced\nlocking can\u0027t be guaranteed, use a new failure kind to punt to\nthe interpreter.\n\nAdd smali tests, with JNI code to check the balanced-locking result.\n\nBug: 23502994\nChange-Id: Icd7db0be20ef2f69f0ac784de43dcba990035cd8\n"
    }
  ]
}