)]}'
{
  "log": [
    {
      "commit": "37dc4df47fec811ea52f7180880961565f013434",
      "tree": "eac308a6c7ef8b7d53f64889ff0a93740a2dc62a",
      "parents": [
        "76754cc816af46b41a8d1f419a38334b5db59b6e"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Jun 28 14:08:00 2017 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Thu Jun 29 11:20:56 2017 -0700"
      },
      "message": "Improved subscript and data dependence analysis.\n\nRationale:\nWe missed vectorizing a simple stencil operation\ndue to inaccurate unit stride analysis and failure\nto detect single runtime data dependence test.\n\nTest: test-art-host, test-art-target\nChange-Id: I07ba03455bfb1c0aff371c1244a1328f885d0916\n"
    },
    {
      "commit": "8523ea11a677b78e1fc05915976c04b1ff081451",
      "tree": "249523715fbba92a524fe01725a218c695c1b19e",
      "parents": [
        "85b82e35d810a21257fec9d850d95f8abc900670"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Thu Jun 01 15:45:09 2017 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Fri Jun 02 08:52:13 2017 -0700"
      },
      "message": "Fixed bug in relying on precise FP in periodic sequence.\n\nRationale:\nFP arithmetic is not always precise, so relying on FP\nperiod sequences to \"compute back\" precisely is not\nvalid; when all values in the period are \"fetches\"\nhowever, the rotation is precise.\n\nBug found by fuzz testing. With regression test.\n\nBug: 62196559\nTest: test-art-host\nChange-Id: Ie8f6b965b1921ff2762b90eebb8c20503c44c6bb\n"
    },
    {
      "commit": "f3e61ee363fe7f82ef56704f06d753e2034a67dd",
      "tree": "a00f1fce4a2e284b0a03f941f530afc5b5c56b59",
      "parents": [
        "741a81af441cbcb7255229bf250bc009d2894e92"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Apr 12 17:09:20 2017 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Apr 19 10:30:57 2017 -0700"
      },
      "message": "Implement halving add idiom (with checker tests).\n\nRationale:\nFirst of several idioms that map to very efficient SIMD instructions.\nNote that the is-zero-ext and is-sign-ext are general-purpose utilities\nthat will be widely used in the vectorizer to detect low precision\nidioms, so expect that code to be shared with many CLs to come.\n\nTest: test-art-host, test-art-target\nChange-Id: If7dc2926c72a2e4b5cea15c44ef68cf5503e9be9\n"
    },
    {
      "commit": "fa76296bc624bc2c879167c260ad6925238efb3d",
      "tree": "045c4488fec66806982a62aa4cfd7d3a2183e958",
      "parents": [
        "6f005931d0a0db65128db803df38d59d205dd218"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Fri Apr 07 11:33:37 2017 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Fri Apr 07 23:03:42 2017 +0000"
      },
      "message": "Fixed missing context while detecting unit strides.\n\nWith regression test (found by fuzz testing).\n\nBug: 37033123\nTest: test-art-target\nChange-Id: Id738b2a3a353985c3d0bf3beeb581a31f1fcbc3f\n"
    },
    {
      "commit": "7adb688fe1213c6dcaf4d2adae3d8b7decf7c28a",
      "tree": "fdafea7c8cdcf9c44825a630e4b1229ece444cd9",
      "parents": [
        "e6a60bff7a78efe7c2c4365984f1f7f036c12158"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Tue Mar 07 13:28:51 2017 -0800"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Tue Mar 07 13:28:51 2017 -0800"
      },
      "message": "Improve isunit computation, some tests edits needed for SIMD.\n\nRationale:\nBreak-out CL of ART Vectorizer: number 4.\nThe purpose is making the original CL smaller\nand easier to review.\n\nBug: 34083438\nTest: test-art-host\nChange-Id: I62a174944bbbe1e08f631f322c513eeaea14de26\n"
    },
    {
      "commit": "b603a5cdff7c1fb300b83175bfb3cd990fdd6265",
      "tree": "523f82d02727bd095c650463e0087af3853f59df",
      "parents": [
        "d88663823f75dbd68cace178ddca4767775d97a1"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Mon Mar 06 18:29:39 2017 -0800"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Tue Mar 07 08:59:08 2017 -0800"
      },
      "message": "Bug fix in overflow detection on pow. With regression test.\n\nRationale:\nMissed potential overflow in base update.\n\nTest: test-art-host\nBug: 35706963\nChange-Id: I117d96c1fa95dde37f5676f9dc750f0a71becafa\n"
    },
    {
      "commit": "8e02e3e77ff8ac7a80e25689750b4d329593e12a",
      "tree": "f8201a2def0d3f9c05b05ddcb8ab338f84169ab7",
      "parents": [
        "e0aa5beef7419b8c9ab5d3dc93553ef2b30d126c"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Tue Feb 28 14:41:55 2017 -0800"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Mar 01 11:14:05 2017 -0800"
      },
      "message": "New utilities for induction variables.\n\nRationale:\nBreak-out CL of ART Vectorizer: 2 OF many.\nThe purpose is making the original CL smaller\nand easier to review.\n\nBug: 34083438\nTest: test-art-host\nChange-Id: I46d297eba504af3850a5998ee279ea9f7b38bed8\n"
    },
    {
      "commit": "d3ba626b424511b91db8e04660e3248cdd597100",
      "tree": "3570f60f42a66fb967bb26cede8bce3c48853781",
      "parents": [
        "fbf47ea64a9f797a82030e919fa4f085c9eb5b28"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Mon Jan 30 14:37:12 2017 -0800"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Mon Jan 30 14:37:12 2017 -0800"
      },
      "message": "Fix bug in geometric last value (found with fuzz testing)\n\nRationale:\nWhen power computation overflows, div should use 0 while mul should\nuse truncated version. Both cases were incorrectly using the latter.\n\nTest: test-art-host\nBug: 34779592\nChange-Id: I9eb8e1280c58b09d57886128f4df4541c143afaa\n"
    },
    {
      "commit": "e761bccf9f0d884cc4d4ec104568cef968296492",
      "tree": "05a2d20d61c0e91270df2747f0c242433b5ce62b",
      "parents": [
        "b0355130e38034db6b904783a00f74a3524e1881"
      ],
      "author": {
        "name": "Nicolas Geoffray",
        "email": "ngeoffray@google.com",
        "time": "Thu Jan 19 08:59:37 2017 +0000"
      },
      "committer": {
        "name": "Nicolas Geoffray",
        "email": "ngeoffray@google.com",
        "time": "Thu Jan 19 09:32:17 2017 +0000"
      },
      "message": "Revert \"Revert \"Load the array class in the compiler for allocations.\"\"\n\nThis reverts commit fee255039e30c1c3dfc70c426c3d176221c3cdf9.\n\nChange-Id: I02b45f9a659d872feeb35df40b42c1be9878413a\n"
    },
    {
      "commit": "fee255039e30c1c3dfc70c426c3d176221c3cdf9",
      "tree": "8207b72cc76513fed9f7b3c01aaa32cd54a87f1c",
      "parents": [
        "cc99df230feb46ba717252f002d0cc2da6828421"
      ],
      "author": {
        "name": "Hiroshi Yamauchi",
        "email": "yamauchi@google.com",
        "time": "Thu Jan 19 02:11:15 2017 +0000"
      },
      "committer": {
        "name": "Hiroshi Yamauchi",
        "email": "yamauchi@google.com",
        "time": "Thu Jan 19 02:11:15 2017 +0000"
      },
      "message": "Revert \"Load the array class in the compiler for allocations.\"\n\nlibcore test fails.\n\nThis reverts commit cc99df230feb46ba717252f002d0cc2da6828421.\n\nChange-Id: I5bac595acd2b240886062e8c1f11f9095ff6a9ed\n"
    },
    {
      "commit": "cc99df230feb46ba717252f002d0cc2da6828421",
      "tree": "73ac045673e150fa367a8da4d46874f28e928491",
      "parents": [
        "4507fdcb70bd570d5f3968061bf991f0a1233a93"
      ],
      "author": {
        "name": "Nicolas Geoffray",
        "email": "ngeoffray@google.com",
        "time": "Tue Jan 17 23:00:24 2017 +0000"
      },
      "committer": {
        "name": "Nicolas Geoffray",
        "email": "ngeoffray@google.com",
        "time": "Wed Jan 18 15:16:32 2017 +0000"
      },
      "message": "Load the array class in the compiler for allocations.\n\nRemoving one other dependency for needing to pass\nthe current method, and having dex_cache_resolved_types_\nin ArtMethod.\n\noat file increase:\n- x64: 0.25%\n- arm32: 0.30%\n- x86: 0.28%\n\ntest: test-art-host, test-art-target\nChange-Id: Ibca4fa00d3e31954db2ccb1f65a584b8c67cb230\n"
    },
    {
      "commit": "6b69e0acb0e4c506ce2587e362c38e36e41e34ab",
      "tree": "976f08c78d3c5efa2dac8ec0409f36fae51456cb",
      "parents": [
        "93939824c7e6e16cf98941cd4724278e87d6259d"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Jan 11 10:20:43 2017 -0800"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Fri Jan 13 10:04:42 2017 -0800"
      },
      "message": "Complete unrolling of loops with small body and trip count one.\n\nRationale:\nAvoids the unnecessary loop control overhead, suspend check,\nand exposes more opportunities for constant folding in the\nresulting loop body. Fully unrolls loop in execute() of\nthe Dhrystone benchmark (3% to 8% improvements).\n\nTest: test-art-host\n\nChange-Id: If30f38caea9e9f87a929df041dfb7ed1c227aba3\n"
    },
    {
      "commit": "e6bd0272b43bf73faabc64abc9c51ab8ed128308",
      "tree": "d75a15baadf77a859cec6540d7cd10ce998e955a",
      "parents": [
        "2c43590dc2bb7fb4a3a015b1b65543bb8705ffe8"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Fri Dec 16 13:57:52 2016 -0800"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Mon Dec 19 10:57:37 2016 -0800"
      },
      "message": "Improved induction var and range analysis around types.\n\nRationale:\nLots of code should not depend on int only. This CL generalizes\nthe kinds of types that can be optimized after analysis. As part\nof the CL, however, a minor cleanup regarding type safety of the\nstored induction var analysis results is required. This further\nimproved our int benchmark, and brings the long benchmark up-to-par.\n\nTest: m test-art-host-run-test\nChange-Id: I5dfb623dabf9113de90c2f6da99328dda8f8b60b\n"
    },
    {
      "commit": "d0a022d76db4b79d105608e7bafab4c17e8ba510",
      "tree": "28fe73f251041ea8dd2ddfbb801d780fbce6968c",
      "parents": [
        "ea63e4d1c3473d1115c1aef674056d035cf67472"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Tue Dec 13 11:22:31 2016 -0800"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Thu Dec 15 11:31:12 2016 -0800"
      },
      "message": "Fine tuning the induction analysis.\n\nRationale:\nBased on some self-imposed \"blind\" testing, improved\nthe induction variable analysis for typical cases\nthat provide a bit more elaborate HIR.\n\nTest: test-art-host\nChange-Id: I6e6bbf99928c29973178fa48f3942b14bf069944\n"
    },
    {
      "commit": "df7822ecf033cecf48d950f3ae34f7043c8df738",
      "tree": "f392a69377e1e281bcd85d811b656c6d14280ab4",
      "parents": [
        "6746874b84a44ab8dff18457eec546a1ebb22e93"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Tue Dec 06 10:05:30 2016 -0800"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Fri Dec 09 08:42:18 2016 -0800"
      },
      "message": "Added polynomial induction variables analysis. With tests.\n\nRationale:\nInformation on polynomial sequences is nice to further enhance\nBCE and last-value assignment. In this case, this CL enables more\nloop optimizations for benchpress\u0027 Sum (80 x speedup). Also\nchanged rem-based geometric induction to wrap-around induction.\n\nTest: test-art-host\n\nChange-Id: Ie4d2659edefb814edda2c971c1f70ba400c31111\n"
    },
    {
      "commit": "c071a01a26013ab6e3dbfc4131efa95a65aeb4ed",
      "tree": "bbe75527b8ee94483e4d797c6b2372adaabd81cf",
      "parents": [
        "5eb1e1e7341f4e7febf77c04f8649a9566b31c03"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Thu Dec 01 10:22:31 2016 -0800"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Mon Dec 05 16:16:42 2016 -0800"
      },
      "message": "Added geometric induction variables analysis.\n\nRationale:\nInformation on geometric and polynomial (coming soon) sequences\nare nice to have to further enhance BCE and last-value assignment.\n\nTest: test-art-host\nChange-Id: Ib5e2998c3eb1009def6fd00b82935da7c3ba7c6e\n"
    },
    {
      "commit": "40fbf748cf56096e35afa48b2a2fd5998491654a",
      "tree": "b648e65792efeedcfe626c34c58f087c6088c0f7",
      "parents": [
        "e7b46e22c7f4f6f503501b3b2ad99113289d142b"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Mon Oct 31 11:02:50 2016 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Nov 02 11:43:18 2016 -0700"
      },
      "message": "Improved range analysis (and thus BCE) around min/max/abs intrinsics.\n\nRationale:\nInspection of some typical bit set utilities revealed that we\nwere missing obvious cases where two or more array lengths\nwere combined using a min.\n\nChange-Id: I3e6463f221c793aaa1d592d4caabef0511754ae9\nTest: test-art-host-run-test-620-checker-bce-intrinsics\n"
    },
    {
      "commit": "639cc8c7bbb7d8c341173bcf24604ccb4328acb8",
      "tree": "ae8181b94ddd145edaff141a2b351478161f1559",
      "parents": [
        "3941c882ea7b54772dec36a9a1b33e0b8a7474f7"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Tue Oct 18 13:03:31 2016 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Thu Oct 20 14:14:17 2016 -0700"
      },
      "message": "Improve recognition of select-based period induction.\n\nRationale:\nSimilar to the previous CL, this helps to eliminate more dead induction.\nNow, CaffeineLogic, when compiled with dx (rather than jack) improves\nby a 1.5 speedup (9000us -\u003e 6000us).\n\nNote:\nWe need to run the simplifier before induction analysis to trigger\nthe select simplification first. Although a bit of a compile-time hit,\nit seems a good idea to run a simplifier here again anyway.\n\nTest: test-art-host\nChange-Id: I93b91ca40a4d64385c64393028e8d213f0c904a8\n"
    },
    {
      "commit": "9abf894ad0e5a6a1594ee1fa3924965e25e5f86f",
      "tree": "5080bd832d4f2234897404195b5d9865f950f47c",
      "parents": [
        "6e5fa09510c7280168e040382d27dd8b55760d9a"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Fri Oct 14 09:49:42 2016 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Tue Oct 18 09:02:47 2016 -0700"
      },
      "message": "Enable last value generation of periodic sequence.\n\nRationale:\nThis helps to eliminate more dead induction. For example,\nCaffeineLogic when compiled with latest Jack improves with\na 1.3 speedup (2900us -\u003e 2200us) due to eliminating first\nloop (second loop can be removed also, but for a later\ncase). The currently benchmarks.dex has a different construct\nfor the periodics, however, still to be recognized.\n\nTest: test-art-host\nChange-Id: Ia81649a207a2b1f03ead0855436862ed4e4f45e0\n"
    },
    {
      "commit": "7dc96932491dde6b5b58998254d5837dbcbbde03",
      "tree": "b2b0c7a133738823eb61871e5482aa657f2f8a2a",
      "parents": [
        "f0ab2ec6008bbd495e59bb9bf81ac399d864f38b"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Oct 12 10:01:05 2016 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Oct 12 14:10:11 2016 -0700"
      },
      "message": "Recognize XOR-based periodic induction.\n\nRationale:\nThis is a commonly used construct (e.g. x \u003d !x for booleans\nand x ^\u003d 1 for integers). This CL prepares some upcoming\noptimizations that exploit such inductions.\n\nChange-Id: I46edffb9de1075a836995daf5c2dfff7891f3034\nTest: 530-checker-loops2 and induction_var_analysis_test\n"
    },
    {
      "commit": "009cace65f649000e46a33db26d14b2af8b5aee4",
      "tree": "47d32693b615e35b3cb60ddd54a47e18cccf8200",
      "parents": [
        "be3a3ee02f148345ba6e1a0361532a3f7e8c0002"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Fri Sep 16 10:15:19 2016 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Fri Sep 16 14:33:00 2016 -0700"
      },
      "message": "Minor induction variable analysis changes.\n\nRationale:\nFew extra comments, added replacement of loop control functionality.\nAll in preparation of bigger loop optimizations CL.\n\nChange-Id: Ie2243c3c8b57b2901654b05fb200e9d80a0cb641\nTest: test-art\n"
    },
    {
      "commit": "16d3a65a25f03a7e447cafc7ab8fbdb52807cae6",
      "tree": "cb22e0f7109b75dee3df0eb57939eb2a5004d854",
      "parents": [
        "32772cbdbcb35f5475b01f31314a3c7289bdb589"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Fri Sep 09 10:33:50 2016 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Thu Sep 15 08:59:36 2016 -0700"
      },
      "message": "Added ability to generate last-value of linear induction.\nAlso added utility to update fetches in induction nodes.\n\nRationale:\nThis is a first step towards the larger CL that introduces\na new loop optimization framework in the optimizing compiler\n(see https://android-review.googlesource.com/#/c/271392/3).\n\nChange-Id: Ibecd674c8146d9665340e68718c498555646129a\nTests: induction_var_range_test\n"
    },
    {
      "commit": "52be7e759acecc3841dca0ac1b703034d8cad60d",
      "tree": "c6ed5e85c33d8938dbbe2b7ca5ecf7e00292884c",
      "parents": [
        "4dd9d82d6c1260b3da908a0d1a22b7d209b87496"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Thu Jun 23 11:20:41 2016 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Jun 29 09:34:28 2016 -0700"
      },
      "message": "Improvements in induction range analysis.\n\nRationale:\nUses range analysis while determining whether trip-counts\nare \"safe\", which improves analysis of triangular loops.\nAlso implements more effective triangular loop analysis\nby evaluating induction information only once and using\na top level hint (instead of the \"iterative refinement\"\nthat was used earlier). Also fixes analysis of triangular\ntrip counts that may wrap-around. All with tests.\n\nTest: see induction_var_range_test/530-checker-loops*\n\nBUG\u003d27151190\n\nChange-Id: I1877c8ce0c9a52005900eb9dfdbb1918df100278\n"
    },
    {
      "commit": "0d345cf8db01f40db250f80de5104e0df24234fa",
      "tree": "ebeac7adff399e908e5dcd3c855505ed02fe72b8",
      "parents": [
        "4485c6964ad414d5c6d0535622cfad1c0a6b640f"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Mar 16 10:49:38 2016 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Mon Mar 21 12:53:33 2016 -0700"
      },
      "message": "Generalize induction and range analysis across type conversions.\n\nRationale:\nThis changelist implements allowing narrowing conversions within\ninductions and loop control. More induction and loops recognized,\nmore bounds eliminated. We all win. The basic idea is pretty simple\n(record type with detected induction) but one has to get all the\ndetails right, as illustrated by the many new unit tests.\n\nBUG\u003d27151098\n\nChange-Id: I254020bfa5fa623799b31bbbb5ccc97d4d5a0100\n"
    },
    {
      "commit": "358af839c60db9e178f0b0bb9d430711c071b82a",
      "tree": "bfeb6c0fe553161a6d3df57f2e35401ea36e94ff",
      "parents": [
        "4e4e64511e530db37b33f450016afe49db3c4b20"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Feb 24 14:17:53 2016 -0800"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Thu Feb 25 09:26:57 2016 -0800"
      },
      "message": "Recognize for (int i \u003d 0; i !\u003d x.length; i++) loops\n\nRationale:\nIdiom occurs in real-life and is very straightforwardly\nrecognized by existing induction/range machinery.\n\nChange-Id: I965a16e9de72f3523ea5023d30ed1c4e47ed5010\n"
    },
    {
      "commit": "97412c92afb3f6630c4f0eafe6d6161862bfb4c1",
      "tree": "124411d0160da1a7f0869c29029ef4dbfac6da97",
      "parents": [
        "b4982aab07ae4cdaba13b4cb99306459d92e52d5"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Fri Feb 19 20:14:38 2016 -0800"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Feb 24 11:08:32 2016 -0800"
      },
      "message": "Use range analysis for better trip count analysis\n\nRationale:\nMarking more loops as always-taken avoids generating\nunnecessary new top tests while marking more loops\nare non-infinite enables more optimizations. This\nCL helps with these improvements. Also, some more\ncode is shared between induction and range analysis\nand a bug with refining ranges has been fixed.\n\nBug: 27151190\n\nChange-Id: Iecc0d7f32ae4779ee5424cda9dcc20816220935e\n"
    },
    {
      "commit": "1fc3afb76dbed78d255db276381df6036db2ee98",
      "tree": "61ee6429e39fad5ae5d5d6d620966ce90fead39c",
      "parents": [
        "97f4bc04b61d5cf78b0820dbf18e999b20d7a108"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Tue Feb 02 13:26:16 2016 -0800"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Feb 03 09:22:27 2016 -0800"
      },
      "message": "Minor improvement on static BCE analysis.\n\nRationale:\nAvoid testing initial range if nothing is known.\n\nChange-Id: I22646a5fd6e4481245d1a2f57891d2805550489f\n"
    },
    {
      "commit": "7d57d7f2f0328241ff07c43a93edadbc1a6697c6",
      "tree": "4a53eb14db06af2e8fc84bbebd48a462e868bfd4",
      "parents": [
        "7d5a577c1f256a7703361afb071dcd4ecc7d275f"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Dec 09 14:39:48 2015 -0800"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Tue Dec 15 11:56:35 2015 -0800"
      },
      "message": "Various induction/range analysis improvements.\n\nRationale: this change list improves analysis of triangular loops\n           both by changing loop order for induction analysis\n           (enabling range analysis in inner loops) and by\n           some symbolic improvements during range analysis;\n           also, a mul/div bug has been fixed (with pass/fail\n           unit tests); lastly this change list prepares some\n           follow up optimizations.\n\nChange-Id: I84a03e848405009541c3fa8e3d3c2f430e100087\n"
    },
    {
      "commit": "b738d4f477a9b6f4c4f69153f9077f1559d2bca1",
      "tree": "53c1100c03c33039e910e0e97ca4ef73d69bb15f",
      "parents": [
        "68cebffee2f44f0345bd7be7e4cbad0f10e64082"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Thu Dec 03 11:23:35 2015 -0800"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Thu Dec 03 16:42:17 2015 -0800"
      },
      "message": "Step-wise improvement of range analysis with outer loop induction.\n\nRationale: Using a step-wise approach (rather than expanding all ranges\n           at once) increases the opportunities for statically removing\n           bound checks, as demonstrated by the new checker tests.\n\nChange-Id: Icbfd9406523a069e1fb7508546ea94f896e5a255\n"
    },
    {
      "commit": "4a34277c55279ba57ab361f7580db846a201d9b1",
      "tree": "31d775a871a351357ecd5d6c76a3a827e20e6d07",
      "parents": [
        "f4c539395244ccfc14aebaf53fdc1122287f65a6"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Mon Nov 30 10:17:46 2015 -0800"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Tue Dec 01 13:26:16 2015 -0800"
      },
      "message": "Dynamic BCE (based on induction range analysis)\n\nRationale:\nA rewritten dynamic BCE that uses induction variable analysis\nto generate the run-time tests before a loop in order to\neliminate bounds-checks from its body. This CL removes now\nobsoleted induction related code inside the BCE module.\nAlso, the dynamic test generation is placed more strategically,\nsince we missed a few cases where static analysis does better.\n\nMost significant performance improvements (filtering noise) is about:\n\nLinpack +20%\nLU    \u003e +10%\n\nChange-Id: I03d7631857154b6a131b132f26a2dc568af1b3a1\n"
    },
    {
      "commit": "d59c70627cc42878cc30b46bd29ff497b4483b22",
      "tree": "b0b7013ea78b10f23fae75ef145e53a696ff797c",
      "parents": [
        "0b5849be045c5683d4a6b6b6c306abadba5f0fcc"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Sat Nov 21 05:21:52 2015 +0000"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Sat Nov 21 05:21:52 2015 +0000"
      },
      "message": "Revert \"Dynamic BCE (based on induction range analysis)\"\n\nThis reverts commit 0b5849be045c5683d4a6b6b6c306abadba5f0fcc.\n\n\nChange-Id: Id33f5da42bbdfb1aff7e2281417c8a7aa492df05\nRationale: so close :-( but bullhead-userdebug (linux) build in git_mnc-dr-dev-plus-aosp reported a breakage with a type inconsistency (long vs int in probably the codegen of dynamic bce); no time to investigate and fix this fully before my trip, so rolling back for now\n"
    },
    {
      "commit": "0b5849be045c5683d4a6b6b6c306abadba5f0fcc",
      "tree": "87245845c05ddef6fb53ac4c3774607a89fa5db4",
      "parents": [
        "3944f7175dcf60316ba58a42698ccf23c65ac57c"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Mon Oct 19 14:59:26 2015 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Fri Nov 20 17:00:04 2015 -0800"
      },
      "message": "Dynamic BCE (based on induction range analysis)\n\nRationale: A rewritten dynamic BCE that uses induction variable analysis\n           to generate the run-time tests before a loop in order to\n           eliminate bounds-checks from its body. This CL removes now\n           obsoleted induction related code inside the BCE module.\n           Also, the dynamic test generation is placed more strategically,\n           since we missed a few cases where static analysis does better.\n\nMost significant performance improvements (after filtering noise) is about:\n  Linpack +20%\n  LU    \u003e +10%\n\nChange-Id: I4e7b8bab0288beff6f98a14856e3536103d32742\n"
    },
    {
      "commit": "389b3dbf5c5390056ff4dacac464219853dd3cda",
      "tree": "85013c1eaf6aaba9ab839f17b978eb689f30e998",
      "parents": [
        "c8894ab5021aecd0fa5eba94af47f732914af33b"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Oct 28 14:23:40 2015 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Nov 04 09:48:26 2015 -0800"
      },
      "message": "Finalized all components of range analysis needed for dynamic bce.\n\nRationale: added ability to generate taken-test, prompt back need\n           for finite-test; cleaned up the API now that bounds\n           check needs are all known.\n\nChange-Id: I3d09b249965d1a980c09381240de175ca4b2e455\n"
    },
    {
      "commit": "22f058726d35dd8f40b3763649e61740b3d22535",
      "tree": "987dd5df7627c2c3e8c798089083cb797627daed",
      "parents": [
        "3dca4c04170bb4c7b26609e55f3f4966d46b681a"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Tue Oct 27 15:56:28 2015 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Oct 28 13:57:14 2015 -0700"
      },
      "message": "Generate taken-test during trip-count analysis.\n\nRationale: for loops that may not be taken, this taken-test\n           can be used by clients of the induction variable\n           analysis to ensure trip-count evaluation is valid.\n\nChange-Id: Ia64749e2389b7224e69d6a49bb604b1964c11068\n"
    },
    {
      "commit": "aec3cce52009afe436a6db0280d6d5aee9b8d4d4",
      "tree": "9fa409750d89871863b10396a2278ebd48f0b5b9",
      "parents": [
        "f69f56e7d4a1e31dfce2a77558c9b7047f82092b"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Oct 14 17:44:55 2015 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Mon Oct 19 12:44:55 2015 -0700"
      },
      "message": "Added ability to generate induction range code.\n\nRationale: used by dynamic BCE (done in another CL).\n\nChange-Id: Ia6ce75da57b5298fba74622822ae0bae69c74188\n"
    },
    {
      "commit": "9401f5397128ddc8dc36de923dd5e6bd4e4b5be4",
      "tree": "4ff8052307da80baa89dfa80a446f48752c0e95c",
      "parents": [
        "931e26843bbb688eacfa67b40414c6b8f221a56a"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Mon Sep 28 16:25:56 2015 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Sep 30 09:58:53 2015 -0700"
      },
      "message": "Implemented trip-count safety information.\n\nAs shown in the induction analysis presentation, trip-counts need to\ndeal with potential taken/not-taken situations (so that trip-count\nis either valid in the full loop or just in the loop-body proper)\nand potential finite/infinite situations (the latter can still be\nanalyzed but may need to run-time test later to guard against the\ninfinite conditions). This CL provides that information.\n\nChange-Id: I0445d8e836b80a3614af217ce3e39d766e77b986\n"
    },
    {
      "commit": "cd26feb7e9f9eb36af9453f3cdb55cfc4e6e5ec4",
      "tree": "39768556d365d9830119f03aae81de1e7ee3efd3",
      "parents": [
        "a6cfe5a3acb3af782322fde734bfcf0c524d6f99"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Sep 23 17:50:50 2015 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Wed Sep 23 18:36:11 2015 -0700"
      },
      "message": "Minor cleanup in range analysis.\n\n(1) replaced min/max macro as previously required.\n(2) removed some redundant code by merging min/max into one.\n\nChange-Id: I610879a06d550346bfac7e6e12ec0299ba226a37\n"
    },
    {
      "commit": "b3365e0c4cda4f8f19284d2d418db158ab78d810",
      "tree": "d76a5ee48912d7fae2c5e612def63000c8284f8a",
      "parents": [
        "6a6916b378c413c2692782e901393607c47f0005"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Mon Sep 21 14:45:05 2015 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Tue Sep 22 16:22:32 2015 +0000"
      },
      "message": "Various improvements in range analysis.\n\nRationale:\nUsing min/max values for \"unknowns\" is a bit wasteful,\nsince it eliminates two useful values. Replaced this\nwith additional boolean to make cases more accurate.\nAdded few cases to handle examples found in real-life.\n\nChange-Id: I211f8d9a28b1ae79abdb55fb4569716f21d8043b\n"
    },
    {
      "commit": "22af3bee34d0ab1a4bd186c71ccab00366882259",
      "tree": "793f358d498142a2e60d7d5131c347b0fd668cbd",
      "parents": [
        "fe9a1b05ea5a21b6d9a2e9e5081f5e80ff8a1ba2"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Thu Sep 10 12:50:58 2015 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Tue Sep 15 17:03:13 2015 -0700"
      },
      "message": "Use induction variable range analysis in BCE (statically).\n\nRationale: Finally! After lots of very large CLs, now a small CL\n           that uses the new induction variable analysis in BCE\n           (statically, using this dynamically with de-opt is TBD).\n           Despite its relative small size, be aware though,\n           since the CL introduces a new phase to the compiler.\n\nChange-Id: If5555a173fd5d55d147c63138ef51fc296fa1414\n"
    },
    {
      "commit": "d14c59564870c910bdc823081f0ed1101f599231",
      "tree": "36a831f8c786823129bebbf03a4ca48ff6ed3c23",
      "parents": [
        "e6576390f957c82c2aede438834d028066757368"
      ],
      "author": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Tue Sep 08 15:25:15 2015 -0700"
      },
      "committer": {
        "name": "Aart Bik",
        "email": "ajcbik@google.com",
        "time": "Thu Sep 10 10:16:04 2015 -0700"
      },
      "message": "Induction variable range analysis.\n\nRationale: by computing an upper bound on the trip count of each\n           loop after induction var analysis has completed, a\n\t   simple range analysis yields lower and upper bounds on\n\t   all induced expressions in a loop; this analysis\n\t   plugs directly into BCE (follow-up CL).\n\nChange-Id: I46a3fe48721ca372547199b39a3498c47992597d\n"
    }
  ]
}