)]}' { "log": [ { "commit": "43c9cd7938a53265ea9899a604b409ce070cc2c5", "tree": "db084fcc115d6a68de25662bf1a5fae030bcf152", "parents": [ "443fc8962db219177c34cd047d2071b5ee775468" ], "author": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Wed Mar 10 15:09:19 2021 +0000" }, "committer": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Wed Mar 24 16:47:38 2021 +0000" }, "message": "Improve nterp -\u003e compiled code transitions.\n\nUse an unused bit in the access flags of an ArtMethod (0x00200000) to store\nthe information a method only takes ints or references and returns an\nint, a reference, or a long. This avoids the need to fetch the shorty in nterp\nwhen doing a call.\n\nTest: test.py\nTest: 821-many-args\nBug: 112676029\n\nChange-Id: Ie657ccf69c17c1097dc2a97f18e3093ef3be391b\n" }, { "commit": "40cd07c3e6717b28c0137e8d72a067a9f8663ec6", "tree": "45cced150c54d6798805882c10607f11b1be9b36", "parents": [ "4924ea9ad98832f0ec7db841defca82331ee1b13" ], "author": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Mon Mar 15 18:33:23 2021 +0000" }, "committer": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Tue Mar 23 15:31:45 2021 +0000" }, "message": "Add a fast path for nterp entrypoint to avoid fetching the shorty.\n\nWhen all parameters are references, we don\u0027t need to look at the shorty.\nUse the 0x00100000 flag in the modifiers which is free for non-native\nmethods.\n\nTest: test.py\nBug: 112676029\nChange-Id: Ied9a253f7f7230045dd13188a5b806fb1d6d019d\n" }, { "commit": "de0d0de0f4a5162a5861515c0ddbbb52e06147d0", "tree": "d5aa36d70696f79a601d2cdcd47d1c5f506bf222", "parents": [ "ec388c026a5709dea339ad43f8bca10b2d991d6e" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Mar 18 14:12:35 2021 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Fri Mar 19 09:11:41 2021 +0000" }, "message": "Remove kAccMiranda and kAccDefaultConflicting.\n\nUse a combination of other existing flags to mark miranda\nand default conflicting methods.\n\nAlso change the value of identical flags kAccCriticalNative\nand kAccPreCompiled to 0x00100000 to have the values of\nkAccFastNative and kAccCriticalNative in consecutive bits.\n\nThe method access flag 0x00200000 is now free for future\nuse. The method access flag 0x00100000 is also potentially\nusable for some methods the same way that the old\nkAccMiranda flag was used despite sharing the same value\nwith both kAccCriticalNative and kAccPreCompiled.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nBug: 112676029\nChange-Id: Ibfd20598f9b4af78e1e72f8331dc647d01dd8eb5\n" }, { "commit": "bc1748f825aed2a59a9bfbb087ab6e29e875d2d1", "tree": "d9599a24114f37850a895cd90fa9ec46d3dd56f3", "parents": [ "157fc969066aa41842c029350e5af780c517fcf6" ], "author": { "name": "David Srbecky", "email": "dsrbecky@google.com", "time": "Mon Feb 22 10:51:32 2021 +0000" }, "committer": { "name": "David Srbecky", "email": "dsrbecky@google.com", "time": "Tue Mar 09 14:26:34 2021 +0000" }, "message": "Refactor statistics printing in oatdump.\n\nSimplify the code so that it is easier to add/remove sections.\n(motivated by the non-trivial removal of DexCache arrays)\n\nTest: m dump-oat\nTest: m test-art-host-gtest-art_oatdump_tests\nChange-Id: I2fe4a3062125d8aa053c1a38eb24a624f95291b0\n" }, { "commit": "cedec9db0a9accfdcf5eb695879e0b2caf2c34cb", "tree": "99fd71a3ef398a471507ed815c4a175805ad51f3", "parents": [ "0ccc970b3c00b172e3cff8e10da13dd6323ccd52" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Feb 08 16:16:13 2021 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Tue Feb 23 14:34:43 2021 +0000" }, "message": "Do not create HandleScope for JNI transitions.\n\nWe previously crated a HandleScope in the JNI transition\nframe to hold references passed as jobject (jclass, etc.)\nto the native function and these references were actually\nspilled twice during the transition.\n\nWe now construct the jobject as a pointer to the reference\nspilled in the reserved out vreg area in the caller\u0027s frame.\nAnd the jclass for static methods is just a pointer to the\nmethod\u0027s declaring class. This reduces the amount of work\nrequired in the JNI transition, both on entry (in compiled\nstubs) and exit (in JniMethodEnd*).\n\nSome additional work is required when GC visits references\nof a native method as we need to walk over the method\u0027s\nshorty which was unnecessary for a HandleScope.\n\nAlso fix Thread::InitStackHwm() to calculate correct stack\nsize needed by the new Thread::IsJniTransitionReference().\n\nThe results for StringToBytesBenchmark on blueline little\ncores running at fixed frequency 1420800 are approximately\narm64 (medians from 3 runs) before after\ntimeGetBytesAscii EMPTY 447.33 436.86\ntimeGetBytesIso88591 EMPTY 440.52 431.13\ntimeGetBytesUtf8 EMPTY 432.31 409.82\narm (medians from 3 runs) before after\ntimeGetBytesAscii EMPTY 500.53 490.87\ntimeGetBytesIso88591 EMPTY 496.45 495.30\ntimeGetBytesUtf8 EMPTY 488.84 472.68\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host\nTest: testrunner.py --host --gcstress\nTest: testrunner.py --host --jit-on-first-use\nTest: testrunner.py --host --jit-on-first-use --gcstress\nTest: run-gtests.sh\nTest: testrunner.py --target --optimizing\nTest: boots.\nBug: 172332525\nChange-Id: I658f9d87071587b3e89f31c65feca976a11e9cc2\n" }, { "commit": "745da80dbf169d83e2acb847b0c9c07985d51bdf", "tree": "a01990e419ee49fe255a6df108027a89b568975a", "parents": [ "16f7f8e384f0ae13c53525ff12affbc00605b0ec" ], "author": { "name": "Nikita Iashchenko", "email": "nikitai@google.com", "time": "Wed Jan 20 21:52:54 2021 +0000" }, "committer": { "name": "Nikita Iashchenko", "email": "nikitai@google.com", "time": "Wed Feb 10 12:33:48 2021 +0000" }, "message": "Add Math.multiplyHigh intrinsic\n\nTest: ./art/test/testrunner/testrunner.py --target --optimizing --64 -t 082-inline-execute\nTest: ./art/test/testrunner/testrunner.py --host --optimizing --64 -t 082-inline-execute\nChange-Id: I4b6cafa8b9e513eca7c5c139440024d87a7ef758\n" }, { "commit": "ac27ac01490f53f9e2413dc9b66fbb2880904c96", "tree": "271018e1ef33667bee8d57c40ffa3f4d9f8cf930", "parents": [ "26bf47a60064fcc42e1b5e7b4b41deb8312d7330" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Feb 01 09:31:02 2021 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Fri Feb 05 10:48:17 2021 +0000" }, "message": "Implement Reference.refersTo() intrinsic.\n\nTest: Added tests to 122-npe and 160-read-barrier-stress\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nTest: run-gtests.sh\nTest: testrunner.py --target --optimizing\nBug: 172573708\nChange-Id: I8342510565289058df218d3249ffac1eb993ca4f\n" }, { "commit": "b1cf83748ce8165fe7e41ce6a57eb5529946a971", "tree": "56fe4acb1a5310cf015467a79427f4de67a53b68", "parents": [ "0f0caf869a9026e723caf79d8f50d8da25bc4185" ], "author": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Tue Feb 02 13:32:20 2021 +0000" }, "committer": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Tue Feb 02 18:08:12 2021 +0000" }, "message": "Fix image bug around clinit and nterp.\n\nnterp trampoline doesn\u0027t do clinit checks, so we should install the\nresolution stub if a method needs a clinit check before being entered.\n\nTest: 818-clinit-nterp\nBug: 178992601\nChange-Id: I9d6b023a6c6620adfcd2a6ed5cd1e042a7cb449b\n" }, { "commit": "c7993d55b2912532c1b0917e6b0ca43097bed21d", "tree": "1145ce1c0ffbbd729d0a6c1b53c608bf7adb3d6a", "parents": [ "22538e4672348c56d4d44d6113bedd6b47d0abd0" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Jan 27 15:20:56 2021 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Jan 28 14:14:16 2021 +0000" }, "message": "Rewrite ClassLinker::LinkFields().\n\nRewrite field offset assignment to avoid heap allocations in\nmost cases. We do one allocation if there are many fields.\n\nRewrite gap filling to prefer small gaps over big ones. This\ncreates more natural field ordering. For example,\n class A { byte unalign; }\n class B extends A {\n long l;\n\tbyte a, b, c, d, e, f;\n }\nwould have previously had offsets\n 12, 10, 14, 9, 11, 13, 15\nfor fields a, b, c, d, e, f, respectively. Now these are\n 9, 10, 11, 12, 13, 14, 15\nin line with their lexicographical order.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nTest: boots.\nBug: 175869411\nChange-Id: I558635ac59c959dd85e8a3b015c26a6d90033853\n" }, { "commit": "c39af9409ed4c8dd711be606f758a9b378cd0ee3", "tree": "43f8cc23a07a41fb8fc03be77a15fa203e27194d", "parents": [ "3a73ffb70151dbc99fa41f300a237f8c29783e0e" ], "author": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Mon Jan 25 08:43:57 2021 +0000" }, "committer": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Wed Jan 27 08:29:40 2021 +0000" }, "message": "Add a nterp trampoline in the oat file, replaced at runtime.\n\nTo avoid checking whether an ArtMethod can run nterp when loading an\nimage, record that information directly in the image, and patch the\nentrypoint so nterp can do direct pointer checks when calling an\nArtMethod.\n\nTest: test.py\nBug: 177444058\nChange-Id: Ia87367c37848e1efe33336b3677490c07d7e1767\n" }, { "commit": "d5a8695394f8f6e8f9a66752441623dc31b97660", "tree": "591ac4892304365282a57262c91e1945ac94c6a9", "parents": [ "2fb37411e42455a00fdad90f0e612c3214e5bd1a" ], "author": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Tue Jan 19 10:35:54 2021 +0000" }, "committer": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Thu Jan 21 14:07:57 2021 +0000" }, "message": "Reland \"Improve invokeinterface for nterp.\"\n\nThis reverts commit f1d06474baa2f7c00761db39099b89ddab71bbe4.\n\nBug: 177554973\nBug: 112676029\n\nTest: test.py\nTest: 815-invokeinterface-default\nTest: enable text-to-speech on device, no crash\n\nReason for revert: Fixed issue with recursive default methods\n\nChange-Id: I2fb9336adb6c4fc920f39aa19bfe7f0a92ce059a\n" }, { "commit": "f1d06474baa2f7c00761db39099b89ddab71bbe4", "tree": "9924ce87e34d1058a0d12e963e3a94fbd3d1539d", "parents": [ "4c176b9de6c430422395017910633bcb001b2e84" ], "author": { "name": "Alex Light", "email": "allight@google.com", "time": "Thu Jan 14 23:40:53 2021 +0000" }, "committer": { "name": "Alex Light", "email": "allight@google.com", "time": "Fri Jan 15 00:36:13 2021 +0000" }, "message": "Revert \"Improve invokeinterface for nterp.\"\n\nThis reverts commit c0d392be9172c4c60b3e81b0123489ec3036c401.\n\nBug: 177554973\nReason for revert: Fix crash when enabling \u0027Select to Speak\u0027\n\nChange-Id: If2865cc1dfc11e6bd27ed533da10f53521e3e613\n" }, { "commit": "c0d392be9172c4c60b3e81b0123489ec3036c401", "tree": "13b77bab3a3a8a1acf26dc02d2ee09bcd3a51c36", "parents": [ "c38d94449a0ccfefd396d2e01b571f3e804d9e48" ], "author": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Tue Dec 29 20:41:43 2020 +0000" }, "committer": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Tue Jan 12 09:39:41 2021 +0000" }, "message": "Improve invokeinterface for nterp.\n\n- Remove bitwise negation of imt_index: the class linker already always\n initializes the imt_index.\n- Special case the imt index of default methods: to simplify invocation\n in nterp, mask the method_index to create the imt index.\n- Add arm64, arm, x64 support in nterp.\n\nTest: test.py\nBug: 112676029\nChange-Id: I815a4a4ec5c219921ab4ed1a20b02586aab19a46\n" }, { "commit": "86d6cd53385eae365f16f45e1a5947cc6595eb63", "tree": "ce94ee1f5d574b2891163f9b9695c97142efd00d", "parents": [ "af17e5f33380ad39d83c31d97a29f3c836150fa1" ], "author": { "name": "David Srbecky", "email": "dsrbecky@google.com", "time": "Wed Dec 02 18:13:10 2020 +0000" }, "committer": { "name": "David Srbecky", "email": "dsrbecky@google.com", "time": "Fri Dec 04 14:31:35 2020 +0000" }, "message": "Remove DexCache arrays from image.\n\nRemove the hashtable storage from the image and allocate it at\nruntime instead (but keep the DexCache object in the image).\n\nFor compiled code, we have largely moved to using .bss, so the\nDexCache just costs us unnecessary extra space and dirty pages.\n\nFor interpreted code, the hashtables are too small and will be\noverridden many times over at run-time regardless.\n\nThe next step will be to make DexCache variable-size so it can\nadapt to both of the extremes (taking minimal amount of memory\nfor compiled code and avoiding cache evictions in interpreter).\n\nTest: test.py --host\nChange-Id: I9f89e8f19829b812cf85dea1a964259ed8b87f4d\n" }, { "commit": "a3bd4ec050e10fbfb6df0fe024bc6f3851bcc326", "tree": "8a373f87b05b3580f564bd8561bf0dd4ebb9161e", "parents": [ "b0f405e1482df602fc57ee2ed4434456c4dc2555" ], "author": { "name": "Artem Serov", "email": "artem.serov@linaro.org", "time": "Thu Aug 27 16:26:17 2020 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Fri Sep 25 09:16:33 2020 +0000" }, "message": "Introduce Long.divideUnsigned() intrinsic on ARM64.\n\nExtends the 082-inline-execute test and reorders\ndivideUnsigned intrinsic code before VarHandle code.\n\nTest: 082-inline-execute.\nTest: testrunner.py --target --optimizing --64\nBug: 156736938\nChange-Id: I8acf3db184da30640aff36cf02570ae0c6d2b88c\n" }, { "commit": "dc787f488ce151478af1149c04a10b21b2b9fc6e", "tree": "38d402813c9c3f541630255e9f9890a9c6842c45", "parents": [ "9a09e7cf11b2653e6da2b179ac03bc6bb71c81e3" ], "author": { "name": "Andra Danciu", "email": "andradanciu@google.com", "time": "Tue Jul 07 14:28:56 2020 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Jul 09 09:44:58 2020 +0000" }, "message": "Make Integer.divideUnsigned intrinsic for x86.\n\nBug: 156736938\nTest: Added a test to 082-inline-execute.\nTest: art/test.py --host --32 -r -t 082-inline-execute\nChange-Id: Id516126fce10f2fa52b95c1b2b107ca7bf45e347\n" }, { "commit": "d0036ac18efcd7774775d521ae11178933041b95", "tree": "175db5c419006e3c146d92ef278100233dd2a827", "parents": [ "69944683c33a316c9ed325ed686d9f81ad3665dc" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Nov 21 11:47:12 2019 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Dec 11 10:03:05 2019 +0000" }, "message": "Single-image boot image extensions.\n\nAllow boot image extensions to be compiled into a single\nart/oat/vdex triplet for any number of input dex files.\nSupport argument --image-fd so that this can be done with\nfile descriptors. This is intended for creating a boot\nimage extension in memory when starting zygote.\n\nRemove the possibility to specify multiple oat or image\nfiles. It was unused and untested.\n\nTest: New test in dex2oat_image_test\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nChange-Id: I78e5410642a5df14c0891d87f27eefcb02156764\n" }, { "commit": "457e9fa3833ef11530056d010f247ad087fd2184", "tree": "54b8a9dcf44646c3e43a9085d581660c5d9a0132", "parents": [ "17a39babb7f42cbe108d6fab2760cbdc68b821a2" ], "author": { "name": "Usama Arif", "email": "usama.arif@linaro.org", "time": "Mon Nov 11 15:29:59 2019 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Nov 28 09:51:05 2019 +0000" }, "message": "ARM64: FP16 greater/less/greaterEquals/lessEquals intrinsics for ARMv8\n\nThis CL implements intrinsics for greater, greaterEquals, less,\nlessEquals methods with ARMv8.2 FP16 instructions. This requires the\nARMv8.2 AArch64 asimd half precision extension.\n\nThe time required in milliseconds to execute the below code for the four\nintrinsics on Pixel3 is (The code below is for FP16.less but is similar\nfor the rest of the intrinsics):\n\n- Java implementation libcore.util.FP16.less():\n - big cluster only: 19876\n - little cluster only: 47525\n- arm64 Intrinisic implementationi for less:\n - big cluster only: 14526 (~27% faster)\n - little cluster only: 45815 (~4% faster)\n\n- Java implementation libcore.util.FP16.lessEquals():\n - big cluster only: 19856\n - little cluster only: 47419\n- arm64 Intrinisic implementation for lessEquals:\n - big cluster only: 14469 (~27% faster)\n - little cluster only: 45762 (~4% faster)\n\n- Java implementation libcore.util.FP16.greater():\n - big cluster only: 19854\n - little cluster only: 47623\n- arm64 Intrinisic implementation for greater:\n - big cluster only: 14519 (~27% faster)\n - little cluster only: 45722 (~4% faster)\n\n- Java implementation libcore.util.FP16.greaterEquals():\n - big cluster only: 19865\n - little cluster only: 47216\n- arm64 Intrinisic implementation for greaterEquals:\n - big cluster only: 14485 (~27% faster)\n - little cluster only: 45729 (~4% faster)\n\npublic static boolean benchmarkComparison(){\n boolean ret \u003d false;\n long before \u003d 0;\n long after \u003d 0;\n before \u003d System.currentTimeMillis();\n for(long i \u003d 0; i \u003c 1e9; i++){\n // FP16.toHalf(12.3) \u003d 0x4a26, FP16.toHalf(12.4) \u003d 0x4a33\n // FP16.toHalf(-12.3) \u003d 0xca26, FP16.toHalf(-12.4) \u003d 0xca33\n ret |\u003d FP16.less((short) 0x4a26,(short) 0x4a33);\n ret |\u003d FP16.less((short) 0x4a33,(short) 0x4a26);\n ret |\u003d FP16.less((short) 0xca26,(short) 0xca33);\n ret |\u003d FP16.less((short) 0xca33,(short) 0xca26);\n }\n after \u003d System.currentTimeMillis();\n System.out.println(\"Time of FP16.less (ms): \" + (after - before));\n System.out.println(ret);\n return ret;\n}\n\nTest: 580-fp16\nTest: art/test/testrunner/run_build_test_target.py -j80 art-test-javac\nChange-Id: Id1a2c3e7328c82c798fcaf1fa74f5908a822cd0b\n" }, { "commit": "681692b6291008caaddf4971eab7ea9f9b25d9ca", "tree": "cad26b5b6016cc27baa480e5f4b83faac3317ef0", "parents": [ "665aac46784684dfb85fe999f6a566ed0cf173ef" ], "author": { "name": "Usama Arif", "email": "usama.arif@linaro.org", "time": "Wed Oct 30 16:23:26 2019 +0000" }, "committer": { "name": "Hans Boehm", "email": "hboehm@google.com", "time": "Tue Nov 12 20:58:17 2019 +0000" }, "message": "ARM64: FP16.rint() intrinsic for ARMv8\n\nThis CL implements an intrinsic for rint() method with\nARMv8.2 FP16 instructions.\n\nThis intrinsic implementation achieves bit-level compatibility with the\noriginal Java implementation android.util.Half.rint().\n\nThe time required in milliseconds to execute the below code on Pixel3:\n- Java implementation android.util.Half.rint():\n - big cluster only: 19828\n - little cluster only: 61457\n- arm64 Intrinisic implementation:\n - big cluster only: 14186 (~28% faster)\n - little cluster only: 54405 (~11% faster)\n\nAnalysis of this function with simpleperf showed that approximately only\n60-65% of the time is spent in libcore.util.FP16.rint. So the percentage\nimprovement using intrinsics is likely to be more than the numbers stated\nabove.\n\nAnother reason that the performance improvement with intrinsic is lower\nthan expected is because the java implementation for values between -1 and\n1 (abs \u003c 0x3c00) only requires a few instructions and should almost give\na similar performance to the intrinsic in this case. In the benchmark function\nbelow, 46.8% of the values tested are between -1 and 1.\n\npublic static short benchmarkrint(){\n short ret \u003d 0;\n long before \u003d 0;\n long after \u003d 0;\n before \u003d System.currentTimeMillis();\n for(int i \u003d 0; i \u003c 50000; i++){\n for (short h \u003d Short.MIN_VALUE; h \u003c Short.MAX_VALUE; h++) {\n ret +\u003d FP16.rint(h);\n }\n }\n after \u003d System.currentTimeMillis();\n System.out.println(\"Time of FP16.rint (ms): \" + (after - before));\n System.out.println(ret);\n return ret;\n}\n\nTest: 580-fp16\nTest: art/test/testrunner/run_build_test_target.py -j80 art-test-javac\nChange-Id: I075c3e85a36fd9bce14deee437c5b961bd667b5d\n" }, { "commit": "665aac46784684dfb85fe999f6a566ed0cf173ef", "tree": "343d6956068b1c21e0fc4af018e1322d7685411b", "parents": [ "b9f02c2f8624bbf0746939e3b2735a1537a567b6" ], "author": { "name": "Usama Arif", "email": "usama.arif@linaro.org", "time": "Tue Oct 29 11:13:18 2019 +0000" }, "committer": { "name": "Hans Boehm", "email": "hboehm@google.com", "time": "Tue Nov 12 20:58:17 2019 +0000" }, "message": "ARM64: FP16.ceil() intrinsic for ARMv8\n\nThis CL implements an intrinsic for ceil() method with\nARMv8.2 FP16 instructions.\n\nThis intrinsic implementation achieves bit-level compatibility with the\noriginal Java implementation android.util.Half.ceil().\n\nThe time required in milliseconds to execute the below code on Pixel3:\n- Java implementation android.util.Half.ceil():\n - big cluster only: 19447\n - little cluster only: 62638\n- arm64 Intrinisic implementation:\n - big cluster only: 14260 (~27% faster)\n - little cluster only: 54387 (~13% faster)\n\nAnalysis of this function with simpleperf showed that approximately only\n60-65% of the time is spent in libcore.util.FP16.ceil. So the percentage\nimprovement using intrinsics is likely to be more than the numbers stated\nabove.\n\nAnother reason that the performance improvement with intrinsic is lower\nthan expected is because the java implementation for values between -1 and\n1 (abs \u003c 0x3c00) only requires a few instructions and should almost give\na similar performance to the intrinsic in this case. In the benchmark function\nbelow, 46.8% of the values tested are between -1 and 1.\n\npublic static short benchmarkCeil(){\n short ret \u003d 0;\n long before \u003d 0;\n long after \u003d 0;\n before \u003d System.currentTimeMillis();\n for(int i \u003d 0; i \u003c 50000; i++){\n for (short h \u003d Short.MIN_VALUE; h \u003c Short.MAX_VALUE; h++) {\n ret +\u003d FP16.ceil(h);\n }\n }\n after \u003d System.currentTimeMillis();\n System.out.println(\"Time of FP16.ceil (ms): \" + (after - before));\n System.out.println(ret);\n return ret;\n}\n\nTest: 580-fp16\nTest: art/test/testrunner/run_build_test_target.py -j80 art-test-javac\nChange-Id: I5474c1d0d7c08ec77a6f82c4fb67f555253bfa67\n" }, { "commit": "b9f02c2f8624bbf0746939e3b2735a1537a567b6", "tree": "ac6ad1f4125bd459a3b424cb5ff8b8029a3d5c7e", "parents": [ "f1b18facd1edd6c8652c42085c5432c878507c8e" ], "author": { "name": "Usama Arif", "email": "usama.arif@linaro.org", "time": "Fri Oct 25 17:37:33 2019 +0100" }, "committer": { "name": "Hans Boehm", "email": "hboehm@google.com", "time": "Tue Nov 12 20:58:17 2019 +0000" }, "message": "ARM64: FP16.floor() intrinsic for ARMv8\n\nThis CL implements an intrinsic for floor() method with ARMv8.2 FP16\ninstructions. This intrinsic calls a template GenerateFP16Round function\nwhich will be used to implement other intrinisics such as ceil and\nrint.\n\nThis intrinsic implementation achieves bit-level compatibility with the\noriginal Java implementation android.util.Half.floor().\n\nThe time required in milliseconds to execute the below code on Pixel3:\n- Java implementation android.util.Half.floor():\n - big cluster only: 18623\n - little cluster only: 60424\n- arm64 Intrinisic implementation:\n - big cluster only: 14213 (~24% faster)\n - little cluster only: 54398 (~10% faster)\n\nAnalysis of this function with simpleperf showed that approximately only\n60-65% of the time is spent in libcore.util.FP16.floor. So the percentage\nimprovement using intrinsics is likely to be more than the numbers stated\nabove.\n\nAnother reason that the performance improvement with intrinsic is lower\nthan expected is because the java implementation for values between -1 and\n1 (abs \u003c 0x3c00) only requires a few instructions and should almost give\na similar performance to the intrinsic in this case. In the benchmark function\nbelow, 46.8% of the values tested are between -1 and 1.\n\npublic static short benchmarkFloor(){\n short ret \u003d 0;\n long before \u003d 0;\n long after \u003d 0;\n before \u003d System.currentTimeMillis();\n for(int i \u003d 0; i \u003c 50000; i++){\n for (short h \u003d Short.MIN_VALUE; h \u003c Short.MAX_VALUE; h++) {\n ret +\u003d FP16.floor(h);\n }\n }\n after \u003d System.currentTimeMillis();\n System.out.println(\"Time of FP16.floor (ms): \" + (after - before));\n System.out.println(ret);\n return ret;\n}\n\nTest: 580-fp16\nTest: art/test/testrunner/run_build_test_target.py -j80 art-test-javac\n\nChange-Id: Iad1dd032d456af54932f13c5cf27228f8652a0b5\n" }, { "commit": "92eec3a90282edf30ca1535d58a4b52294ba7e76", "tree": "27b05f588a5286f5b08b33f7a57b20544841f5ef", "parents": [ "4f7c0cc6b272f9998074d0cde36b24bdd7b06bf9" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Tue Nov 05 10:59:36 2019 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Fri Nov 08 14:46:52 2019 +0000" }, "message": "Add boot image checksums to image header.\n\nThis adds an early validity check, so that we do not need to\nload an image extension and oat file before we can discard\nthem as out-of-date based on the checksums from oat header.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing --interp-ac\nBug: 119800099\nChange-Id: I951e8429f8dde73f3fc41c36dffe5a657a0db62b\n" }, { "commit": "2191069047034ad891ea15f60a217246edc38d53", "tree": "5fcdb37aa8d1e052979631adac01fe74794748b3", "parents": [ "add69acb7db965bd8d817a185e962e5474f690d1" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Nov 06 13:27:03 2019 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Nov 07 14:04:57 2019 +0000" }, "message": "Revert^4 \"Boot image extension.\"\n\nThis reverts commit f5c5eb30fc71e0c305d678bd3c1c995a5c36d508.\n\nRemoved wrong check for no dex files. Dex files can also be\nspecified with zip-location and file descriptors. Added a\nregression test to dex2oat_test.\n\nFixed the image checksums validation to use the appropriate\npart of BCP rather then full BCP. This fixes errorneous\nrejection of oat files compiled against partial BCP.\n\nChanged dex2oat_image_test to use ART_BASE_ADDRESS to try\nand avoid failures when reserving space for loading the\ncompiled images.\n\nTest: Additional tests in dex2oat_{,image_}_test\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nBug: 119800099\nBug: 143492855\nBug: 144001974\nChange-Id: I062cbecd0020f5c24353eb75643ea5905cb6f4fd\n" }, { "commit": "f5c5eb30fc71e0c305d678bd3c1c995a5c36d508", "tree": "2f08c5562d3b35b7f0139f772b6324a9689d01ea", "parents": [ "acdb9a350fede0776e1f4d6d04dc3246249345e3" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Nov 06 12:25:10 2019 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Nov 06 13:25:07 2019 +0000" }, "message": "Revert^3 \"Boot image extension.\"\n\nThis reverts commit 02820f424714e711bbd4cb4b04a109416eb0c8b8.\nAlso reverts commit beb66b38dcce937d7eee9ef2d07b6402c720f8ee\nthat contained a follow-up fix.\n\nReason for revert: b/144001974\n\nBug: 119800099\nBug: 143492855\nBug: 144001974\nChange-Id: I4da5330c3efa9f0c3508e85344c031d3b360ca0a\n" }, { "commit": "beb66b38dcce937d7eee9ef2d07b6402c720f8ee", "tree": "4bdf0034b5fbd4aa6bb07c73f8c38f9eac03669e", "parents": [ "a48f0bbb2939fa743904f7553a060b12846f100e" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Fri Nov 01 14:59:36 2019 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Tue Nov 05 09:42:18 2019 +0000" }, "message": "Fix boot image extension relocation.\n\nStore boot image begin and size to the extension header as\nit is needed for relocation; this requires rewriting the\nImageHeader::IsAppImage(). Also mark the class roots as\nvisited to satisfy a DCHECK() for extension relocation.\n\nAnd re-enable loading boot image extensions for tests.\n\nTest: Add relocation tests to dex2oat_image_test.\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing --relocate --no-relocate\nTest: testrunner.py --target --64 --optimizing --relocate\nBug: 119800099\nChange-Id: Ie7e883beaa927ca4dc91f0ae660c8de74f7d4ddb\n" }, { "commit": "7f958e36546bc0ebf92573da24ff620179526243", "tree": "a712a5527fdb5fe0b3e14b6f17fea90a89f55ecc", "parents": [ "7814c122dd35ac5c79e7b76b12f1ae95201739a1" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Oct 24 09:03:58 2019 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Fri Oct 25 08:16:27 2019 +0000" }, "message": "Revert^2 \"ARM64: toHalf() intrinsic for ARMv8\"\n\nThis reverts commit 67bf99b8a575b15c117a3fbf7aae421330e4795a.\n\nReason for revert: The original CL was reverted because\nof build breakages not directly related to the CL itself.\n\nChange-Id: Ic98e9912701d81d73bc3af719a7e3a8e44e8c058\n" }, { "commit": "67bf99b8a575b15c117a3fbf7aae421330e4795a", "tree": "2077b459c5b43c3b1399544faf07a326e253681a", "parents": [ "2cc0c0f4b6a76dfb1ad205cfd79efe7efe2904d6" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Oct 23 13:29:10 2019 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Oct 23 13:29:10 2019 +0000" }, "message": "Revert \"ARM64: toHalf() intrinsic for ARMv8\"\n\nThis reverts commit 2cc0c0f4b6a76dfb1ad205cfd79efe7efe2904d6.\n\nReason for revert: Broke some builds. To be investigated.\n\nBug: 143205070\nChange-Id: Ib36e149cc5ed7f53c932c0b611b43c28d19f22dc\n" }, { "commit": "2cc0c0f4b6a76dfb1ad205cfd79efe7efe2904d6", "tree": "5ff5f5f7c5c25f441a36506a84988fa95d2dbd46", "parents": [ "b8c884e5f22390386b202459ab55ef3046631e42" ], "author": { "name": "Usama Arif", "email": "usama.arif@linaro.org", "time": "Tue Oct 15 15:36:51 2019 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Oct 23 12:04:00 2019 +0000" }, "message": "ARM64: toHalf() intrinsic for ARMv8\n\nThis CL implements an intrinsic for toHalf() method with\nARMv8.2 FP16 instructions.\n\nThis intrinsic implementation achieves bit-level compatibility with the\noriginal Java implementation android.util.Half.toFloat().\n\nThe time required to execute the below code on Pixel3:\n- Java implementation android.util.Half.toFloat():\n - big cluster only: 2136ms\n - little cluster only: 6442ms\n- arm64 Intrinisic implementation:\n - big cluster only: 1347ms (~37% faster)\n - little cluster only: 4937ms (~ 23% faster)\n\nint benchmarkToHalf() {\n int result \u003d 0;\n // 5.9605E-8 is the smallest positive subnormal number that can be\n // represented by FP16. This is 0x33800032 in float bits.\n int raw_input \u003d 0x33800032;\n long before \u003d 0;\n long after \u003d 0;\n before \u003d System.currentTimeMillis();\n do {\n float input \u003d Float.intBitsToFloat(raw_input);\n short output \u003d FP16.toHalf(input);\n result +\u003d output;\n } while (++raw_input !\u003d 0x477fff00);\n // 65535 is the max possible integer that can be represented by FP16.\n //This is 0x477fff00 in float bits.\n after \u003d System.currentTimeMillis();\n System.out.println(\"Time of FP16.toHalf (ms): \" + (after - before));\n return result;\n}\n\nTest: 580-fp16\nTest: art/test/testrunner/run_build_test_target.py -j80 art-test-javac\nTest: test-art-host, test-art-target\n\nChange-Id: I69b152682390e5ffa5b3fdca60b496261191655d\n" }, { "commit": "9ce340f829f836560278ecd078fbefcf19c9d629", "tree": "3327d6d7dd3c0ff86861cdab67791ee7c8dbe3b8", "parents": [ "d4fc62c66328c0944348a314e3770b4f2b8006ce" ], "author": { "name": "xueliang.zhong", "email": "xueliang.zhong@linaro.org", "time": "Tue Jan 22 17:46:09 2019 +0000" }, "committer": { "name": "Hans Boehm", "email": "hboehm@google.com", "time": "Thu Aug 15 20:35:11 2019 +0000" }, "message": "ARM64: toFloat() intrinsics with ARMv8 FP16.\n\nThis CL intrinsifies toFloat() method with ARMv8.2 FP16 instructions.\n\nThis CL depends on the android framework and libcore changes:\nmoving FP16 implementations into libcore.\n\nTested with local micro benchmark on Pixel 3, compared to original\nandroid.util.Half.toFloat() Java implementation, this intrinsic is\n50% faster.\n\nIn real-life case, the FP16 toFloat() intrinsic can help\naccelerate ColorLong ARGB decoding in Android framework.\n\nThis intrinsic implementation archieves bit-level compatibility with the\noriginal Java implementation android.util.Half.toFloat().\n\nTest: 580-fp16\nTest: art/test/testrunner/run_build_test_target.py -j80 art-test-javac\nTest: test-art-host, test-art-target\n\nChange-Id: I059c69747067b84f2c532465e32a1dcd3c25269f\n" }, { "commit": "c0b30c9cd2553f58035853919a9820b0b54db783", "tree": "7aa919f061ae4f30f55bc908cd234c3876dd6942", "parents": [ "672c08092a7d95bb504bceb712812f72524d90da" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Tue Jul 23 14:58:25 2019 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Jul 29 12:37:41 2019 +0000" }, "message": "Implement boot image extension relocation.\n\nThe creation of boot image extensions and other aspects\nof loading them shall be implemented in separate CLs.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nTest: aosp_taimen-userdebug boots.\nChange-Id: I1630d6cc44550bc0db72cb5c4a8e08573cf3b92c\n" }, { "commit": "024d69fb9936ca5a0031d35c9f248853cbc25d3f", "tree": "c1eeddf91ea15eda5d139d4592ac7f0df80e9be0", "parents": [ "43ae4acf219fe25a56e2055ebcebc4d08020a25d" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Jun 13 10:52:32 2019 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Jun 17 09:30:53 2019 +0000" }, "message": "Use cleared JNI weak sentinel from boot image.\n\nWe were already adding the sentinel to the boot image,\nso we may as well reuse the boot image copy.\n\nAlso move pre-allocated objects from class roots to the\nboot image live objects.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nTest: aosp_taimen-userdebug boots.\nChange-Id: I635dcdd146ca2c6b55d187e9a545a9990b0b35ca\n" }, { "commit": "552a13415573da19eafa46e1ac00fb0eb68f2b23", "tree": "8cae5f3602d8f8e65cd3cbc349af17d785128605", "parents": [ "0dda8c84938d6bb4ce5a1707e5e109ea187fc33d" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Tue Oct 31 10:56:47 2017 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu May 16 14:44:09 2019 +0000" }, "message": "ART: Optimize StringBuilder append pattern.\n\nRecognize appending with StringBuilder and replace the\nentire expression with a runtime call that perfoms the\nappend in a more efficient manner.\n\nFor now, require the entire pattern to be in a single block\nand be very strict about the StringBuilder environment uses.\nAlso, do not accept StringBuilder/char[]/Object/float/double\narguments as they throw non-OOME exceptions and/or require a\ncall from the entrypoint back to a helper function in Java;\nthese shall be implemented later.\n\nBoot image size for aosp_taimen-userdebug:\n - before:\n arm/boot*.oat: 19653872\n arm64/boot*.oat: 23292784\n oat/arm64/services.odex: 22408664\n - after:\n arm/boot*.oat: 19432184 (-216KiB)\n arm64/boot*.oat: 22992488 (-293KiB)\n oat/arm64/services.odex: 22376776 (-31KiB)\nNote that const-string in compiled boot image methods cannot\nthrow, but for apps it can and therefore its environment can\nprevent the optimization for apps. We could implement either\na simple carve-out for const-string or generic environment\npruning to allow this pattern to be applied more often.\n\nResults for the new StringBuilderAppendBenchmark on taimen:\n timeAppendLongStrings: ~700ns -\u003e ~200ns\n timeAppendStringAndInt: ~220ns -\u003e ~140ns\n timeAppendStrings: ~200ns -\u003e 130ns\n\nBug: 19575890\nTest: 697-checker-string-append\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nTest: aosp_taimen-userdebug boots.\nTest: run-gtests.sh\nTest: testrunner.py --target --optimizing\nTest: vogar --benchmark art/benchmark/stringbuilder-append/src/StringBuilderAppendBenchmark.java\nChange-Id: I51789bf299f5219f68ada4c077b6a1d3fe083964\n" }, { "commit": "d456117108c1efad790808698787868813b428e2", "tree": "d5785c40ad689dbd158fc52eed8b32c9ced967d5", "parents": [ "ba2531826613a21736dd170d09877d3eb7990c5d" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Oct 30 17:48:25 2017 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon May 13 15:36:23 2019 +0000" }, "message": "ART: Mark StringBuilder.append()s as intrinsics.\n\nWe do not actually emit intrinsic implementations for these\nmethods but we use this designation to simplify their use\nknowing that they \"return this\". This also prepares for\nrecognizing and optimizing the StringBuilder append pattern.\n\nNote that intrinsics are never inlined but these methods are\ndefined in core-oj and therefore the rest of the boot image\nand prebuilts are not allowed to inline them anyway.\n\nBoot image sizes for aosp_taimen-userdebug:\n - before:\n arm/boot*.oat: 19756400\n arm64/boot*.oat: 23415656\n - after:\n arm/boot*.oat: 19653872 (-100KiB)\n arm64/boot*.oat: 23292784 (-120KiB)\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nTest: aosp_taimen-userdebug boots\nTest: run-gtests.sh\nTest: testrunner.py --target --optimizing\nBug: 19575890\nChange-Id: I2f3bac17830ac1c51287b090a96facddd3d09f5c\n" }, { "commit": "776a7c294f6356b314505369679bffc73f4a345e", "tree": "ec1d678a0923d2c7ec869f806f5b4860e1a1a721", "parents": [ "8ceeb2ede9a7280fab618dfcb55f424a79dc00ca" ], "author": { "name": "Evgeny Astigeevich", "email": "evgeny.astigeevich@linaro.org", "time": "Mon Dec 17 11:40:34 2018 +0000" }, "committer": { "name": "Evgeny Astigeevich", "email": "evgeny.astigeevich@linaro.org", "time": "Wed Dec 19 23:01:02 2018 +0000" }, "message": "ART: Add CRC32.updateByteBuffer intrinsic for ARM64\n\nUse crc32 instructions for\njava.util.zip.CRC32.updateByteBuffer(int, long, int, int).\n\nNote that CRC32 is an optional feature in ARMv8, this intrinsic\nis only enabled for devices supporting the CRC32 instructions.\n\nThe performance of the intrinsic is the same as the performance of the\nCRC32.updateBytes intrinsic. However the intrinsic does not have a\nrestriction on the size of the byte buffer. For big input data\nthe intrinsic will be faster than CRC32.updateBytes.\n\nTest: m test-art-target-gtest\nTest: m test-art-host-gtest\nTest: art/test.py --target --optimizing\nTest: art/test.py --host --optimizing\nTest: 580-crc32\n\nChange-Id: I6b6160b56e829731402bd5fc56bcac59664f634d\n" }, { "commit": "7391c8ce016c9bcd9e7cec91fb8dcf29f5ca5909", "tree": "fca941e7b216abd17a51d9ea0ba3d8aab059a05d", "parents": [ "5ab44e73e5f9e2403e0cac0a448b35378dae1de8" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Nov 21 17:58:44 2018 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Dec 10 11:31:57 2018 +0000" }, "message": "Store additional information in image header.\n\nStore the reservation needed for loading the boot image and\nthe number of boot class path components. The former helps\navoid one extra file open+read+close when loading the boot\nimage and the latter shall be used in subsequent CLs to load\nboot image for a partial boot class path.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nTest: Pixel 2 XL boots.\nBug: 119868597\nChange-Id: I01dba923cfa3f8faf0e41a4139b8913c78404d54\n" }, { "commit": "15c5b97d0093a804881acf3a53bda8b6b4c1521d", "tree": "ad6befd10a3d6dee0f1667d254374e267ad89ce7", "parents": [ "cfc4b3ac67b364d753d970d1adcf97668114e959" ], "author": { "name": "Evgeny Astigeevich", "email": "evgeny.astigeevich@linaro.org", "time": "Tue Nov 20 13:41:40 2018 +0000" }, "committer": { "name": "Evgeny Astigeevich", "email": "evgeny.astigeevich@linaro.org", "time": "Thu Dec 06 22:19:23 2018 +0000" }, "message": "ART: Add CRC32.updateBytes intrinsic for ARM64\n\nUse crc32 instructions for\njava.util.zip.CRC32.updateBytes(int,byte[],int,int).\n\nThe intrinsic is used if a number of processed bytes is less or equal to\nkCRC32UpdateBytesThreshold. If it exceeds kCRC32UpdateBytesThreshold the\ncore library provided function is used.\n\nNote that CRC32 is an optional feature in ARMv8, this intrinsic\nis only enabled for devices supporting the CRC32 instructions.\n\nThe CL is based on code from tim.zhang@linaro.org.\n\nPerformance improvements - speedup:\narray size | Cortex-A53 | Cortex-A57\n------------------------------------\n128 | 14x | 20x\n256 | 10x | 14x\n512 | 8x | 11x\n1024 | 7x | 9x\n2048 | 6x | 8x\n4096 | 5x | 7x\n8192 | 5x | 7x\n16384 | 5x | 7x\n32768 | 5x | 7x\n65536 | 5x | 7x\n\nTest: m test-art-target-gtest\nTest: m test-art-host-gtest\nTest: art/test/testrunner/testrunner.py --target --optimizing --interpreter\nTest: art/test/testrunner/testrunner.py --target --jit\nTest: art/test/testrunner/testrunner.py --host --optimizing --interpreter\nTest: art/test/testrunner/testrunner.py --host --jit\nTest: 580-crc32\n\nChange-Id: I0054cea41b5fc3e712e18b0afc7e3eacbf41feb6\n" }, { "commit": "1a8429680f9d08d5f2b49fd93f9ad4df81b7cb66", "tree": "75bb0ef44acf50a3108e85529f0495176afbced9", "parents": [ "f2970cd870948a6ee1c8ecd30c9c3147d05aa0be" ], "author": { "name": "Mathieu Chartier", "email": "mathieuc@google.com", "time": "Tue Nov 13 15:09:51 2018 -0800" }, "committer": { "name": "Mathieu Chartier", "email": "mathieuc@google.com", "time": "Thu Nov 29 17:01:44 2018 -0800" }, "message": "Add image compressed blocks\n\nAdd support for splitting the image into a set of solid blocks.\n\nAdded dex2oat option --max-image-block-size and correspodning image\nunit test.\n\nMotivation: Enable parallel image decompression in the future.\n\nBug: 116052292\nTest: test-art-host\n\nChange-Id: I37c6c6a43ef94c4a62bf38a0cf51f26ce06347ac\n" }, { "commit": "e36e7f2226e4e08b7a7094f78cb80bbe0e729c2b", "tree": "0787ae19a232728121a60791f3a18094529120f1", "parents": [ "244470afafdb1c5aba17507ef793d316b9c4d038" ], "author": { "name": "David Srbecky", "email": "dsrbecky@google.com", "time": "Wed Nov 14 14:21:23 2018 +0000" }, "committer": { "name": "David Srbecky", "email": "dsrbecky@google.com", "time": "Tue Nov 27 17:31:01 2018 +0000" }, "message": "Store ImtIndex in ArtMethod.\n\nThis avoids recalculation and reduces pressure on the thread local cache.\n\nThis halves the time we spend hashing from 2% to 1% (maps on device).\n\nTest: ./art/test.py -b --host --64\nChange-Id: I2407bd9c222de4ddc6eea938908a1ac6d7abc35b\n" }, { "commit": "0c78ef7f2e8f010311ce9c0715b1ab3fe28f68f6", "tree": "97755cc52e33418aacfd1cc5ebbf061085dec780", "parents": [ "5a78e8dbf068c544b84e98f636b30045b09451c4" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Nov 21 14:09:35 2018 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Nov 21 16:47:03 2018 +0000" }, "message": "Remove boot image begin/size from image header.\n\nUse a single boot image range for both the heap (*.art) and\ncode (*.oat) boot files.\n\nThis is done in preparation for boot image extensions. We do\nnot want to encode multiple heap and code ranges and shall\nkeep a single boot image range for the boot image including\nextensions.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nBug: 119868597\nChange-Id: I8baec8361602c445652e69bdfbb57e7a7c7c6419\n" }, { "commit": "c10a0c60ca388ea5f45c11dd86ca0af11191015a", "tree": "8befa85940f28ca1c764e211f6f269daf184d28c", "parents": [ "d52a6f402fac600e7432cfd59a1298659625a6d3" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Fri Nov 16 11:39:22 2018 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Nov 21 11:23:38 2018 +0000" }, "message": "ART: Add boot/app image checksums.\n\nAnd check app oat files against the boot image checksum\ninstead of checking the oat checksum. The oat checksums are\nincluded in image checksum calculations and the primary\nimage checksum calculation includes the checksums of the\nsecondary images.\n\nAlso remove the obsolete \"patch delta\" to keep the size\nof the ImageHeader 8-byte aligned, remove the key-value\nstore from secondary oat files and move the oat checksum\nupdate code from oat.cc to the oat_writer.cc.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host\nTest: Pixel 2 XL boots.\nTest: m test-art-target-gtest\nChange-Id: If74f5f18479c44ede0503bf1911ddb9ff8f3c4f8\n" }, { "commit": "cb58b07d233c5f8e910e6e5946da8ac4ec3099f5", "tree": "bd19cb9a44555adf51e2d145c322a592c6cd3527", "parents": [ "e2acc339ede3c620a476827880cf68d044dcd53e" ], "author": { "name": "xueliang.zhong", "email": "xueliang.zhong@linaro.org", "time": "Fri Oct 13 12:06:56 2017 +0100" }, "committer": { "name": "Evgeny Astigeevich", "email": "evgeny.astigeevich@linaro.org", "time": "Fri Nov 02 10:07:12 2018 +0000" }, "message": "ART: Add CRC32.udate(int,int) intrinsic for ARM64\n\nUse crc32 instructions for java.util.zip.CRC32.update(int,int).\n\nNote that CRC32 is an optional feature on ARMv8, this intrinsic\nis only enabled for devices with CRC32 intruction support.\n\nOriginal author: tim.zhang@linaro.org\n\nPerformance improvements in CRC32Bench.UpdateInt:\nPixel 2: 22.8x\nNexus 6P:\n little core: 28.3x\n big core : 21.6x\n\nTest: m test-art-target-gtest\nTest: testrunner.py --target --optimizing --jit --interpreter\nTest: 580-crc32\n\nChange-Id: I1a9bc2befd2934b04103a27ce05806e919874070\n" }, { "commit": "1ca718ec2de53fb5ef2a9211bfcc921311775c90", "tree": "f5ce8c3f4edd022765c343cc5a9df8ebd292f0ec", "parents": [ "ce2a00daa92670a4fc01ef59fdbc3769a846f69c" ], "author": { "name": "Mathieu Chartier", "email": "mathieuc@google.com", "time": "Tue Oct 23 12:55:34 2018 -0700" }, "committer": { "name": "Mathieu Chartier", "email": "mathieuc@google.com", "time": "Tue Oct 30 16:03:00 2018 +0000" }, "message": "Add PreResolved strings dex cache array\n\nFor app images, this dex cache may be prepopulated to speed up\napplication startup. This new dex cache array is created when\n--resolve-startup-const-strings\u003dtrue.\n\nTest: test-art-host\nBug: 116059983\nBug: 118385560\n\nChange-Id: I379dc15174281665d7f4ceb106f7fbf51f89b921\n" }, { "commit": "d68ab24b556fca8b89695d4e199a906a611dae64", "tree": "88b5135de38d51b4df837a58f509ce2fd7d6950c", "parents": [ "62ada4cd2c9a7ef5d1ab4c538d660f8a6a7f7571" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Oct 18 16:07:10 2018 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Tue Oct 23 11:11:52 2018 +0100" }, "message": "Remove the relocations section from the image.\n\nThis section was unused since\n https://android-review.googlesource.com/771001 .\n\nThis partially reverts commits\n 6121aa69098e3496cf1a81bf3e5e7ae70f66eacb\n ca8de0a70eab62707f3c71a211093f340fdcd5f4\nand cleans up.\n\nPrebuilt sizes for aosp_taimen-userdebug:\n - before:\n arm/boot*.art: 14951820\n arm64/boot*.art: 19398428\n oat/arm64/services.art: 526138\n - after:\n arm/boot*.art: 14503936 (-437KiB)\n arm64/boot*.art: 18817024 (-568KiB)\n oat/arm64/services.art: 499712 (-26KiB)\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nTest: Pixel 2 XL boots.\nTest: m test-art-target-gtest\nTest: testrunner.py --target --optimizing\nBug: 77856493\nChange-Id: Ib44cd00cf3944e93e2d61ca8df381993cfeff130\n" }, { "commit": "a2da9b99fa1ea3d25d52da71308a623b2aae216c", "tree": "5533be23eee7c24b68b1b72272cbae3f35a708cb", "parents": [ "dc3b4670b170b39a8bd6498d4de69c1513af1db2" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Oct 10 14:21:55 2018 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Oct 10 15:25:26 2018 +0100" }, "message": "ART: Completely remove the --compile-pic option.\n\nAnd the PIC-related fields from image header.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nTest: Pixel 2 XL boots.\nTest: testrunner.py --target --optimizing\nBug: 77856493\nChange-Id: I3787369378f12d8cd9003bebeae62830a67def33\n" }, { "commit": "4df2d8041f5dcc7af8c3b3b60b0ea87a1e0d3b94", "tree": "0273072a2b65d6c0cf692a3e7f8eab9814d9d23d", "parents": [ "233b572a940431a94a1790750afdceab2d6f4fde" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Sep 27 16:42:44 2018 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Fri Sep 28 09:12:06 2018 +0100" }, "message": "Revert^2 \"Load boot image at a random address.\"\n\nThis reverts commit f3d077373536c54824e4449759dff2f18369eab3.\n\nFixed Heap constructor to reserve extra space for GSS.\n\nChange-Id: I6a65be35f4aa183304db5491da4a4810d8e3b266\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing --relocate --no-relocate\nTest: Pixel 2 XL boots.\nTest: m test-art-target-gtest\nTest: testrunner.py --target --optimizing --relocate --no-relocate\nTest: art/test/testrunner/run_build_test_target.py -j48 art-gtest-gss-gc-tlab\nBug: 77856493\n" }, { "commit": "f3d077373536c54824e4449759dff2f18369eab3", "tree": "e00d19a342ff429fc514bf70bf52ff005a00911c", "parents": [ "5ad79d85d77a42456728897ac3e2e7d4530e618e" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Sep 27 16:24:17 2018 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Sep 27 16:24:17 2018 +0000" }, "message": "Revert \"Load boot image at a random address.\"\n\nThis reverts commit 5ad79d85d77a42456728897ac3e2e7d4530e618e.\n\nReason for revert: Breaks GSS garbage collection config.\n\nBug: 77856493\nChange-Id: Ifa39966ac2470154f8ba093de4804689d545219b\n" }, { "commit": "5ad79d85d77a42456728897ac3e2e7d4530e618e", "tree": "ff70d1f12904c718f2f3f721d3846f4305eae53d", "parents": [ "5d7015cd64085068b1685d44339b4b705ef3f065" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Sep 03 09:54:09 2018 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Sep 26 13:08:36 2018 +0100" }, "message": "Load boot image at a random address.\n\nAnd perform in-place fixup of references and pointers. This\ndirties all the boot image memory loaded by zygote, so there\nshall be no \"shared clean\" boot image pages anymore, these\nshall change to \"shared dirty\". However, as we\u0027re using a\nprofile-based boot image, these pages are presumably used\noften enough and unlikely to be paged out anyway.\n\nThe in-place fixup takes around 60-120ms when starting the\nzygote on aosp_taimen-userdebug. However, an experiment with\nMAP_POPULATE pushes the raw fixup down to around 12-15ms.\nIf we used compressed images, this would be the actual time\nfor fixup as the data would be already present in memory.\nIf we keep using uncompressed images, we shall need to tune\nthe loading with MAP_POPULATE or MADV_WILLNEED.\n\nThe -Xrelocate/-Xno-relocate option is re-interpreted from\n\"use patchoat if needed\" to \"relocate the boot image in\nmemory if possible\". We do not allow relocation for the AOT\ncompilation to speed up dex2oat execution and help producing\ndeterministic output.\n\nThe patchoat tool shall be removed in a follow-up CL.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing --relocate --no-relocate\nTest: Pixel 2 XL boots.\nTest: m test-art-target-gtest\nTest: testrunner.py --target --optimizing --relocate --no-relocate\nBug: 77856493\nChange-Id: I2db1fabefb5d4b85c798cd51e04c78cb232bff4a\n" }, { "commit": "990fc44436ab2e19d42b869ef6db5745eb298783", "tree": "65701383bf2fb341de2d76d7e7287de769309fb0", "parents": [ "7786176c18d63db66684d336d647bef4a0aeb02b" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Aug 29 13:58:23 2018 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Aug 29 13:58:49 2018 +0000" }, "message": "Revert \"Reduce DexCache field/method array size to 512.\"\n\nThis reverts commit b1e46ff03798fbb6745c9825b02a3e468cd8ace0.\n\nReason for revert: Large instant app startup time regression.\nBug: 112901509\n\nChange-Id: Ic6fc9a82a92c5138f917f4389c7d334c640f3892\n" }, { "commit": "b1e46ff03798fbb6745c9825b02a3e468cd8ace0", "tree": "45fd0d8cbaf614b56bf83f48c6dc19b15894b8f9", "parents": [ "4075f08db80c7c0475f583caae2aa625d53e1a48" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon May 14 18:06:36 2018 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Tue Aug 07 18:31:00 2018 +0100" }, "message": "Reduce DexCache field/method array size to 512.\n\nNo significant change in interpreter benchmarks on Golem.\nReduces the maximum size of each of these arrays from\n8KiB/16KiB to 4KiB/8KiB for 32-/64-bit architectures.\n\nTest: Rely on TreeHugger.\nChange-Id: I78833b82a672453abc4fffc6be1022b4afd7c784\n" }, { "commit": "6121aa69098e3496cf1a81bf3e5e7ae70f66eacb", "tree": "4a09aaee7669821725b33f8d7c5a7fe087b5f324", "parents": [ "6250dfa22577160b270f3f3165260739707cf7d3" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Fri Jul 06 10:04:35 2018 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Fri Jul 20 13:55:09 2018 +0100" }, "message": "Move .art.rel data to a section in .art, part 2.\n\nAnd change the encoding to reduce the size of the data.\nKeep the checksum in the .art.rel file, it shall be removed\nlater together with patchoat.\n\nBoot image sizes for aosp_taimen-userdebug:\n - before:\n arm/boot*.art: 9216000\n arm/boot*.art.rel: 700767\n arm64/boot*.art: 11399168\n arm64/boot*.art.rel: 700808\n oat/arm64/services.art: 192512\n - after:\n arm/boot*.art: 9499351 (+276.7KiB)\n arm/boot*.art.rel: 480 (-683.9KiB)\n arm64/boot*.art: 11750203 (+342.8KiB)\n arm64/boot*.art.rel: 480 (-683.9KiB)\n oat/arm64/services.art: 202466 (+9.7KiB)\nNote that the new section is currently uncompressed in the\nboot image but we have the ability to compress it in the\nfuture using the same compression as the heap data.\n\nThe extra data we now encode in app images is unused so far\nbut it shall permit fast in-memory patching without looking\nat object types.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nTest: Pixel 2 XL boots.\nTest: testrunner.py --target --optimizing\nBug: 77856493\nChange-Id: I20c9bed9797ce0f23f39c2fb0d64320b457e18d4\n" }, { "commit": "ca8de0a70eab62707f3c71a211093f340fdcd5f4", "tree": "a391c1307e87edad399418236a2a1842a8f06f34", "parents": [ "7f7f9d3991f3a55da8934a3b72890d4776373598" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Jul 04 11:56:08 2018 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Jul 19 13:42:20 2018 +0100" }, "message": "Move .art.rel data to a section in .art, part 1.\n\nRefactor the code so that we can collect all relocation\noffsets while writing the image.\n\nTest: Manually verify that the aosp_taimen-userdebug boot\n image is identical before and after this change;\n using WIP follow-up code, check that we record the\n same locations as the patchoat-generated .rel files.\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nBug: 77856493\nChange-Id: Id8c78e417e77e859a82a680504253f41e84560d6\n" }, { "commit": "f75613cb7c5cda6a4603ab7041d6f98ec62a80cd", "tree": "5d253108db708da20e0bbf55a5f3d8e2f8bce4d7", "parents": [ "2b40dd35c65ad644d448611750f5b577e97594a1" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Tue Jun 05 12:51:04 2018 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Jun 07 14:24:02 2018 +0100" }, "message": "Keep objects for Integer.valueOf() intrinsic alive.\n\nKeep boot image objects referenced by the intrinsic alive\nthrough boot image roots, so that they can never be dead as\nthe intrinsic would still be able to resurrect them later.\n\nNote that currently the GC considers all boot image objects\nlive forever. That risks leaking memory and should be fixed.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nBug: 71526895\nChange-Id: Iec25cc27e9c5c4bd3c5711991e3111bfb19ef182\n" }, { "commit": "c13fbd8596988f1daf71197008007c2eaa380585", "tree": "4c02277989abc7ac1081698269504ab329ddab8f", "parents": [ "026105570d8fcd4b3cb60c3d6b7bb7141cd8146c" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Jun 04 16:16:28 2018 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Jun 04 18:33:38 2018 +0100" }, "message": "Use pre-allocated Throwables from the boot image.\n\nThe pre-allocated OOMEs and NoClassDefFoundError were stored\nin the boot image but they were not used, we instead used to\nallocate and use new objects. This change adds references to\nthe image roots, so that these Throwables can be used when\nstarting the runtime using the boot image.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nBug: 77947463\nChange-Id: I2079344dee61242bf0bef5c32770c33ac8a6b7a4\n" }, { "commit": "c7aa87e1666ac48ddf9149cfdfd64b026b3969e5", "tree": "32d5d74718cc558e13642873e55724782ac9df22", "parents": [ "0278be74269fcfe4f2517d449f2bd53472f9b2f9" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu May 24 15:19:52 2018 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Fri May 25 11:37:45 2018 +0100" }, "message": "Remove static_class_ from Method/VarHandle and CallSite.\n\nAnd add MethodHandle to the class roots to avoid extra\nindirection through MethodHandleImpl.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nBug: 31113334\nChange-Id: Iaf172f3732677f2b4509e8297e6e9af5fb81a89f\n" }, { "commit": "c7b28de9f8bf407d91cff22de782d022492b45f7", "tree": "7d85879a528f21e25b5ab36f20de716b4cf5892d", "parents": [ "8fd8cdc43e10a421f6f63afb87f6f99c086058de" ], "author": { "name": "Hans Boehm", "email": "hboehm@hboehm.mtv.corp.google.com", "time": "Fri Mar 09 17:05:28 2018 -0800" }, "committer": { "name": "Hans Boehm", "email": "hboehm@google.com", "time": "Thu Mar 29 10:42:58 2018 -0700" }, "message": "Add reachabilityFence intrinsics\n\nAdd intrinsics that generate no code or do nothing for all architectures\nand for the interpreter. The only impact is to keep the argument live at\nall suspend points preceding the call. We ensure that the code is not\nmoved across other memory accesses by declaring it to have write side-effects.\n\nAdd a minimal test.\n\nModify 036-finalizer to use a reachabilityFence, hopefully making it\nmore robust to dead refererence elimination.\n\nBug: 72698200\n\nTest: Build and boot AOSP.\n art/test.py --host -r -t 072-reachability-fence\n Look at generated code.\n\nChange-Id: I0f298bf5cc375d8ebc19bb791cc05a8490d55430\n" }, { "commit": "bfd3bc801beebc023acb21ae24ab8d8cacf8a4c7", "tree": "e5d17aad9bdce6970fd0d897a9ec2282b7f54925", "parents": [ "41a10575b319ca9549d7b3e8a9fe91c1cc602659" ], "author": { "name": "Andreas Gampe", "email": "agampe@google.com", "time": "Mon Mar 26 14:19:32 2018 -0700" }, "committer": { "name": "Andreas Gampe", "email": "agampe@google.com", "time": "Mon Mar 26 14:20:14 2018 -0700" }, "message": "ART: Bump image version\n\nBug: 26687569\nBug: 64692057\nBug: 76420366\nTest: m\nChange-Id: I3e67e750d99362accfc8cb35a19510638b217907\n" }, { "commit": "4b2b54701cbe2e9b4250ebe72a0780245454e75f", "tree": "634c050466cf544f32376e95bd2311f1dc6ab29f", "parents": [ "7075583146f52a754c6d7c91757a6333bf1b95da" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Mar 26 11:41:51 2018 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Mar 26 11:43:35 2018 +0100" }, "message": "Bump image version for bitstring type check.\n\nEnabling the bitstring type check requires rebuilding\nthe boot image. This should have been done in\n https://android-review.googlesource.com/615367 .\n\nTest: Rely on TreeHugger.\nBug: 64692057\nChange-Id: I7c0cdbb1cd1e1b0b30469b9f808842b4c38585c7\n" }, { "commit": "9992e095643f6746361df03c4c98e742d9ad5899", "tree": "8abf49af54ee57fc0acebf2a3d9cafd87d6ec48e", "parents": [ "a5867bfeb34529dad71220046e7327cef23af207", "e47f60c482648172334aaca59e6c1ab7a3d42610" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Fri Mar 09 10:17:11 2018 +0000" }, "committer": { "name": "Gerrit Code Review", "email": "noreply-gerritcodereview@google.com", "time": "Fri Mar 09 10:17:11 2018 +0000" }, "message": "Merge \"Retrieve String/Class references from .data.bimg.rel.ro.\"" }, { "commit": "e47f60c482648172334aaca59e6c1ab7a3d42610", "tree": "ae0672b12a6ad200e1c38962c77bccfc3e5cb531", "parents": [ "b066d43b1d9184899aff32b1f243d092611ad9c6" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Feb 21 13:43:28 2018 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Mar 08 10:40:12 2018 +0000" }, "message": "Retrieve String/Class references from .data.bimg.rel.ro.\n\nFor PIC AOT-compiled app, use the .data.bimg.rel.ro to load\nthe boot image String/Class references instead of using the\nmmapped boot image ClassTable and InternTable.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing --pictest --npictest\nTest: Pixel 2 XL boots.\nTest: testrunner.py --target --optimizing --pictest --npictest\nBug: 71526895\nChange-Id: Id5703229777aecb589a933a41f92e44d3ec02a3d\n" }, { "commit": "c431b9dc4b23cc950eb313695258df5d89f53b22", "tree": "422273559c3ae52caff0c6b1cf1a62a8312f0e26", "parents": [ "f46f46cf5bd32788d5252b7107628a66594a5e98" ], "author": { "name": "David Sehr", "email": "sehr@google.com", "time": "Fri Mar 02 12:01:51 2018 -0800" }, "committer": { "name": "David Sehr", "email": "sehr@google.com", "time": "Mon Mar 05 13:58:20 2018 -0800" }, "message": "Move most of runtime/base to libartbase/base\n\nEnforce the layering that code in runtime/base should not depend on\nruntime by separating it into libartbase. Some of the code in\nruntime/base depends on the Runtime class, so it cannot be moved yet.\nAlso, some of the tests depend on CommonRuntimeTest, which itself needs\nto be factored (in a subsequent CL).\n\nBug: 22322814\nTest: make -j 50 checkbuild\n make -j 50 test-art-host\n\nChange-Id: I8b096c1e2542f829eb456b4b057c71421b77d7e2\n" }, { "commit": "305c38b395eddff7ca001fa7bdb621397033184d", "tree": "809965b36e35d8b5a06229a4904c7943525ea54c", "parents": [ "3bffbe8a8288cf01b7248fffdd320bf688529bcb" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Feb 14 11:50:07 2018 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Feb 14 16:16:05 2018 +0000" }, "message": "Disable bitstring type check runtime hooks.\n\nIntroduce a build flag for the bitstring type check, put\nruntime hooks behind the flag and set the flag to false.\n\nAlso add bitstring initialization for proxy classes, a test\nand a benchmark for the type checks.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --interpreter\nTest: Repeat with kBitstringSubtypeCheckEnabled \u003d true.\nBug: 73299705\nChange-Id: Ibcd88a828c7addc0473d8c428818734f80226b19\n" }, { "commit": "4d17987da58d9411adbed1a18203d76d6119612d", "tree": "f2953a0eb3ebc3f8533d22c14f4a09d7f0d4168d", "parents": [ "e57043081e6b091a9fd23a84043373148ae72f1f" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Fri Jan 19 14:50:10 2018 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Jan 22 18:35:16 2018 +0000" }, "message": "ART: Add entrypoint and intrinsic for Math.pow().\n\nMathBenchmarks.java#timePow results on taimen\u0027s little cores\nfixed at frequency 1401600 with forced JIT compilation:\n - before:\n - X32: 356.33 (@FastNative), 315.39 (@CriticalNative)\n - X64: 357.31 (@FastNative), 315.37 (@CriticalNative)\n - after (LICM defeats the benchmark):\n - X32: 2.88\n - X64: 2.87\n - after but with kAllSideEffects to prevent LICM:\n - X32: 275.42\n - X64: 275.67\n\nTest: Rely on TreeHugger.\nBug: 70727450\nChange-Id: Iaa31f70acabbd57c163cfeafe02eed67c1348861\n" }, { "commit": "dc682aa9d0eae1a851af059434adb6f6cf8f06f8", "tree": "f93f00493ee5887b05b42a6a5dd99eb6794daad4", "parents": [ "d6b7e8c63f8eca25460f56f66dcae15eaa897ff0" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Jan 04 18:42:57 2018 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Jan 10 14:30:26 2018 +0000" }, "message": "Use 28 bits for type check bit string.\n\nAnd reverse the order of fields in the Class::status_. This\navoids generated code size increase:\n - ClassStatus in high bits allows class initialization\n check using \"status_high_byte \u003c (kInitialized \u003c\u003c 4)\"\n which is unaffected by the low 4 bits of LHS instead of\n needing to extract the status bits,\n - the type check bit string in the bottom bits instead of\n somewehere in the middle allows the comparison on ARM\n to be done using the same code size as with the old\n layout in most cases (except when the compared value is\n 9-16 bits and not a modified immediate: 2 bytes less for\n 9-12 bits and sometimes 2 bytes more for 13-16 bits; the\n latter could be worked around using LDRH if the second\n character\u0027s boundary is at 16 bits).\n\nAdd one of the extra bits to the 2nd character to push its\nboundary to 16 bits so that we can test an implementation\nusing 16-bit loads in a subsequent CL, arbitrarily add the\nother three bits to the 3rd character. This CL is only\nabout making those bits available and allowing testing, the\ndetermination of how to use the additonal bits for the best\nimpact (whether to have a 4th character or distribute them\ndifferently among the three characters) shall be done later.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nTest: Pixel 2 XL boots.\nTest: testrunner.py --target --optimizing\nBug: 64692057\nChange-Id: I38c59837e3df3accb813fb1e04dc42e9afcd2d73\n" }, { "commit": "2c64a837e62c2839521c89060b5bb0dcb237ddda", "tree": "65475ed2e313ff17354e741bac7e9c85739b8b95", "parents": [ "6cd0005698181e4cef2247b632d396e605d58fa3" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Jan 04 11:31:56 2018 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Fri Jan 05 11:27:50 2018 +0000" }, "message": "Change ClassStatus to fit into 4 bits.\n\nIn preparation for extending the type check bit string from\n24 to 28 bits, rewrite ClassStatus to fit into 4 bits. Also\nperform a proper cleanup of the ClassStatus, i.e. change it\nto an enum class, remove the \"Status\" word from enumerator\nnames, replace \"Max\" with \"Last\" in line with other\nenumerations and remove aliases from mirror::Class.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nTest: Pixel 2 XL boots.\nTest: testrunner.py --target --optimizing\nBug: 64692057\nBug: 65318848\nChange-Id: Iec1610ba5dac2c527b36c12819f132e1a77f2d45\n" }, { "commit": "b0a6aeee250945b1d156ebab94053380f2e5a3c5", "tree": "7ddb52bb20997b8f1c81be3df2250949a088e697", "parents": [ "0db16e00e3927445585a588499731c58c1ae1bef" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Fri Oct 27 10:34:04 2017 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Fri Nov 10 08:30:36 2017 +0000" }, "message": "Record @{Fast,Critical}Native in method\u0027s access flags.\n\nRepurpose the old kAccFastNative flag (which wasn\u0027t actually\nused for some time) and define a new kAccCriticalNative flag\nto record the native method\u0027s annotation-based kind. This\navoids repeated determination of the kind from GenericJNI.\nAnd making two transitions to runnable and back (using the\nScopedObjectAccess) from GenericJniMethodEnd() for normal\nnative methods just to determine that we need to transition\nto runnable was really weird.\n\nSince the IsFastNative() function now records the presence\nof the @FastNative annotation, synchronized @FastNative\nmethod calls now avoid thread state transitions.\n\nWhen initializing the Runtime without a boot image, the\nWellKnowClasses may not yet be initialized, so relax the\nDCheckNativeAnnotation() to take that into account.\n\nAlso revert\n https://android-review.googlesource.com/509715\nas the annotation checks are now much faster.\n\nBug: 65574695\nBug: 35644369\nTest: m test-art-host-gtest\nTest: testrunner.py --host\nChange-Id: I2fc5ba192b9ce710a0e9202977b4f9543e387efe\n" }, { "commit": "0db16e00e3927445585a588499731c58c1ae1bef", "tree": "9d151a03534703642d20cd2265260ff13c7592a4", "parents": [ "a3c21d9c0552a04dfa69817b991b8e902cd0d6d6" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Nov 08 14:32:33 2017 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Nov 08 14:50:56 2017 +0000" }, "message": "Use strcmp() for checking @{Fast,Critical}Native annotations.\n\nInstead of looking up the descriptor in boot class path\nloader (where @{Fast,Critical}Native are guaranteed to\nbe already resolved) and then checking if it\u0027s the\n@{Fast,Critical}Native annotation, just check the\ndescriptor with strcmp().\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --optimizing\nBug: 65574695\nChange-Id: I765590d039981d169fb3c606b6166580a84303b6\n" }, { "commit": "4a4610a438ff2b836f6fe07839a0689ce618863a", "tree": "afb5c6303c6c8eb68f81752a642a5b036049b28d", "parents": [ "865a0503515f78611dce103bd04b7c66c051d304" ], "author": { "name": "Orion Hodson", "email": "oth@google.com", "time": "Thu Sep 28 16:57:55 2017 +0100" }, "committer": { "name": "Orion Hodson", "email": "oth@google.com", "time": "Mon Oct 02 14:09:21 2017 +0100" }, "message": "ART: Add VarHandle fence intrinsics\n\nBug: 65872996\nTest: art/test/run-test --host --64 709\nTest: art/test.py --host -j32\nChange-Id: I6fa399bb00f0c83048ac2b4372b08e4b4b29ce7f\n" }, { "commit": "cfcc9cfb44bab79f7381bcc4bfd9bf2d4435f734", "tree": "6637daf0cde4f14afd02793ad268f22f9257bd4c", "parents": [ "f67f115423c9ef5aa62a33c12670cd8f89457c9c" ], "author": { "name": "Orion Hodson", "email": "oth@google.com", "time": "Fri Sep 29 15:07:27 2017 +0100" }, "committer": { "name": "Orion Hodson", "email": "oth@google.com", "time": "Sun Oct 01 12:08:08 2017 +0100" }, "message": "ART: Increase the number of potential instrinsics\n\nThe new limit is 256 intrinsics.\n\nAdds additional sanity checks.\n\nAvoids setting the kAccPreviouslyWarm bit for intrinics (defaults to\ntrue).\n\nBug: 65872996\nTest: art/test.py --host -j32\nChange-Id: I33ea67c9b6b8500b3ceb8a085358f075f6fcbb82\n" }, { "commit": "a49e053faa4a254df1f21e055cb73555de241cfc", "tree": "bed0e8f0934550d4329c0aed10f016c884c6750f", "parents": [ "47d00f6c4657af33c3cc29783c0f2eb156aa1449" ], "author": { "name": "David Sehr", "email": "sehr@google.com", "time": "Fri Aug 25 08:05:29 2017 -0700" }, "committer": { "name": "David Sehr", "email": "sehr@google.com", "time": "Mon Sep 11 15:31:00 2017 -0700" }, "message": "Show ArtMethods in imgdiag\n\nSince ArtMethods were moved out of mirror:: classes imgdiag does not\nshow information about them. Diff ArtMethods to facilitate finding\ndirty memory there.\n\nBug: 38173645\nTest: imgdiag --boot-image\u003d/system/framework/boot.art --image-diff-pid\u003d`pid system_server`\nChange-Id: Icd86a9ef14d5177a297026c22c81c080f5c85fc1\n" }, { "commit": "6cfbdbc359ec5414d3e49f70d28f8c0e65b98d63", "tree": "f92b309ddc43c2254b6067346a653170fbbf7316", "parents": [ "0f3c7003e08a42a4ed8c9f8dfffb1bee1118de59" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Tue Jul 25 13:26:39 2017 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Sep 07 17:52:35 2017 +0100" }, "message": "Use mmapped boot image intern table for PIC app HLoadString.\n\nImplement new HLoadString load kind for boot image strings\nreferenced by PIC-compiled apps (i.e. prebuilts) that uses\nPC-relative load from a boot image InternTable mmapped into\nthe apps .bss. This reduces the size of the PIC prebuilts\nthat reference boot image strings compared to the kBssEntry\nas we can completely avoid the slow path and stack map.\n\nWe separate the InternedStrings and ClassTable sections of\nthe boot image (.art) file from the rest, aligning the\nstart of the InternedStrings section to a page boundary.\nThis may actually increase the size of the boot image file\nby a page but it also allows mprotecting() these tables as\nread-only. The ClassTable section is included in\nanticipation of a similar load kind for HLoadClass.\n\nPrebuilt services.odex for aosp_angler-userdebug (arm64):\n - before: 20862776\n - after: 20308512 (-541KiB)\nNote that 92KiB savings could have been achieved by simply\navoiding the read barrier, similar to the HLoadClass flag\nIsInBootImage(). Such flag is now unnecessary.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host\nTest: testrunner.py --host --pictest\nTest: testrunner.py --target on Nexus 6P.\nTest: testrunner.py --target --pictest on Nexus 6P.\nTest: Nexus 6P boots.\nBug: 31951624\nChange-Id: I5f2bf1fc0bb36a8483244317cfdfa69e192ef6c5\n" }, { "commit": "cd87c3ee6fefc505010c795284cfc4bca076edf2", "tree": "074c0326206fe4a52dc8ebfa09657ef040d88022", "parents": [ "39d44ba52b512cdfa7a0a2fccc7f77d28a15329e" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Tue Sep 05 13:11:57 2017 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Tue Sep 05 13:26:54 2017 +0100" }, "message": "Clean up ImageHeader section getters.\n\nTest: m art-test-host-gtest\nTest: testrunner.py --host\nChange-Id: Iaea7ce6f9bc3ff3c6d9bb6fb598aa62333ebf30c\n" }, { "commit": "5122e6ba34d46851cd89f2ad55bf6bb067e038d6", "tree": "e96ba37b6451be7a06d930b0274251cac35ce05e", "parents": [ "02cb397857c979dffae95e2db2678a72ec407cf0" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Aug 17 16:10:09 2017 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Tue Aug 29 11:29:31 2017 +0100" }, "message": "ART: Remove ArtMethod::dex_cache_resolved_methods_.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host\nTest: testrunner.py --target on Nexus 6P\nTest: Repeat the above tests with ART_HEAP_POISONING\u003dtrue\nTest: Build aosp_mips64-eng\nChange-Id: I9cd0b8aa5001542b0863cccfca4f9c1cd4d25396\n" }, { "commit": "8cf9cb386cd9286d67e879f1ee501ec00d72a4e1", "tree": "88e86e214b425e444760fe4e0ffeee677e1558a2", "parents": [ "914b7b6a6c9f399b26b41e9160e9871ef749e0db" ], "author": { "name": "Andreas Gampe", "email": "agampe@google.com", "time": "Wed Jul 19 09:28:38 2017 -0700" }, "committer": { "name": "Andreas Gampe", "email": "agampe@google.com", "time": "Mon Jul 24 16:07:10 2017 -0700" }, "message": "ART: Include cleanup\n\nLet clang-format reorder the header includes.\n\nDerived with:\n\n* .clang-format:\n BasedOnStyle: Google\n IncludeIsMainRegex: \u0027(_test|-inl)?$\u0027\n\n* Steps:\n find . -name \u0027*.cc\u0027 -o -name \u0027*.h\u0027 | xargs sed -i.bak -e \u0027s/^#include/ #include/\u0027 ; git commit -a -m \u0027ART: Include cleanup\u0027\n git-clang-format -style\u003dfile HEAD^\n manual inspection\n git commit -a --amend\n\nTest: mmma art\nChange-Id: Ia963a8ce3ce5f96b5e78acd587e26908c7a70d02\n" }, { "commit": "07bfbace6f835e6c748fd68ec7624992478b16c1", "tree": "5d094a00fbc90455bd9b53e042cf8b4fe8433462", "parents": [ "ba118827465d12177f3996e50133960087b1c916" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Jul 06 14:55:02 2017 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Jul 20 16:33:00 2017 +0100" }, "message": "Hash-based DexCache methods array.\n\nTotal boot*.art size for aosp_angler-userdebug:\n - arm64:\n - before: 11603968\n - after: 10129408 (-1.4MiB, -12.7%)\n - arm:\n - before: 8626176\n - after: 7888896 (-0.7MiB, -8.5%)\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host\nTest: Nexus 6P boots.\nTest: testrunner.py --target\nTest: Build aosp_mips64-eng\nBug: 30627598\nChange-Id: I7f858605de5f074cbd7f0d9c4c072fbd44aee28f\n" }, { "commit": "2b0dfe7b2d6539a8dded06104b11f5ddca3e277c", "tree": "1afb090cb2211e4f0aaae5d5782d5f950c3a0627", "parents": [ "6375a04cae864416499865453fecd2b50706b3b2" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Jul 10 15:31:30 2017 +0100" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Jul 10 16:45:30 2017 +0100" }, "message": "Fix over-allocation of DexCache field array.\n\nFix a facepalm in DexCacheArraysLayout that causes\nover-allocation of the DexCache fields array.\nWe should not waste valuable flash storage and\nvirtual address space with all these zeros.\n\nTotal boot*.art size for aosp_angler-userdebug:\n - arm:\n - before: 13316096\n - after: 11603968 (1.6MiB, -13%)\n - arm64:\n - before: 9486336\n - after: 8626176 (-0.8MiB, 9%)\nWe should have saved half of this with the original\nchange rather than regressing the other half.\n\nTest: m test-art-host-gtest\nTest: testrunner.py --host --interp-ac\nTest: testtunner.py --host --interpreter\nBug: 30627598\nChange-Id: Ied2ffc4fc8e569141fd6db45841163f7f071ddd0\n" }, { "commit": "365719c23d809e595cf320bfba40e76bb4e87940", "tree": "0939f0d8dc47723978a665fa11dd637f6976d521", "parents": [ "d6705a0586377f1b0d7d14d3abe2b270bb0adb18" ], "author": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Wed Mar 08 13:11:50 2017 +0000" }, "committer": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Tue May 16 01:58:03 2017 +0100" }, "message": "No need to lock when calling Thread.interrupted.\n\nAlso intrinsify the Thread.interrupted call.\n\nThe rationale behind this optimization is that the flag can only\nhave two values, and only self can set it to false.\n\nTest: libcore, jdwp, run-tests, 050-sync-test\nChange-Id: I5c2b43bf872ba0bfafcb54b2cfcd19181864bc4c\n" }, { "commit": "c6ea7d00ad069a2736f603daa3d8eaa9a1f8ea11", "tree": "a6d3332a8592fb806841314d55b206b573d37d86", "parents": [ "d68677c5fde1ace16ea58d65733776c954e7acb4" ], "author": { "name": "Andreas Gampe", "email": "agampe@google.com", "time": "Wed Feb 01 16:46:28 2017 -0800" }, "committer": { "name": "Andreas Gampe", "email": "agampe@google.com", "time": "Fri Apr 21 08:40:33 2017 -0700" }, "message": "ART: Clean up art_method.h\n\nClean up the header. Fix up other headers including the -inl file,\nin an effort to prune the include graph. Fix broken transitive\nincludes by making includes explicit. Introduce new -inl files\nfor method handles and reference visiting.\n\nTest: source build/envsetup.sh \u0026\u0026 lunch aosp_angler-userdebug \u0026\u0026 mmma art\nTest: source build/envsetup.sh \u0026\u0026 lunch aosp_mips64-userdebug \u0026\u0026 mmma art\nChange-Id: I8f60f1160c2a702fdf3598149dae38f6fa6bc851\n" }, { "commit": "f44d36c8423f81cbb5e9f55d8813e26ffa1a7f3b", "tree": "324b41485ce6c414c1a006c72cbcc5ed9f466138", "parents": [ "8d6768d47b66a688d35399d524ad5a5450e9d9d4" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Tue Mar 14 14:18:46 2017 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Tue Mar 14 19:03:20 2017 +0000" }, "message": "Revert^2 \"Hash-based DexCache field array.\"\n\nTest: testrunner.py --host --interpreter\nBug: 30627598\n\nThis reverts commit 6374c58f2ea403b3a05fb27376110fe4d0fc8e3f.\n\nChange-Id: I275508e288a85d3aa08f7405a1a4f362af43b775\n" }, { "commit": "8d6768d47b66a688d35399d524ad5a5450e9d9d4", "tree": "0b8733baef378d9e060c8e74319a3846489590b2", "parents": [ "fbe1516b91f2117bb9c7282aaea5537572f62dd7" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Tue Mar 14 10:13:21 2017 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Tue Mar 14 13:55:13 2017 +0000" }, "message": "Revert^6 \"Hash-based dex cache type array.\"\n\nFixed ImageWriter to write class table also if it contains\nonly boot class loader classes. Added a regression test and\nadded extra checks for debug-build to verify that dex cache\ntypes from app image are also in the class table. Removed\nsome unnecessary debug output.\n\nTest: 158-app-image-class-table\nBug: 34839984\nBug: 30627598\nBug: 34659969\n\nThis reverts commit 0b66d6174bf1f6023f9d36dda8538490b79c2e9f.\n\nChange-Id: I6a747904940c6ebc297f4946feef99dc0adf930c\n" }, { "commit": "0b66d6174bf1f6023f9d36dda8538490b79c2e9f", "tree": "1cc4d2ae868745a65fd0489a6fb2f5f2fc9e880f", "parents": [ "6374c58f2ea403b3a05fb27376110fe4d0fc8e3f" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Mar 13 14:50:04 2017 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Mar 13 16:08:01 2017 +0000" }, "message": "Revert^5 \"Hash-based dex cache type array.\"\n\nFor app images, ImageWriter does not add boot image\nclasses to the app image class table even though it\nkeeps them in the dex caches. The reason for that is\nunknown, the code looks OK.\n\nBug: 34839984\nBug: 30627598\nBug: 34659969\n\nAlso reverts \"Improve debugging output for a crash.\"\n\nThis reverts commits\n bfb80d25eaeb7a604d5dd25a370e3869e96a33ab,\n 8dd56fcb3196f466ecaffd445397cb11ef85f89f.\n\nTest: testrunner.py --host\nChange-Id: Ic8db128207c07588c7f11563208ae1e85c8b0e84\n" }, { "commit": "6374c58f2ea403b3a05fb27376110fe4d0fc8e3f", "tree": "0aacaaf1f193e96b55e0b203b678ae724c0fb6f5", "parents": [ "8f323e09e692ff4f95f40300391fe41fb96a6c49" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Mar 13 14:51:19 2017 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Mar 13 14:52:17 2017 +0000" }, "message": "Revert \"Hash-based DexCache field array.\"\n\nReverting to allow rebasing the revert\n https://android-review.googlesource.com/351689\nwithout too many conflicts.\n\nBug: 30627598\n\nThis reverts commit 1aea3510b8dd0c512cec61c91c5ef1f1e5d53d64.\n\nChange-Id: I4af65e9f41c8bad8106c028947eca7c5a9534c53\n" }, { "commit": "1aea3510b8dd0c512cec61c91c5ef1f1e5d53d64", "tree": "94cbab7c3097ce7d3a1feb1a69f28406644af085", "parents": [ "d1d4530ffa97729aa8944932a7ac2009ae51c7e3" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Dec 08 11:39:42 2016 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Fri Mar 10 12:59:37 2017 +0000" }, "message": "Hash-based DexCache field array.\n\nTest: m test-art-host, both AOT and interpreter\nTest: m test-art-target, both AOT and interpreter\nTest: m valgrind-test-art-host\nBug: 30627598\nChange-Id: If992f091aadd862d17b09928d21659573dd285a0\n" }, { "commit": "331605a7ba842573b3876e14c933175382b923c8", "tree": "8a660cc7d1563272e747ac051b3e3e3dd2b90070", "parents": [ "ec280c9a8f9c192eb6ea74fba40bfa1b668c14c1" ], "author": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Wed Mar 01 11:01:41 2017 +0000" }, "committer": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Wed Mar 01 11:32:02 2017 +0000" }, "message": "Revert \"Revert \"Intrinsify Integer.valueOf.\"\"\n\nFix heap poisoning.\nLOG INFO instead of ERROR to avoid run-test failures with --no-image.\n\nbug:30933338\nTest: ART_HEAP_POISONING\u003dtrue test-art-host test-art-target\n\nThis reverts commit db7b44ac3ea80a722aaed12e913ebc1661a57998.\n\nChange-Id: I0b7d4f1eb11c62c9a3df8e0de0b1a5d8af760181\n" }, { "commit": "db7b44ac3ea80a722aaed12e913ebc1661a57998", "tree": "ef5e6236203e04b59151b2e1b1529f9b389957b4", "parents": [ "cd0b27287843cfd904dd163056322579ab4bbf27" ], "author": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Tue Feb 28 17:04:50 2017 +0000" }, "committer": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Tue Feb 28 17:04:50 2017 +0000" }, "message": "Revert \"Intrinsify Integer.valueOf.\"\n\nHeap poisoning missing\njit-gcstress not optimizing it.\n\nbug:30933338\n\nThis reverts commit cd0b27287843cfd904dd163056322579ab4bbf27.\n\nChange-Id: I5ece1818afbca5214babb6803f62614a649aedeb\n" }, { "commit": "cd0b27287843cfd904dd163056322579ab4bbf27", "tree": "e5e1f0a8cae1f8d604123a72e3377528e4e4f333", "parents": [ "0fc3e418428f9f29a52c1dd60b1d86c71eb00d18" ], "author": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Thu Feb 23 16:18:41 2017 +0000" }, "committer": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Mon Feb 27 23:13:20 2017 +0000" }, "message": "Intrinsify Integer.valueOf.\n\nImproves performance of ArrayListStress and Ritz by ~10% and ~3%.\n\nTest: test-art-host test-art-target\nbug: 30933338\n\nChange-Id: I639046e3a18dae50069d3a7ecb538a900bb590a1\n" }, { "commit": "bcf716f3e0c72aa8bb75b15e5b273a813658fdb5", "tree": "92d55371dc3859f4be6c05eb3fcd2b4670057b47", "parents": [ "09c8235da86ce1d01b4dfb38a5af1481d13efe80" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Feb 23 10:43:09 2017 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Fri Feb 24 11:07:53 2017 +0000" }, "message": "Revert^2 \"Enable String compression.\"\n\nTest: m test-art-host (interpreter, JIT, AOT)\nTest: m test-art-target on Nexus 6P (interpreter, JIT, AOT)\nTest: Nexus 6P boots\nTest: m ahat-test\nBug: 31040547\n\nThis reverts commit afcb365b2d3ae152989826c2afb0d98a9cf3f6be.\n\nChange-Id: Id7e27d717c8e36d3c7abdde59ca2c8fd4fc221c9\n" }, { "commit": "bfb80d25eaeb7a604d5dd25a370e3869e96a33ab", "tree": "96068dd9b9b3bc2e1bcc25d38bbd2420d5b0f1bb", "parents": [ "7a49d93ee265b978ab29cae659e46d41da78459a" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Tue Feb 14 14:08:12 2017 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Feb 20 14:13:14 2017 +0000" }, "message": "Revert^4 \"Hash-based dex cache type array.\"\n\nAdded extra output to the abort message to collect more data\nwhen we hit the crash. Added extra check when loading an app\nimage to verify that the class table isn\u0027t already broken.\n\nTest: testrunner.py --host\nBug: 34839984\nBug: 30627598\nBug: 34659969\n\nThis reverts commit 5812e20ff7cbc8efa0b8d7486ada2f58840a6ad5.\n\nChange-Id: I9bb442a184c236dcb75b3e42a095f39cd6bee59d\n" }, { "commit": "afcb365b2d3ae152989826c2afb0d98a9cf3f6be", "tree": "abba45a6314d4b83a1e1d004019b5c9bc360b34d", "parents": [ "24bbf98cec9f39a2592892adb245ce15d58cab19" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Feb 16 21:46:27 2017 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Feb 16 21:46:27 2017 +0000" }, "message": "Revert \"Enable String compression.\"\n\nThe internal API String.setCharAt() is broken\nwith string compression.\n\nBug: 31040547\n\nThis reverts commit 24bbf98cec9f39a2592892adb245ce15d58cab19.\n\nChange-Id: If2c2fa0eb42a8613994dfa5299febf7a5de2bf31\n" }, { "commit": "24bbf98cec9f39a2592892adb245ce15d58cab19", "tree": "176def9d02c8bb1d995b69e9b23f845aeeb96c15", "parents": [ "ab58a629d6cfe75f0ed34326c2ca6364b2203885" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Thu Feb 02 17:48:20 2017 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Feb 13 13:31:04 2017 +0000" }, "message": "Enable String compression.\n\nTest: m test-art-host (interpreter, JIT, AOT)\nTest: m test-art-target on Nexus 6P (interpreter, JIT, AOT)\nTest: Nexus 6P boots\nTest: m ahat-test\nBug: 31040547\nChange-Id: I2c2f74ca147b3ea98ef06f9c64553737d17d42fd\n" }, { "commit": "d16363a93053de0f32252c7897d839a46aff14ae", "tree": "8823c6dd4641eae333aa7001203460757a515010", "parents": [ "ac240396125911b52437a460a400c8b38be43a96" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Feb 01 14:09:13 2017 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Wed Feb 01 15:27:37 2017 +0000" }, "message": "Revert \"Hash-based dex cache type array.\"\n\nReverting to work around some programs crashing with\n Check failed: handle_scope_iface.Get() !\u003d nullptr.\nthough the reason for the failure not yet understood.\n\nTest: m test-art-host\nBug: 34839984\nBug: 30627598\nBug: 34659969\nBug: 30419309\n\nThis reverts commit ec7862283dd49f5a58d0ac45960ce27c2f7671b8.\n\nChange-Id: Ifded663633082f1e59e5b6ff2e026dc559bd6b82\n" }, { "commit": "ec7862283dd49f5a58d0ac45960ce27c2f7671b8", "tree": "26d6dcc1d5ed4f0ba5ac15f17ef7377215684bf6", "parents": [ "c01d49091f4588777db5bf45345f388058caa99f" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Tue Dec 20 16:24:13 2016 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Mon Jan 30 10:21:16 2017 +0000" }, "message": "Hash-based dex cache type array.\n\nTest: m test-art-host (Interpreter, Optimizing, JIT)\nTest: m test-art-target on Nexus 6P (Interpreter, Optimizing, JIT)\nTest: Nexus 6P boots\nTest: m valgrind-test-art-host\nBug: 30627598\nBug: 34659969\nBug: 30419309\nChange-Id: Ic00eda89e58088a3573fc9ec0ad04c0e69e161d1\n" }, { "commit": "72ab684871f870aead76b23cb67deb046107b380", "tree": "d4bb40cab65df449246b35ae582234238676b7bb", "parents": [ "e36c51aee58e61e9fc89851b767379c587f050e3" ], "author": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Fri Jan 20 19:32:50 2017 +0000" }, "committer": { "name": "Vladimir Marko", "email": "vmarko@google.com", "time": "Tue Jan 24 12:05:02 2017 +0000" }, "message": "Add class status for resolved erroneous classes.\n\nSplit the old ambiguous status mirror::Class::kStatusError\ninto kStatusErrorUnresolved and kStatusErrorResolved. Once\na class has been resolved, IsResolved() shall return true\neven if the class later becomes erroneous. Allow returning\nerroneous class from ClassLinker::EnsureResolved() if it has\nbeen previously resolved. This allows consistent behavior\nfor retrieving classes, immune to multi-threaded races and\nmulti-dex weirdness. It also allows JVMTI to properly report\n\"prepared\" (i.e. resolved) classes that are also erroneous.\n\nThe new behavior is consistent with the RI.\n\nAdd regression tests to 008-exceptions for inconsistent\nbehavior for multi-dex retrieval of erroneous resolved class\n(wrapping or not wrapping the old exception based on which\ndex file is used for lookup) and for a CHECK(IsResolved())\ncrash in ClassLinker::LoadSuperAndInterfaces() (without any\ntests for similar checks that could have previously failed\nonly due to extremely unlikely race conditions; these should\nnow also be fixed).\n\nInconsistency still remains for class verification as shown\nby the new exceptionsForSuperClassInitFailure() test in\n008-exceptions, where interpreter and Optimizing still\ncause different exceptions to be thrown.\n\nNote: This is partially changing behavior implemented for\nbug 28787733. Since we allow the class loader to retrieve an\nerroneous resolved class, the ExceptionInInitializerError is\nnot thrown at all from VMClassLoader_findLoadedClass(), so\nthere is nothing to wrap in ClassNotFoundException.\n\nTest: m test-art-host\nBug: 30627598\nBug: 28787733\nChange-Id: I86cdca00f35a0d6221d2559e3026ac0428a3613c\n" }, { "commit": "9cd522b50a5f0209e97a2e6bcf8c9c350d3f3f83", "tree": "503af07a0a6122d161a81dfed2d2e64f8e501d5a", "parents": [ "c3b7bf3bcc55ec3c684210176ff6ddeb9d33ac19" ], "author": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Wed Jan 18 17:05:59 2017 +0000" }, "committer": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Wed Jan 18 17:08:57 2017 +0000" }, "message": "Remove now unused dex_cache_resolved_types_ from ArtMethod.\n\nTest: test-art-host test-art-target\nChange-Id: I004c040da165934f74eb8b8e85e99e7c6621c2df\n" }, { "commit": "f9bf250cc642c4e3e0c6904550fe922d3359a737", "tree": "953351c7e13b45e9246625ad81bce708788d96f9", "parents": [ "b73e659be3fb6474ff3a993cd25ecbfa1cc4715c" ], "author": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Wed Dec 14 14:59:04 2016 +0000" }, "committer": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Thu Dec 15 14:52:22 2016 +0000" }, "message": "Remove dex cache string from mirror::Class.\n\nThe compiled code does not need it anymore.\n\nTest: test-art-host test-art-target\nChange-Id: I71c08ce12d450ab2b2cd018330d42a8b8c9e3da2\n" }, { "commit": "71bf7b43380eb445973f32a7f789d9670f8cc97d", "tree": "76425f8578b247a845fe61724d71efc63059760f", "parents": [ "52f52361c2ea37941fbda71e3c653bb8096bf516" ], "author": { "name": "Aart Bik", "email": "ajcbik@google.com", "time": "Wed Nov 16 10:17:46 2016 -0800" }, "committer": { "name": "Aart Bik", "email": "ajcbik@google.com", "time": "Wed Nov 30 13:21:21 2016 -0800" }, "message": "Optimizations around escape analysis. With tests.\n\nDetails:\n(1) added new intrinsics\n(2) implemented optimizations\n more !can be null information\n more null check removals\n replace return-this uses with incoming parameter\n remove dead StringBuffer/Builder calls (with escape analysis)\n(3) Fixed exposed bug in CanBeMoved()\n\nPerformance gain:\nThis improves CafeineString by about 360%\n(removes null check from first loop, eliminates second loop completely)\n\nTest: test-art-host\n\nChange-Id: Iaf16a1b9cab6a7386f43d71c6b51dd59600e81c1\n" }, { "commit": "ff7d89c0364f6ebd0f0798eb18ef8bd62917de6a", "tree": "9a8401416b499f815731fd38dec219a1e49b623d", "parents": [ "8acdebc1a1b5821d3bb8e9461f7877fc234a37ff" ], "author": { "name": "Aart Bik", "email": "ajcbik@google.com", "time": "Mon Nov 07 08:49:28 2016 -0800" }, "committer": { "name": "Aart Bik", "email": "ajcbik@google.com", "time": "Wed Nov 09 09:32:36 2016 -0800" }, "message": "Allow read side effects for removing dead instructions.\n\nRationale:\nInstructions that only have the harmless read side effect may\nbe removed when dead as well, we were too strict previously.\nAs proof of concept, this cl also provides more accurate information\non a few string related intrinsics. This removes the dead indexOf\nfrom CaffeineString (17% performance improvement, big bottleneck\nof the StringBuffer\u0027s toString() still remains in loop).\n\nTest: test-art-host\nChange-Id: Id835a8e287e13e1f09be6b46278a039b8865802e\n" }, { "commit": "762869dee6e0eadab5be1c606792d6693bbabf4e", "tree": "8c986c621e8a5f3cf4e4e3b2cc13b400401ad89b", "parents": [ "b4cf427734c6839b46d0d6037e3189a5e8aa1bdb" ], "author": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Fri Jul 15 15:28:35 2016 +0100" }, "committer": { "name": "Nicolas Geoffray", "email": "ngeoffray@google.com", "time": "Fri Sep 30 10:44:31 2016 +0100" }, "message": "Simplify our intrinsic recognizer.\n\n- Use the modifiers for storing the intrinsic kind.\n- Delete dex_file_method_inliner and its associated map.\n\nThis work was also motivated by the fact that the inline\nmethod analyzer leaks intrinsic tables, and even worse, might re-use\na table from one dex file to another unrelated dex file in the presence\nof class unloading and the unlikely event of the dex files getting\nthe same address.\n\ntest: m test-art-host m test-art-target\n\nChange-Id: Ia653d2c72df13889dc85dd8c84997582c034ea4b\n" }, { "commit": "bda1d606f2d31086874b68edd9254e3817d8049c", "tree": "db07417935fe72e99c3da60152e13f0620c7d8d7", "parents": [ "d14d515df39cd963179088b8721768f9645243aa" ], "author": { "name": "Andreas Gampe", "email": "agampe@google.com", "time": "Mon Aug 29 17:43:45 2016 -0700" }, "committer": { "name": "Andreas Gampe", "email": "agampe@google.com", "time": "Thu Sep 08 10:13:47 2016 -0700" }, "message": "ART: Detach libart-disassembler from libart\n\nSome more intrusive changes than I would have liked, as long as\nART logging is different from libbase logging.\n\nFix up some includes.\n\nBug: 15436106\nBug: 31338270\nTest: m test-art-host\nChange-Id: I9fbe4b85b2d74e079a4981f3aec9af63b163a461\n" }, { "commit": "542451cc546779f5c67840e105c51205a1b0a8fd", "tree": "11e09bb5abaee12dddffefbe7e425291076dfa7a", "parents": [ "85c4a4b8c9eabfe16e4e49f9b4aa78c1bf4be023" ], "author": { "name": "Andreas Gampe", "email": "agampe@google.com", "time": "Tue Jul 26 09:02:02 2016 -0700" }, "committer": { "name": "Andreas Gampe", "email": "agampe@google.com", "time": "Mon Aug 01 18:54:48 2016 -0700" }, "message": "ART: Convert pointer size to enum\n\nMove away from size_t to dedicated enum (class).\n\nBug: 30373134\nBug: 30419309\nTest: m test-art-host\nChange-Id: Id453c330f1065012e7d4f9fc24ac477cc9bb9269\n" }, { "commit": "a62cb9bb6cb2278cb41ab0664191623e178c6a4f", "tree": "62263dac644e2d80a34a04f2649e4741a1bed6f4", "parents": [ "bb8d501c9bb882a8927c6ceda07bf9577e06c3e1" ], "author": { "name": "Artem Udovichenko", "email": "artem.u@samsung.com", "time": "Thu Jun 30 09:18:25 2016 +0000" }, "committer": { "name": "Artem Udovichenko", "email": "artem.u@samsung.com", "time": "Thu Jul 07 10:01:04 2016 +0000" }, "message": "Revert \"Revert \"Optimize IMT\"\"\n\nThis reverts commit 88f288e3564d79d87c0cd8bb831ec5a791ba4861.\n\nChange-Id: I49605d53692cbec1e2622e23ff2893fc51ed4115\n" } ], "next": "88f288e3564d79d87c0cd8bb831ec5a791ba4861" }