)]}'
{
  "log": [
    {
      "commit": "58f5ce146a8c2b25cc3bbd18df6e264addda2684",
      "tree": "cf2893c9d5da9aa989c88cf2fa8b14dad6d3861a",
      "parents": [
        "320910fcf1e7d7729f37f95f591093191414e73f"
      ],
      "author": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Wed Aug 12 14:29:10 2020 +0000"
      },
      "committer": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Thu Aug 20 01:03:06 2020 +0000"
      },
      "message": "aaudio: prevent onAudioDeviceUpdate past close\n\nTo prevent late device update callbacks, the AudioStream\nwas made into a RefBase AudioDeviceCallback.\nSo now we can use a smart pointer to delay deletion of the\nstream if any callbacks are in flight while the stream is closed.\n\nSo an AudioStream is now a RefBased object and can be used with\nandroid::sp. That required some changes to the way it is referenced\nin several places.\n\nMyPlayerBase was modified to use a weak pointer to the parent stream.\n\nBug: 161914201\nBug: 163165126\nBug: 164411271\nTest: see bug for repro of the crash\nTest: atest CtsNativeMediaAAudioTestCases\nChange-Id: Ic24aca3eaf5d1610504bb89c06001a37d0d4a1c3\n"
    },
    {
      "commit": "320910fcf1e7d7729f37f95f591093191414e73f",
      "tree": "e7052b3542898e68b5dbe5f5785ef0a19946bc2b",
      "parents": [
        "f356189f05884926290b1f7920241ef15674cbc6"
      ],
      "author": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Wed Aug 12 14:29:10 2020 +0000"
      },
      "committer": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Tue Aug 18 18:51:03 2020 +0000"
      },
      "message": "aaudio: fix crash from callbacks during close\n\nMove some code from release_l() into a new close_l().\nAlso stop callbacks before freeing memory in the\nFixedBlockReader.\n\nThe AudioTrack and AudioRecord callbacks can occur\nup until their destructors are called. This can lead\nto race conditions if the AAudio stream is dismantled\nwhile the AudioTrack or AudioRecord is still alive.\nThe AudioRecord was being deleted but not the AudioTrack.\nThat caused some streams to fail if they were using\na FixedBlockReader, which is used when the app\ncalls AAudioStreamBuilder_setFramesPerDataCallback().\n\nThere was also a problem with a few functions like\nAAudioStream_getFramesPerBurst() or AAudioStream_getTimestamp(),\nwhich would crash if called after AAudioStream_release( for\nINPUT streams.\n\nBug: 161914201\nBug: 163165126\nTest: see bug for repro of the crash\nTest: atest CtsNativeMediaAAudioTestCases\nChange-Id: If8f6f6f17ffe06eae98eb8b3930bca08c49a15f8\n"
    },
    {
      "commit": "7ebbc11051f1af7a520e9ae7e582125c63e92f95",
      "tree": "7605285baef0ae3c9e08a2a887af0fa4b4739628",
      "parents": [
        "05be7e510992ebf41919c84a26b38717a7f3211f"
      ],
      "author": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Wed May 13 15:55:17 2020 -0700"
      },
      "committer": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Mon May 18 16:23:28 2020 -0700"
      },
      "message": "aaudio: lock transport methods\n\nThe start/pause/stop/flush/close and other binder methods\nneed to be thread safe. They do not need to run\nin parallel. So a lock was added for each.\n\nWhere virtual methods are needed, the locked method calls\na corresponding _l submethod, eg. stop() calls stop_l().\n\nThe close logic was also simplified because the \"pending\"\ntechnique is not needed now that we have the locks.\nIt was only needed because a close could have occured\nwhile in the middle of another method.\n\nThis CL was merged with changes in RVC-DEV.\n\nBug: 153358911\nTest: adb logcat *:F\nTest: in another window:  test_binder_attack\nTest: There should be no fatal error in the logcat.\nTest: atest CtsNativeMediaAAudioTestCases\nChange-Id: I5920cf78af4501856756c5c2fc8e77758232508a\n"
    },
    {
      "commit": "d1f1cb68f567363a0b1e4cd4db6286afc6c8a1e2",
      "tree": "3b32985218612d05bf95fed08c18025d1735b2d4",
      "parents": [
        "06675ff30f63323905ff18c3c4e0d34a9fee4c60"
      ],
      "author": {
        "name": "jiabin",
        "email": "jiabin@google.com",
        "time": "Tue Mar 24 11:57:57 2020 -0700"
      },
      "committer": {
        "name": "jiabin",
        "email": "jiabin@google.com",
        "time": "Fri Mar 27 11:33:09 2020 -0700"
      },
      "message": "Add audio attributes info when starting aaudio tracks on MMAP mix.\n\nIn aaudio, when adding tracks to mixed stream, it is needed to update\naudio attributes to HAL whenever the tracks are added or removed.\nCurrently, only the audio attributes used to open the stream will be\nsent to the HAL. In that case, adding audio attributes of the clients\nwhen starting the stream can help solve the problem. In audio flinger,\nthe client\u0027s audio attributes will be used to create MmapTrack.\n\nTest: play multiple aaudio tracks, add log\nBug: 77279923\nChange-Id: Ic1c536049e194a2bb7513425ee4828d52769d27f\n"
    },
    {
      "commit": "8b4e05ecee60bf08c9121c61d69526e076495766",
      "tree": "9879d2a21175f35e47545c03ec1a2a2ba46de933",
      "parents": [
        "0bc6f60e1f036380bd48d022631e6561b90f210c"
      ],
      "author": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Tue Dec 17 12:12:09 2019 -0800"
      },
      "committer": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Wed Jan 22 13:31:58 2020 -0800"
      },
      "message": "aaudio: add AAudioStream_release()\n\nSplit close into close and release phases.\nRelease() will release hardware and service resources.\nClose() will delete the stream object.\nThis allows us to defer the deletion and avoid race conditions.\n\nThe CLOSING state is used to indicate that a stream has been released.\n\nAdded some test cases to test_various.cpp.\nA lone call to close() will automatically call release() so we should\nhave good code coverage for testing.\n\nBug: 136288001\nTest: adb shell test_various\nTest: atest CtsNativeMediaAAudioTestCases\nChange-Id: Ia648838f30c521ba101f55259dbcd4594c1263cd\n"
    },
    {
      "commit": "a53ffa67b35a7e57efec57811e4f915d1caac1eb",
      "tree": "bb234b23e9f5a749e490a84d9fe161195af8ece7",
      "parents": [
        "9f81192d77a62564dc1e2c7dad1afe601784a7b9"
      ],
      "author": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Wed Oct 10 16:21:37 2018 -0700"
      },
      "committer": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Sat Jan 19 15:43:55 2019 +0000"
      },
      "message": "aaudio: rename timestamp variables\n\nThere were several different timestamp variables with slightly\ndifferent meaning but with the same name.\nThat made it hard to debug the code.\n\nTest: atest CtsNativeMediaAAudioTestCases\nChange-Id: Iebf438676925c698377abe02c81f3d559829318a\n"
    },
    {
      "commit": "56d34f252ad803873fa4798b4cd0f4c6d28d2670",
      "tree": "9967806e97967f0ce9f06cabcb97ca0d2eea75d5",
      "parents": [
        "4cc873cc1dcf9e69572c2e52b68af1aebea0ad98"
      ],
      "author": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Thu Oct 11 13:30:56 2018 -0700"
      },
      "committer": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Thu Oct 11 13:38:12 2018 -0700"
      },
      "message": "aaudio: fix race condition in timestamp path\n\nThere was a potential race between setting the first timestamp\nand reading the first timestamp that could cause the timestamp service\nthread to exit. This would cause the client to not get any timestamps.\nThen the timing model would not advance and the stream would stall.\nThis could result in read() or write() returning zero,\nor a TIMEOUT error.\n\nThis was rare except for some devices that had different\ntiming in the HAL.\n\nBug: 68271891\nBug: 117557847\nTest: adb shell write_sine_callback -pl -m2 -n2 -s2 -c2 -l100\nTest: atest CtsNativeMediaAAudioTestCases\nChange-Id: Ib0974a576aaa17ff8de40899e60ed56015027bd7\n"
    },
    {
      "commit": "0127c1be07d63e046e136126cd465c3fa8977768",
      "tree": "4eb611ec5d010becd3583f93c640cc9468acb3b3",
      "parents": [
        "fdfae2898f0785f19de198ed00488e03bf52bca9"
      ],
      "author": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Thu Mar 29 13:48:06 2018 -0700"
      },
      "committer": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Fri Sep 07 11:08:24 2018 -0700"
      },
      "message": "aaudio: use new flowgraph to simplify processing\n\nConstruct a flowgraph based on the source and destination\nformat and channelCount. This is groundwork for supporting 24-bit\nPCM formats.\n\nAlso cleaned up handling of device related format.\n\nThis CL removes more code than it adds.\n\nBug: 65067568\nTest: write_sine_callback.cpp -pl\nTest: write_sine_callback.cpp -pl -x\nTest: input_monitor -pl\nTest: input_monitor -pl -x\nChange-Id: Ia155bff0164912011d09b61b54f983ccf4490bd1\n"
    },
    {
      "commit": "55e5eabf96af5a2964bfab1e64d81ad9be9ac975",
      "tree": "46127c87c1c2100d983c4bfbb611ec85bc6067b8",
      "parents": [
        "b8bddf5ece276f01052ad9b706417fd8cfe72d90"
      ],
      "author": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Tue Apr 10 15:16:38 2018 -0700"
      },
      "committer": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Fri Apr 13 18:14:20 2018 +0000"
      },
      "message": "aaudio: cleanup logs and add counters to dumpsys\n\nCounters will give us some history for AAudio.\nRemoved some unneeded logs.\nAdded log to help debug intermittent volume problem.\n\nBug: 64310586\nTest: adb shell logcat\nTest: adb shell dumpsys media.aaudio\nChange-Id: I012361bb4d0d0bfdedfe5ceef856b59282c8d3f0\n"
    },
    {
      "commit": "be645507a3ded6a481375419c8185e17c66dcb37",
      "tree": "d938ff05ee88f29968a6de963584e7b235f5f01a",
      "parents": [
        "142a00d21641393bb65a2484c0306f7c49ae9cb8"
      ],
      "author": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Thu Mar 22 13:48:18 2018 -0700"
      },
      "committer": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Fri Mar 23 11:10:30 2018 -0700"
      },
      "message": "aaudio: pass smart pointer to mixer thread\n\nBefore this, the mixer thread could close the stream using a stale pointer.\nThis could cause timeouts when plugging headphones in or out.\n\nBug: 76092458\nTest: write_sine_callback -m3 -pl -s1000\nTest: unplug and replug headphones many times on Marlin\nChange-Id: I437eb756686446a75fc5d04bc09d9408816585de\n"
    },
    {
      "commit": "4e1af9fc9c1108d4514e92774f750bcd434dbbcc",
      "tree": "03dc781c49ae6b37c4661c591ece2c5a4eed8cfe",
      "parents": [
        "a11f9357196db29e6f8f05b91334524de4f3207e"
      ],
      "author": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Wed Jan 03 15:54:35 2018 -0800"
      },
      "committer": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Thu Jan 18 19:48:46 2018 -0800"
      },
      "message": "aaudio: implement sessionId for effects\n\nIf a sessionID other than AAUDIO_SESSION_ID_NONE is requested\nthen legacy data path will be forced.\n\nSessionID is implemented for MMAP data path but not used.\nIt is easier to implement it now then try to figure\nit out later.\n\nBug: 33268927\nTest: test_session_id.cpp will be moved to CTS when stable\nChange-Id: I3ea67f57c1cbe24a512980f1352b917ab3cb6387\n"
    },
    {
      "commit": "a62fb9565d73d9ec9b56da65929afc0f686c696e",
      "tree": "f0291746c32af771178c350696f98bbac0f34c1f",
      "parents": [
        "d6be64a3f22be3eceb0d6209c88dcad573c08191"
      ],
      "author": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Tue Jan 16 12:44:06 2018 -0800"
      },
      "committer": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Tue Jan 16 17:14:59 2018 -0800"
      },
      "message": "aaudio: fix USAGE, INPUT_PRESET and CONTENT_TYPE\n\nThe attributes were not propagating all the way down for MMAP\nstreams.\n\nBug: 72046612\nBug: 72049948\nTest: adb shell aaudio_loopback -te -pl -n1 -x -X -i6\nChange-Id: I41b592d0c3a9daef5070ad299198c8d6cd890cfc\n"
    },
    {
      "commit": "2329638e097622653cb2c4acc145b996e317a38b",
      "tree": "d25dccdc184ac7ce8b7aceb227411592ff256484",
      "parents": [
        "10edfcf70857cc0a22afbbd3fc8f57f0c8a2953a"
      ],
      "author": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Mon Nov 20 15:45:11 2017 -0800"
      },
      "committer": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Tue Nov 21 10:30:58 2017 -0800"
      },
      "message": "aaudio: use xruns detected in the service\n\nThe service knows when an actual xrun occurs so we pass\nthat value to the client.\n\nBug: 38512417\nTest: write_sine_callback -pl -n1 -s20 -c2 -u\nChange-Id: I9e5ff1edd2e9378957309280162d7cda855f5adb\n"
    },
    {
      "commit": "be0a5b6e2c204446274248e1fa058e30f46b4b71",
      "tree": "ef61e6bc874fd89e0f2e5b60ffa96555a0656000",
      "parents": [
        "7de1920b41a7c5a6aa466dcd4fffc936193328b1"
      ],
      "author": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Thu Oct 12 15:56:00 2017 -0700"
      },
      "committer": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Tue Oct 17 12:15:07 2017 -0700"
      },
      "message": "aaudio service: prevent disconnected endpoints from being used\n\nAlso close the internal stream used by AAudioServiceEndpointShared.\n\nThis prevents a resource leak caused by an app not closing a\ndisconnected stream.\nIt also prevents an app from reopening a device that was just\ndisconnected.\n\nBug: 67664976\nTest: test_bad_disconnect.cpp should get new device ID after disconnect\nChange-Id: I9a234b5704d62af788064fdd352b63181ad7e559\n"
    },
    {
      "commit": "b92c988ffc48f093ab8615896b02b8c8e10e19b2",
      "tree": "32472c5cce0546223e63dcc76a6be9add080cc43",
      "parents": [
        "523b3047af443d0a51141ddabd328d52be5a23b0"
      ],
      "author": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Fri Sep 15 11:39:15 2017 -0700"
      },
      "committer": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Fri Sep 15 23:29:48 2017 +0000"
      },
      "message": "aaudio: fix audioserver crash when starting a stream is rejected\n\nBug: 65738271\nTest: repeat when not root:   input_monitor  -m2 -pl\nChange-Id: I98659dbd78a75785565b1b6f0254a48220253af6\n"
    },
    {
      "commit": "bcc3674648bc6f554d89a2a5d7721ed41c53f83b",
      "tree": "ae3f8fcecce7f06b44088ff4c8adf3248f287f4b",
      "parents": [
        "7328a80ffc83729cc6252c79e0851a31af59c653"
      ],
      "author": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Thu Aug 31 17:24:51 2017 -0700"
      },
      "committer": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Wed Sep 06 17:21:34 2017 -0700"
      },
      "message": "aaudio: fix timestamps and underruns\n\nStart the client after we get valid timing data from the server.\nThat can take a while because of the long cold start times.\nThe client is synced with the current position of the service.\nNow the client can start clean with no underruns.\n\nBug: 63918065\nTest: test_timestamps.cpp\nChange-Id: I5d01eb844e4b14cd5477d56ea1dd9e309abc1c52\n"
    },
    {
      "commit": "39f02ddfbfd9313370d862a6c4727826379a319a",
      "tree": "aa5c61a50fff59d92db99619edd67b31e9d938c3",
      "parents": [
        "ee079fe295c41f89c2ac6316e99ee10847fb78fd"
      ],
      "author": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Fri Aug 04 09:13:31 2017 -0700"
      },
      "committer": {
        "name": "Phil Burk",
        "email": "philburk@google.com",
        "time": "Thu Aug 24 14:59:25 2017 -0700"
      },
      "message": "aaudio: fix EXCLUSIVE mode interfering with SHARED\n\nThe MMAP endpoints were not tracked by the EndpointManager\nso it could not broker EXCLUSIVE access. To fix this the MMAP stream\nwas refactored to use a per-client stream and a per-device endpoint.\n\nCleanup close() of MMAP stream.\nAdd AAudioServiceEndpointShared.cpp.\nExtract AAudioServiceEndpointMMAP from AAudioServiceStreamMMAP.\nTrack MMAP endpoints so we can manage EXCLUSIVE and SHARED access.\n\nBug: 64494572\nBug: 64310586\nTest: see bug, use write_sine to play a shared stream and a excl stream\nChange-Id: I5053193abfd9b8a69a2f7e1110739d65e2af5d64\nMerged-In: I5053193abfd9b8a69a2f7e1110739d65e2af5d64\n"
    }
  ]
}