)]}'
{
  "log": [
    {
      "commit": "7fd803cdfa873c01138dcbd614ec92418169b1c2",
      "tree": "7cf2172c26930b783a5e19106b35364c8964504d",
      "parents": [
        "0a2cb815974ea96af664fa966079966a08916722"
      ],
      "author": {
        "name": "Elliott Hughes",
        "email": "enh@google.com",
        "time": "Thu Feb 14 16:33:52 2013 -0800"
      },
      "committer": {
        "name": "Elliott Hughes",
        "email": "enh@google.com",
        "time": "Thu Feb 14 16:35:58 2013 -0800"
      },
      "message": "Fix the stack protector death test.\n\nNow __stack_chk_fail calls abort(3) directly, we terminate with\nSIGSEGV rather than SIGABRT. (Because of the workaround for the\ndebuggerd lossage in the abort(3) implementation, which was the\nmotivation for switching __stack_chk_fail over to abort(3).)\n\nAlso clarify the comment on the weird pthread death test, so it\ndoesn\u0027t get copied and pasted onto real death tests.\n\nChange-Id: Ie832eaded61359c99e7a10db65e28f35e8f63eed\n"
    },
    {
      "commit": "fb7eb5e07f43587c2bedf2aaa53b21fa002417bb",
      "tree": "2854e0357877216fe1d7e8722085a561b22b123b",
      "parents": [
        "c2d26ce7452e69b9492cd28b7c0271866798658b"
      ],
      "author": {
        "name": "Elliott Hughes",
        "email": "enh@google.com",
        "time": "Thu Feb 14 14:37:34 2013 -0800"
      },
      "committer": {
        "name": "Elliott Hughes",
        "email": "enh@google.com",
        "time": "Thu Feb 14 14:37:34 2013 -0800"
      },
      "message": "Simplify __stack_chk_fail, and fix it so we get debuggerd stack traces.\n\nBug: 2487269\nChange-Id: Iec5e470fc22cd9108404f634a9d4baa2c7b7f58f\n"
    },
    {
      "commit": "d3920b3a996b358e48232f417aa0a1e44a60f155",
      "tree": "b0520d0d300dbca1e6e54a9a7c26e6d2cd81ed08",
      "parents": [
        "f6afd3b670e23f56bf341d12136416aee17ea249"
      ],
      "author": {
        "name": "Elliott Hughes",
        "email": "enh@google.com",
        "time": "Thu Feb 07 18:39:34 2013 -0800"
      },
      "committer": {
        "name": "Elliott Hughes",
        "email": "enh@google.com",
        "time": "Fri Feb 08 11:16:13 2013 -0800"
      },
      "message": "Switch to using AT_RANDOM for the stack guards.\n\nBug: 7959813\nChange-Id: I8db4b8912ba649bfe668c6f22aa44690ddd401a2\n"
    },
    {
      "commit": "36bd371e26c716cbc18e11801b13eff0352d91b0",
      "tree": "b63e06f7a65f3ac2df09449bb65172644f230386",
      "parents": [
        "ba117e4172fe6f160bf5f4d58b37e12c08c34245"
      ],
      "author": {
        "name": "Nick Kralevich",
        "email": "nnk@google.com",
        "time": "Wed Jan 16 13:13:22 2013 -0800"
      },
      "committer": {
        "name": "Nick Kralevich",
        "email": "nnk@google.com",
        "time": "Wed Jan 16 13:16:42 2013 -0800"
      },
      "message": "Revert \"stack protector: use AT_RANDOM\"\n\nThe AT_RANDOM changes broke setuid / setgid executables\nsuch as \"ping\". When the linker executes a setuid program,\nit cleans the environment, removing any invalid environment\nentries, and adding \"NULL\"s to the end of the environment\narray for each removed variable. Later on, we try to determine\nthe location of the aux environment variable, and get tripped\nup by these extra NULLs.\n\nReverting this patch will get setuid executables working again,\nbut getauxval() is still broken for setuid programs because of\nthis bug.\n\nThis reverts commit e3a49a8661125f24aec8a1453e54b3b78005e21e.\n\nChange-Id: I05c58a896b1fe32cfb5d95d43b096045cda0aa4a\n"
    },
    {
      "commit": "e3a49a8661125f24aec8a1453e54b3b78005e21e",
      "tree": "a6a1be7dbdf1901db051fcfa1bd19f823e7a7f2b",
      "parents": [
        "14e1975e13c197180ed0481f305f83a362b16a24"
      ],
      "author": {
        "name": "Nick Kralevich",
        "email": "nnk@google.com",
        "time": "Mon Jan 14 14:46:26 2013 -0800"
      },
      "committer": {
        "name": "Nick Kralevich",
        "email": "nnk@google.com",
        "time": "Wed Jan 16 10:09:52 2013 -0800"
      },
      "message": "stack protector: use AT_RANDOM\n\nPopulate the stack canaries from the kernel supplied\nAT_RANDOM value, which doesn\u0027t involve any system calls.\nThis is slightly faster (6 fewer syscalls) and avoids\nunnecessarily reading /dev/urandom, which depletes entropy.\n\nBug: 7959813\n\nChange-Id: If2b43100a2a9929666df3de56b6139fed969e0f1\n"
    },
    {
      "commit": "dcab1b2c76a498c56bc00024613386de8b4b2aae",
      "tree": "6d7eb1ea166186df3092552b190b4485658b3e68",
      "parents": [
        "bfde0b6fd9e5de545746ab963d3a05ed2a8014f6"
      ],
      "author": {
        "name": "Nick Kralevich",
        "email": "nnk@google.com",
        "time": "Thu Jan 10 17:12:29 2013 -0800"
      },
      "committer": {
        "name": "Nick Kralevich",
        "email": "nnk@google.com",
        "time": "Fri Jan 11 10:52:36 2013 -0800"
      },
      "message": "Add stack canaries / strcpy tests.\n\nAdd a test to ensure that stack canaries are working\ncorrectly. Since stack canaries aren\u0027t normally generated\non non-string functions, we have to enable stack-protector-all.\n\nAdd a test to ensure that an out of bounds strcpy generates\na runtime failure.\n\nChange-Id: Id0d3e59fc4b9602da019e4d35c5c653e1a57fae4\n"
    },
    {
      "commit": "ad88a0863110798cef5169dcf917e18b967a7cf6",
      "tree": "9b5fcbe6ebc81b4fb781dd788b8795a334def653",
      "parents": [
        "d0f2b7e7e65f19f978c59abcbb522c08e76b1508"
      ],
      "author": {
        "name": "Elliott Hughes",
        "email": "enh@google.com",
        "time": "Wed Oct 24 18:37:21 2012 -0700"
      },
      "committer": {
        "name": "Elliott Hughes",
        "email": "enh@google.com",
        "time": "Thu Oct 25 12:04:03 2012 -0700"
      },
      "message": "Per-thread -fstack-protector guards for x86.\n\nBased on a pair of patches from Intel:\n\n  https://android-review.googlesource.com/#/c/43909/\n  https://android-review.googlesource.com/#/c/44903/\n\nFor x86, this patch supports _both_ the global that ARM/MIPS use\nand the per-thread TLS entry (%gs:20) that GCC uses by default. This\nlets us support binaries built with any x86 toolchain (right now,\nthe NDK is emitting x86 code that uses the global).\n\nI\u0027ve also extended the original tests to cover ARM/MIPS too, and\nbe a little more thorough for x86.\n\nChange-Id: I02f279a80c6b626aecad449771dec91df235ad01\n"
    }
  ]
}