)]}'
{
  "log": [
    {
      "commit": "7f0849fd113691e62af0400989936d3eff151e37",
      "tree": "166fa5afeed341e2685091c89d9d8418acb66e5d",
      "parents": [
        "a2b947e0a5ff4106e36f8110232131dab1356c25"
      ],
      "author": {
        "name": "Elliott Hughes",
        "email": "enh@google.com",
        "time": "Fri Aug 26 16:17:17 2016 -0700"
      },
      "committer": {
        "name": "Elliott Hughes",
        "email": "enh@google.com",
        "time": "Wed Sep 07 15:01:54 2016 -0700"
      },
      "message": "Fix sscanf/wcstod parsing of NaNs.\n\nThe parsefloat routines -- which let us pass NaNs and infinities on to\nstrto(f|d|ld) -- come from NetBSD.\n\nAlso fix LP64\u0027s strtold to return a NaN, and fix all the architectures\nto return quiet NaNs.\n\nAlso fix wcstof/wcstod/wcstold to use parsefloat so they support hex\nfloats.\n\nLots of new tests.\n\nBug: http://b/31101647\nChange-Id: Id7d46ac2d8acb8770b5e8c445e87cfabfde6f111\n"
    },
    {
      "commit": "f3ea093ebbfd9bd348ac16a46c96b637e3696c82",
      "tree": "2151b3c55819f47f9e61d7e7bd007de1047e10a9",
      "parents": [
        "b83bd7bc256d5121ec3ff1babc3bc08d7f04eea2"
      ],
      "author": {
        "name": "jzha136",
        "email": "jingwei.zhang@intel.com",
        "time": "Fri Jun 12 09:15:02 2015 -0700"
      },
      "committer": {
        "name": "Elliott Hughes",
        "email": "enh@google.com",
        "time": "Mon Mar 07 18:31:26 2016 -0800"
      },
      "message": "Add x86 optimization of rint functions and tests\n\nChange-Id: I5e7696ff9bcb1efc2625100ef8565b68dca2326c\nSigned-off-by: Jingwei Zhang \u003cjingwei.zhang@intel.com\u003e\n"
    },
    {
      "commit": "56b2b2916b72eb21352c7aed529e0deeb5582072",
      "tree": "1299d64a93aec1d48f3abe304a8db75668a9dad1",
      "parents": [
        "f9fb52ab309c97d5b2ef4f8e1daff3eed4cf4024"
      ],
      "author": {
        "name": "Jingwei Zhang",
        "email": "jingwei.zhang@intel.com",
        "time": "Tue Sep 02 21:39:14 2014 +0800"
      },
      "committer": {
        "name": "Mingwei Shi",
        "email": "mingwei.shi@intel.com",
        "time": "Sat Feb 14 00:56:46 2015 +0800"
      },
      "message": "Accuracy tests for libm\n\nThis patch adds more tests for math functions to address coverage\nissue of math functions discussed in:\nhttps://android-review.googlesource.com/#/c/49653/\nhttps://android-review.googlesource.com/#/c/94780/\n\nThese are data sets used in regression tests for the Intel the math library (libm). They were collected over a long period of testing various libm implementations.\nThe data sets contain function specific data (special and corner cases such as +/-0, maximum/minimum normalized numbers, +/-infinity, QNaN/SNaN, maximum/minimum denormal numbers, arguments that would produce close to overflow/underflow results, known hard-to-round cases, etc), implementation specific data (arguments close to table look-up values for different polynomial approximations, worst cases for range reduction algorithms) and other data with interesting bit patterns.\nThe reference values are computed with Maple and were converted into hexadecimal format.\n\nChange-Id: I7177c282937369eae98f25d02134e4fc3beadde8\nSigned-off-by: Jingwei Zhang \u003cjingwei.zhang@intel.com\u003e\nSigned-off-by: Mingwei Shi \u003cmingwei.shi@intel.com\u003e\n"
    },
    {
      "commit": "1b37ba2178d618221905e17436f38e0c5a8397f3",
      "tree": "bdd2e0a562113f74cedabc1e8428a43a201bfaae",
      "parents": [
        "695781b6f0419f82939176a6ec1a240300d9f036"
      ],
      "author": {
        "name": "Elliott Hughes",
        "email": "enh@google.com",
        "time": "Mon Nov 03 17:03:20 2014 -0800"
      },
      "committer": {
        "name": "Elliott Hughes",
        "email": "enh@google.com",
        "time": "Tue Nov 04 18:00:02 2014 -0800"
      },
      "message": "Improve math tests to allow a specific ulp bound.\n\nAt the moment our libm is only good enough for a 1 ulp bound on these tests,\nbut that\u0027s better than the 4 ulp bound you get from gtest by default.\n\nI\u0027m not really happy with the multiple structures and corresponding functions,\nbut at least they mean there\u0027s no duplication in the tests themselves, and it\nshould be easy enough for us to make further improvements in future.\n\nChange-Id: I004e12970332e1d9531721361d6c34f908cfcecc\n"
    }
  ]
}