)]}'
{
  "commit": "a6dfdb4e1cd8ea0cf3e8588b97a0c33f8ad3fa4b",
  "tree": "51ea5a0c4efcccf7549a221241ef37a6d84329af",
  "parents": [
    "149d6ad83663b4820ca09c9d40b1eea7f5c22c2b",
    "e4788b800f66718f0736bca385da548e48d39b8e"
  ],
  "author": {
    "name": "David S. Miller",
    "email": "davem@davemloft.net",
    "time": "Wed Nov 09 21:20:01 2016 -0500"
  },
  "committer": {
    "name": "David S. Miller",
    "email": "davem@davemloft.net",
    "time": "Wed Nov 09 21:20:01 2016 -0500"
  },
  "message": "Merge branch \u0027PHC-freq-fine-tuning\u0027\n\nRichard Cochran says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nPHC frequency fine tuning\n\nThis series expands the PTP Hardware Clock subsystem by adding a\nmethod that passes the frequency tuning word to the the drivers\nwithout dropping the low order bits.  Keeping those bits is useful for\ndrivers whose frequency resolution is higher than 1 ppb.\n\nThe appended script (below) runs a simple demonstration of the\nimprovement.  This test needs two Intel i210 PCIe cards installed in\nthe same PC, with their SDP0 pins connected by copper wire.  Measuring\nthe estimated offset (from the ptp4l servo) and the true offset (from\nthe PPS) over one hour yields the following statistics.\n\n|        |   Est. Before |    Est. After |   True Before |    True After |\n|--------+---------------+---------------+---------------+---------------|\n| min    | -5.200000e+01 | -1.600000e+01 | -3.100000e+01 | -1.000000e+00 |\n| max    | +5.700000e+01 | +2.500000e+01 | +8.500000e+01 | +4.000000e+01 |\n| pk-pk: | +1.090000e+02 | +4.100000e+01 | +1.160000e+02 | +4.100000e+01 |\n| mean   | +6.472222e-02 | +1.277778e-02 | +2.422083e+01 | +1.826083e+01 |\n| stddev | +1.158006e+01 | +4.581982e+00 | +1.207708e+01 | +4.981435e+00 |\n\nHere the numbers in units of nanoseconds, and the ~20 nanosecond PPS\noffset is due to input/output delays on the i210\u0027s external interface\nlogic.\n\nWith the series applied, both the peak to peak error and the standard\ndeviation improve by a factor of more than two.  These two graphs show\nthe improvement nicely.\n\n  http://linuxptp.sourceforge.net/fine-tuning/fine-est.png\n\n  http://linuxptp.sourceforge.net/fine-tuning/fine-tru.png\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n",
  "tree_diff": []
}