)]}'
{
  "commit": "301efc8cea4ca41b830314245005c505bf4eb0af",
  "tree": "a5ceff8dd83eda62e52d9bb9982807352768bae9",
  "parents": [
    "e2136d9c287f6c640441c539a60e17d92a977560"
  ],
  "author": {
    "name": "Brian Smith",
    "email": "brian@briansmith.org",
    "time": "Thu Nov 12 18:31:35 2015 -1000"
  },
  "committer": {
    "name": "Adam Langley",
    "email": "agl@google.com",
    "time": "Thu Nov 19 00:52:33 2015 +0000"
  },
  "message": "Fix error handling in |p256-x86_64|.\n\nThis makes similar fixes as were done in the following OpenSSL commits:\n\n    c028254b12a8ea0d0f8a677172eda2e2d78073f3: Correctly set Z_is_one on\n    the return value in the NISTZ256 implementation.\n\n    e22d2199e2a5cc9b243f45c2b633d1e31fadecd7: Error checking and memory\n    leak leak fixes in NISTZ256.\n\n    4446044a793a9103a4bc70c0214005e6a4463767: NISTZ256: set Z_is_one to\n    boolean 0/1 as is customary.\n\n    a4d5269e6d0dba0c276c968448a3576f7604666a: NISTZ256: don\u0027t swallow\n    malloc errors.\n\nThe fixes aren\u0027t exactly the same. In particular, the comments \"This is\nan unusual input, we don\u0027t guarantee constant-timeness\" and the changes\nto |ecp_nistz256_mult_precompute| (which isn\u0027t in BoringSSL) were\nomitted.\n\nChange-Id: Ia7bb982daa62fb328e8bd2d4dd49a8857e104096\nReviewed-on: https://boringssl-review.googlesource.com/6492\nReviewed-by: Adam Langley \u003cagl@google.com\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "e40374fe0b7c93f73c278cae66ca4c3c5aa0ecd5",
      "old_mode": 33188,
      "old_path": "crypto/ec/p256-x86_64.c",
      "new_id": "52dd634afab8990bc9057f18c1d7797f8f3b7d82",
      "new_mode": 33188,
      "new_path": "crypto/ec/p256-x86_64.c"
    }
  ]
}