)]}'
{
  "commit": "4dcb05729fc4e0ea93326805bc5ddc2c2953530f",
  "tree": "2deb36e13f9652b763211d0943f29378862c00f7",
  "parents": [
    "25bde4c1f9b025deedfd3e70bf3f6f7f7694405e"
  ],
  "author": {
    "name": "Eric Roman",
    "email": "eroman@chromium.org",
    "time": "Tue Jan 20 12:19:13 2015 -0800"
  },
  "committer": {
    "name": "Adam Langley",
    "email": "agl@google.com",
    "time": "Tue Jan 20 23:44:05 2015 +0000"
  },
  "message": "Remove support for passing a C-String to PKCS5_PBKDF2_HMAC*().\n\nBefore it was possible to pass a NULL-terminated C-string to the PBKDF2\nfunctions, and indicate the parameter was a C-string by passing a length\nof -1.\n\nThis is not relied on anywhere in the BoringSSL code, and the API contract is\npossible to misuse as it is not the common way of doing things.\n\n(A problem would arise when passing in a large unsigned length that\nsubsequently gets interpreted as -1).\n\nChange-Id: Ifbd31ff76e183fa74e9fa346908daf4bfb8fc3da\nReviewed-on: https://boringssl-review.googlesource.com/2953\nReviewed-by: David Benjamin \u003cdavidben@chromium.org\u003e\nReviewed-by: Adam Langley \u003cagl@google.com\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "4ab10f45fc965386011f700f56a7ce61745a898f",
      "old_mode": 33188,
      "old_path": "crypto/evp/pbkdf.c",
      "new_id": "4a18b08731eaa696ee22427ad744793a08adea74",
      "new_mode": 33188,
      "new_path": "crypto/evp/pbkdf.c"
    },
    {
      "type": "modify",
      "old_id": "4859b8a357d7e2bb44e1a12f54a49fcb27ca18e3",
      "old_mode": 33188,
      "old_path": "include/openssl/evp.h",
      "new_id": "39da689394e0cbc3d6b7fe4be48b3fc2d1105f95",
      "new_mode": 33188,
      "new_path": "include/openssl/evp.h"
    }
  ]
}