)]}'
{
  "commit": "014a602b86f08de96fc80ef3f96a87db6cccad56",
  "tree": "186b5eac9979ff0b46bd88d08b8f0d4331fe621c",
  "parents": [
    "2a2b2372420bf0373b932734e80a80c10adfdf14"
  ],
  "author": {
    "name": "Alan Modra",
    "email": "amodra@gmail.com",
    "time": "Wed May 24 12:47:09 2023 +0930"
  },
  "committer": {
    "name": "Alan Modra",
    "email": "amodra@gmail.com",
    "time": "Wed May 24 19:04:48 2023 +0930"
  },
  "message": "Don\u0027t optimise bfd_seek to same position\n\nIt\u0027s not worth avoiding an fseek to the same position, and can cause\nproblems if the linker\u0027s output file (which is opened \"w+\") is read,\nbecause that can result in writing, reading, then writing again.\n\nPOSIX.1-2017 (IEEE Std 1003.1) says of fopen:\n\"When a file is opened with update mode (\u0027+\u0027 as the second or third\ncharacter in the mode argument), both input and output may be\nperformed on the associated stream. However, the application shall\nensure that output is not directly followed by input without an\nintervening call to fflush() or to a file positioning function\n(fseek(), fsetpos(), or rewind()), and input is not directly followed\nby output without an intervening call to a file positioning function,\nunless the input operation encounters end-of-file.\"\n\n\t* bfdio.c (bfd_seek): Always call iovec-\u003ebseek.\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "75a3309c582079c93c3ae040b728e6b950415ede",
      "old_mode": 33188,
      "old_path": "bfd/bfdio.c",
      "new_id": "22c39a7b0ccfb0bd9ef29cea58ce1d699dec9d45",
      "new_mode": 33188,
      "new_path": "bfd/bfdio.c"
    }
  ]
}
