)]}'
{
  "commit": "9858893ea8457d9d9d822f47eab08081f4caebbe",
  "tree": "b924a8865fa6bd89049a6d8c8566c02de85271bd",
  "parents": [
    "cfba66aa78491265735e80f4114ff274ad765cff"
  ],
  "author": {
    "name": "Eugeniy Paltsev",
    "email": "Eugeniy.Paltsev@synopsys.com",
    "time": "Sat Nov 28 21:16:32 2020 +0300"
  },
  "committer": {
    "name": "Anas Nashif",
    "email": "anas.nashif@intel.com",
    "time": "Wed Dec 02 14:08:26 2020 -0500"
  },
  "message": "ARC: runner: mdb: tweak searching for cld process pid\n\nmdb binary starts several subproceses and one of them is cld process.\nIn runners/mdb.py we record process id of cld on each mdb launch\nto terminate simulator correctly later. However we can finish test\nand terminate mdb before the cld process was found (so cld won\u0027t\nbe terminated correctly by sanitycheck infrastructure). It may happen\nif we launch mdb on fast host machine.\n\nThat leads to several issues. First of all we get ugly error in\nsanitycheck output:\n------------------------\u003e8--------------------------------\nFileNotFoundError: [Errno 2] No such file or directory: \u0027/xxxx/mdb.pid\u0027\n------------------------\u003e8--------------------------------\n\nSecondly (and it\u0027s more important) we terminate simulator incorrectly.\nWe terminate mdb leaving cld process alive, running and consuming one\ncpu core permanently (until we kill it manually)\n\nSo, let\u0027s increase granularity of lookups and don\u0027t wait extra 0.5\nseconds before the first lookup.\n\nSigned-off-by: Eugeniy Paltsev \u003cEugeniy.Paltsev@synopsys.com\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "1e6df58c346c0f56b710e3ebb9d0f79c6eabcc8a",
      "old_mode": 33188,
      "old_path": "scripts/west_commands/runners/mdb.py",
      "new_id": "12c70e0cd8b908132a0fc8ca5f8427b2a5018f91",
      "new_mode": 33188,
      "new_path": "scripts/west_commands/runners/mdb.py"
    }
  ]
}