)]}'
{
  "commit": "dbbfefda4e54165a7f8d1d565ea84ca8577dc413",
  "tree": "7e284e33a5b9fa868b1817eb6248a17a6c386a6e",
  "parents": [
    "4d2434a32e97097ceefd93f9d84af3477167bb9f"
  ],
  "author": {
    "name": "Ali Zhang",
    "email": "alizhang@fuchsia.infra.roller.google.com",
    "time": "Thu Oct 31 22:52:08 2024 +0000"
  },
  "committer": {
    "name": "Copybara-Service",
    "email": "copybara-worker@google.com",
    "time": "Thu Oct 31 15:53:43 2024 -0700"
  },
  "message": "[roll] Roll fuchsia [sestarnix] Add self profiling instrumentation\n\nAdd `profile_duration!(...)` instrumentation to the following SELinux\nentry points.\n\n1. SELinuxFS mounting and file-based APIs\n2. SELinux policy loading\n3. All LSM/security hooks, minus a few that are too trivial to\n   benchmark\n\nWhen used with profiled variant of the Starnix boot test, these\ninstrumentations provide the raw data to be analyzed to better\nunderstand the SELinux overhead incurred by the representative\nCUJ of booting unmodified Android.\n\nOriginal-Bug: b/366533140\nTest: Manually ran the profiled variant of the starnix boot test 5x\n      each with and without this CL and observed negligible profiling\n      overhead (+44ms of total ~20s boot time, or ~0.2%).\n\nOriginal-Reviewed-on: https://fuchsia-review.googlesource.com/c/fuchsia/+/1145956\nOriginal-Revision: 17fa0e8d3459ad3b62531df6afba62a06387ba7b\nGitOrigin-RevId: 825d570f06d5c7687b3860b04380c2b08abfd0f3\nChange-Id: Iad289fa1d39fcc5aeb522139ca3d40193ff622ae\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "e0fa037766e429a6381b41cd748966177d41ac36",
      "old_mode": 33188,
      "old_path": "stem",
      "new_id": "84032b3eda034c3558c470631f90fa7529fa6e7b",
      "new_mode": 33188,
      "new_path": "stem"
    }
  ]
}