)]}'
{
  "commit": "84732c8bb2efe96bc87766737d29456a01b1d7e2",
  "tree": "d0483c9b49117f9d06f124e0a86d31870dd8485f",
  "parents": [
    "78a9d85dfa12ca6ce5c05613e622321c626f49ae"
  ],
  "author": {
    "name": "Peter Collingbourne",
    "email": "pcc@google.com",
    "time": "Wed Feb 11 02:07:17 2026 -0800"
  },
  "committer": {
    "name": "GitHub",
    "email": "noreply@github.com",
    "time": "Wed Feb 11 10:07:17 2026 +0000"
  },
  "message": "Replicate generic hardware events on all CPU PMUs (#2123)\n\nOn systems with more than one PMU for the CPUs (e.g. Apple M series SOCs),\ngeneric hardware events are only created for an arbitrary PMU. Usually\nthis is the big cluster\u0027s PMU, which can cause inaccuracies when the\nprocess is scheduled onto a little core. To fix this, teach PerfCounters\nto register generic hardware events on all CPU PMUs.\n\nCPU PMUs are identified using the same method as perf.",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "f47aa7b42d04014dc374a25559a8d83df472ddaf",
      "old_mode": 33188,
      "old_path": "src/perf_counters.cc",
      "new_id": "e6f220921c5ebcc3516ff26b77295b34fa7693ad",
      "new_mode": 33188,
      "new_path": "src/perf_counters.cc"
    },
    {
      "type": "modify",
      "old_id": "bf5eb6bc3aec9d82c17ec74015f6507661d854ee",
      "old_mode": 33188,
      "old_path": "src/perf_counters.h",
      "new_id": "4e45344318f10f7a7f739dd9bf71c2967bfe02af",
      "new_mode": 33188,
      "new_path": "src/perf_counters.h"
    },
    {
      "type": "modify",
      "old_id": "5de262fa2b44cd0a46ce7b95288f91252d572003",
      "old_mode": 33188,
      "old_path": "test/perf_counters_gtest.cc",
      "new_id": "6c923be897cd7a5c65f4057802cbf0829c0e1e04",
      "new_mode": 33188,
      "new_path": "test/perf_counters_gtest.cc"
    }
  ]
}