)]}'
{
  "commit": "edc01ab10f52135ec80e3589b6b4e0a9c65b27fd",
  "tree": "cdaa41759620d9d2bfb3a3c6fbde81bc148b6dd7",
  "parents": [
    "affc5009c36da4ac11accf49e08586537952f451"
  ],
  "author": {
    "name": "Martin Pluskal",
    "email": "martin@pluskal.org",
    "time": "Thu Jul 02 01:26:29 2026 +0200"
  },
  "committer": {
    "name": "GitHub",
    "email": "noreply@github.com",
    "time": "Wed Jul 01 16:26:29 2026 -0700"
  },
  "message": "test: make ValueTest/objects float check robust against x87 excess precision (#1700)\n\nValueTest/objects stores the float literal 0.12345f into a Json::Value\n(widened to the stored double) and then asserts equality against the\noriginal 0.12345f. On 32-bit x86 targets where GCC defaults to x87 math\n(-mfpmath\u003d387, 80-bit excess precision) and the baseline lacks SSE2, the\nround-trip yields the exact double 0.12345 rather than the float-widened\n0.12345000356435776, so the exact double comparison fails even though the\nlibrary itself is correct.\n\nCompare the stored value narrowed back to float via asFloat(): both the\nexpected literal and asFloat() collapse to the same float value on every\narchitecture, keeping the numeric round-trip check while making it\nprecision-robust.\n\nCo-authored-by: Jordan Bayles \u003cbayles.jordan@gmail.com\u003e",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "1c0377dee35e607174169608cb9873623d2e2bc3",
      "old_mode": 33188,
      "old_path": "src/test_lib_json/main.cpp",
      "new_id": "0d1c33064d9580f3f2715142611d9726a03171b3",
      "new_mode": 33188,
      "new_path": "src/test_lib_json/main.cpp"
    }
  ]
}