)]}'
{
  "commit": "75e613cdc7bb2ba3795b1bc3ddf19476c767ba68",
  "tree": "14864c1ad72554fa1061dbefd2e97b549d0a2507",
  "parents": [
    "f79b1b146b52765ee38bfb91bb14eb850fa98017"
  ],
  "author": {
    "name": "Yinghai Lu",
    "email": "yinghai@kernel.org",
    "time": "Wed Jun 03 00:13:13 2009 -0700"
  },
  "committer": {
    "name": "Jesse Barnes",
    "email": "jbarnes@virtuousgeek.org",
    "time": "Thu Jun 04 11:31:13 2009 +0100"
  },
  "message": "x86/pci: fix mmconfig detection with 32bit near 4g\n\nPascal reported and bisected a commit:\n|\tx86/PCI: don\u0027t call e820_all_mapped with -1 in the mmconfig case\n\nwhich broke one system system.\n\nACPI: Using IOAPIC for interrupt routing\nPCI: MCFG configuration 0: base f0000000 segment 0 buses 0 - 255\nPCI: MCFG area at f0000000 reserved in ACPI motherboard resources\nPCI: Using MMCONFIG for extended config space\n\nit didn\u0027t have\nPCI: updated MCFG configuration 0: base f0000000 segment 0 buses 0 - 63\nanymore, and try to use 0xf000000 - 0xffffffff for mmconfig\n\nFor 32bit, mcfg_res-\u003eend could be 32bit only (if 64 resources aren\u0027t used)\nSo use end - 1 to pass the value in mcfg-\u003eend to avoid overflow.\n\nWe don\u0027t need to worry about the e820 path, they are always 64 bit.\n\nReported-by: Pascal Terjan \u003cpterjan@mandriva.com\u003e\nBisected-by: Pascal Terjan \u003cpterjan@mandriva.com\u003e\nTested-by: Pascal Terjan \u003cpterjan@mandriva.com\u003e\nSigned-off-by: Yinghai Lu \u003cyinghai@kernel.org\u003e\nCc: stable@kernel.org\nSigned-off-by: Jesse Barnes \u003cjbarnes@virtuousgeek.org\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "5fa10bb9604f5efe9c18f58fc21b95bea5c5bed2",
      "old_mode": 33188,
      "old_path": "arch/x86/pci/mmconfig-shared.c",
      "new_id": "8766b0e216c5f1a139a7094630933670dd3f1cc4",
      "new_mode": 33188,
      "new_path": "arch/x86/pci/mmconfig-shared.c"
    }
  ]
}