)]}'
{
  "commit": "ddb9fa78b5d97e4107a9404f4146140ef8bf9a5a",
  "tree": "98b6388643bdb4d90034fa4d2804745f64c2e837",
  "parents": [
    "7d5c5ebd6717378e58405a002b8c3e63f532e99b"
  ],
  "author": {
    "name": "Barret Rhoden",
    "email": "brho@cs.berkeley.edu",
    "time": "Wed Apr 01 19:55:53 2015 -0400"
  },
  "committer": {
    "name": "Barret Rhoden",
    "email": "brho@cs.berkeley.edu",
    "time": "Tue Apr 07 15:06:59 2015 -0400"
  },
  "message": "x86: EPT and KPT are contiguous\n\nRon had a great idea: alloc two contig pages for every KPT intermediate\npage table, and have the corresponding EPT page be the adjacent physical\npage.  The downside is the memory waste for non-VMM processes and the\npressure on the memory allocator.  The upside is the code is much\nsimpler and memwalks are faster, since we can easily find an EPTE given\na KPTE.\n\nThis commit does the two-page alloc for the intermediate pages.  The\nexisting EPT code won\u0027t actually work right now, esp since the pm ops\ndon\u0027t do EPT work.\n\nThere\u0027s some ugliness with the boot page directories.  I can\u0027t put and\nalign 8192 in entry64.S and have grub boot the damn thing.  So I put in\nsome dirty hacks to special case the boottime kernel mappings.\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "a82c6f0c9d1531583a0b36dbc1aaaf3470107ed6",
      "old_mode": 33188,
      "old_path": "kern/arch/x86/pmap64.c",
      "new_id": "658df50026155f1a448e4f5048e149b1a180a75e",
      "new_mode": 33188,
      "new_path": "kern/arch/x86/pmap64.c"
    },
    {
      "type": "modify",
      "old_id": "ca0fba57b92820b984e347ddf0da76a03865e0dc",
      "old_mode": 33188,
      "old_path": "kern/arch/x86/vmm/vmm.c",
      "new_id": "53961be3bdd108e1868ec61ca130146ff192cf67",
      "new_mode": 33188,
      "new_path": "kern/arch/x86/vmm/vmm.c"
    }
  ]
}