)]}'
{
  "commit": "f7e6779c382d2f6f68488662a6cbd6093764f821",
  "tree": "c55077887a961776bcf70c30967fde0146f5162d",
  "parents": [
    "6e1b0386ffa810c58aaed22ac24a0e187bc9d8dc"
  ],
  "author": {
    "name": "Barret Rhoden",
    "email": "brho@cs.berkeley.edu",
    "time": "Wed Mar 06 11:47:29 2019 -0500"
  },
  "committer": {
    "name": "Barret Rhoden",
    "email": "brho@cs.berkeley.edu",
    "time": "Wed Mar 06 11:47:29 2019 -0500"
  },
  "message": "Check event_queue pointer addresses\n\nWe had a couple checks in send_event(), mostly warnings and prints.\nThis commit changes those warnings to one panic, and pushes the error\nhandling closer to userspace.\n\nNote that there\u0027s not much we can do about a bad pointer in a syscall\nstruct, other than \"silently\" fail.  It\u0027s almost always a bad user bug,\nhence the informational print.\n\nAlso note that the kernel can still fault on an unmapped, paged out\naddress in userspace.  None of that is sorted.  This is just the sanity\nchecks to make sure we aren\u0027t give a kernel pointer.  It also catches\nany very low pointers, below the MMAP_LIM.  (So null pointers and\nfriends).\n\nReported-by: syzbot+ab67ae285c93a8d288a7@syzkaller.appspotmail.com\nSigned-off-by: Barret Rhoden \u003cbrho@cs.berkeley.edu\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "ed5fe3943ff2e8a198e718529b07101a40fb0504",
      "old_mode": 33188,
      "old_path": "kern/src/event.c",
      "new_id": "91243aa67fdd3a178a7b2f5d877bbc16cac1f80a",
      "new_mode": 33188,
      "new_path": "kern/src/event.c"
    },
    {
      "type": "modify",
      "old_id": "1b9161a3bf7de39b2925559f0125cdabbddf7155",
      "old_mode": 33188,
      "old_path": "kern/src/syscall.c",
      "new_id": "a596f753d79d2b84866e1d7c728ed6a854df7ef8",
      "new_mode": 33188,
      "new_path": "kern/src/syscall.c"
    }
  ]
}