)]}'
{
  "commit": "edec47cfef96209987cb7922286cb384916aae02",
  "tree": "72ba93a5e646a5b83526ecda1c5b16006c10c9b1",
  "parents": [
    "3737129917c918767cdb8acd8ca6b342c45fa154"
  ],
  "author": {
    "name": "Paolo Bonzini",
    "email": "pbonzini@redhat.com",
    "time": "Fri Jul 24 13:42:55 2015 +0200"
  },
  "committer": {
    "name": "Peter Maydell",
    "email": "peter.maydell@linaro.org",
    "time": "Mon Jul 27 17:12:19 2015 +0100"
  },
  "message": "main-loop: fix qemu_notify_event for aio_notify optimization\n\naio_notify can be optimized away, and in fact almost always will.  However,\nqemu_notify_event is used in places where this is incorrect---most notably,\nwhen handling SIGTERM.  When aio_notify is optimized away, it is possible that\nQEMU enters a blocking ppoll immediately afterwards and stays there, without\nreaching main_loop_should_exit().\n\nFix this by using a bottom half.  The bottom half can be optimized too, but\nscheduling it is enough for the ppoll not to block.  The hang is thus avoided.\n\nReported-by: Peter Maydell \u003cpeter.maydell@linaro.org\u003e\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\nReviewed-by: Stefan Hajnoczi \u003cstefanha@redhat.com\u003e\nMessage-id: 1437738175-23624-1-git-send-email-pbonzini@redhat.com\nSigned-off-by: Peter Maydell \u003cpeter.maydell@linaro.org\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "82875a4dfd5488dbc1755f9f13fe5d4945bc61d8",
      "old_mode": 33188,
      "old_path": "main-loop.c",
      "new_id": "39970437f89c6df99f89747c9f5d39b2478f24c3",
      "new_mode": 33188,
      "new_path": "main-loop.c"
    }
  ]
}