)]}'
{
  "commit": "ffd71da4e3f323b7673b061e6f7e0d0c12dc2b49",
  "tree": "edebe057b20e6684ab3dac06e53050595195fa63",
  "parents": [
    "cd80a8142efa3468c2cd9fb52845f334c3220d54"
  ],
  "author": {
    "name": "Ingo Molnar",
    "email": "mingo@elte.hu",
    "time": "Fri Mar 13 10:54:24 2009 +0100"
  },
  "committer": {
    "name": "Ingo Molnar",
    "email": "mingo@elte.hu",
    "time": "Fri Mar 13 11:06:47 2009 +0100"
  },
  "message": "panic: decrease oops_in_progress only after having done the panic\n\nImpact: eliminate secondary warnings during panic()\n\nWe can panic() in a number of difficult, atomic contexts, hence\nwe use bust_spinlocks(1) in panic() to increase oops_in_progress,\nwhich prevents various debug checks we have in place.\n\nBut in practice this protection only covers the first few printk\u0027s\ndone by panic() - it does not cover the later attempt to stop all\nother CPUs and kexec(). If a secondary warning triggers in one of\nthose facilities that can make the panic message scroll off.\n\nSo do bust_spinlocks(0) only much later in panic(). (which code\nis only reached if panic policy is relaxed that it can return\nafter a warning message)\n\nReported-by: Jan Beulich \u003cjbeulich@novell.com\u003e\nLKML-Reference: \u003c49B91A7E.76E4.0078.0@novell.com\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "32fe4eff1b89ecf75af7c229fc8db2e46179c924",
      "old_mode": 33188,
      "old_path": "kernel/panic.c",
      "new_id": "57fb005de546e46d454590c1246edb8921bcef44",
      "new_mode": 33188,
      "new_path": "kernel/panic.c"
    }
  ]
}