)]}'
{
  "log": [
    {
      "commit": "800d4d30c8f20bd728e5741a3b77c4859a613f7c",
      "tree": "1eb329fb8a1dce1c1432c07e19494a3e1fb68499",
      "parents": [
        "1747b21fecbfb63fbf6b9624e8b92707960d5a97"
      ],
      "author": {
        "name": "Yong Zhang",
        "email": "yong.zhang0@gmail.com",
        "time": "Sun Feb 20 15:08:14 2011 +0800"
      },
      "committer": {
        "name": "Ingo Molnar",
        "email": "mingo@elte.hu",
        "time": "Wed Feb 23 11:33:59 2011 +0100"
      },
      "message": "sched, autogroup: Stop going ahead if autogroup is disabled\n\nwhen autogroup is disable from the beginning,\nsched_autogroup_create_attach()\n  autogroup_move_group()                    \u003c\u003d\u003d 1\n    sched_move_task()                       \u003c\u003d\u003d 2\n      task_move_group_fair()\n        set_task_rq()\n          task_group()\n            autogroup_task_group()\n\nWe go the whole path without doing anything useful.\n\nThen stop going further if autogroup is disabled.\n\nBut there will be a race window between 1 and 2, in which\nsysctl_sched_autogroup_enabled is enabled. This issue\nwill be toke by following patch.\n\nSigned-off-by: Yong Zhang \u003cyong.zhang0@gmail.com\u003e\nSigned-off-by: Peter Zijlstra \u003ca.p.zijlstra@chello.nl\u003e\nCc: Mike Galbraith \u003cefault@gmx.de\u003e\nLKML-Reference: \u003c1298185696-4403-4-git-send-email-yong.zhang0@gmail.com\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\n"
    },
    {
      "commit": "f44937718ce3b8360f72f6c68c9481712517a867",
      "tree": "72b07f4daa686e10f90b666f5b61b66b30bbed2d",
      "parents": [
        "8ecedd7a06d27a31dbb36fab88e2ba6e6edd43ca"
      ],
      "author": {
        "name": "Mike Galbraith",
        "email": "efault@gmx.de",
        "time": "Thu Jan 13 04:54:50 2011 +0100"
      },
      "committer": {
        "name": "Ingo Molnar",
        "email": "mingo@elte.hu",
        "time": "Tue Jan 18 15:09:42 2011 +0100"
      },
      "message": "sched, autogroup: Fix CONFIG_RT_GROUP_SCHED sched_setscheduler() failure\n\nIf CONFIG_RT_GROUP_SCHED is set, __sched_setscheduler() fails due to autogroup\nnot allocating rt_runtime.  Free unused/unusable rt_se and rt_rq, redirect RT\ntasks to the root task group, and tell __sched_setscheduler() that it\u0027s ok.\n\nReported-and-tested-by: Bharata B Rao \u003cbharata@linux.vnet.ibm.com\u003e\nSigned-off-by: Mike Galbraith \u003cefault@gmx.de\u003e\nSigned-off-by: Peter Zijlstra \u003ca.p.zijlstra@chello.nl\u003e\nLKML-Reference: \u003c1294890890.8089.39.camel@marge.simson.net\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\n"
    },
    {
      "commit": "5091faa449ee0b7d73bc296a93bca9540fc51d0a",
      "tree": "55f5e96e189af65c85c769fce48627b8a5abb86b",
      "parents": [
        "822bc180a7f7a7bc5fcaaea195f41b487cc8cae8"
      ],
      "author": {
        "name": "Mike Galbraith",
        "email": "efault@gmx.de",
        "time": "Tue Nov 30 14:18:03 2010 +0100"
      },
      "committer": {
        "name": "Ingo Molnar",
        "email": "mingo@elte.hu",
        "time": "Tue Nov 30 16:03:35 2010 +0100"
      },
      "message": "sched: Add \u0027autogroup\u0027 scheduling feature: automated per session task groups\n\nA recurring complaint from CFS users is that parallel kbuild has\na negative impact on desktop interactivity.  This patch\nimplements an idea from Linus, to automatically create task\ngroups.  Currently, only per session autogroups are implemented,\nbut the patch leaves the way open for enhancement.\n\nImplementation: each task\u0027s signal struct contains an inherited\npointer to a refcounted autogroup struct containing a task group\npointer, the default for all tasks pointing to the\ninit_task_group.  When a task calls setsid(), a new task group\nis created, the process is moved into the new task group, and a\nreference to the preveious task group is dropped.  Child\nprocesses inherit this task group thereafter, and increase it\u0027s\nrefcount.  When the last thread of a process exits, the\nprocess\u0027s reference is dropped, such that when the last process\nreferencing an autogroup exits, the autogroup is destroyed.\n\nAt runqueue selection time, IFF a task has no cgroup assignment,\nits current autogroup is used.\n\nAutogroup bandwidth is controllable via setting it\u0027s nice level\nthrough the proc filesystem:\n\n  cat /proc/\u003cpid\u003e/autogroup\n\nDisplays the task\u0027s group and the group\u0027s nice level.\n\n  echo \u003cnice level\u003e \u003e /proc/\u003cpid\u003e/autogroup\n\nSets the task group\u0027s shares to the weight of nice \u003clevel\u003e task.\nSetting nice level is rate limited for !admin users due to the\nabuse risk of task group locking.\n\nThe feature is enabled from boot by default if\nCONFIG_SCHED_AUTOGROUP\u003dy is selected, but can be disabled via\nthe boot option noautogroup, and can also be turned on/off on\nthe fly via:\n\n  echo [01] \u003e /proc/sys/kernel/sched_autogroup_enabled\n\n... which will automatically move tasks to/from the root task group.\n\nSigned-off-by: Mike Galbraith \u003cefault@gmx.de\u003e\nAcked-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nAcked-by: Peter Zijlstra \u003ca.p.zijlstra@chello.nl\u003e\nCc: Markus Trippelsdorf \u003cmarkus@trippelsdorf.de\u003e\nCc: Mathieu Desnoyers \u003cmathieu.desnoyers@efficios.com\u003e\nCc: Paul Turner \u003cpjt@google.com\u003e\nCc: Oleg Nesterov \u003coleg@redhat.com\u003e\n[ Removed the task_group_path() debug code, and fixed !EVENTFD build failure. ]\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\nLKML-Reference: \u003c1290281700.28711.9.camel@maggy.simson.net\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@elte.hu\u003e\n"
    }
  ]
}