)]}'
{
  "commit": "78944e549d36673eb6265a2411574e79c28e23dc",
  "tree": "5e080e775c21f6d7af69f37f37d25b7680cc46b8",
  "parents": [
    "38cbcdc0a7be69a15462dc49512d43353f34b43b"
  ],
  "author": {
    "name": "Antonino A. Daplas",
    "email": "adaplas@gmail.com",
    "time": "Sat Aug 05 12:14:16 2006 -0700"
  },
  "committer": {
    "name": "Linus Torvalds",
    "email": "torvalds@g5.osdl.org",
    "time": "Sun Aug 06 08:57:47 2006 -0700"
  },
  "message": "[PATCH] vt: printk: Fix framebuffer console triggering might_sleep assertion\n\nReported by: Dave Jones\n\nWhilst printk\u0027ing to both console and serial console, I got this...\n(2.6.18rc1)\n\nBUG: sleeping function called from invalid context at kernel/sched.c:4438\nin_atomic():0, irqs_disabled():1\n\nCall Trace:\n [\u003cffffffff80271db8\u003e] show_trace+0xaa/0x23d\n [\u003cffffffff80271f60\u003e] dump_stack+0x15/0x17\n [\u003cffffffff8020b9f8\u003e] __might_sleep+0xb2/0xb4\n [\u003cffffffff8029232e\u003e] __cond_resched+0x15/0x55\n [\u003cffffffff80267eb8\u003e] cond_resched+0x3b/0x42\n [\u003cffffffff80268c64\u003e] console_conditional_schedule+0x12/0x14\n [\u003cffffffff80368159\u003e] fbcon_redraw+0xf6/0x160\n [\u003cffffffff80369c58\u003e] fbcon_scroll+0x5d9/0xb52\n [\u003cffffffff803a43c4\u003e] scrup+0x6b/0xd6\n [\u003cffffffff803a4453\u003e] lf+0x24/0x44\n [\u003cffffffff803a7ff8\u003e] vt_console_print+0x166/0x23d\n [\u003cffffffff80295528\u003e] __call_console_drivers+0x65/0x76\n [\u003cffffffff80295597\u003e] _call_console_drivers+0x5e/0x62\n [\u003cffffffff80217e3f\u003e] release_console_sem+0x14b/0x232\n [\u003cffffffff8036acd6\u003e] fb_flashcursor+0x279/0x2a6\n [\u003cffffffff80251e3f\u003e] run_workqueue+0xa8/0xfb\n [\u003cffffffff8024e5e0\u003e] worker_thread+0xef/0x122\n [\u003cffffffff8023660f\u003e] kthread+0x100/0x136\n [\u003cffffffff8026419e\u003e] child_rip+0x8/0x12\n\nThis can occur when release_console_sem() is called but the log\nbuffer still has contents that need to be flushed. The console drivers\nare called while the console_may_schedule flag is still true. The\nmight_sleep() is triggered when fbcon calls console_conditional_schedule().\n\nFix by setting console_may_schedule to zero earlier, before the call to the\nconsole drivers.\n\nSigned-off-by: Antonino Daplas \u003cadaplas@pol.net\u003e\nSigned-off-by: Andrew Morton \u003cakpm@osdl.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "65ca0688f86f1b78fafd144671a17015b2fd1525",
      "old_mode": 33188,
      "old_path": "kernel/printk.c",
      "new_id": "1149365e989edee40639bda0cd6865204e05cc89",
      "new_mode": 33188,
      "new_path": "kernel/printk.c"
    }
  ]
}