arch/powerpc/platforms/pseries/eeh_event.c

   1 /*
   2  * eeh_event.c
   3  *
   4  * This program is free software; you can redistribute it and/or modify
   5  * it under the terms of the GNU General Public License as published by
   6  * the Free Software Foundation; either version 2 of the License, or
   7  * (at your option) any later version.
   8  *
   9  * This program is distributed in the hope that it will be useful,
  10  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  11  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12  * GNU General Public License for more details.
  13  *
  14  * You should have received a copy of the GNU General Public License
  15  * along with this program; if not, write to the Free Software
  16  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
  17  *
  18  * Copyright (c) 2005 Linas Vepstas <linas@linas.org>
  19  */
  20
  21 #include <linux/list.h>
  22 #include <linux/pci.h>
  23 #include <asm/eeh_event.h>
  24 #include <asm/ppc-pci.h>
  25
  26 /** Overview:
  27  *  EEH error states may be detected within exception handlers;
  28  *  however, the recovery processing needs to occur asynchronously
  29  *  in a normal kernel context and not an interrupt context.
  30  *  This pair of routines creates an event and queues it onto a
  31  *  work-queue, where a worker thread can drive recovery.
  32  */
  33
  34 /* EEH event workqueue setup. */
  35 static spinlock_t eeh_eventlist_lock = SPIN_LOCK_UNLOCKED;
  36 LIST_HEAD(eeh_eventlist);
  37 static void eeh_thread_launcher(void *);
  38 DECLARE_WORK(eeh_event_wq, eeh_thread_launcher, NULL);
  39
  40 /**
  41  * eeh_event_handler - dispatch EEH events.  The detection of a frozen
  42  * slot can occur inside an interrupt, where it can be hard to do
  43  * anything about it.  The goal of this routine is to pull these
  44  * detection events out of the context of the interrupt handler, and
  45  * re-dispatch them for processing at a later time in a normal context.
  46  *
  47  * @dummy - unused
  48  */
  49 static int eeh_event_handler(void * dummy)
  50 {
  51         unsigned long flags;
  52         struct eeh_event        *event;
  53
  54         daemonize ("eehd");
  55
  56         while (1) {
  57                 set_current_state(TASK_INTERRUPTIBLE);
  58
  59                 spin_lock_irqsave(&eeh_eventlist_lock, flags);
  60                 event = NULL;
  61
  62                 /* Unqueue the event, get ready to process. */
  63                 if (!list_empty(&eeh_eventlist)) {
  64                         event = list_entry(eeh_eventlist.next, struct eeh_event, list);
  65                         list_del(&event->list);
  66                 }
  67
  68                 if (event)
  69                         eeh_mark_slot(event->dn, EEH_MODE_RECOVERING);
  70
  71                 spin_unlock_irqrestore(&eeh_eventlist_lock, flags);
  72                 if (event == NULL)
  73                         break;
  74
  75                 printk(KERN_INFO "EEH: Detected PCI bus error on device %s\n",
  76                        pci_name(event->dev));
  77
  78                 handle_eeh_events(event);
  79
  80                 eeh_clear_slot(event->dn, EEH_MODE_RECOVERING);
  81
  82                 pci_dev_put(event->dev);
  83                 kfree(event);
  84         }
  85
  86         return 0;
  87 }
  88
  89 /**
  90  * eeh_thread_launcher
  91  *
  92  * @dummy - unused
  93  */
  94 static void eeh_thread_launcher(void *dummy)
  95 {
  96         if (kernel_thread(eeh_event_handler, NULL, CLONE_KERNEL) < 0)
  97                 printk(KERN_ERR "Failed to start EEH daemon\n");
  98 }
  99
 100 /**
 101  * eeh_send_failure_event - generate a PCI error event
 102  * @dev pci device
 103  *
 104  * This routine can be called within an interrupt context;
 105  * the actual event will be delivered in a normal context
 106  * (from a workqueue).
 107  */
 108 int eeh_send_failure_event (struct device_node *dn,
 109                             struct pci_dev *dev,
 110                             enum pci_channel_state state,
 111                             int time_unavail)
 112 {
 113         unsigned long flags;
 114         struct eeh_event *event;
 115
 116         event = kmalloc(sizeof(*event), GFP_ATOMIC);
 117         if (event == NULL) {
 118                 printk (KERN_ERR "EEH: out of memory, event not handled\n");
 119                 return 1;
 120         }
 121
 122         if (dev)
 123                 pci_dev_get(dev);
 124
 125         event->dn = dn;
 126         event->dev = dev;
 127         event->state = state;
 128         event->time_unavail = time_unavail;
 129
 130         /* We may or may not be called in an interrupt context */
 131         spin_lock_irqsave(&eeh_eventlist_lock, flags);
 132         list_add(&event->list, &eeh_eventlist);
 133         spin_unlock_irqrestore(&eeh_eventlist_lock, flags);
 134
 135         schedule_work(&eeh_event_wq);
 136
 137         return 0;
 138 }
 139
 140 /********************** END OF FILE ******************************/