eeh_event.c 4.1 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146
  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. #include <linux/list.h>
  21. #include <linux/mutex.h>
  22. #include <linux/pci.h>
  23. #include <linux/workqueue.h>
  24. #include <asm/eeh_event.h>
  25. #include <asm/ppc-pci.h>
  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. /* EEH event workqueue setup. */
  34. static spinlock_t eeh_eventlist_lock = SPIN_LOCK_UNLOCKED;
  35. LIST_HEAD(eeh_eventlist);
  36. static void eeh_thread_launcher(void *);
  37. DECLARE_WORK(eeh_event_wq, eeh_thread_launcher, NULL);
  38. /* Serialize reset sequences for a given pci device */
  39. DEFINE_MUTEX(eeh_event_mutex);
  40. /**
  41. * eeh_event_handler - dispatch EEH events.
  42. * @dummy - unused
  43. *
  44. * The detection of a frozen slot can occur inside an interrupt,
  45. * where it can be hard to do anything about it. The goal of this
  46. * routine is to pull these detection events out of the context
  47. * of the interrupt handler, and re-dispatch them for processing
  48. * at a later time in a normal context.
  49. */
  50. static int eeh_event_handler(void * dummy)
  51. {
  52. unsigned long flags;
  53. struct eeh_event *event;
  54. daemonize ("eehd");
  55. while (1) {
  56. set_current_state(TASK_INTERRUPTIBLE);
  57. spin_lock_irqsave(&eeh_eventlist_lock, flags);
  58. event = NULL;
  59. /* Unqueue the event, get ready to process. */
  60. if (!list_empty(&eeh_eventlist)) {
  61. event = list_entry(eeh_eventlist.next, struct eeh_event, list);
  62. list_del(&event->list);
  63. }
  64. spin_unlock_irqrestore(&eeh_eventlist_lock, flags);
  65. if (event == NULL)
  66. break;
  67. /* Serialize processing of EEH events */
  68. mutex_lock(&eeh_event_mutex);
  69. eeh_mark_slot(event->dn, EEH_MODE_RECOVERING);
  70. printk(KERN_INFO "EEH: Detected PCI bus error on device %s\n",
  71. pci_name(event->dev));
  72. handle_eeh_events(event);
  73. eeh_clear_slot(event->dn, EEH_MODE_RECOVERING);
  74. pci_dev_put(event->dev);
  75. kfree(event);
  76. mutex_unlock(&eeh_event_mutex);
  77. }
  78. return 0;
  79. }
  80. /**
  81. * eeh_thread_launcher
  82. * @dummy - unused
  83. */
  84. static void eeh_thread_launcher(void *dummy)
  85. {
  86. if (kernel_thread(eeh_event_handler, NULL, CLONE_KERNEL) < 0)
  87. printk(KERN_ERR "Failed to start EEH daemon\n");
  88. }
  89. /**
  90. * eeh_send_failure_event - generate a PCI error event
  91. * @dev pci device
  92. *
  93. * This routine can be called within an interrupt context;
  94. * the actual event will be delivered in a normal context
  95. * (from a workqueue).
  96. */
  97. int eeh_send_failure_event (struct device_node *dn,
  98. struct pci_dev *dev,
  99. enum pci_channel_state state,
  100. int time_unavail)
  101. {
  102. unsigned long flags;
  103. struct eeh_event *event;
  104. event = kmalloc(sizeof(*event), GFP_ATOMIC);
  105. if (event == NULL) {
  106. printk (KERN_ERR "EEH: out of memory, event not handled\n");
  107. return 1;
  108. }
  109. if (dev)
  110. pci_dev_get(dev);
  111. event->dn = dn;
  112. event->dev = dev;
  113. event->state = state;
  114. event->time_unavail = time_unavail;
  115. /* We may or may not be called in an interrupt context */
  116. spin_lock_irqsave(&eeh_eventlist_lock, flags);
  117. list_add(&event->list, &eeh_eventlist);
  118. spin_unlock_irqrestore(&eeh_eventlist_lock, flags);
  119. schedule_work(&eeh_event_wq);
  120. return 0;
  121. }
  122. /********************** END OF FILE ******************************/