arch/powerpc/platforms/powernv/opal-memory-errors.c

   1 /*
   2  * OPAL asynchronus Memory error handling support in PowerNV.
   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 2013 IBM Corporation
  19  * Author: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
  20  */
  21
  22 #undef DEBUG
  23
  24 #include <linux/kernel.h>
  25 #include <linux/init.h>
  26 #include <linux/of.h>
  27 #include <linux/mm.h>
  28 #include <linux/slab.h>
  29
  30 #include <asm/machdep.h>
  31 #include <asm/opal.h>
  32 #include <asm/cputable.h>
  33
  34 static int opal_mem_err_nb_init;
  35 static LIST_HEAD(opal_memory_err_list);
  36 static DEFINE_SPINLOCK(opal_mem_err_lock);
  37
  38 struct OpalMsgNode {
  39         struct list_head list;
  40         struct opal_msg msg;
  41 };
  42
  43 static void handle_memory_error_event(struct OpalMemoryErrorData *merr_evt)
  44 {
  45         uint64_t paddr_start, paddr_end;
  46
  47         pr_debug("%s: Retrieved memory error event, type: 0x%x\n",
  48                   __func__, merr_evt->type);
  49         switch (merr_evt->type) {
  50         case OPAL_MEM_ERR_TYPE_RESILIENCE:
  51                 paddr_start = be64_to_cpu(merr_evt->u.resilience.physical_address_start);
  52                 paddr_end = be64_to_cpu(merr_evt->u.resilience.physical_address_end);
  53                 break;
  54         case OPAL_MEM_ERR_TYPE_DYN_DALLOC:
  55                 paddr_start = be64_to_cpu(merr_evt->u.dyn_dealloc.physical_address_start);
  56                 paddr_end = be64_to_cpu(merr_evt->u.dyn_dealloc.physical_address_end);
  57                 break;
  58         default:
  59                 return;
  60         }
  61
  62         for (; paddr_start < paddr_end; paddr_start += PAGE_SIZE) {
  63                 memory_failure(paddr_start >> PAGE_SHIFT, 0);
  64         }
  65 }
  66
  67 static void handle_memory_error(void)
  68 {
  69         unsigned long flags;
  70         struct OpalMemoryErrorData *merr_evt;
  71         struct OpalMsgNode *msg_node;
  72
  73         spin_lock_irqsave(&opal_mem_err_lock, flags);
  74         while (!list_empty(&opal_memory_err_list)) {
  75                  msg_node = list_entry(opal_memory_err_list.next,
  76                                            struct OpalMsgNode, list);
  77                 list_del(&msg_node->list);
  78                 spin_unlock_irqrestore(&opal_mem_err_lock, flags);
  79
  80                 merr_evt = (struct OpalMemoryErrorData *)
  81                                         &msg_node->msg.params[0];
  82                 handle_memory_error_event(merr_evt);
  83                 kfree(msg_node);
  84                 spin_lock_irqsave(&opal_mem_err_lock, flags);
  85         }
  86         spin_unlock_irqrestore(&opal_mem_err_lock, flags);
  87 }
  88
  89 static void mem_error_handler(struct work_struct *work)
  90 {
  91         handle_memory_error();
  92 }
  93
  94 static DECLARE_WORK(mem_error_work, mem_error_handler);
  95
  96 /*
  97  * opal_memory_err_event - notifier handler that queues up the opal message
  98  * to be preocessed later.
  99  */
 100 static int opal_memory_err_event(struct notifier_block *nb,
 101                           unsigned long msg_type, void *msg)
 102 {
 103         unsigned long flags;
 104         struct OpalMsgNode *msg_node;
 105
 106         if (msg_type != OPAL_MSG_MEM_ERR)
 107                 return 0;
 108
 109         msg_node = kzalloc(sizeof(*msg_node), GFP_ATOMIC);
 110         if (!msg_node) {
 111                 pr_err("MEMORY_ERROR: out of memory, Opal message event not"
 112                        "handled\n");
 113                 return -ENOMEM;
 114         }
 115         memcpy(&msg_node->msg, msg, sizeof(struct opal_msg));
 116
 117         spin_lock_irqsave(&opal_mem_err_lock, flags);
 118         list_add(&msg_node->list, &opal_memory_err_list);
 119         spin_unlock_irqrestore(&opal_mem_err_lock, flags);
 120
 121         schedule_work(&mem_error_work);
 122         return 0;
 123 }
 124
 125 static struct notifier_block opal_mem_err_nb = {
 126         .notifier_call  = opal_memory_err_event,
 127         .next           = NULL,
 128         .priority       = 0,
 129 };
 130
 131 static int __init opal_mem_err_init(void)
 132 {
 133         int ret;
 134
 135         if (!opal_mem_err_nb_init) {
 136                 ret = opal_message_notifier_register(
 137                                         OPAL_MSG_MEM_ERR, &opal_mem_err_nb);
 138                 if (ret) {
 139                         pr_err("%s: Can't register OPAL event notifier (%d)\n",
 140                                __func__, ret);
 141                         return ret;
 142                 }
 143                 opal_mem_err_nb_init = 1;
 144         }
 145         return 0;
 146 }
 147 machine_device_initcall(powernv, opal_mem_err_init);