drivers/infiniband/hw/mthca/mthca_catas.c

   1 /*
   2  * Copyright (c) 2005 Cisco Systems.  All rights reserved.
   3  *
   4  * This software is available to you under a choice of one of two
   5  * licenses.  You may choose to be licensed under the terms of the GNU
   6  * General Public License (GPL) Version 2, available from the file
   7  * COPYING in the main directory of this source tree, or the
   8  * OpenIB.org BSD license below:
   9  *
  10  *     Redistribution and use in source and binary forms, with or
  11  *     without modification, are permitted provided that the following
  12  *     conditions are met:
  13  *
  14  *      - Redistributions of source code must retain the above
  15  *        copyright notice, this list of conditions and the following
  16  *        disclaimer.
  17  *
  18  *      - Redistributions in binary form must reproduce the above
  19  *        copyright notice, this list of conditions and the following
  20  *        disclaimer in the documentation and/or other materials
  21  *        provided with the distribution.
  22  *
  23  * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
  24  * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
  25  * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
  26  * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
  27  * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
  28  * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
  29  * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  30  * SOFTWARE.
  31  */
  32
  33 #include <linux/jiffies.h>
  34 #include <linux/timer.h>
  35 #include <linux/workqueue.h>
  36
  37 #include "mthca_dev.h"
  38
  39 enum {
  40         MTHCA_CATAS_POLL_INTERVAL       = 5 * HZ,
  41
  42         MTHCA_CATAS_TYPE_INTERNAL       = 0,
  43         MTHCA_CATAS_TYPE_UPLINK         = 3,
  44         MTHCA_CATAS_TYPE_DDR            = 4,
  45         MTHCA_CATAS_TYPE_PARITY         = 5,
  46 };
  47
  48 static DEFINE_SPINLOCK(catas_lock);
  49
  50 static LIST_HEAD(catas_list);
  51 static struct workqueue_struct *catas_wq;
  52 static struct work_struct catas_work;
  53
  54 static int catas_reset_disable;
  55 module_param_named(catas_reset_disable, catas_reset_disable, int, 0644);
  56 MODULE_PARM_DESC(catas_reset_disable, "disable reset on catastrophic event if nonzero");
  57
  58 static void catas_reset(struct work_struct *work)
  59 {
  60         struct mthca_dev *dev, *tmpdev;
  61         LIST_HEAD(tlist);
  62         int ret;
  63
  64         mutex_lock(&mthca_device_mutex);
  65
  66         spin_lock_irq(&catas_lock);
  67         list_splice_init(&catas_list, &tlist);
  68         spin_unlock_irq(&catas_lock);
  69
  70         list_for_each_entry_safe(dev, tmpdev, &tlist, catas_err.list) {
  71                 struct pci_dev *pdev = dev->pdev;
  72                 ret = __mthca_restart_one(dev->pdev);
  73                 /* 'dev' now is not valid */
  74                 if (ret)
  75                         printk(KERN_ERR "mthca %s: Reset failed (%d)\n",
  76                                pci_name(pdev), ret);
  77                 else {
  78                         struct mthca_dev *d = pci_get_drvdata(pdev);
  79                         mthca_dbg(d, "Reset succeeded\n");
  80                 }
  81         }
  82
  83         mutex_unlock(&mthca_device_mutex);
  84 }
  85
  86 static void handle_catas(struct mthca_dev *dev)
  87 {
  88         struct ib_event event;
  89         unsigned long flags;
  90         const char *type;
  91         int i;
  92
  93         event.device = &dev->ib_dev;
  94         event.event  = IB_EVENT_DEVICE_FATAL;
  95         event.element.port_num = 0;
  96         dev->active = false;
  97
  98         ib_dispatch_event(&event);
  99
 100         switch (swab32(readl(dev->catas_err.map)) >> 24) {
 101         case MTHCA_CATAS_TYPE_INTERNAL:
 102                 type = "internal error";
 103                 break;
 104         case MTHCA_CATAS_TYPE_UPLINK:
 105                 type = "uplink bus error";
 106                 break;
 107         case MTHCA_CATAS_TYPE_DDR:
 108                 type = "DDR data error";
 109                 break;
 110         case MTHCA_CATAS_TYPE_PARITY:
 111                 type = "internal parity error";
 112                 break;
 113         default:
 114                 type = "unknown error";
 115                 break;
 116         }
 117
 118         mthca_err(dev, "Catastrophic error detected: %s\n", type);
 119         for (i = 0; i < dev->catas_err.size; ++i)
 120                 mthca_err(dev, "  buf[%02x]: %08x\n",
 121                           i, swab32(readl(dev->catas_err.map + i)));
 122
 123         if (catas_reset_disable)
 124                 return;
 125
 126         spin_lock_irqsave(&catas_lock, flags);
 127         list_add(&dev->catas_err.list, &catas_list);
 128         queue_work(catas_wq, &catas_work);
 129         spin_unlock_irqrestore(&catas_lock, flags);
 130 }
 131
 132 static void poll_catas(unsigned long dev_ptr)
 133 {
 134         struct mthca_dev *dev = (struct mthca_dev *) dev_ptr;
 135         int i;
 136
 137         for (i = 0; i < dev->catas_err.size; ++i)
 138                 if (readl(dev->catas_err.map + i)) {
 139                         handle_catas(dev);
 140                         return;
 141                 }
 142
 143         mod_timer(&dev->catas_err.timer,
 144                   round_jiffies(jiffies + MTHCA_CATAS_POLL_INTERVAL));
 145 }
 146
 147 void mthca_start_catas_poll(struct mthca_dev *dev)
 148 {
 149         unsigned long addr;
 150
 151         init_timer(&dev->catas_err.timer);
 152         dev->catas_err.map  = NULL;
 153
 154         addr = pci_resource_start(dev->pdev, 0) +
 155                 ((pci_resource_len(dev->pdev, 0) - 1) &
 156                  dev->catas_err.addr);
 157
 158         dev->catas_err.map = ioremap(addr, dev->catas_err.size * 4);
 159         if (!dev->catas_err.map) {
 160                 mthca_warn(dev, "couldn't map catastrophic error region "
 161                            "at 0x%lx/0x%x\n", addr, dev->catas_err.size * 4);
 162                 return;
 163         }
 164
 165         dev->catas_err.timer.data     = (unsigned long) dev;
 166         dev->catas_err.timer.function = poll_catas;
 167         dev->catas_err.timer.expires  = jiffies + MTHCA_CATAS_POLL_INTERVAL;
 168         INIT_LIST_HEAD(&dev->catas_err.list);
 169         add_timer(&dev->catas_err.timer);
 170 }
 171
 172 void mthca_stop_catas_poll(struct mthca_dev *dev)
 173 {
 174         del_timer_sync(&dev->catas_err.timer);
 175
 176         if (dev->catas_err.map)
 177                 iounmap(dev->catas_err.map);
 178
 179         spin_lock_irq(&catas_lock);
 180         list_del(&dev->catas_err.list);
 181         spin_unlock_irq(&catas_lock);
 182 }
 183
 184 int __init mthca_catas_init(void)
 185 {
 186         INIT_WORK(&catas_work, catas_reset);
 187
 188         catas_wq = create_singlethread_workqueue("mthca_catas");
 189         if (!catas_wq)
 190                 return -ENOMEM;
 191
 192         return 0;
 193 }
 194
 195 void mthca_catas_cleanup(void)
 196 {
 197         destroy_workqueue(catas_wq);
 198 }