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 #define LINUXKPI_PARAM_PREFIX mthca_ 34 35 #include <linux/jiffies.h> 36 #include <linux/module.h> 37 #include <linux/timer.h> 38 #include <linux/workqueue.h> 39 40 #include "mthca_dev.h" 41 42 #define MTHCA_CATAS_POLL_INTERVAL (5 * HZ) 43 44 enum { 45 MTHCA_CATAS_TYPE_INTERNAL = 0, 46 MTHCA_CATAS_TYPE_UPLINK = 3, 47 MTHCA_CATAS_TYPE_DDR = 4, 48 MTHCA_CATAS_TYPE_PARITY = 5, 49 }; 50 51 static DEFINE_SPINLOCK(catas_lock); 52 53 static LIST_HEAD(catas_list); 54 static struct workqueue_struct *catas_wq; 55 static struct work_struct catas_work; 56 57 static int catas_reset_disable; 58 module_param_named(catas_reset_disable, catas_reset_disable, int, 0644); 59 MODULE_PARM_DESC(catas_reset_disable, "disable reset on catastrophic event if nonzero"); 60 61 static void catas_reset(struct work_struct *work) 62 { 63 struct mthca_dev *dev, *tmpdev; 64 LIST_HEAD(tlist); 65 int ret; 66 67 mutex_lock(&mthca_device_mutex); 68 69 spin_lock_irq(&catas_lock); 70 list_splice_init(&catas_list, &tlist); 71 spin_unlock_irq(&catas_lock); 72 73 list_for_each_entry_safe(dev, tmpdev, &tlist, catas_err.list) { 74 struct pci_dev *pdev = dev->pdev; 75 ret = __mthca_restart_one(dev->pdev); 76 /* 'dev' now is not valid */ 77 if (ret) 78 printk(KERN_ERR "mthca %s: Reset failed (%d)\n", 79 pci_name(pdev), ret); 80 else { 81 struct mthca_dev *d = pci_get_drvdata(pdev); 82 mthca_dbg(d, "Reset succeeded\n"); 83 } 84 } 85 86 mutex_unlock(&mthca_device_mutex); 87 } 88 89 static void handle_catas(struct mthca_dev *dev) 90 { 91 struct ib_event event; 92 unsigned long flags; 93 const char *type; 94 int i; 95 96 event.device = &dev->ib_dev; 97 event.event = IB_EVENT_DEVICE_FATAL; 98 event.element.port_num = 0; 99 dev->active = false; 100 101 ib_dispatch_event(&event); 102 103 switch (swab32(readl(dev->catas_err.map)) >> 24) { 104 case MTHCA_CATAS_TYPE_INTERNAL: 105 type = "internal error"; 106 break; 107 case MTHCA_CATAS_TYPE_UPLINK: 108 type = "uplink bus error"; 109 break; 110 case MTHCA_CATAS_TYPE_DDR: 111 type = "DDR data error"; 112 break; 113 case MTHCA_CATAS_TYPE_PARITY: 114 type = "internal parity error"; 115 break; 116 default: 117 type = "unknown error"; 118 break; 119 } 120 121 mthca_err(dev, "Catastrophic error detected: %s\n", type); 122 for (i = 0; i < dev->catas_err.size; ++i) 123 mthca_err(dev, " buf[%02x]: %08x\n", 124 i, swab32(readl(dev->catas_err.map + i))); 125 126 if (catas_reset_disable) 127 return; 128 129 spin_lock_irqsave(&catas_lock, flags); 130 list_add(&dev->catas_err.list, &catas_list); 131 queue_work(catas_wq, &catas_work); 132 spin_unlock_irqrestore(&catas_lock, flags); 133 } 134 135 static void poll_catas(unsigned long dev_ptr) 136 { 137 struct mthca_dev *dev = (struct mthca_dev *) dev_ptr; 138 int i; 139 140 for (i = 0; i < dev->catas_err.size; ++i) 141 if (readl(dev->catas_err.map + i)) { 142 handle_catas(dev); 143 return; 144 } 145 146 mod_timer(&dev->catas_err.timer, 147 round_jiffies(jiffies + MTHCA_CATAS_POLL_INTERVAL)); 148 } 149 150 void mthca_start_catas_poll(struct mthca_dev *dev) 151 { 152 phys_addr_t addr; 153 154 init_timer(&dev->catas_err.timer); 155 dev->catas_err.map = NULL; 156 157 addr = pci_resource_start(dev->pdev, 0) + 158 ((pci_resource_len(dev->pdev, 0) - 1) & 159 dev->catas_err.addr); 160 161 dev->catas_err.map = ioremap(addr, dev->catas_err.size * 4); 162 if (!dev->catas_err.map) { 163 mthca_warn(dev, "couldn't map catastrophic error region " 164 "at 0x%llx/0x%x\n", (unsigned long long) addr, 165 dev->catas_err.size * 4); 166 return; 167 } 168 169 dev->catas_err.timer.data = (unsigned long) dev; 170 dev->catas_err.timer.function = poll_catas; 171 dev->catas_err.timer.expires = jiffies + MTHCA_CATAS_POLL_INTERVAL; 172 INIT_LIST_HEAD(&dev->catas_err.list); 173 add_timer(&dev->catas_err.timer); 174 } 175 176 void mthca_stop_catas_poll(struct mthca_dev *dev) 177 { 178 del_timer_sync(&dev->catas_err.timer); 179 180 if (dev->catas_err.map) 181 iounmap(dev->catas_err.map); 182 183 spin_lock_irq(&catas_lock); 184 list_del(&dev->catas_err.list); 185 spin_unlock_irq(&catas_lock); 186 } 187 188 int __init mthca_catas_init(void) 189 { 190 INIT_WORK(&catas_work, catas_reset); 191 192 catas_wq = alloc_ordered_workqueue("mthca_catas", WQ_MEM_RECLAIM); 193 if (!catas_wq) 194 return -ENOMEM; 195 196 return 0; 197 } 198 199 void mthca_catas_cleanup(void) 200 { 201 destroy_workqueue(catas_wq); 202 } 203