1 /* 2 * Copyright (c) 2007 Cisco Systems, Inc. All rights reserved. 3 * Copyright (c) 2007, 2008 Mellanox Technologies. All rights reserved. 4 * 5 * This software is available to you under a choice of one of two 6 * licenses. You may choose to be licensed under the terms of the GNU 7 * General Public License (GPL) Version 2, available from the file 8 * COPYING in the main directory of this source tree, or the 9 * OpenIB.org BSD license below: 10 * 11 * Redistribution and use in source and binary forms, with or 12 * without modification, are permitted provided that the following 13 * conditions are met: 14 * 15 * - Redistributions of source code must retain the above 16 * copyright notice, this list of conditions and the following 17 * disclaimer. 18 * 19 * - Redistributions in binary form must reproduce the above 20 * copyright notice, this list of conditions and the following 21 * disclaimer in the documentation and/or other materials 22 * provided with the distribution. 23 * 24 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, 25 * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF 26 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND 27 * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS 28 * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN 29 * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN 30 * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE 31 * SOFTWARE. 32 */ 33 34 #include <linux/workqueue.h> 35 #include <linux/module.h> 36 37 #include "mlx4.h" 38 39 enum { 40 MLX4_CATAS_POLL_INTERVAL = 5 * HZ, 41 }; 42 43 static DEFINE_SPINLOCK(catas_lock); 44 45 static LIST_HEAD(catas_list); 46 static struct work_struct catas_work; 47 48 static int internal_err_reset = 1; 49 module_param(internal_err_reset, int, 0644); 50 MODULE_PARM_DESC(internal_err_reset, 51 "Reset device on internal errors if non-zero (default 1)"); 52 53 static void dump_err_buf(struct mlx4_dev *dev) 54 { 55 struct mlx4_priv *priv = mlx4_priv(dev); 56 57 int i; 58 59 mlx4_err(dev, "Internal error detected:\n"); 60 for (i = 0; i < priv->fw.catas_size; ++i) 61 mlx4_err(dev, " buf[%02x]: %08x\n", 62 i, swab32(readl(priv->catas_err.map + i))); 63 } 64 65 static void poll_catas(unsigned long dev_ptr) 66 { 67 struct mlx4_dev *dev = (struct mlx4_dev *) dev_ptr; 68 struct mlx4_priv *priv = mlx4_priv(dev); 69 70 if (readl(priv->catas_err.map)) { 71 dump_err_buf(dev); 72 73 mlx4_dispatch_event(dev, MLX4_DEV_EVENT_CATASTROPHIC_ERROR, 0); 74 75 if (internal_err_reset) { 76 spin_lock(&catas_lock); 77 list_add(&priv->catas_err.list, &catas_list); 78 spin_unlock(&catas_lock); 79 80 queue_work(mlx4_wq, &catas_work); 81 } 82 } else 83 mod_timer(&priv->catas_err.timer, 84 round_jiffies(jiffies + MLX4_CATAS_POLL_INTERVAL)); 85 } 86 87 static void catas_reset(struct work_struct *work) 88 { 89 struct mlx4_priv *priv, *tmppriv; 90 struct mlx4_dev *dev; 91 92 LIST_HEAD(tlist); 93 int ret; 94 95 spin_lock_irq(&catas_lock); 96 list_splice_init(&catas_list, &tlist); 97 spin_unlock_irq(&catas_lock); 98 99 list_for_each_entry_safe(priv, tmppriv, &tlist, catas_err.list) { 100 struct pci_dev *pdev = priv->dev.pdev; 101 102 ret = mlx4_restart_one(priv->dev.pdev); 103 /* 'priv' now is not valid */ 104 if (ret) 105 pr_err("mlx4 %s: Reset failed (%d)\n", 106 pci_name(pdev), ret); 107 else { 108 dev = pci_get_drvdata(pdev); 109 mlx4_dbg(dev, "Reset succeeded\n"); 110 } 111 } 112 } 113 114 void mlx4_start_catas_poll(struct mlx4_dev *dev) 115 { 116 struct mlx4_priv *priv = mlx4_priv(dev); 117 phys_addr_t addr; 118 119 INIT_LIST_HEAD(&priv->catas_err.list); 120 init_timer(&priv->catas_err.timer); 121 priv->catas_err.map = NULL; 122 123 addr = pci_resource_start(dev->pdev, priv->fw.catas_bar) + 124 priv->fw.catas_offset; 125 126 priv->catas_err.map = ioremap(addr, priv->fw.catas_size * 4); 127 if (!priv->catas_err.map) { 128 mlx4_warn(dev, "Failed to map internal error buffer at 0x%llx\n", 129 (unsigned long long) addr); 130 return; 131 } 132 133 priv->catas_err.timer.data = (unsigned long) dev; 134 priv->catas_err.timer.function = poll_catas; 135 priv->catas_err.timer.expires = 136 round_jiffies(jiffies + MLX4_CATAS_POLL_INTERVAL); 137 add_timer(&priv->catas_err.timer); 138 } 139 140 void mlx4_stop_catas_poll(struct mlx4_dev *dev) 141 { 142 struct mlx4_priv *priv = mlx4_priv(dev); 143 144 del_timer_sync(&priv->catas_err.timer); 145 146 if (priv->catas_err.map) 147 iounmap(priv->catas_err.map); 148 149 spin_lock_irq(&catas_lock); 150 list_del(&priv->catas_err.list); 151 spin_unlock_irq(&catas_lock); 152 } 153 154 void __init mlx4_catas_init(void) 155 { 156 INIT_WORK(&catas_work, catas_reset); 157 } 158