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" 52 " (default 1, in SRIOV mode default is 0)"); 53 54 static void dump_err_buf(struct mlx4_dev *dev) 55 { 56 struct mlx4_priv *priv = mlx4_priv(dev); 57 58 int i; 59 60 mlx4_err(dev, "Internal error detected:\n"); 61 for (i = 0; i < priv->fw.catas_size; ++i) 62 mlx4_err(dev, " buf[%02x]: %08x\n", 63 i, swab32(readl(priv->catas_err.map + i))); 64 } 65 66 static void poll_catas(unsigned long dev_ptr) 67 { 68 struct mlx4_dev *dev = (struct mlx4_dev *) dev_ptr; 69 struct mlx4_priv *priv = mlx4_priv(dev); 70 71 if (readl(priv->catas_err.map)) { 72 /* If the device is off-line, we cannot try to recover it */ 73 if (pci_channel_offline(dev->pdev)) 74 mod_timer(&priv->catas_err.timer, 75 round_jiffies(jiffies + MLX4_CATAS_POLL_INTERVAL)); 76 else { 77 dump_err_buf(dev); 78 mlx4_dispatch_event(dev, MLX4_DEV_EVENT_CATASTROPHIC_ERROR, 0); 79 80 if (internal_err_reset) { 81 spin_lock(&catas_lock); 82 list_add(&priv->catas_err.list, &catas_list); 83 spin_unlock(&catas_lock); 84 85 queue_work(mlx4_wq, &catas_work); 86 } 87 } 88 } else 89 mod_timer(&priv->catas_err.timer, 90 round_jiffies(jiffies + MLX4_CATAS_POLL_INTERVAL)); 91 } 92 93 static void catas_reset(struct work_struct *work) 94 { 95 struct mlx4_priv *priv, *tmppriv; 96 struct mlx4_dev *dev; 97 98 LIST_HEAD(tlist); 99 int ret; 100 101 spin_lock_irq(&catas_lock); 102 list_splice_init(&catas_list, &tlist); 103 spin_unlock_irq(&catas_lock); 104 105 list_for_each_entry_safe(priv, tmppriv, &tlist, catas_err.list) { 106 struct pci_dev *pdev = priv->dev.pdev; 107 108 /* If the device is off-line, we cannot reset it */ 109 if (pci_channel_offline(pdev)) 110 continue; 111 112 ret = mlx4_restart_one(priv->dev.pdev); 113 /* 'priv' now is not valid */ 114 if (ret) 115 pr_err("mlx4 %s: Reset failed (%d)\n", 116 pci_name(pdev), ret); 117 else { 118 dev = pci_get_drvdata(pdev); 119 mlx4_dbg(dev, "Reset succeeded\n"); 120 } 121 } 122 } 123 124 void mlx4_start_catas_poll(struct mlx4_dev *dev) 125 { 126 struct mlx4_priv *priv = mlx4_priv(dev); 127 phys_addr_t addr; 128 129 /*If we are in SRIOV the default of the module param must be 0*/ 130 if (mlx4_is_mfunc(dev)) 131 internal_err_reset = 0; 132 133 INIT_LIST_HEAD(&priv->catas_err.list); 134 init_timer(&priv->catas_err.timer); 135 priv->catas_err.map = NULL; 136 137 addr = pci_resource_start(dev->pdev, priv->fw.catas_bar) + 138 priv->fw.catas_offset; 139 140 priv->catas_err.map = ioremap(addr, priv->fw.catas_size * 4); 141 if (!priv->catas_err.map) { 142 mlx4_warn(dev, "Failed to map internal error buffer at 0x%llx\n", 143 (unsigned long long) addr); 144 return; 145 } 146 147 priv->catas_err.timer.data = (unsigned long) dev; 148 priv->catas_err.timer.function = poll_catas; 149 priv->catas_err.timer.expires = 150 round_jiffies(jiffies + MLX4_CATAS_POLL_INTERVAL); 151 add_timer(&priv->catas_err.timer); 152 } 153 154 void mlx4_stop_catas_poll(struct mlx4_dev *dev) 155 { 156 struct mlx4_priv *priv = mlx4_priv(dev); 157 158 del_timer_sync(&priv->catas_err.timer); 159 160 if (priv->catas_err.map) 161 iounmap(priv->catas_err.map); 162 163 spin_lock_irq(&catas_lock); 164 list_del(&priv->catas_err.list); 165 spin_unlock_irq(&catas_lock); 166 } 167 168 void __init mlx4_catas_init(void) 169 { 170 INIT_WORK(&catas_work, catas_reset); 171 } 172