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 36#include "mlx4.h" 37 38#define MLX4_CATAS_POLL_INTERVAL (5 * HZ) 39 40static DEFINE_SPINLOCK(catas_lock); 41 42static LIST_HEAD(catas_list); 43static struct work_struct catas_work; 44 45static int internal_err_reset = 1; 46module_param(internal_err_reset, int, 0644); 47MODULE_PARM_DESC(internal_err_reset, 48 "Reset device on internal errors if non-zero (default 1)"); 49 50static void dump_err_buf(struct mlx4_dev *dev) 51{ 52 struct mlx4_priv *priv = mlx4_priv(dev); 53 54 int i; 55 56 mlx4_err(dev, "Internal error detected:\n"); 57 for (i = 0; i < priv->fw.catas_size; ++i) 58 mlx4_err(dev, " buf[%02x]: %08x\n", 59 i, swab32(readl(priv->catas_err.map + i))); 60} 61 62static void poll_catas(unsigned long dev_ptr) 63{ 64 struct mlx4_dev *dev = (struct mlx4_dev *) dev_ptr; 65 struct mlx4_priv *priv = mlx4_priv(dev); 66 67 if (readl(priv->catas_err.map)) { 68 dump_err_buf(dev); 69 70 mlx4_dispatch_event(dev, MLX4_DEV_EVENT_CATASTROPHIC_ERROR, 0); 71 72 if (internal_err_reset) { 73 spin_lock(&catas_lock); 74 list_add(&priv->catas_err.list, &catas_list); 75 spin_unlock(&catas_lock); 76 77 queue_work(mlx4_wq, &catas_work); 78 } 79 } else 80 mod_timer(&priv->catas_err.timer, 81 round_jiffies(jiffies + MLX4_CATAS_POLL_INTERVAL)); 82} 83 84static void catas_reset(struct work_struct *work) 85{ 86 struct mlx4_priv *priv, *tmppriv; 87 struct mlx4_dev *dev; 88 89 LIST_HEAD(tlist); 90 int ret; 91 92 if (!mutex_trylock(&drv_mutex)) 93 return; 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 printk(KERN_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 mutex_unlock(&drv_mutex); 113} 114 115void mlx4_start_catas_poll(struct mlx4_dev *dev) 116{ 117 struct mlx4_priv *priv = mlx4_priv(dev); 118 unsigned long addr; 119 120 INIT_LIST_HEAD(&priv->catas_err.list); 121 init_timer(&priv->catas_err.timer); 122 priv->catas_err.map = NULL; 123 124 addr = pci_resource_start(dev->pdev, priv->fw.catas_bar) + 125 priv->fw.catas_offset; 126 127 priv->catas_err.map = ioremap(addr, priv->fw.catas_size * 4); 128 if (!priv->catas_err.map) { 129 mlx4_warn(dev, "Failed to map internal error buffer at 0x%lx\n", 130 addr); 131 return; 132 } 133 134 priv->catas_err.timer.data = (unsigned long) dev; 135 priv->catas_err.timer.function = poll_catas; 136 priv->catas_err.timer.expires = 137 round_jiffies(jiffies + MLX4_CATAS_POLL_INTERVAL); 138 add_timer(&priv->catas_err.timer); 139} 140 141void mlx4_stop_catas_poll(struct mlx4_dev *dev) 142{ 143 struct mlx4_priv *priv = mlx4_priv(dev); 144 145 del_timer_sync(&priv->catas_err.timer); 146 147 if (priv->catas_err.map) 148 iounmap(priv->catas_err.map); 149 150 spin_lock_irq(&catas_lock); 151 list_del(&priv->catas_err.list); 152 spin_unlock_irq(&catas_lock); 153} 154 155void __init mlx4_catas_init(void) 156{ 157 INIT_WORK(&catas_work, catas_reset); 158} 159