2 * Copyright (c) 2007 Cisco Systems, Inc. All rights reserved.
3 * Copyright (c) 2007, 2008 Mellanox Technologies. All rights reserved.
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:
11 * Redistribution and use in source and binary forms, with or
12 * without modification, are permitted provided that the following
15 * - Redistributions of source code must retain the above
16 * copyright notice, this list of conditions and the following
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.
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
34 #include <linux/workqueue.h>
38 #define MLX4_CATAS_POLL_INTERVAL (5 * HZ)
40 static DEFINE_SPINLOCK(catas_lock);
42 static LIST_HEAD(catas_list);
43 static struct work_struct catas_work;
45 static int internal_err_reset = 1;
46 module_param(internal_err_reset, int, 0644);
47 MODULE_PARM_DESC(internal_err_reset,
48 "Reset device on internal errors if non-zero (default 1)");
50 static void dump_err_buf(struct mlx4_dev *dev)
52 struct mlx4_priv *priv = mlx4_priv(dev);
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)));
62 static void poll_catas(unsigned long dev_ptr)
64 struct mlx4_dev *dev = (struct mlx4_dev *) dev_ptr;
65 struct mlx4_priv *priv = mlx4_priv(dev);
67 if (readl(priv->catas_err.map)) {
70 mlx4_dispatch_event(dev, MLX4_DEV_EVENT_CATASTROPHIC_ERROR, 0);
72 if (internal_err_reset) {
73 spin_lock(&catas_lock);
74 list_add(&priv->catas_err.list, &catas_list);
75 spin_unlock(&catas_lock);
77 queue_work(mlx4_wq, &catas_work);
80 mod_timer(&priv->catas_err.timer,
81 round_jiffies(jiffies + MLX4_CATAS_POLL_INTERVAL));
84 static void catas_reset(struct work_struct *work)
86 struct mlx4_priv *priv, *tmppriv;
92 if (!mutex_trylock(&drv_mutex))
95 spin_lock_irq(&catas_lock);
96 list_splice_init(&catas_list, &tlist);
97 spin_unlock_irq(&catas_lock);
99 list_for_each_entry_safe(priv, tmppriv, &tlist, catas_err.list) {
100 struct pci_dev *pdev = priv->dev.pdev;
102 ret = mlx4_restart_one(priv->dev.pdev);
103 /* 'priv' now is not valid */
105 printk(KERN_ERR "mlx4 %s: Reset failed (%d)\n",
106 pci_name(pdev), ret);
108 dev = pci_get_drvdata(pdev);
109 mlx4_dbg(dev, "Reset succeeded\n");
112 mutex_unlock(&drv_mutex);
115 void mlx4_start_catas_poll(struct mlx4_dev *dev)
117 struct mlx4_priv *priv = mlx4_priv(dev);
120 INIT_LIST_HEAD(&priv->catas_err.list);
121 init_timer(&priv->catas_err.timer);
122 priv->catas_err.map = NULL;
124 addr = pci_resource_start(dev->pdev, priv->fw.catas_bar) +
125 priv->fw.catas_offset;
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",
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);
141 void mlx4_stop_catas_poll(struct mlx4_dev *dev)
143 struct mlx4_priv *priv = mlx4_priv(dev);
145 del_timer_sync(&priv->catas_err.timer);
147 if (priv->catas_err.map)
148 iounmap(priv->catas_err.map);
150 spin_lock_irq(&catas_lock);
151 list_del(&priv->catas_err.list);
152 spin_unlock_irq(&catas_lock);
155 void __init mlx4_catas_init(void)
157 INIT_WORK(&catas_work, catas_reset);