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>
39 MLX4_CATAS_POLL_INTERVAL = 5 * HZ,
42 static DEFINE_SPINLOCK(catas_lock);
44 static LIST_HEAD(catas_list);
45 static struct workqueue_struct *catas_wq;
46 static struct work_struct catas_work;
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)");
53 static void dump_err_buf(struct mlx4_dev *dev)
55 struct mlx4_priv *priv = mlx4_priv(dev);
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)));
65 static void poll_catas(unsigned long dev_ptr)
67 struct mlx4_dev *dev = (struct mlx4_dev *) dev_ptr;
68 struct mlx4_priv *priv = mlx4_priv(dev);
70 if (readl(priv->catas_err.map)) {
73 mlx4_dispatch_event(dev, MLX4_DEV_EVENT_CATASTROPHIC_ERROR, 0);
75 if (internal_err_reset) {
76 spin_lock(&catas_lock);
77 list_add(&priv->catas_err.list, &catas_list);
78 spin_unlock(&catas_lock);
80 queue_work(catas_wq, &catas_work);
83 mod_timer(&priv->catas_err.timer,
84 round_jiffies(jiffies + MLX4_CATAS_POLL_INTERVAL));
87 static void catas_reset(struct work_struct *work)
89 struct mlx4_priv *priv, *tmppriv;
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 ret = mlx4_restart_one(priv->dev.pdev);
103 mlx4_err(dev, "Reset failed (%d)\n", ret);
105 mlx4_dbg(dev, "Reset succeeded\n");
109 void mlx4_start_catas_poll(struct mlx4_dev *dev)
111 struct mlx4_priv *priv = mlx4_priv(dev);
114 INIT_LIST_HEAD(&priv->catas_err.list);
115 init_timer(&priv->catas_err.timer);
116 priv->catas_err.map = NULL;
118 addr = pci_resource_start(dev->pdev, priv->fw.catas_bar) +
119 priv->fw.catas_offset;
121 priv->catas_err.map = ioremap(addr, priv->fw.catas_size * 4);
122 if (!priv->catas_err.map) {
123 mlx4_warn(dev, "Failed to map internal error buffer at 0x%lx\n",
128 priv->catas_err.timer.data = (unsigned long) dev;
129 priv->catas_err.timer.function = poll_catas;
130 priv->catas_err.timer.expires =
131 round_jiffies(jiffies + MLX4_CATAS_POLL_INTERVAL);
132 add_timer(&priv->catas_err.timer);
135 void mlx4_stop_catas_poll(struct mlx4_dev *dev)
137 struct mlx4_priv *priv = mlx4_priv(dev);
139 del_timer_sync(&priv->catas_err.timer);
141 if (priv->catas_err.map)
142 iounmap(priv->catas_err.map);
144 spin_lock_irq(&catas_lock);
145 list_del(&priv->catas_err.list);
146 spin_unlock_irq(&catas_lock);
149 int __init mlx4_catas_init(void)
151 INIT_WORK(&catas_work, catas_reset);
153 catas_wq = create_singlethread_workqueue("mlx4_err");
160 void mlx4_catas_cleanup(void)
162 destroy_workqueue(catas_wq);