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 work_struct catas_work;
47 static int internal_err_reset = 1;
48 module_param(internal_err_reset, int, 0644);
49 MODULE_PARM_DESC(internal_err_reset,
50 "Reset device on internal errors if non-zero (default 1)");
52 static void dump_err_buf(struct mlx4_dev *dev)
54 struct mlx4_priv *priv = mlx4_priv(dev);
58 mlx4_err(dev, "Internal error detected:\n");
59 for (i = 0; i < priv->fw.catas_size; ++i)
60 mlx4_err(dev, " buf[%02x]: %08x\n",
61 i, swab32(readl(priv->catas_err.map + i)));
64 static void poll_catas(unsigned long dev_ptr)
66 struct mlx4_dev *dev = (struct mlx4_dev *) dev_ptr;
67 struct mlx4_priv *priv = mlx4_priv(dev);
69 if (readl(priv->catas_err.map)) {
72 mlx4_dispatch_event(dev, MLX4_DEV_EVENT_CATASTROPHIC_ERROR, 0);
74 if (internal_err_reset) {
75 spin_lock(&catas_lock);
76 list_add(&priv->catas_err.list, &catas_list);
77 spin_unlock(&catas_lock);
79 queue_work(mlx4_wq, &catas_work);
82 mod_timer(&priv->catas_err.timer,
83 round_jiffies(jiffies + MLX4_CATAS_POLL_INTERVAL));
86 static void catas_reset(struct work_struct *work)
88 struct mlx4_priv *priv, *tmppriv;
94 spin_lock_irq(&catas_lock);
95 list_splice_init(&catas_list, &tlist);
96 spin_unlock_irq(&catas_lock);
98 list_for_each_entry_safe(priv, tmppriv, &tlist, catas_err.list) {
99 ret = mlx4_restart_one(priv->dev.pdev);
102 mlx4_err(dev, "Reset failed (%d)\n", ret);
104 mlx4_dbg(dev, "Reset succeeded\n");
108 void mlx4_start_catas_poll(struct mlx4_dev *dev)
110 struct mlx4_priv *priv = mlx4_priv(dev);
113 INIT_LIST_HEAD(&priv->catas_err.list);
114 init_timer(&priv->catas_err.timer);
115 priv->catas_err.map = NULL;
117 addr = pci_resource_start(dev->pdev, priv->fw.catas_bar) +
118 priv->fw.catas_offset;
120 priv->catas_err.map = ioremap(addr, priv->fw.catas_size * 4);
121 if (!priv->catas_err.map) {
122 mlx4_warn(dev, "Failed to map internal error buffer at 0x%lx\n",
127 priv->catas_err.timer.data = (unsigned long) dev;
128 priv->catas_err.timer.function = poll_catas;
129 priv->catas_err.timer.expires =
130 round_jiffies(jiffies + MLX4_CATAS_POLL_INTERVAL);
131 add_timer(&priv->catas_err.timer);
134 void mlx4_stop_catas_poll(struct mlx4_dev *dev)
136 struct mlx4_priv *priv = mlx4_priv(dev);
138 del_timer_sync(&priv->catas_err.timer);
140 if (priv->catas_err.map)
141 iounmap(priv->catas_err.map);
143 spin_lock_irq(&catas_lock);
144 list_del(&priv->catas_err.list);
145 spin_unlock_irq(&catas_lock);
148 void __init mlx4_catas_init(void)
150 INIT_WORK(&catas_work, catas_reset);