This patch adds the Advantech WDT driver (ported from 2.2.19pre,
only minimal changes) to the 2.4.2 kernel.
Please Cc: any replies to me (I'm not subscribed to linux-kernel).
Thanks,
Marek
diff -urN v2.4.2/linux/drivers/char/Config.in linux/drivers/char/Config.in
--- v2.4.2/linux/drivers/char/Config.in Tue Feb 13 23:13:43 2001
+++ linux/drivers/char/Config.in Tue Feb 27 10:33:31 2001
@@ -133,6 +133,7 @@
fi
tristate ' Berkshire Products PC Watchdog' CONFIG_PCWATCHDOG
tristate ' Acquire SBC Watchdog Timer' CONFIG_ACQUIRE_WDT
+ tristate ' Advantech SBC Watchdog Timer' CONFIG_ADVANTECH_WDT
tristate ' SBC-60XX Watchdog Timer' CONFIG_60XX_WDT
tristate ' Mixcom Watchdog' CONFIG_MIXCOMWD
tristate ' Intel i810 TCO timer / Watchdog' CONFIG_I810_TCO
diff -urN v2.4.2/linux/drivers/char/Makefile linux/drivers/char/Makefile
--- v2.4.2/linux/drivers/char/Makefile Thu Jan 4 22:00:55 2001
+++ linux/drivers/char/Makefile Tue Feb 27 10:34:28 2001
@@ -177,6 +177,7 @@
obj-$(CONFIG_PCWATCHDOG) += pcwd.o
obj-$(CONFIG_ACQUIRE_WDT) += acquirewdt.o
+obj-$(CONFIG_ADVANTECH_WDT) += advantechwdt.o
obj-$(CONFIG_MIXCOMWD) += mixcomwd.o
obj-$(CONFIG_60XX_WDT) += sbc60xxwdt.o
obj-$(CONFIG_WDT) += wdt.o
diff -urN v2.4.2/linux/drivers/char/advantechwdt.c linux/drivers/char/advantechwdt.c
--- v2.4.2/linux/drivers/char/advantechwdt.c Thu Jan 1 01:00:00 1970
+++ linux/drivers/char/advantechwdt.c Tue Feb 27 11:07:22 2001
@@ -0,0 +1,241 @@
+/*
+ * Advantech Single Board Computer WDT driver for Linux 2.4.x
+ *
+ * (c) Copyright 2000-2001 Marek Michalkiewicz <[email protected]>
+ *
+ * Based on acquirewdt.c which is based on wdt.c.
+ * Original copyright messages:
+ *
+ * (c) Copyright 1996 Alan Cox <[email protected]>, All Rights Reserved.
+ * http://www.redhat.com
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version
+ * 2 of the License, or (at your option) any later version.
+ *
+ * Neither Alan Cox nor CymruNet Ltd. admit liability nor provide
+ * warranty for any of this software. This material is provided
+ * "AS-IS" and at no charge.
+ *
+ * (c) Copyright 1995 Alan Cox <[email protected]>
+ *
+ */
+
+#include <linux/config.h>
+#include <linux/module.h>
+#include <linux/version.h>
+#include <linux/types.h>
+#include <linux/errno.h>
+#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/miscdevice.h>
+#include <linux/watchdog.h>
+#include <linux/slab.h>
+#include <linux/ioport.h>
+#include <linux/fcntl.h>
+#include <asm/io.h>
+#include <asm/uaccess.h>
+#include <asm/system.h>
+#include <linux/notifier.h>
+#include <linux/reboot.h>
+#include <linux/init.h>
+#include <linux/spinlock.h>
+#include <linux/smp_lock.h>
+
+static int advwdt_is_open;
+static spinlock_t advwdt_lock;
+
+/*
+ * You must set these - there is no sane way to probe for this board.
+ *
+ * To enable or restart, write the timeout value in seconds (1 to 63)
+ * to I/O port WDT_START. To disable, read I/O port WDT_STOP.
+ * Both are 0x443 for most boards (tested on a PCA-6276VE-00B1), but
+ * check your manual (at least the PCA-6159 seems to be different -
+ * the manual says WDT_STOP is 0x43, not 0x443).
+ * (0x43 is also a write-only control register for the 8254 timer!)
+ *
+ * TODO: module parameters to set the I/O port addresses and NOWAYOUT
+ * option at load time.
+ */
+
+#define WDT_STOP 0x443
+#define WDT_START 0x443
+
+#define WD_TIMO 60 /* 1 minute */
+
+/*
+ * Kernel methods.
+ */
+
+static void
+advwdt_ping(void)
+{
+ /* Write a watchdog value */
+ outb_p(WD_TIMO, WDT_START);
+}
+
+static ssize_t
+advwdt_write(struct file *file, const char *buf, size_t count, loff_t *ppos)
+{
+ /* Can't seek (pwrite) on this device */
+ if (ppos != &file->f_pos)
+ return -ESPIPE;
+
+ if (count) {
+ advwdt_ping();
+ return 1;
+ }
+ return 0;
+}
+
+static ssize_t
+advwdt_read(struct file *file, char *buf, size_t count, loff_t *ppos)
+{
+ return -EINVAL;
+}
+
+static int
+advwdt_ioctl(struct inode *inode, struct file *file, unsigned int cmd,
+ unsigned long arg)
+{
+ static struct watchdog_info ident = {
+ WDIOF_KEEPALIVEPING, 1, "Advantech WDT"
+ };
+
+ switch (cmd) {
+ case WDIOC_GETSUPPORT:
+ if (copy_to_user((struct watchdog_info *)arg, &ident, sizeof(ident)))
+ return -EFAULT;
+ break;
+
+ case WDIOC_GETSTATUS:
+ if (copy_to_user((int *)arg, &advwdt_is_open, sizeof(int)))
+ return -EFAULT;
+ break;
+
+ case WDIOC_KEEPALIVE:
+ advwdt_ping();
+ break;
+
+ default:
+ return -ENOIOCTLCMD;
+ }
+ return 0;
+}
+
+static int
+advwdt_open(struct inode *inode, struct file *file)
+{
+ switch (MINOR(inode->i_rdev)) {
+ case WATCHDOG_MINOR:
+ spin_lock(&advwdt_lock);
+ if (advwdt_is_open) {
+ spin_unlock(&advwdt_lock);
+ return -EBUSY;
+ }
+ /*
+ * Activate
+ */
+
+ advwdt_is_open = 1;
+ advwdt_ping();
+ spin_unlock(&advwdt_lock);
+ return 0;
+ default:
+ return -ENODEV;
+ }
+}
+
+static int
+advwdt_close(struct inode *inode, struct file *file)
+{
+ lock_kernel();
+ if (MINOR(inode->i_rdev) == WATCHDOG_MINOR) {
+ spin_lock(&advwdt_lock);
+#ifndef CONFIG_WATCHDOG_NOWAYOUT
+ inb_p(WDT_STOP);
+#endif
+ advwdt_is_open = 0;
+ spin_unlock(&advwdt_lock);
+ }
+ unlock_kernel();
+ return 0;
+}
+
+/*
+ * Notifier for system down
+ */
+
+static int
+advwdt_notify_sys(struct notifier_block *this, unsigned long code,
+ void *unused)
+{
+ if (code == SYS_DOWN || code == SYS_HALT) {
+ /* Turn the WDT off */
+ inb_p(WDT_STOP);
+ }
+ return NOTIFY_DONE;
+}
+
+/*
+ * Kernel Interfaces
+ */
+
+static struct file_operations advwdt_fops = {
+ owner: THIS_MODULE,
+ read: advwdt_read,
+ write: advwdt_write,
+ ioctl: advwdt_ioctl,
+ open: advwdt_open,
+ release: advwdt_close,
+};
+
+static struct miscdevice advwdt_miscdev = {
+ WATCHDOG_MINOR,
+ "watchdog",
+ &advwdt_fops
+};
+
+/*
+ * The WDT needs to learn about soft shutdowns in order to
+ * turn the timebomb registers off.
+ */
+
+static struct notifier_block advwdt_notifier = {
+ advwdt_notify_sys,
+ NULL,
+ 0
+};
+
+static int __init
+advwdt_init(void)
+{
+ printk("WDT driver for Advantech single board computer initialising.\n");
+
+ spin_lock_init(&advwdt_lock);
+ misc_register(&advwdt_miscdev);
+#if WDT_START != WDT_STOP
+ request_region(WDT_STOP, 1, "Advantech WDT");
+#endif
+ request_region(WDT_START, 1, "Advantech WDT");
+ register_reboot_notifier(&advwdt_notifier);
+ return 0;
+}
+
+static void __exit
+advwdt_exit(void)
+{
+ misc_deregister(&advwdt_miscdev);
+ unregister_reboot_notifier(&advwdt_notifier);
+#if WDT_START != WDT_STOP
+ release_region(WDT_STOP,1);
+#endif
+ release_region(WDT_START,1);
+}
+
+module_init(advwdt_init);
+module_exit(advwdt_exit);
+
+/* end of advantechwdt.c */
Marek Michalkiewicz wrote:
> +static int
> +advwdt_close(struct inode *inode, struct file *file)
> +{
> + lock_kernel();
> + if (MINOR(inode->i_rdev) == WATCHDOG_MINOR) {
> + spin_lock(&advwdt_lock);
> +#ifndef CONFIG_WATCHDOG_NOWAYOUT
> + inb_p(WDT_STOP);
> +#endif
> + advwdt_is_open = 0;
> + spin_unlock(&advwdt_lock);
> + }
> + unlock_kernel();
> + return 0;
> +}
Why is lock_kernel necessary?
> +static int __init
> +advwdt_init(void)
> +{
> + printk("WDT driver for Advantech single board computer initialising.\n");
> +
> + spin_lock_init(&advwdt_lock);
> + misc_register(&advwdt_miscdev);
check return code for error
> +#if WDT_START != WDT_STOP
> + request_region(WDT_STOP, 1, "Advantech WDT");
> +#endif
> + request_region(WDT_START, 1, "Advantech WDT");
check return..
> + register_reboot_notifier(&advwdt_notifier);
ditto
--
Jeff Garzik | "You see, in this world there's two kinds of
Building 1024 | people, my friend: Those with loaded guns
MandrakeSoft | and those who dig. You dig." --Blondie
> Why is lock_kernel necessary?
Well, it is there in 2.4.2 acquirewdt.c (which this driver is based on -
really only minimal changes, the hardware is only slightly different).
I can remove it if you tell me it's really not necessary.
> > + spin_lock_init(&advwdt_lock);
> > + misc_register(&advwdt_miscdev);
>
> check return code for error
Again, acquirewdt.c doesn't do it with the misc_register, request_region
and register_reboot_notifier calls. (Should it? Cc: Alan Cox)
Thanks,
Marek
Marek Michalkiewicz wrote:
>
> > Why is lock_kernel necessary?
>
> Well, it is there in 2.4.2 acquirewdt.c (which this driver is based on -
> really only minimal changes, the hardware is only slightly different).
> I can remove it if you tell me it's really not necessary.
Sounds like it got caught in an audit and is necessary, I didn't look
closely...
> > > + spin_lock_init(&advwdt_lock);
> > > + misc_register(&advwdt_miscdev);
> >
> > check return code for error
>
> Again, acquirewdt.c doesn't do it with the misc_register, request_region
> and register_reboot_notifier calls. (Should it? Cc: Alan Cox)
Yes, the return codes should be checked. :)
--
Jeff Garzik | "You see, in this world there's two kinds of
Building 1024 | people, my friend: Those with loaded guns
MandrakeSoft | and those who dig. You dig." --Blondie