-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
I've created a patch that monitors changes to the network statistics variables
and keeps internal 64-bit counter. I decided to split it into two parts
(patches are to follow in next emails):
1) generic variable monitoring system (watch64)
The watch64 system allows the programmer to specify the approximate interval
at which he wants his variables checked. If he tries to specify shorter
interval than the minimum a default value of HZ/10 is used. To minimize
locking, RCU and seqlock are used. On 64-bit systems, all is optimized away.
2) network statistics specific patch (64network)
Upon registration of a network device, all the statistics variables are
registered with watch64. Additionally, a new proc file is
created /proc/net/dev64 displays the 64-bit values as supposed
to /proc/net/dev which is left to display the original 32-bit variables for
backward compatibility. The sysfs interface
(/sys/class/net/<interface>/statistics/*) displays the 64-bit values only. On
64-bit systems, all is optimized away through watch64.
Josef "Jeff" Sipek.
- --
*NOTE: This message is ROT-13 encrypted twice for extra protection*
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)
iD8DBQFBOKSzwFP0+seVj/4RAkz7AJ0Ut21nPMkHGKv1dXK17yoA5hQ1+ACglpMq
IHh+tYW3innmwjlA7EU2x78=
=LnHg
-----END PGP SIGNATURE-----
The watch64 system allows the programmer to specify the approximate interval
at which he wants his variables checked. If he tries to specify shorter
interval than the minimum a default value of HZ/10 is used. To minimize
locking, RCU and seqlock are used. On 64-bit systems, all is optimized away.
The following patch can be also pulled from http://jeffpc.bkbits.net/watch64-2.6
Josef "Jeff" Sipek.
Signed-off-by: Josef "Jeff" Sipek <[email protected]>
diff -Nru a/Documentation/00-INDEX b/Documentation/00-INDEX
--- a/Documentation/00-INDEX 2004-09-03 12:21:17 -04:00
+++ b/Documentation/00-INDEX 2004-09-03 12:21:17 -04:00
@@ -250,6 +250,8 @@
- directory with info regarding video/TV/radio cards and linux.
vm/
- directory with info on the Linux vm code.
+watch64.txt
+ - watch64 API description
watchdog/
- how to auto-reboot Linux if it has "fallen and can't get up". ;-)
x86_64/
diff -Nru a/Documentation/watch64.txt b/Documentation/watch64.txt
--- /dev/null Wed Dec 31 16:00:00 196900
+++ b/Documentation/watch64.txt 2004-09-03 12:21:17 -04:00
@@ -0,0 +1,35 @@
+int watch64_register(unsigned long* ptr, unsigned int interval);
+
+ - Registers *ptr to be monitored every interval jiffies.
+ - If interval==0, WATCH64_INTERVAL will be used (HZ/10 by default)
+
+int watch64_unregister(unsigned long* ptr, struct watch64* st);
+
+ - Unregister *ptr
+ - st is optional pointer to the struct containing the registration
+ information
+ - if st==NULL, it will be looked up automatically
+
+struct watch64* watch64_find(unsigned long* ptr);
+
+ - Return struct with registration information of *ptr
+
+int watch64_disable(unsigned long* ptr, struct watch64* st);
+
+ - Disable *ptr from being monitored, without removing it from the list
+ - st is optional (see watch64_unregister for more information)
+
+int watch64_enable(unsigned long* ptr, struct watch64* st);
+
+ - Enable *ptr from being monitored (opposite of watch64_disable)
+ - st is optional (see watch64_unregister for more information)
+
+int watch64_toggle(unsigned long* ptr, struct watch64* st);
+
+ - Toggle the enable/disable status
+ - st is optional (see watch64_unregister for more information)
+
+inline u_int64_t watch64_getval(unsigned long* ptr, struct watch64* st);
+
+ - Return the whole 64-bit counter
+ - st is optional (see watch64_unregister for more information)
diff -Nru a/include/linux/watch64.h b/include/linux/watch64.h
--- /dev/null Wed Dec 31 16:00:00 196900
+++ b/include/linux/watch64.h 2004-09-03 12:21:17 -04:00
@@ -0,0 +1,63 @@
+/*
+ * inclue/linux/watch64.h
+ *
+ * Copyright (C) 2003 Josef "Jeff" Sipek <[email protected]>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ */
+
+#ifndef _LINUX_64WATCH_H
+#define _LINUX_64WATCH_H
+
+#include <linux/list.h>
+#include <linux/timer.h>
+#include <linux/kernel.h>
+#include <linux/seqlock.h>
+#include <linux/rcupdate.h>
+
+#define WATCH64_INTERVAL (HZ/10)
+#define WATCH64_MINIMUM (HZ/20)
+#define WATCH64_MAGIC 0x573634
+
+#if (BITS_PER_LONG == 64)
+
+struct watch64 {
+};
+
+#else
+
+struct watch64 {
+ struct list_head list;
+ unsigned long *ptr;
+ unsigned long oldval;
+ u_int64_t total;
+ unsigned int interval;
+ int active;
+ seqlock_t lock;
+ struct rcu_head rcuhead;
+};
+
+#endif /* (BITS_PER_LONG == 64) */
+
+/*
+ * Prototypes
+ */
+
+void watch64_init(void);
+void watch64_run(unsigned long var);
+int watch64_register(unsigned long* ptr, unsigned int interval);
+int watch64_unregister(unsigned long* ptr, struct watch64* st);
+void watch64_rcufree(struct rcu_head* p);
+struct watch64* watch64_find(unsigned long* ptr);
+inline struct watch64* __watch64_find(unsigned long* ptr);
+int watch64_disable(unsigned long* ptr, struct watch64* st);
+inline int __watch64_disable(unsigned long* ptr, struct watch64* st);
+int watch64_enable(unsigned long* ptr, struct watch64* st);
+inline int __watch64_enable(unsigned long* ptr, struct watch64* st);
+int watch64_toggle(unsigned long* ptr, struct watch64* st);
+inline u_int64_t watch64_getval(unsigned long* ptr, struct watch64* st);
+
+#endif /* _LINUX_WATCH64_H */
diff -Nru a/kernel/Makefile b/kernel/Makefile
--- a/kernel/Makefile 2004-09-03 12:21:17 -04:00
+++ b/kernel/Makefile 2004-09-03 12:21:17 -04:00
@@ -7,7 +7,7 @@
sysctl.o capability.o ptrace.o timer.o user.o \
signal.o sys.o kmod.o workqueue.o pid.o \
rcupdate.o intermodule.o extable.o params.o posix-timers.o \
- kthread.o
+ kthread.o watch64.o
obj-$(CONFIG_FUTEX) += futex.o
obj-$(CONFIG_GENERIC_ISA_DMA) += dma.o
diff -Nru a/kernel/watch64.c b/kernel/watch64.c
--- /dev/null Wed Dec 31 16:00:00 196900
+++ b/kernel/watch64.c 2004-09-03 12:21:17 -04:00
@@ -0,0 +1,392 @@
+/*
+ * kernel/watch64.c
+ *
+ * Copyright (C) 2003 Josef "Jeff" Sipek <[email protected]>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ */
+
+#include <asm/param.h>
+#include <linux/slab.h>
+#include <linux/errno.h>
+#include <linux/jiffies.h>
+#include <linux/module.h>
+#include <linux/list.h>
+#include <linux/timer.h>
+#include <linux/kernel.h>
+#include <linux/seqlock.h>
+#include <linux/spinlock.h>
+#include <linux/rcupdate.h>
+#include <linux/watch64.h>
+
+/*
+ * Watch64 global variables
+ */
+
+spinlock_t watch64_biglock = SPIN_LOCK_UNLOCKED;
+LIST_HEAD(watch64_head);
+struct timer_list watch64_timer;
+int watch64_setup;
+
+#if (BITS_PER_LONG == 64)
+
+void watch64_init(void)
+{
+}
+
+void watch64_run(unsigned long var)
+{
+}
+
+int watch64_register(unsigned long* ptr, unsigned int interval)
+{
+ return 0;
+}
+
+int watch64_unregister(unsigned long* ptr, struct watch64* st)
+{
+ return 0;
+}
+
+void watch64_rcufree(void* p)
+{
+}
+
+struct watch64* watch64_find(unsigned long* ptr)
+{
+ return NULL;
+}
+
+struct watch64* __watch64_find(unsigned long* ptr)
+{
+ return NULL;
+}
+
+int watch64_disable(unsigned long* ptr, struct watch64* st)
+{
+ return 0;
+}
+
+int __watch64_disable(unsigned long* ptr, struct watch64* st)
+{
+ return 0;
+}
+
+int watch64_enable(unsigned long* ptr, struct watch64* st)
+{
+ return 0;
+}
+
+int __watch64_enable(unsigned long* ptr, struct watch64* st)
+{
+ return 0;
+}
+
+int watch64_toggle(unsigned long* ptr, struct watch64* st)
+{
+ return 0;
+}
+
+inline u_int64_t watch64_getval(unsigned long* ptr, struct watch64* st)
+{
+ return (u_int64_t) *ptr;
+}
+
+#else
+
+/*
+ * Initiate watch64 system
+ */
+
+void watch64_init(void)
+{
+ spin_lock(&watch64_biglock);
+
+ if (watch64_setup==WATCH64_MAGIC) {
+ spin_unlock(&watch64_biglock);
+ return;
+ }
+
+ printk(KERN_WARNING "watch64: 2003/08/22 Josef 'Jeff' Sipek <[email protected]>\n");
+ printk(KERN_WARNING "watch64: Enabling Watch64 extensions...");
+
+ init_timer(&watch64_timer);
+ watch64_timer.function = watch64_run;
+ watch64_timer.data = (unsigned long) NULL;
+ watch64_timer.expires = jiffies + WATCH64_MINIMUM;
+ add_timer(&watch64_timer);
+
+ printk("done.\n");
+
+ watch64_setup = WATCH64_MAGIC;
+
+ spin_unlock(&watch64_biglock);
+}
+
+/*
+ * Go through the list of registered variables and check them for changes
+ */
+
+void watch64_run(unsigned long var)
+{
+ struct list_head* entry;
+ struct watch64* watch_struct;
+ unsigned long tmp;
+
+ rcu_read_lock();
+ list_for_each_rcu(entry, &watch64_head) {
+ watch_struct = list_entry(entry, struct watch64, list);
+ if (*watch_struct->ptr != watch_struct->oldval) {
+ tmp = *watch_struct->ptr;
+ if (tmp > watch_struct->oldval) {
+ write_seqlock(&watch_struct->lock);
+ watch_struct->total += tmp - watch_struct->oldval;
+ write_sequnlock(&watch_struct->lock);
+ } else if (tmp < watch_struct->oldval) {
+ write_seqlock(&watch_struct->lock);
+ watch_struct->total += ((u_int64_t) 1<<BITS_PER_LONG) - watch_struct->oldval + tmp;
+ write_sequnlock(&watch_struct->lock);
+ }
+ watch_struct->oldval = tmp;
+ }
+ }
+ rcu_read_unlock();
+
+ mod_timer(&watch64_timer, jiffies + WATCH64_MINIMUM);
+}
+
+/*
+ * Register a new variable with watch64
+ */
+
+int watch64_register(unsigned long* ptr, unsigned int interval)
+{
+ struct watch64* temp;
+
+ temp = (struct watch64*) kmalloc(sizeof(struct watch64),GFP_ATOMIC);
+
+ if (!temp)
+ return -ENOMEM;
+
+ if (watch64_setup!=WATCH64_MAGIC)
+ watch64_init();
+
+ temp->ptr = ptr;
+ temp->oldval = 0;
+ temp->total = 0;
+ if (interval==0)
+ temp->interval = WATCH64_INTERVAL;
+ else if (interval<WATCH64_MINIMUM) {
+ temp->interval = WATCH64_MINIMUM;
+ printk("watch64: attempted to add new watch with interval below %d jiffies",WATCH64_MINIMUM);
+ } else
+ temp->interval = interval;
+
+ temp->active = 0;
+
+ seqlock_init(&temp->lock);
+
+ list_add_rcu(&temp->list, &watch64_head);
+
+ return 0;
+}
+
+/*
+ * Unregister a variable with watch64
+ */
+
+int watch64_unregister(unsigned long* ptr, struct watch64* st)
+{
+ rcu_read_lock();
+ if (!st)
+ st = __watch64_find(ptr);
+
+ if (!st)
+ return -EINVAL;
+
+ __watch64_disable(ptr, st);
+ list_del_rcu(&st->list);
+
+ call_rcu(&st->rcuhead, watch64_rcufree);
+ rcu_read_unlock();
+
+ return 0;
+}
+
+/*
+ * Free memory via RCU
+ */
+
+void watch64_rcufree(struct rcu_head* p)
+{
+ kfree(container_of(p, struct watch64, rcuhead));
+}
+
+/*
+ * Find watch64 structure with RCU lock
+ */
+
+struct watch64* watch64_find(unsigned long* ptr)
+{
+ struct watch64* tmp;
+
+ rcu_read_lock();
+ tmp = __watch64_find(ptr);
+ rcu_read_unlock();
+
+ return tmp;
+}
+
+/*
+ * Find watch64 structure without RCU lock
+ */
+
+inline struct watch64* __watch64_find(unsigned long* ptr)
+{
+ struct list_head* tmp;
+ struct watch64* watch64_struct;
+
+ list_for_each_rcu(tmp, &watch64_head) {
+ watch64_struct = list_entry(tmp, struct watch64, list);
+ if (watch64_struct->ptr==ptr)
+ return watch64_struct;
+ }
+
+ return NULL;
+}
+
+/*
+ * Disable a variable watch with RCU lock
+ */
+
+int watch64_disable(unsigned long* ptr, struct watch64* st)
+{
+ int tmp;
+
+ rcu_read_lock();
+ tmp = __watch64_disable(ptr,st);
+ rcu_read_unlock();
+
+ return tmp;
+}
+
+/*
+ * Disable a variable watch without RCU lock
+ */
+
+inline int __watch64_disable(unsigned long* ptr, struct watch64* st)
+{
+ if (!st)
+ st = watch64_find(ptr);
+
+ if (!st)
+ return -EINVAL;
+
+ st->active = 0;
+
+ return 0;
+}
+
+/*
+ * Enable a variable watch with RCU lock
+ */
+
+int watch64_enable(unsigned long* ptr, struct watch64* st)
+{
+ int tmp;
+
+ rcu_read_lock();
+ tmp = __watch64_enable(ptr,st);
+ rcu_read_unlock();
+
+ return tmp;
+}
+
+/*
+ * Enable a variable watch without RCU lock
+ */
+
+inline int __watch64_enable(unsigned long* ptr, struct watch64* st)
+{
+ if (!st)
+ st = __watch64_find(ptr);
+
+ if (!st)
+ return -EINVAL;
+
+ st->oldval = *ptr;
+ write_seqlock(&st->lock);
+ st->total = (u_int64_t) st->oldval;
+ write_sequnlock(&st->lock);
+ st->active = 1;
+
+ return 0;
+}
+
+/*
+ * Toggle a variable watch
+ */
+
+int watch64_toggle(unsigned long* ptr, struct watch64* st)
+{
+ rcu_read_lock();
+ if (!st)
+ st = __watch64_find(ptr);
+
+ if (!st) {
+ rcu_read_unlock();
+ return -EINVAL;
+ }
+
+ if (st->active)
+ __watch64_disable(ptr,st);
+ else
+ __watch64_enable(ptr,st);
+ rcu_read_unlock();
+
+ return 0;
+}
+
+/*
+ * Return the total 64-bit value
+ */
+
+inline u_int64_t watch64_getval(unsigned long* ptr, struct watch64* st)
+{
+ unsigned int seq;
+ u_int64_t total;
+
+ rcu_read_lock();
+ if (!st)
+ st = __watch64_find(ptr);
+
+ if (!st) {
+ rcu_read_unlock();
+ return *ptr;
+ }
+
+ do {
+ seq = read_seqbegin(&st->lock);
+ total = st->total;
+ } while (read_seqretry(&st->lock, seq));
+ rcu_read_unlock();
+
+ return total;
+}
+
+#endif /* (BITS_PER_LONG == 64) */
+
+/*
+ * Export all the necessary symbols
+ */
+
+EXPORT_SYMBOL(watch64_register);
+EXPORT_SYMBOL(watch64_unregister);
+EXPORT_SYMBOL(watch64_find);
+EXPORT_SYMBOL(watch64_disable);
+EXPORT_SYMBOL(watch64_enable);
+EXPORT_SYMBOL(watch64_toggle);
+EXPORT_SYMBOL(watch64_getval);
Upon registration of a network device, all the statistics variables are
registered with watch64. Additionally, a new proc file is
created /proc/net/dev64 displays the 64-bit values as supposed
to /proc/net/dev which is left to display the original 32-bit variables for
backward compatibility. The sysfs interface
(/sys/class/net/<interface>/statistics/*) displays the 64-bit values only. On
64-bit systems, all is optimized away through watch64.
Requires: watch64
The following patch can be also pulled from http://jeffpc.bkbits.net/64network-2.6 (includes watch64)
Josef "Jeff" Sipek
Signed-off-by: Josef "Jeff" Sipek <[email protected]>
diff -Nru a/include/linux/netdevice.h b/include/linux/netdevice.h
--- a/include/linux/netdevice.h 2004-09-03 12:22:08 -04:00
+++ b/include/linux/netdevice.h 2004-09-03 12:22:08 -04:00
@@ -14,6 +14,7 @@
* Alan Cox, <[email protected]>
* Bjorn Ekwall. <[email protected]>
* Pekka Riikonen <[email protected]>
+ * Josef "Jeff" Sipek <[email protected]>
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
@@ -945,6 +946,10 @@
#ifdef CONFIG_SYSCTL
extern char *net_sysctl_strdup(const char *s);
#endif
+
+/* Register/unregister all the members of struct net_device_stats with watch64 */
+inline void net_register_stats64(struct net_device_stats* stats);
+inline void net_unregister_stats64(struct net_device_stats* stats);
#endif /* __KERNEL__ */
diff -Nru a/net/core/dev.c b/net/core/dev.c
--- a/net/core/dev.c 2004-09-03 12:22:08 -04:00
+++ b/net/core/dev.c 2004-09-03 12:22:08 -04:00
@@ -18,6 +18,7 @@
* Alexey Kuznetsov <[email protected]>
* Adam Sulmicki <[email protected]>
* Pekka Riikonen <[email protected]>
+ * Josef "Jeff" Sipek <[email protected]>
*
* Changes:
* D.J. Barrow : Fixed bug where dev->refcnt gets set
@@ -70,6 +71,7 @@
* indefinitely on dev->refcnt
* J Hadi Salim : - Backlog queue sampling
* - netif_rx() feedback
+ * Josef "Jeff" Sipek : Added watch64 calls for network statistics
*/
#include <asm/uaccess.h>
@@ -108,6 +110,7 @@
#include <linux/kallsyms.h>
#include <linux/netpoll.h>
#include <linux/rcupdate.h>
+#include <linux/watch64.h>
#ifdef CONFIG_NET_RADIO
#include <linux/wireless.h> /* Note : will define WIRELESS_EXT */
#include <net/iw_handler.h>
@@ -2110,6 +2113,49 @@
seq_printf(seq, "%6s: No statistics available.\n", dev->name);
}
+static void dev_seq_printf_stats64(struct seq_file *seq, struct net_device *dev)
+{
+ if (dev->get_stats) {
+ struct net_device_stats *stats = dev->get_stats(dev);
+
+ seq_printf(seq, "%6s:%8llu %7llu %4llu %4llu %4llu %5llu %10llu %9llu "
+ "%8llu %7llu %4llu %4llu %4llu %5llu %7llu %10llu\n",
+ dev->name, watch64_getval(&stats->rx_bytes,NULL),
+ watch64_getval(&stats->rx_packets,NULL),
+ watch64_getval(&stats->rx_errors,NULL),
+ watch64_getval(&stats->rx_dropped,NULL) +
+ watch64_getval(&stats->rx_missed_errors,NULL),
+ watch64_getval(&stats->rx_fifo_errors,NULL),
+ watch64_getval(&stats->rx_length_errors,NULL) +
+ watch64_getval(&stats->rx_over_errors,NULL) +
+ watch64_getval(&stats->rx_crc_errors,NULL) +
+ watch64_getval(&stats->rx_frame_errors,NULL),
+ watch64_getval(&stats->rx_compressed,NULL),
+ watch64_getval(&stats->multicast,NULL),
+ watch64_getval(&stats->tx_bytes,NULL),
+ watch64_getval(&stats->tx_packets,NULL),
+ watch64_getval(&stats->tx_errors,NULL),
+ watch64_getval(&stats->tx_dropped,NULL),
+ watch64_getval(&stats->tx_fifo_errors,NULL),
+ watch64_getval(&stats->collisions,NULL),
+ watch64_getval(&stats->tx_carrier_errors,NULL) +
+ watch64_getval(&stats->tx_aborted_errors,NULL) +
+ watch64_getval(&stats->tx_window_errors,NULL) +
+ watch64_getval(&stats->tx_heartbeat_errors,NULL),
+ watch64_getval(&stats->tx_compressed,NULL));
+ } else
+ seq_printf(seq, "%6s: No statistics available.\n", dev->name);
+}
+
+static void dev_seq_show_header(struct seq_file *seq)
+{
+ seq_puts(seq, "Inter-| Receive "
+ " | Transmit\n"
+ " face |bytes packets errs drop fifo frame "
+ "compressed multicast|bytes packets errs "
+ "drop fifo colls carrier compressed\n");
+}
+
/*
* Called from the PROCfs module. This now uses the new arbitrary sized
* /proc/net interface to create /proc/net/dev
@@ -2117,16 +2163,21 @@
static int dev_seq_show(struct seq_file *seq, void *v)
{
if (v == SEQ_START_TOKEN)
- seq_puts(seq, "Inter-| Receive "
- " | Transmit\n"
- " face |bytes packets errs drop fifo frame "
- "compressed multicast|bytes packets errs "
- "drop fifo colls carrier compressed\n");
+ dev_seq_show_header(seq);
else
dev_seq_printf_stats(seq, v);
return 0;
}
+static int dev_seq_show64(struct seq_file *seq, void *v)
+{
+ if (v == SEQ_START_TOKEN)
+ dev_seq_show_header(seq);
+ else
+ dev_seq_printf_stats64(seq, v);
+ return 0;
+}
+
static struct netif_rx_stats *softnet_get_online(loff_t *pos)
{
struct netif_rx_stats *rc = NULL;
@@ -2179,11 +2230,23 @@
.show = dev_seq_show,
};
+static struct seq_operations dev_seq_ops64 = {
+ .start = dev_seq_start,
+ .next = dev_seq_next,
+ .stop = dev_seq_stop,
+ .show = dev_seq_show64,
+};
+
static int dev_seq_open(struct inode *inode, struct file *file)
{
return seq_open(file, &dev_seq_ops);
}
+static int dev_seq_open64(struct inode *inode, struct file *file)
+{
+ return seq_open(file, &dev_seq_ops64);
+}
+
static struct file_operations dev_seq_fops = {
.owner = THIS_MODULE,
.open = dev_seq_open,
@@ -2192,6 +2255,14 @@
.release = seq_release,
};
+static struct file_operations dev_seq_fops64 = {
+ .owner = THIS_MODULE,
+ .open = dev_seq_open64,
+ .read = seq_read,
+ .llseek = seq_lseek,
+ .release = seq_release,
+};
+
static struct seq_operations softnet_seq_ops = {
.start = softnet_seq_start,
.next = softnet_seq_next,
@@ -2224,8 +2295,10 @@
if (!proc_net_fops_create("dev", S_IRUGO, &dev_seq_fops))
goto out;
- if (!proc_net_fops_create("softnet_stat", S_IRUGO, &softnet_seq_fops))
+ if (!proc_net_fops_create("dev64", S_IRUGO, &dev_seq_fops64))
goto out_dev;
+ if (!proc_net_fops_create("softnet_stat", S_IRUGO, &softnet_seq_fops))
+ goto out_dev64;
if (wireless_proc_init())
goto out_softnet;
rc = 0;
@@ -2233,6 +2306,8 @@
return rc;
out_softnet:
proc_net_remove("softnet_stat");
+out_dev64:
+ proc_net_remove("dev64");
out_dev:
proc_net_remove("dev");
goto out;
@@ -2910,6 +2985,9 @@
* device is present.
*/
+ if (dev->get_stats)
+ net_register_stats64(dev->get_stats(dev));
+
set_bit(__LINK_STATE_PRESENT, &dev->state);
dev->next = NULL;
@@ -2922,7 +3000,7 @@
dev_hold(dev);
dev->reg_state = NETREG_REGISTERING;
write_unlock_bh(&dev_base_lock);
-
+
/* Notify protocols, that a new device appeared. */
notifier_call_chain(&netdev_chain, NETDEV_REGISTER, dev);
@@ -3145,6 +3223,9 @@
/* If device is running, close it first. */
if (dev->flags & IFF_UP)
dev_close(dev);
+
+ if (dev->get_stats)
+ net_unregister_stats64(dev->get_stats(dev));
/* And unlink it from device chain. */
for (dp = &dev_base; (d = *dp) != NULL; dp = &d->next) {
@@ -3246,6 +3327,98 @@
}
#endif /* CONFIG_HOTPLUG_CPU */
+/*
+ * Register all the members of the net_device_stats structure
+ *
+ */
+
+inline void net_register_stats64(struct net_device_stats* stats)
+{
+ if (!stats)
+ return;
+
+ watch64_register(&stats->tx_packets,0);
+ watch64_enable (&stats->tx_packets,NULL);
+ watch64_register(&stats->rx_packets,0);
+ watch64_enable (&stats->rx_packets,NULL);
+ watch64_register(&stats->tx_bytes,0);
+ watch64_enable (&stats->tx_bytes,NULL);
+ watch64_register(&stats->rx_bytes,0);
+ watch64_enable (&stats->rx_bytes,NULL);
+ watch64_register(&stats->tx_errors,0);
+ watch64_enable (&stats->tx_errors,NULL);
+ watch64_register(&stats->rx_errors,0);
+ watch64_enable (&stats->rx_errors,NULL);
+ watch64_register(&stats->tx_dropped,0);
+ watch64_enable (&stats->tx_dropped,NULL);
+ watch64_register(&stats->rx_dropped,0);
+ watch64_enable (&stats->rx_dropped,NULL);
+ watch64_register(&stats->multicast,0);
+ watch64_enable (&stats->multicast,NULL);
+ watch64_register(&stats->collisions,0);
+ watch64_enable (&stats->collisions,NULL);
+ watch64_register(&stats->rx_length_errors,0);
+ watch64_enable (&stats->rx_length_errors,NULL);
+ watch64_register(&stats->rx_over_errors,0);
+ watch64_enable (&stats->rx_over_errors,NULL);
+ watch64_register(&stats->rx_crc_errors,0);
+ watch64_enable (&stats->rx_crc_errors,NULL);
+ watch64_register(&stats->rx_frame_errors,0);
+ watch64_enable (&stats->rx_frame_errors,NULL);
+ watch64_register(&stats->rx_fifo_errors,0);
+ watch64_enable (&stats->rx_fifo_errors,NULL);
+ watch64_register(&stats->rx_missed_errors,0);
+ watch64_enable (&stats->rx_missed_errors,NULL);
+ watch64_register(&stats->tx_aborted_errors,0);
+ watch64_enable (&stats->tx_aborted_errors,NULL);
+ watch64_register(&stats->tx_carrier_errors,0);
+ watch64_enable (&stats->tx_carrier_errors,NULL);
+ watch64_register(&stats->tx_fifo_errors,0);
+ watch64_enable (&stats->tx_fifo_errors,NULL);
+ watch64_register(&stats->tx_heartbeat_errors,0);
+ watch64_enable (&stats->tx_heartbeat_errors,NULL);
+ watch64_register(&stats->tx_window_errors,0);
+ watch64_enable (&stats->tx_window_errors,NULL);
+ watch64_register(&stats->rx_compressed,0);
+ watch64_enable (&stats->rx_compressed,NULL);
+ watch64_register(&stats->tx_compressed,0);
+ watch64_enable (&stats->tx_compressed,NULL);
+}
+
+/*
+ * Unregister all the members of the net_device_stats structure
+ *
+ */
+
+inline void net_unregister_stats64(struct net_device_stats* stats)
+{
+ if (!stats)
+ return;
+
+ watch64_unregister(&stats->tx_packets,0);
+ watch64_unregister(&stats->rx_packets,0);
+ watch64_unregister(&stats->tx_bytes,0);
+ watch64_unregister(&stats->rx_bytes,0);
+ watch64_unregister(&stats->tx_errors,0);
+ watch64_unregister(&stats->rx_errors,0);
+ watch64_unregister(&stats->tx_dropped,0);
+ watch64_unregister(&stats->rx_dropped,0);
+ watch64_unregister(&stats->multicast,0);
+ watch64_unregister(&stats->collisions,0);
+ watch64_unregister(&stats->rx_length_errors,0);
+ watch64_unregister(&stats->rx_over_errors,0);
+ watch64_unregister(&stats->rx_crc_errors,0);
+ watch64_unregister(&stats->rx_frame_errors,0);
+ watch64_unregister(&stats->rx_fifo_errors,0);
+ watch64_unregister(&stats->rx_missed_errors,0);
+ watch64_unregister(&stats->tx_aborted_errors,0);
+ watch64_unregister(&stats->tx_carrier_errors,0);
+ watch64_unregister(&stats->tx_fifo_errors,0);
+ watch64_unregister(&stats->tx_heartbeat_errors,0);
+ watch64_unregister(&stats->tx_window_errors,0);
+ watch64_unregister(&stats->rx_compressed,0);
+ watch64_unregister(&stats->tx_compressed,0);
+}
/*
* Initialize the DEV module. At boot time this walks the device list and
diff -Nru a/net/core/net-sysfs.c b/net/core/net-sysfs.c
--- a/net/core/net-sysfs.c 2004-09-03 12:22:08 -04:00
+++ b/net/core/net-sysfs.c 2004-09-03 12:22:08 -04:00
@@ -16,6 +16,7 @@
#include <net/sock.h>
#include <linux/rtnetlink.h>
#include <linux/wireless.h>
+#include <linux/watch64.h>
#define to_class_dev(obj) container_of(obj,struct class_device,kobj)
#define to_net_dev(class) container_of(class, struct net_device, class_dev)
@@ -23,6 +24,7 @@
static const char fmt_hex[] = "%#x\n";
static const char fmt_dec[] = "%d\n";
static const char fmt_ulong[] = "%lu\n";
+static const char fmt_ullong[] = "%llu\n";
static inline int dev_isalive(const struct net_device *dev)
{
@@ -204,8 +206,8 @@
read_lock(&dev_base_lock);
if (dev_isalive(dev) && dev->get_stats &&
(stats = (*dev->get_stats)(dev)))
- ret = sprintf(buf, fmt_ulong,
- *(unsigned long *)(((u8 *) stats) + offset));
+ ret = sprintf(buf, fmt_ullong,
+ watch64_getval((unsigned long *)(((u8 *) stats) + offset),NULL));
read_unlock(&dev_base_lock);
return ret;
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Friday 03 September 2004 13:06, josef Jeff Sipek wrote:
> I've created a patch that monitors changes to the network statistics
> variables and keeps internal 64-bit counter. I decided to split it into two
> parts (patches are to follow in next emails):
> 1) generic variable monitoring system (watch64)
> 2) network statistics specific patch (64network)
Btw, both of these patches apply cleanly against 2.6.9-rc1-bk10.
Jeff.
- --
bad pun of the week: the formula 1 control computer suffered from a race
condition
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)
iD8DBQFBOKjQwFP0+seVj/4RApg/AKDEFSTVOMSvVh9zVU65o/P6ZcfBxgCffeId
QddOVsR+uHdkV2D4/U8QVO4=
=jQIT
-----END PGP SIGNATURE-----
In article <[email protected]> (at Fri, 03 Sep 2004 13:19:24 -0400), "Josef 'Jeff' Sipek" <[email protected]> says:
> The watch64 system allows the programmer to specify the approximate interval
> at which he wants his variables checked. If he tries to specify shorter
> interval than the minimum a default value of HZ/10 is used. To minimize
> locking, RCU and seqlock are used. On 64-bit systems, all is optimized away.
I agree with the basic principle; it is very similar to mine.
However, it is too complicated isn't it?
I would do per-"table" registration (instead of per-variable one);
watch64_getval() seems very ugly to me...
--
Hideaki YOSHIFUJI @ USAGI Project <[email protected]>
GPG FP: 9022 65EB 1ECF 3AD1 0BDF 80D8 4807 F894 E062 0EEA
On Fri, 03 Sep 2004 13:19:24 -0400
"Josef 'Jeff' Sipek" <[email protected]> wrote:
> The watch64 system allows the programmer to specify the approximate interval
> at which he wants his variables checked. If he tries to specify shorter
> interval than the minimum a default value of HZ/10 is used. To minimize
> locking, RCU and seqlock are used. On 64-bit systems, all is optimized away.
>
> The following patch can be also pulled from http://jeffpc.bkbits.net/watch64-2.6
>
> Josef "Jeff" Sipek.
>
> Signed-off-by: Josef "Jeff" Sipek <[email protected]>
- Seems like a big interface for a simple problem.
- Code doesn't match the kernel style (read Documentation/CodingStyle)
- Does it really need RCU and seqlock, wouldn't one suffice
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Friday 03 September 2004 15:16, you wrote:
> - Seems like a big interface for a simple problem.
I agree that the interface could be simplified a lot, however I decided to
make it as generic as possible so that any part of the kernel can use it. I
also am contemplating whether it would make sense to remove the capability of
setting the interval to any value, and just have all variables checked every
WATCH64_INTERVAL.
> - Code doesn't match the kernel style (read Documentation/CodingStyle)
Agreed. I'm going submit a correction patch.
> - Does it really need RCU and seqlock, wouldn't one suffice
As far as I can tell, it does. RCU to protect the linked list, and seqlock to
protect the internal 64-bit counter.
Jeff.
- --
Research, n.:
Consider Columbus:
He didn't know where he was going.
When he got there he didn't know where he was.
When he got back he didn't know where he had been.
And he did it all on someone else's money.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)
iD8DBQFBONGlwFP0+seVj/4RAlAFAJ9OiipyTpCGljodcwYgDSlXaWB3LACgnWLk
ZZLfkwnTFdsihHvTmfR+AuA=
=g5hE
-----END PGP SIGNATURE-----
The following fixes watch64 patch previously submitted to follow CodingStyle
guidelines. BK repo is up to date as well
Jeff.
Signed-off-by: Josef "Jeff" Sipek <[email protected]>
--- 1.7/kernel/watch64.c 2004-07-14 16:41:26 -04:00
+++ edited/watch64.c 2004-09-03 16:12:39 -04:00
@@ -110,7 +110,8 @@
return;
}
- printk(KERN_WARNING "watch64: 2003/08/22 Josef 'Jeff' Sipek <[email protected]>\n");
+ printk(KERN_WARNING "watch64: 2003/08/22 Josef 'Jeff' Sipek "
+ "<[email protected]>\n");
printk(KERN_WARNING "watch64: Enabling Watch64 extensions...");
init_timer(&watch64_timer);
@@ -139,19 +140,21 @@
rcu_read_lock();
list_for_each_rcu(entry, &watch64_head) {
watch_struct = list_entry(entry, struct watch64, list);
- if (*watch_struct->ptr != watch_struct->oldval) {
- tmp = *watch_struct->ptr;
- if (tmp > watch_struct->oldval) {
- write_seqlock(&watch_struct->lock);
- watch_struct->total += tmp - watch_struct->oldval;
- write_sequnlock(&watch_struct->lock);
- } else if (tmp < watch_struct->oldval) {
- write_seqlock(&watch_struct->lock);
- watch_struct->total += ((u_int64_t) 1<<BITS_PER_LONG) - watch_struct->oldval + tmp;
- write_sequnlock(&watch_struct->lock);
- }
- watch_struct->oldval = tmp;
+ if (*watch_struct->ptr == watch_struct->oldval)
+ continue;
+
+ tmp = *watch_struct->ptr;
+ if (tmp > watch_struct->oldval) {
+ write_seqlock(&watch_struct->lock);
+ watch_struct->total += tmp - watch_struct->oldval;
+ write_sequnlock(&watch_struct->lock);
+ } else if (tmp < watch_struct->oldval) {
+ write_seqlock(&watch_struct->lock);
+ watch_struct->total += ((u_int64_t) 1<<BITS_PER_LONG)
+ - watch_struct->oldval + tmp;
+ write_sequnlock(&watch_struct->lock);
}
+ watch_struct->oldval = tmp;
}
rcu_read_unlock();
@@ -181,7 +184,8 @@
temp->interval = WATCH64_INTERVAL;
else if (interval<WATCH64_MINIMUM) {
temp->interval = WATCH64_MINIMUM;
- printk("watch64: attempted to add new watch with interval below %d jiffies",WATCH64_MINIMUM);
+ printk("watch64: attempted to add new watch with "
+ "interval below %d jiffies",WATCH64_MINIMUM);
} else
temp->interval = interval;
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Friday 03 September 2004 15:16, Stephen Hemminger wrote:
> - Code doesn't match the kernel style (read Documentation/CodingStyle)
Sorry about the white space, KMail apparently likes to butcher the text. These
are the same patches with the little cleanup update.
Jeff.
- --
Reality is merely an illusion, albeit a very persistent one.
- Albert Einstein
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)
iD4DBQFBOOW+wFP0+seVj/4RAgSiAJj54qcqdEx66lbMW9ik0XviupTNAKC82an1
R0pGX0pTBZ78NWrZpxJm+w==
=EesC
-----END PGP SIGNATURE-----
I have a feeling this was discussed somewhere(other than netdev) and i
missed it. Why isnt this watch64 being done in user space?
cheers,
jamal
On Fri, 2004-09-03 at 17:44, Josef 'Jeff' Sipek wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> On Friday 03 September 2004 15:16, Stephen Hemminger wrote:
> > - Code doesn't match the kernel style (read Documentation/CodingStyle)
>
> Sorry about the white space, KMail apparently likes to butcher the text. These
> are the same patches with the little cleanup update.
>
> Jeff.
>
> - --
> Reality is merely an illusion, albeit a very persistent one.
> - Albert Einstein
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.2.5 (GNU/Linux)
>
> iD4DBQFBOOW+wFP0+seVj/4RAgSiAJj54qcqdEx66lbMW9ik0XviupTNAKC82an1
> R0pGX0pTBZ78NWrZpxJm+w==
> =EesC
> -----END PGP SIGNATURE-----
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Saturday 04 September 2004 09:19, jamal wrote:
> I have a feeling this was discussed somewhere(other than netdev) and i
> missed it. Why isnt this watch64 being done in user space?
There was a discussion about 64-bit network statistics about a year ago on
lkml.
watch64 is a generic so that anyone in the kernel can use it.
Jeff.
- --
Mankind invented the atomic bomb, but no mouse would ever construct a
mousetrap.
- Albert Einstein
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)
iD8DBQFBOzybwFP0+seVj/4RArmhAKC3ddX4ZGoAMQKxGplXqqbER9BBMQCfencW
wDt06dC8MifG9NU3xWx0ULo=
=z9kC
-----END PGP SIGNATURE-----
On Sun, 2004-09-05 at 12:19, Jeff Sipek wrote:
> On Saturday 04 September 2004 09:19, jamal wrote:
> > I have a feeling this was discussed somewhere(other than netdev) and i
> > missed it. Why isnt this watch64 being done in user space?
>
> There was a discussion about 64-bit network statistics about a year ago on
> lkml.
Sorry unsubscribed from lkml since summer of '94. [net related
discussions should really happen on netdev].
> watch64 is a generic so that anyone in the kernel can use it.
Ok - so why does this have to be in the kernel?
cheers,
jamal
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Sorry, I missed your email.
On Tuesday 07 September 2004 06:18, jamal wrote:
> On Sun, 2004-09-05 at 12:19, Jeff Sipek wrote:
> > There was a discussion about 64-bit network statistics about a year ago
> > on lkml.
>
> Sorry unsubscribed from lkml since summer of '94. [net related
> discussions should really happen on netdev].
netdev was CC'd during this whole discussion.
> > watch64 is a generic so that anyone in the kernel can use it.
>
> Ok - so why does this have to be in the kernel?
I think it is convenient to have the 64 bit net stats reported by the kernel.
Jeff.
- --
My public GPG key can be found at
http://shells.gnugeneration.com/~jeffpc/gpg/public-0xC7958FFE.txt
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)
iD8DBQFBRHENwFP0+seVj/4RAuzcAKCdBAooPae8pTaMEHbWmVDKAO7C5ACeLi21
cen/Ag4bH5Dm9xkQiXj+d0Q=
=BrV+
-----END PGP SIGNATURE-----