1 files changed, 203 insertions, 87 deletions

diff --git a/net/ipv4/tcp_probe.c b/net/ipv4/tcp_probe.c
index f230eeecf09..3b66610d415 100644
--- a/net/ipv4/tcp_probe.c
+++ b/net/ipv4/tcp_probe.c
@@ -6,8 +6,7 @@
  *
  * 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.
+ * the Free Software Foundation; either version 2 of the License.
  *
  * This program is distributed in the hope that it will be useful,
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
@@ -19,168 +18,285 @@
  * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  */
 
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
 #include <linux/kernel.h>
 #include <linux/kprobes.h>
 #include <linux/socket.h>
 #include <linux/tcp.h>
+#include <linux/slab.h>
 #include <linux/proc_fs.h>
 #include <linux/module.h>
-#include <linux/kfifo.h>
-#include <linux/vmalloc.h>
+#include <linux/ktime.h>
+#include <linux/time.h>
+#include <net/net_namespace.h>
 
 #include <net/tcp.h>
 
-MODULE_AUTHOR("Stephen Hemminger <shemminger@osdl.org>");
+MODULE_AUTHOR("Stephen Hemminger <shemminger@linux-foundation.org>");
 MODULE_DESCRIPTION("TCP cwnd snooper");
 MODULE_LICENSE("GPL");
+MODULE_VERSION("1.1");
 
-static int port = 0;
+static int port __read_mostly;
 MODULE_PARM_DESC(port, "Port to match (0=all)");
 module_param(port, int, 0);
 
-static int bufsize = 64*1024;
-MODULE_PARM_DESC(bufsize, "Log buffer size (default 64k)");
-module_param(bufsize, int, 0);
+static unsigned int bufsize __read_mostly = 4096;
+MODULE_PARM_DESC(bufsize, "Log buffer size in packets (4096)");
+module_param(bufsize, uint, 0);
+
+static unsigned int fwmark __read_mostly;
+MODULE_PARM_DESC(fwmark, "skb mark to match (0=no mark)");
+module_param(fwmark, uint, 0);
+
+static int full __read_mostly;
+MODULE_PARM_DESC(full, "Full log (1=every ack packet received,  0=only cwnd changes)");
+module_param(full, int, 0);
 
 static const char procname[] = "tcpprobe";
 
-struct {
-	struct kfifo  *fifo;
-	spinlock_t    lock;
+struct tcp_log {
+	ktime_t tstamp;
+	union {
+		struct sockaddr		raw;
+		struct sockaddr_in	v4;
+		struct sockaddr_in6	v6;
+	}	src, dst;
+	u16	length;
+	u32	snd_nxt;
+	u32	snd_una;
+	u32	snd_wnd;
+	u32	rcv_wnd;
+	u32	snd_cwnd;
+	u32	ssthresh;
+	u32	srtt;
+};
+
+static struct {
+	spinlock_t	lock;
 	wait_queue_head_t wait;
-	struct timeval tstart;
-} tcpw;
+	ktime_t		start;
+	u32		lastcwnd;
+
+	unsigned long	head, tail;
+	struct tcp_log	*log;
+} tcp_probe;
 
-static void printl(const char *fmt, ...)
-{
-	va_list args;
-	int len;
-	struct timeval now;
-	char tbuf[256];
-
-	va_start(args, fmt);
-	do_gettimeofday(&now);
-
-	now.tv_sec -= tcpw.tstart.tv_sec;
-	now.tv_usec -= tcpw.tstart.tv_usec;
-	if (now.tv_usec < 0) {
-		--now.tv_sec;
-		now.tv_usec += 1000000;
-	}
 
-	len = sprintf(tbuf, "%lu.%06lu ",
-		      (unsigned long) now.tv_sec,
-		      (unsigned long) now.tv_usec);
-	len += vscnprintf(tbuf+len, sizeof(tbuf)-len, fmt, args);
-	va_end(args);
+static inline int tcp_probe_used(void)
+{
+	return (tcp_probe.head - tcp_probe.tail) & (bufsize - 1);
+}
 
-	kfifo_put(tcpw.fifo, tbuf, len);
-	wake_up(&tcpw.wait);
+static inline int tcp_probe_avail(void)
+{
+	return bufsize - tcp_probe_used() - 1;
 }
 
-static int jtcp_sendmsg(struct kiocb *iocb, struct sock *sk,
-			struct msghdr *msg, size_t size)
+#define tcp_probe_copy_fl_to_si4(inet, si4, mem)		\
+	do {							\
+		si4.sin_family = AF_INET;			\
+		si4.sin_port = inet->inet_##mem##port;		\
+		si4.sin_addr.s_addr = inet->inet_##mem##addr;	\
+	} while (0)						\
+
+
+/*
+ * Hook inserted to be called before each receive packet.
+ * Note: arguments must match tcp_rcv_established()!
+ */
+static void jtcp_rcv_established(struct sock *sk, struct sk_buff *skb,
+				 const struct tcphdr *th, unsigned int len)
 {
 	const struct tcp_sock *tp = tcp_sk(sk);
 	const struct inet_sock *inet = inet_sk(sk);
 
-	if (port == 0 || ntohs(inet->dport) == port ||
-	    ntohs(inet->sport) == port) {
-		printl("%d.%d.%d.%d:%u %d.%d.%d.%d:%u %d %#x %#x %u %u %u\n",
-		       NIPQUAD(inet->saddr), ntohs(inet->sport),
-		       NIPQUAD(inet->daddr), ntohs(inet->dport),
-		       size, tp->snd_nxt, tp->snd_una,
-		       tp->snd_cwnd, tcp_current_ssthresh(sk),
-		       tp->snd_wnd);
+	/* Only update if port or skb mark matches */
+	if (((port == 0 && fwmark == 0) ||
+	     ntohs(inet->inet_dport) == port ||
+	     ntohs(inet->inet_sport) == port ||
+	     (fwmark > 0 && skb->mark == fwmark)) &&
+	    (full || tp->snd_cwnd != tcp_probe.lastcwnd)) {
+
+		spin_lock(&tcp_probe.lock);
+		/* If log fills, just silently drop */
+		if (tcp_probe_avail() > 1) {
+			struct tcp_log *p = tcp_probe.log + tcp_probe.head;
+
+			p->tstamp = ktime_get();
+			switch (sk->sk_family) {
+			case AF_INET:
+				tcp_probe_copy_fl_to_si4(inet, p->src.v4, s);
+				tcp_probe_copy_fl_to_si4(inet, p->dst.v4, d);
+				break;
+			case AF_INET6:
+				memset(&p->src.v6, 0, sizeof(p->src.v6));
+				memset(&p->dst.v6, 0, sizeof(p->dst.v6));
+#if IS_ENABLED(CONFIG_IPV6)
+				p->src.v6.sin6_family = AF_INET6;
+				p->src.v6.sin6_port = inet->inet_sport;
+				p->src.v6.sin6_addr = inet6_sk(sk)->saddr;
+
+				p->dst.v6.sin6_family = AF_INET6;
+				p->dst.v6.sin6_port = inet->inet_dport;
+				p->dst.v6.sin6_addr = sk->sk_v6_daddr;
+#endif
+				break;
+			default:
+				BUG();
+			}
+
+			p->length = skb->len;
+			p->snd_nxt = tp->snd_nxt;
+			p->snd_una = tp->snd_una;
+			p->snd_cwnd = tp->snd_cwnd;
+			p->snd_wnd = tp->snd_wnd;
+			p->rcv_wnd = tp->rcv_wnd;
+			p->ssthresh = tcp_current_ssthresh(sk);
+			p->srtt = tp->srtt_us >> 3;
+
+			tcp_probe.head = (tcp_probe.head + 1) & (bufsize - 1);
+		}
+		tcp_probe.lastcwnd = tp->snd_cwnd;
+		spin_unlock(&tcp_probe.lock);
+
+		wake_up(&tcp_probe.wait);
 	}
 
 	jprobe_return();
-	return 0;
 }
 
-static struct jprobe tcp_send_probe = {
+static struct jprobe tcp_jprobe = {
 	.kp = {
-		.symbol_name	= "tcp_sendmsg",
+		.symbol_name	= "tcp_rcv_established",
 	},
-	.entry	= JPROBE_ENTRY(jtcp_sendmsg),
+	.entry	= jtcp_rcv_established,
 };
 
-
-static int tcpprobe_open(struct inode * inode, struct file * file)
+static int tcpprobe_open(struct inode *inode, struct file *file)
 {
-	kfifo_reset(tcpw.fifo);
-	do_gettimeofday(&tcpw.tstart);
+	/* Reset (empty) log */
+	spin_lock_bh(&tcp_probe.lock);
+	tcp_probe.head = tcp_probe.tail = 0;
+	tcp_probe.start = ktime_get();
+	spin_unlock_bh(&tcp_probe.lock);
+
 	return 0;
 }
 
+static int tcpprobe_sprint(char *tbuf, int n)
+{
+	const struct tcp_log *p
+		= tcp_probe.log + tcp_probe.tail;
+	struct timespec tv
+		= ktime_to_timespec(ktime_sub(p->tstamp, tcp_probe.start));
+
+	return scnprintf(tbuf, n,
+			"%lu.%09lu %pISpc %pISpc %d %#x %#x %u %u %u %u %u\n",
+			(unsigned long) tv.tv_sec,
+			(unsigned long) tv.tv_nsec,
+			&p->src, &p->dst, p->length, p->snd_nxt, p->snd_una,
+			p->snd_cwnd, p->ssthresh, p->snd_wnd, p->srtt, p->rcv_wnd);
+}
+
 static ssize_t tcpprobe_read(struct file *file, char __user *buf,
 			     size_t len, loff_t *ppos)
 {
-	int error = 0, cnt = 0;
-	unsigned char *tbuf;
+	int error = 0;
+	size_t cnt = 0;
 
-	if (!buf || len < 0)
+	if (!buf)
 		return -EINVAL;
 
-	if (len == 0)
-		return 0;
+	while (cnt < len) {
+		char tbuf[256];
+		int width;
+
+		/* Wait for data in buffer */
+		error = wait_event_interruptible(tcp_probe.wait,
+						 tcp_probe_used() > 0);
+		if (error)
+			break;
 
-	tbuf = vmalloc(len);
-	if (!tbuf)
-		return -ENOMEM;
+		spin_lock_bh(&tcp_probe.lock);
+		if (tcp_probe.head == tcp_probe.tail) {
+			/* multiple readers race? */
+			spin_unlock_bh(&tcp_probe.lock);
+			continue;
+		}
 
-	error = wait_event_interruptible(tcpw.wait,
-					 __kfifo_len(tcpw.fifo) != 0);
-	if (error)
-		goto out_free;
+		width = tcpprobe_sprint(tbuf, sizeof(tbuf));
 
-	cnt = kfifo_get(tcpw.fifo, tbuf, len);
-	error = copy_to_user(buf, tbuf, cnt);
+		if (cnt + width < len)
+			tcp_probe.tail = (tcp_probe.tail + 1) & (bufsize - 1);
 
-out_free:
-	vfree(tbuf);
+		spin_unlock_bh(&tcp_probe.lock);
 
-	return error ? error : cnt;
+		/* if record greater than space available
+		   return partial buffer (so far) */
+		if (cnt + width >= len)
+			break;
+
+		if (copy_to_user(buf + cnt, tbuf, width))
+			return -EFAULT;
+		cnt += width;
+	}
+
+	return cnt == 0 ? error : cnt;
 }
 
-static struct file_operations tcpprobe_fops = {
+static const struct file_operations tcpprobe_fops = {
 	.owner	 = THIS_MODULE,
 	.open	 = tcpprobe_open,
 	.read    = tcpprobe_read,
+	.llseek  = noop_llseek,
 };
 
 static __init int tcpprobe_init(void)
 {
 	int ret = -ENOMEM;
 
-	init_waitqueue_head(&tcpw.wait);
-	spin_lock_init(&tcpw.lock);
-	tcpw.fifo = kfifo_alloc(bufsize, GFP_KERNEL, &tcpw.lock);
-	if (IS_ERR(tcpw.fifo))
-		return PTR_ERR(tcpw.fifo);
+	/* Warning: if the function signature of tcp_rcv_established,
+	 * has been changed, you also have to change the signature of
+	 * jtcp_rcv_established, otherwise you end up right here!
+	 */
+	BUILD_BUG_ON(__same_type(tcp_rcv_established,
+				 jtcp_rcv_established) == 0);
+
+	init_waitqueue_head(&tcp_probe.wait);
+	spin_lock_init(&tcp_probe.lock);
+
+	if (bufsize == 0)
+		return -EINVAL;
 
-	if (!proc_net_fops_create(procname, S_IRUSR, &tcpprobe_fops))
+	bufsize = roundup_pow_of_two(bufsize);
+	tcp_probe.log = kcalloc(bufsize, sizeof(struct tcp_log), GFP_KERNEL);
+	if (!tcp_probe.log)
 		goto err0;
 
-	ret = register_jprobe(&tcp_send_probe);
+	if (!proc_create(procname, S_IRUSR, init_net.proc_net, &tcpprobe_fops))
+		goto err0;
+
+	ret = register_jprobe(&tcp_jprobe);
 	if (ret)
 		goto err1;
 
-	pr_info("TCP watch registered (port=%d)\n", port);
+	pr_info("probe registered (port=%d/fwmark=%u) bufsize=%u\n",
+		port, fwmark, bufsize);
 	return 0;
  err1:
-	proc_net_remove(procname);
+	remove_proc_entry(procname, init_net.proc_net);
  err0:
-	kfifo_free(tcpw.fifo);
+	kfree(tcp_probe.log);
 	return ret;
 }
 module_init(tcpprobe_init);
 
 static __exit void tcpprobe_exit(void)
 {
-	kfifo_free(tcpw.fifo);
-	proc_net_remove(procname);
-	unregister_jprobe(&tcp_send_probe);
-
+	remove_proc_entry(procname, init_net.proc_net);
+	unregister_jprobe(&tcp_jprobe);
+	kfree(tcp_probe.log);
 }
 module_exit(tcpprobe_exit);