[PATCH] remove superfluous diverter printk'ing

[PATCH] remove superfluous diverter printk'ing

[Lennert Buytenhek]

Hi Benoit,

You're not listed as diverter maintainer but I'm sending this via
you anyway.  This patch removes some printk'ing from the diverter
driver (and cleans up the code a bit), because after a while it gets
really annoying to have log messages like this one for every single
tunnel you create or tear down.

	divert: not allocating divert_blk for non-ethernet device tunl0

Please consider sending upstream.


cheers,
Lennert


--- linux-2.6.10/net/core/dv.c.orig	2005-01-14 13:36:44.507808269 +0100
+++ linux-2.6.10/net/core/dv.c	2005-01-14 13:38:18.808399830 +0100
@@ -52,26 +52,20 @@
 {
 	int alloc_size = (sizeof(struct divert_blk) + 3) & ~3;
 
+	dev->divert = NULL;
 	if (dev->type == ARPHRD_ETHER) {
-		printk(KERN_DEBUG "divert: allocating divert_blk for %s\n",
-		       dev->name);
-
 		dev->divert = (struct divert_blk *)
 			kmalloc(alloc_size, GFP_KERNEL);
 		if (dev->divert == NULL) {
-			printk(KERN_DEBUG "divert: unable to allocate divert_blk for %s\n",
+			printk(KERN_INFO "divert: unable to allocate divert_blk for %s\n",
 			       dev->name);
 			return -ENOMEM;
-		} else {
-			memset(dev->divert, 0, sizeof(struct divert_blk));
 		}
-		dev_hold(dev);
-	} else {
-		printk(KERN_DEBUG "divert: not allocating divert_blk for non-ethernet device %s\n",
-		       dev->name);
 
-		dev->divert = NULL;
+		memset(dev->divert, 0, sizeof(struct divert_blk));
+		dev_hold(dev);
 	}
+
 	return 0;
 } 
 
@@ -85,11 +79,6 @@
 		kfree(dev->divert);
 		dev->divert=NULL;
 		dev_put(dev);
-		printk(KERN_DEBUG "divert: freeing divert_blk for %s\n",
-		       dev->name);
-	} else {
-		printk(KERN_DEBUG "divert: no divert_blk to free, %s not ethernet\n",
-		       dev->name);
 	}
 }
 
@@ -192,8 +181,12 @@
 /*
  * control function of the diverter
  */
+#if 0
 #define	DVDBG(a)	\
 	printk(KERN_DEBUG "divert_ioctl() line %d %s\n", __LINE__, (a))
+#else
+#define	DVDBG(a)
+#endif
 
 int divert_ioctl(unsigned int cmd, struct divert_cf __user *arg)
 {

Re: [PATCH] remove superfluous diverter printk'ing

[David S. Miller]

On Fri, 14 Jan 2005 13:47:00 +0100
Lennert Buytenhek <buytenh@wantstofly.org> wrote:

> You're not listed as diverter maintainer but I'm sending this via
> you anyway.  This patch removes some printk'ing from the diverter
> driver (and cleans up the code a bit), because after a while it gets
> really annoying to have log messages like this one for every single
> tunnel you create or tear down.
> 
> 	divert: not allocating divert_blk for non-ethernet device tunl0
> 
> Please consider sending upstream.

Those diverter messages drive me crazy too, I think I'll
add your patch Lennert.

Re: [PATCH] remove superfluous diverter printk'ing

[Lennert Buytenhek]

On Fri, Jan 14, 2005 at 01:46:42PM -0800, David S. Miller wrote:

> > You're not listed as diverter maintainer but I'm sending this via
> > you anyway.  This patch removes some printk'ing from the diverter
> > driver (and cleans up the code a bit), because after a while it gets
> > really annoying to have log messages like this one for every single
> > tunnel you create or tear down.
> > 
> > 	divert: not allocating divert_blk for non-ethernet device tunl0
> > 
> > Please consider sending upstream.
> 
> Those diverter messages drive me crazy too, I think I'll
> add your patch Lennert.

Benoit's email address bounces, and it's the latest address of his I
could find, so if he doesn't read this list you might as well go ahead.


cheers,
Lennert

Re: [PATCH] remove superfluous diverter printk'ing

[jamal]

Diverter could simply be killed now that tc actions exist if theres no
maintainer or user.
Should be able to divert any packet of choce to any interface.

cheers,
jamal


On Sat, 2005-01-15 at 05:47, Lennert Buytenhek wrote:
> On Fri, Jan 14, 2005 at 01:46:42PM -0800, David S. Miller wrote:
> 
> > > You're not listed as diverter maintainer but I'm sending this via
> > > you anyway.  This patch removes some printk'ing from the diverter
> > > driver (and cleans up the code a bit), because after a while it gets
> > > really annoying to have log messages like this one for every single
> > > tunnel you create or tear down.
> > > 
> > > 	divert: not allocating divert_blk for non-ethernet device tunl0
> > > 
> > > Please consider sending upstream.
> > 
> > Those diverter messages drive me crazy too, I think I'll
> > add your patch Lennert.
> 
> Benoit's email address bounces, and it's the latest address of his I
> could find, so if he doesn't read this list you might as well go ahead.
> 
> 
> cheers,
> Lennert
> 
> 

Re: [PATCH] remove superfluous diverter printk'ing

[Lennert Buytenhek]

On Sun, Jan 16, 2005 at 12:10:21PM -0500, jamal wrote:

> Diverter could simply be killed now that tc actions exist if theres no
> maintainer or user.
> Should be able to divert any packet of choce to any interface.

The diverter is not used for redirecting a packet to another interface.

What it does (IIRC) is overwrite the destination MAC address on selected
packets so that the local host will process them.  You can do the same
thing with iptables' "-j REDIRECT" but the diverter was made for use on
ethernet bridges (REDIRECTing bridged packets) when the bridge-nf stuff
didn't exist yet.

Nowadays we have bridge-nf and you should be able to do the same thing
by just using '-j REDIRECT' as you would do with routing (Bart, please
correct me if I'm wrong there.)


--L

Re: [PATCH] remove superfluous diverter printk'ing

[jamal]

On Sun, 2005-01-16 at 12:22, Lennert Buytenhek wrote:
> On Sun, Jan 16, 2005 at 12:10:21PM -0500, jamal wrote:
> 
> > Diverter could simply be killed now that tc actions exist if theres no
> > maintainer or user.
> > Should be able to divert any packet of choce to any interface.
> 
> The diverter is not used for redirecting a packet to another interface.
> 
> What it does (IIRC) is overwrite the destination MAC address on selected
> packets so that the local host will process them.  You can do the same
> thing with iptables' "-j REDIRECT" but the diverter was made for use on
> ethernet bridges (REDIRECTing bridged packets) when the bridge-nf stuff
> didn't exist yet.

the action code is hit before bridging, so something along the lines of:

#attach to ingress of eth0
tc qdisc add dev eth0 ingress 
# Munge MAC address if coming from 192.168.200.200/32
tc filter add dev eth0 parent ffff: protocol ip prio 10 u32 \
match ip src 192.168.200.200/32 flowid 1:2 \
action pedit munge offset -16 u16 set 0x0000 \
munge offset -12 u32 set 0x00020200 \
munge offset -8 u32 set 0x0001AF0A \
munge offset -4 u32 set 0x06EC0800

pedit could be taught better (macros needed) to do:

action pedit munge set src MAC  00:01:AF:0A:06:EC 
instead of all that raw formating

Or if you want to redirect it after rewriting MAC to eth1:
tc filter add dev eth0 parent ffff: protocol ip prio 10 u32 \
match ip src 192.168.200.200/32 flowid 1:2 \
action pedit munge set src MAC  00:01:AF:0A:06:EC \
mirred egress redirect dev eth1

cheers,
jamal