Debian Bug report logs - #254243
netstat: chops IPv6 addresses

Toggle useless messages

Message #5 received at submit@bugs.debian.org (full text, mbox, reply):

From: martin f krafft <madduck@debian.org>

To: Debian Bug Tracking System <submit@bugs.debian.org>

Subject: netstat: chops IPv6 addresses

Date: Sun, 13 Jun 2004 22:18:58 +0200

[Message part 1 (text/plain, inline)]

Package: net-tools
Version: 1.60-10
Severity: minor

I could not find a reference to this in the docs, but it's rather
annoying that netstat can't be made to display full IPv6 addresses.
SSH uses tcp6 by default, so IPv4 addresses are shown with the IPv6
counterparts, but the last octet is chopped off:

tcp6  0  1216 ::ffff:130.60.xx.xx:22  ::ffff:217.233.53:36428 ESTABLISHED
                                             ^^^^^^^^^^
It would be nice to see the full address!

Thanks for consideration!

-- System Information:
Debian Release: testing/unstable
  APT prefers testing
  APT policy: (600, 'testing'), (98, 'unstable')
Architecture: i386 (i686)
Kernel: Linux 2.6.6-1-k7
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8

Versions of packages net-tools depends on:
ii  libc6                       2.3.2.ds1-13 GNU C Library: Shared libraries an

-- no debconf information

-- 
Please do not CC me when replying to lists; I read them!
 
 .''`.     martin f. krafft <madduck@debian.org>
: :'  :    proud Debian developer, admin, and user
`. `'`
  `-  Debian - when you have better things to do than fixing a system
 
Invalid/expired PGP subkeys? Use subkeys.pgp.net as keyserver!

[signature.asc (application/pgp-signature, inline)]

Message #12 received at 254243-quiet@bugs.debian.org (full text, mbox, reply):

From: martin f krafft <madduck@debian.org>

To: 254243-quiet@bugs.debian.org

Subject: Fwd: net-tools : netstat ipv6 display bug

Date: Sun, 24 Jul 2005 13:03:32 +0200

[Message part 1 (text/plain, inline)]

tags 254243 + patch
thanks

----- Forwarded message from Omry Yadan <omry_y@inter.net.il> -----

790,791c790,797
<       if ((strlen(rem_addr) + strlen(buffer)) > 22)
<           rem_addr[22 - strlen(buffer)] = '\0';
---
>       // THIS IS VERY WRONG FOR IPV6 ADDRESSES AS IT TRIMS THE END OF
>       // THE ADDRESS.
>       // COMMENTED OUT TO FIX IT.
>       // A MORE SERIOUS FIX IS IN ORDER.
>       // Omry Yadan (omry_y@inter.net.il)
>       //
>       //if ((strlen(rem_addr) + strlen(buffer)) > 22)
>       //    rem_addr[22 - strlen(buffer)] = '\0';

----- End forwarded message -----

[signature.asc (application/pgp-signature, inline)]

Message #17 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Ian Jackson <ian@davenant.greenend.org.uk>

To: 254243@bugs.debian.org, Phil Blundell <philb@gnu.org>, Bernd Eckenfels <net-tools@lina.inka.de>

Subject: netstat IPv6 address truncation patch

Date: Tue, 20 Sep 2005 16:05:45 +0100

I am about to apply the attached patch to Ubuntu's version of
net-tools.  Here is the debian/changelog entry:

net-tools (1.60-15ubuntu1) breezy; urgency=low

  * Do not truncate numeric IPv6 addresses in netstat TCP and UDP.
    The new output format is not parseable with the same algorithms as
    before because it will be flexible about field widths, but it will not
    be used unless it is necessary to avoid truncating numeric addresses,
    so no existing non-broken netstat-output parsing code will be broken.
    (Also: made 4 copies(!) of endpoint formatting code into one
    common function.  Urgh.)  Ubuntu 7294, Debian #254243.

 -- Ian Jackson <iwj@ubuntu.com>  Tue, 20 Sep 2005 15:28:53 +0100

I think that applying this patch to Debian's net-tools should fix
#254243, and it would probably be useful upstream too.

Looking at the code it looks like some serious overhaul is required.
Unfortunately I don't think that I have time to do this now,
especially at this point in Ubuntu's release cycle.

Thanks,
Ian.

diff -u net-tools-1.60/netstat.c net-tools-1.60/netstat.c
--- net-tools-1.60/netstat.c
+++ net-tools-1.60/netstat.c
@@ -85,6 +85,7 @@
 #include <sys/ioctl.h>
 #include <net/if.h>
 #include <dirent.h>
+#include <assert.h>
 
 #include "net-support.h"
 #include "pathnames.h"
@@ -704,11 +705,71 @@
 	       igmp_do_one);
 }
 
+static void tcpudp_one_end(char *result_buf, size_t result_buf_size,
+			   struct aftype *ap, int end_port,
+			   struct sockaddr *endaddr)
+{
+    char buffer[8192];
+    
+    snprintf(buffer, sizeof(buffer), "%s",
+	     get_sname(htons(end_port), "udp",
+		       flag_not & FLAG_NUM_PORT));
+
+    assert(result_buf_size > 23);
+
+    safe_strncpy(result_buf, ap->sprint(endaddr, flag_not), result_buf_size);
+    if ((strlen(result_buf) + strlen(buffer)) > 22 &&
+	!(flag_not & FLAG_NUM_HOST))
+	result_buf[22 - strlen(buffer)] = '\0';
+
+    assert(strlen(result_buf) + strlen(buffer) + 1 < sizeof(buffer));
+    strcat(result_buf, ":");
+    strcat(result_buf, buffer);
+}
+
+static void reduce_deficit(int *deficit, int *spc, int min_spc)
+{
+    int reduction;
+    reduction= netmin(*deficit, *spc - min_spc);
+    if (reduction <= 0) return;
+    *spc -= reduction;
+    *deficit -= reduction;
+}
+
+static void tcpudp_write(const char *protname,
+			 unsigned long rxq, unsigned long txq,
+			 const char *local_addr, const char *rem_addr,
+			 const char *state) {
+    int rxq_spc, txq_spc, local_addr_spc, rem_addr_spc, protname_spc;
+    int local_addr_len, rem_addr_len, deficit;
+
+    protname_spc = 5;
+    rxq_spc = 6;
+    txq_spc = 6;
+    local_addr_spc = 23;
+    local_addr_len = strlen(local_addr);
+    rem_addr_spc = 23;
+    rem_addr_len = strlen(rem_addr);
+    deficit = netmax(local_addr_len - local_addr_spc, 0) +
+	netmax(rem_addr_len - rem_addr_spc, 0);
+    
+    reduce_deficit(&deficit, &rem_addr_spc, rem_addr_len);
+    reduce_deficit(&deficit, &local_addr_spc, local_addr_len);
+    reduce_deficit(&deficit, &txq_spc, 1);
+    reduce_deficit(&deficit, &rxq_spc, 1);
+    reduce_deficit(&deficit, &protname_spc, strlen(protname));
+		       
+    printf("%-*s %*ld %*ld %-*s %-*s %-11s",
+	   protname_spc, protname, rxq_spc, rxq, txq_spc, txq,
+	   local_addr_spc, local_addr,
+	   rem_addr_spc, rem_addr, state);
+}
+
 static void tcp_do_one(int lnr, const char *line)
 {
     unsigned long rxq, txq, time_len, retr, inode;
     int num, local_port, rem_port, d, state, uid, timer_run, timeout;
-    char rem_addr[128], local_addr[128], timers[64], buffer[1024], more[512];
+    char rem_addr[128], local_addr[128], timers[64], more[512];
     char *protname;
     struct aftype *ap;
 #if HAVE_AFINET6
@@ -775,25 +836,13 @@
     safe_strncpy(rem_addr, ap->sprint((struct sockaddr *) &remaddr, flag_not),
 		 sizeof(rem_addr));
     if (flag_all || (flag_lst && !rem_port) || (!flag_lst && rem_port)) {
-	snprintf(buffer, sizeof(buffer), "%s",
-		 get_sname(htons(local_port), "tcp",
-			   flag_not & FLAG_NUM_PORT));
-
-	if ((strlen(local_addr) + strlen(buffer)) > 22)
-	    local_addr[22 - strlen(buffer)] = '\0';
-
-	strcat(local_addr, ":");
-	strcat(local_addr, buffer);
-	snprintf(buffer, sizeof(buffer), "%s",
-		 get_sname(htons(rem_port), "tcp", flag_not & FLAG_NUM_PORT));
+	tcpudp_one_end(local_addr, sizeof(local_addr),
+		       ap, local_port, (struct sockaddr *) &localaddr);
 
-	if ((strlen(rem_addr) + strlen(buffer)) > 22)
-	    rem_addr[22 - strlen(buffer)] = '\0';
+	tcpudp_one_end(rem_addr, sizeof(rem_addr),
+		       ap, rem_port, (struct sockaddr *) &remaddr);
 
-	strcat(rem_addr, ":");
-	strcat(rem_addr, buffer);
 	timers[0] = '\0';
-
 	if (flag_opt)
 	    switch (timer_run) {
 	    case 0:
@@ -820,8 +869,11 @@
 			 timer_run, (double) time_len / HZ, retr, timeout);
 		break;
 	    }
-	printf("%-4s  %6ld %6ld %-*s %-*s %-11s",
-	       protname, rxq, txq, netmax(23,strlen(local_addr)), local_addr, netmax(23,strlen(rem_addr)), rem_addr, _(tcp_state[state]));
+
+	tcpudp_write(protname,
+		     rxq, txq,
+		     local_addr, rem_addr,
+		     _(tcp_state[state]));
 
 	finish_this_one(uid,inode,timers);
     }
@@ -835,7 +887,7 @@
 
 static void udp_do_one(int lnr, const char *line)
 {
-    char buffer[8192], local_addr[64], rem_addr[64];
+    char local_addr[64], rem_addr[64];
     char *udp_state, timers[64], more[512];
     int num, local_port, rem_port, d, state, timer_run, uid, timeout;
     char *protname;
@@ -927,24 +979,11 @@
 
     if (flag_all || (notnull(remaddr) && !flag_lst) || (!notnull(remaddr) && flag_lst))
     {
-        safe_strncpy(local_addr, ap->sprint((struct sockaddr *) &localaddr, 
-					    flag_not), sizeof(local_addr));
-	snprintf(buffer, sizeof(buffer), "%s",
-		 get_sname(htons(local_port), "udp",
-			   flag_not & FLAG_NUM_PORT));
-	if ((strlen(local_addr) + strlen(buffer)) > 22)
-	    local_addr[22 - strlen(buffer)] = '\0';
-	strcat(local_addr, ":");
-	strcat(local_addr, buffer);
-
-	snprintf(buffer, sizeof(buffer), "%s",
-		 get_sname(htons(rem_port), "udp", flag_not & FLAG_NUM_PORT));
-        safe_strncpy(rem_addr, ap->sprint((struct sockaddr *) &remaddr, 
-					  flag_not), sizeof(rem_addr));
-	if ((strlen(rem_addr) + strlen(buffer)) > 22)
-	    rem_addr[22 - strlen(buffer)] = '\0';
-	strcat(rem_addr, ":");
-	strcat(rem_addr, buffer);
+	tcpudp_one_end(local_addr, sizeof(local_addr),
+		       ap, local_port, (struct sockaddr *) &localaddr);
+
+	tcpudp_one_end(rem_addr, sizeof(rem_addr),
+		       ap, rem_port, (struct sockaddr *) &remaddr);
 
 	timers[0] = '\0';
 	if (flag_opt)
@@ -963,8 +1002,11 @@
 			 retr, timeout);
 		break;
 	    }
-	printf("%-4s  %6ld %6ld %-23s %-23s %-11s",
-	       protname, rxq, txq, local_addr, rem_addr, udp_state);
+
+	tcpudp_write(protname,
+		     rxq, txq,
+		     local_addr, rem_addr,
+		     udp_state);
 
 	finish_this_one(uid,inode,timers);
     }
diff -u net-tools-1.60/debian/changelog net-tools-1.60/debian/changelog
--- net-tools-1.60/debian/changelog
+++ net-tools-1.60/debian/changelog
@@ -1,3 +1,15 @@
+net-tools (1.60-15ubuntu1) breezy; urgency=low
+
+  * Do not truncate numeric IPv6 addresses in netstat TCP and UDP.
+    The new output format is not parseable with the same algorithms as
+    before because it will be flexible about field widths, but it will not
+    be used unless it is necessary to avoid truncating numeric addresses,
+    so no existing non-broken netstat-output parsing code will be broken.
+    (Also: made 4 copies(!) of endpoint formatting code into one
+    common function.  Urgh.)  Ubuntu 7294, Debian #254243.
+
+ -- Ian Jackson <iwj@ubuntu.com>  Tue, 20 Sep 2005 15:28:53 +0100
+
 net-tools (1.60-15) unstable; urgency=low
 
   * minor formating fix to ifconfig(8)

Message #22 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Bernd Eckenfels <be-mail2005@lina.inka.de>

To: Ian Jackson <ian@davenant.greenend.org.uk>

Cc: 254243@bugs.debian.org, Phil Blundell <philb@gnu.org>, Bernd Eckenfels <net-tools@lina.inka.de>

Subject: Re: Bug#254243: netstat IPv6 address truncation patch

Date: Tue, 20 Sep 2005 21:19:39 +0200

Hello Ian,

On Tue, Sep 20, 2005 at 04:05:45PM +0100, Ian Jackson wrote:
> Looking at the code it looks like some serious overhaul is required.
> Unfortunately I don't think that I have time to do this now,
> especially at this point in Ubuntu's release cycle.

yes I agree, and you might want to join upstream berlios project anyway:

https://developer.berlios.de/projects/net-tools/

So you can subscribe to the CVS commit info and even get access to commit.

Currently the net-tools 1.65 tag about to be massaged on the cvs repository
to be the debian source (it is not yet the case). some patches are missing
on the main branch and some patches are in the head but not yet tagged, some
changes on the head are not (yet) part of debian because too invassive. I do
take care of that, however you can work on the head.

Gruss
Bernd
-- 
  (OO)     -- Bernd_Eckenfels@Mörscher_Strasse_8.76185Karlsruhe.de --
 ( .. )    ecki@{inka.de,linux.de,debian.org}  http://www.eckes.org/
  o--o   1024D/E383CD7E  eckes@IRCNet  v:+497211603874  f:+49721151516129
(O____O)  When cryptography is outlawed, bayl bhgynjf jvyy unir cevinpl!

Message #27 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Olaf van der Spek <olafvdspek@gmail.com>

To: ian@davenant.greenend.org.uk

Cc: 254243@bugs.debian.org, Phil Blundell <philb@gnu.org>, Bernd Eckenfels <net-tools@lina.inka.de>

Subject: Re: netstat IPv6 address truncation patch

Date: Thu, 29 Sep 2005 11:47:52 +0200

Hi Ian,

Did you see my related patch at
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=222324

It should also solve the truncation issue but I think it's more general.

Message #32 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Olaf van der Spek <olafvdspek@gmail.com>

To: ian@davenant.greenend.org.uk

Cc: 254243@bugs.debian.org, Phil Blundell <philb@gnu.org>, Bernd Eckenfels <net-tools@lina.inka.de>

Subject: Re: netstat IPv6 address truncation patch

Date: Thu, 29 Sep 2005 12:00:14 +0200

> +static void tcpudp_one_end(char *result_buf, size_t result_buf_size,
> +			   struct aftype *ap, int end_port,
> +			   struct sockaddr *endaddr)
> +{
> +    char buffer[8192];
> +
> +    snprintf(buffer, sizeof(buffer), "%s",
> +	     get_sname(htons(end_port), "udp",

Are you using "udp" for tcp too?

> +		       flag_not & FLAG_NUM_PORT));

Message #37 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Ian Jackson <ian@davenant.greenend.org.uk>

To: Olaf van der Spek <olafvdspek@gmail.com>

Cc: 254243@bugs.debian.org, Phil Blundell <philb@gnu.org>, Bernd Eckenfels <net-tools@lina.inka.de>

Subject: Re: netstat IPv6 address truncation patch

Date: Thu, 29 Sep 2005 12:31:33 +0100

Olaf van der Spek writes ("Re: netstat IPv6 address truncation patch"):
> > +static void tcpudp_one_end(char *result_buf, size_t result_buf_size,
> > +			   struct aftype *ap, int end_port,
> > +			   struct sockaddr *endaddr)
> > +{
> > +    char buffer[8192];
> > +
> > +    snprintf(buffer, sizeof(buffer), "%s",
> > +	     get_sname(htons(end_port), "udp",
> 
> Are you using "udp" for tcp too?

Oops.

Ian.

Message #42 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Ian Jackson <ian@davenant.greenend.org.uk>

To: Olaf van der Spek <olafvdspek@gmail.com>

Cc: 254243@bugs.debian.org, Phil Blundell <philb@gnu.org>, Bernd Eckenfels <net-tools@lina.inka.de>

Subject: Re: netstat IPv6 address truncation patch

Date: Thu, 29 Sep 2005 12:35:09 +0100

Olaf van der Spek writes ("Re: netstat IPv6 address truncation patch"):
> Did you see my related patch at
> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=222324
> 
> It should also solve the truncation issue but I think it's more general.

The thing you sent to the Debian BTS is an ed-format diff !  Looking
at it it seems to just add a new command-line flag for not truncating
the address.  That's wrong, IMO.

I think my patch (with the obvious fix to the tcp/udp string) is far
better.  It never truncates an IPv6 address when -n is requested,
which I think is necessary to preserve the behaviour that people
(including scripts) rely on.

Ian.

Message #47 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Olaf van der Spek <olafvdspek@gmail.com>

To: Ian Jackson <ian@davenant.greenend.org.uk>

Cc: 254243@bugs.debian.org, Phil Blundell <philb@gnu.org>, Bernd Eckenfels <net-tools@lina.inka.de>

Subject: Re: netstat IPv6 address truncation patch

Date: Thu, 29 Sep 2005 13:44:34 +0200

On 9/29/05, Ian Jackson <ian@davenant.greenend.org.uk> wrote:
> Olaf van der Spek writes ("Re: netstat IPv6 address truncation patch"):
> > Did you see my related patch at
> > http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=222324
> >
> > It should also solve the truncation issue but I think it's more general.
>
> The thing you sent to the Debian BTS is an ed-format diff !  Looking

There are two patches. The second version is in another format.

> at it it seems to just add a new command-line flag for not truncating
> the address.  That's wrong, IMO.

--wide is enabled by default on interactive terminals
Bernd prefered that way to avoid breaking any scripts.

> I think my patch (with the obvious fix to the tcp/udp string) is far
> better.  It never truncates an IPv6 address when -n is requested,
> which I think is necessary to preserve the behaviour that people
> (including scripts) rely on.

I haven't looked at it in detail but I guess it just uses wider fields
in one row if necessary?

Message #52 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Ian Jackson <ian@davenant.greenend.org.uk>

To: Olaf van der Spek <olafvdspek@gmail.com>

Cc: 254243@bugs.debian.org, Phil Blundell <philb@gnu.org>, Bernd Eckenfels <net-tools@lina.inka.de>

Subject: Re: netstat IPv6 address truncation patch

Date: Thu, 29 Sep 2005 12:51:47 +0100

Olaf van der Spek writes ("Re: netstat IPv6 address truncation patch"):
> There are two patches. The second version is in another format.

Ah, yes.

> > at it it seems to just add a new command-line flag for not truncating
> > the address.  That's wrong, IMO.
> 
> --wide is enabled by default on interactive terminals
> Bernd prefered that way to avoid breaking any scripts.

?

My patch very carefully (a) breaks no scripts that aren't already
broken and nevertheless (b) has a good stab at producing good output.

I don't think inventing a command-line option to switch between one
kind and another kind of unhelpful output is the right answer.

> I haven't looked at it in detail but I guess it just uses wider fields
> in one row if necessary?

Yes, but only if -n is used.  Otherwise people are presumably already
used to truncation.

Ian.

Message #57 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Olaf van der Spek <olafvdspek@gmail.com>

To: Ian Jackson <ian@davenant.greenend.org.uk>

Cc: 254243@bugs.debian.org, Phil Blundell <philb@gnu.org>, Bernd Eckenfels <net-tools@lina.inka.de>

Subject: Re: netstat IPv6 address truncation patch

Date: Thu, 29 Sep 2005 13:56:12 +0200

On 9/29/05, Ian Jackson <ian@davenant.greenend.org.uk> wrote:
> > > at it it seems to just add a new command-line flag for not truncating
> > > the address.  That's wrong, IMO.
> >
> > --wide is enabled by default on interactive terminals
> > Bernd prefered that way to avoid breaking any scripts.
>
> ?
>
> My patch very carefully (a) breaks no scripts that aren't already
> broken and nevertheless (b) has a good stab at producing good output.

Doesn't my patch do both too?

> I don't think inventing a command-line option to switch between one
> kind and another kind of unhelpful output is the right answer.

How is the output of my patch unhelpful?
And the option is only necessary to enable it on non-interactive 'stdout'.

> > I haven't looked at it in detail but I guess it just uses wider fields
> > in one row if necessary?
>
> Yes, but only if -n is used.  Otherwise people are presumably already
> used to truncation.

I'm not sure about that.

Message #62 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Ian Jackson <ian@davenant.greenend.org.uk>

To: Olaf van der Spek <olafvdspek@gmail.com>

Cc: 254243@bugs.debian.org, Phil Blundell <philb@gnu.org>, Bernd Eckenfels <net-tools@lina.inka.de>

Subject: Re: netstat IPv6 address truncation patch

Date: Thu, 29 Sep 2005 13:46:19 +0100

Olaf van der Spek writes ("Re: netstat IPv6 address truncation patch"):
> On 9/29/05, Ian Jackson <ian@davenant.greenend.org.uk> wrote:
> > Yes, but only if -n is used.  Otherwise people are presumably already
> > used to truncation.
> 
> I'm not sure about that.

Without -n, addresses are turned into domain names, which are often
truncated by the narrow field width.

Ian.

Message #67 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Olaf van der Spek <olafvdspek@gmail.com>

To: Ian Jackson <ian@davenant.greenend.org.uk>

Cc: 254243@bugs.debian.org, Phil Blundell <philb@gnu.org>, Bernd Eckenfels <net-tools@lina.inka.de>

Subject: Re: netstat IPv6 address truncation patch

Date: Thu, 29 Sep 2005 14:48:35 +0200

On 9/29/05, Ian Jackson <ian@davenant.greenend.org.uk> wrote:
> Olaf van der Spek writes ("Re: netstat IPv6 address truncation patch"):
> > On 9/29/05, Ian Jackson <ian@davenant.greenend.org.uk> wrote:
> > > Yes, but only if -n is used.  Otherwise people are presumably already
> > > used to truncation.
> >
> > I'm not sure about that.
>
> Without -n, addresses are turned into domain names, which are often
> truncated by the narrow field width.

But I think that's not desired by users.

Message #72 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Ian Jackson <ian@davenant.greenend.org.uk>

To: Olaf van der Spek <olafvdspek@gmail.com>

Cc: 254243@bugs.debian.org, Phil Blundell <philb@gnu.org>, Bernd Eckenfels <net-tools@lina.inka.de>

Subject: Re: netstat IPv6 address truncation patch

Date: Thu, 29 Sep 2005 13:52:39 +0100

Olaf van der Spek writes ("Re: netstat IPv6 address truncation patch"):
> On 9/29/05, Ian Jackson <ian@davenant.greenend.org.uk> wrote:
> > Without -n, addresses are turned into domain names, which are often
> > truncated by the narrow field width.
> 
> But I think that's not desired by users.

I agree that it might be useful to have an option which prevents any
field truncation, but that ought not to be the default and we need the
logic in my patch anyway.

Also, -$ as a short option name is rather odd.

Ian.

Message #77 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Olaf van der Spek <olafvdspek@gmail.com>

To: Ian Jackson <ian@davenant.greenend.org.uk>

Cc: 254243@bugs.debian.org, Phil Blundell <philb@gnu.org>, Bernd Eckenfels <net-tools@lina.inka.de>

Subject: Re: netstat IPv6 address truncation patch

Date: Thu, 29 Sep 2005 14:56:20 +0200

On 9/29/05, Ian Jackson <ian@davenant.greenend.org.uk> wrote:
> Olaf van der Spek writes ("Re: netstat IPv6 address truncation patch"):
> > On 9/29/05, Ian Jackson <ian@davenant.greenend.org.uk> wrote:
> > > Without -n, addresses are turned into domain names, which are often
> > > truncated by the narrow field width.
> >
> > But I think that's not desired by users.
>
> I agree that it might be useful to have an option which prevents any
> field truncation, but that ought not to be the default and we need the

Why should it not be the default on interactive stdout?

> logic in my patch anyway.
>
> Also, -$ as a short option name is rather odd.

Only the long version is supposed to be used.

Message #82 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Ian Jackson <ian@davenant.greenend.org.uk>

To: Olaf van der Spek <olafvdspek@gmail.com>, 254243@bugs.debian.org, Phil Blundell <philb@gnu.org>, Bernd Eckenfels <net-tools@lina.inka.de>

Subject: Re: netstat IPv6 address truncation patch

Date: Fri, 30 Sep 2005 15:02:52 +0100

Here is an updated version of my patch which fixes the "tcp" vs. "udp"
string in the protocol lookup bit of the unified tcpudp_one_end.

Ian.

diff -Nru /tmp/BEfybKtWqd/net-tools-1.60/debian/changelog /tmp/BQGUuO9HrM/net-tools-1.60/debian/changelog
--- /tmp/BEfybKtWqd/net-tools-1.60/debian/changelog	2005-09-30 14:14:34.000000000 +0100
+++ /tmp/BQGUuO9HrM/net-tools-1.60/debian/changelog	2005-09-29 18:02:28.000000000 +0100
@@ -1,3 +1,23 @@
+net-tools (1.60-15ubuntu2) breezy; urgency=low
+
+  * Look up TCP/UDP protocol numbers using correct protocol, not
+    always UDP.  (Bug introduced in 1.60-15ubuntu1.)
+    Ubuntu 7294, Debian #254243.
+
+ -- Ian Jackson <iwj@ubuntu.com>  Thu, 29 Sep 2005 18:02:23 +0100
+
+net-tools (1.60-15ubuntu1) breezy; urgency=low
+
+  * Do not truncate numeric IPv6 addresses in netstat TCP and UDP.
+    The new output format is not parseable with the same algorithms as
+    before because it will be flexible about field widths, but it will not
+    be used unless it is necessary to avoid truncating numeric addresses,
+    so no existing non-broken netstat-output parsing code will be broken.
+    (Also: made 4 copies(!) of endpoint formatting code into one
+    common function.  Urgh.)  Ubuntu 7294, Debian #254243.
+
+ -- Ian Jackson <iwj@ubuntu.com>  Tue, 20 Sep 2005 15:28:53 +0100
+
 net-tools (1.60-15) unstable; urgency=low
 
   * minor formating fix to ifconfig(8)
diff -Nru /tmp/BEfybKtWqd/net-tools-1.60/netstat.c /tmp/BQGUuO9HrM/net-tools-1.60/netstat.c
--- /tmp/BEfybKtWqd/net-tools-1.60/netstat.c	2005-09-30 14:14:34.000000000 +0100
+++ /tmp/BQGUuO9HrM/net-tools-1.60/netstat.c	2005-09-29 17:59:55.000000000 +0100
@@ -85,6 +85,7 @@
 #include <sys/ioctl.h>
 #include <net/if.h>
 #include <dirent.h>
+#include <assert.h>
 
 #include "net-support.h"
 #include "pathnames.h"
@@ -704,11 +705,71 @@
 	       igmp_do_one);
 }
 
+static void tcpudp_one_end(char *result_buf, size_t result_buf_size,
+			   struct aftype *ap, int end_port,
+			   struct sockaddr *endaddr, const char *protostr)
+{
+    char buffer[8192];
+    
+    snprintf(buffer, sizeof(buffer), "%s",
+	     get_sname(htons(end_port), protostr,
+		       flag_not & FLAG_NUM_PORT));
+
+    assert(result_buf_size > 23);
+
+    safe_strncpy(result_buf, ap->sprint(endaddr, flag_not), result_buf_size);
+    if ((strlen(result_buf) + strlen(buffer)) > 22 &&
+	!(flag_not & FLAG_NUM_HOST))
+	result_buf[22 - strlen(buffer)] = '\0';
+
+    assert(strlen(result_buf) + strlen(buffer) + 1 < sizeof(buffer));
+    strcat(result_buf, ":");
+    strcat(result_buf, buffer);
+}
+
+static void reduce_deficit(int *deficit, int *spc, int min_spc)
+{
+    int reduction;
+    reduction= netmin(*deficit, *spc - min_spc);
+    if (reduction <= 0) return;
+    *spc -= reduction;
+    *deficit -= reduction;
+}
+
+static void tcpudp_write(const char *protname,
+			 unsigned long rxq, unsigned long txq,
+			 const char *local_addr, const char *rem_addr,
+			 const char *state) {
+    int rxq_spc, txq_spc, local_addr_spc, rem_addr_spc, protname_spc;
+    int local_addr_len, rem_addr_len, deficit;
+
+    protname_spc = 5;
+    rxq_spc = 6;
+    txq_spc = 6;
+    local_addr_spc = 23;
+    local_addr_len = strlen(local_addr);
+    rem_addr_spc = 23;
+    rem_addr_len = strlen(rem_addr);
+    deficit = netmax(local_addr_len - local_addr_spc, 0) +
+	netmax(rem_addr_len - rem_addr_spc, 0);
+    
+    reduce_deficit(&deficit, &rem_addr_spc, rem_addr_len);
+    reduce_deficit(&deficit, &local_addr_spc, local_addr_len);
+    reduce_deficit(&deficit, &txq_spc, 1);
+    reduce_deficit(&deficit, &rxq_spc, 1);
+    reduce_deficit(&deficit, &protname_spc, strlen(protname));
+		       
+    printf("%-*s %*ld %*ld %-*s %-*s %-11s",
+	   protname_spc, protname, rxq_spc, rxq, txq_spc, txq,
+	   local_addr_spc, local_addr,
+	   rem_addr_spc, rem_addr, state);
+}
+
 static void tcp_do_one(int lnr, const char *line)
 {
     unsigned long rxq, txq, time_len, retr, inode;
     int num, local_port, rem_port, d, state, uid, timer_run, timeout;
-    char rem_addr[128], local_addr[128], timers[64], buffer[1024], more[512];
+    char rem_addr[128], local_addr[128], timers[64], more[512];
     char *protname;
     struct aftype *ap;
 #if HAVE_AFINET6
@@ -775,25 +836,15 @@
     safe_strncpy(rem_addr, ap->sprint((struct sockaddr *) &remaddr, flag_not),
 		 sizeof(rem_addr));
     if (flag_all || (flag_lst && !rem_port) || (!flag_lst && rem_port)) {
-	snprintf(buffer, sizeof(buffer), "%s",
-		 get_sname(htons(local_port), "tcp",
-			   flag_not & FLAG_NUM_PORT));
-
-	if ((strlen(local_addr) + strlen(buffer)) > 22)
-	    local_addr[22 - strlen(buffer)] = '\0';
+	tcpudp_one_end(local_addr, sizeof(local_addr),
+		       ap, local_port, (struct sockaddr *) &localaddr,
+		       "tcp");
+
+	tcpudp_one_end(rem_addr, sizeof(rem_addr),
+		       ap, rem_port, (struct sockaddr *) &remaddr,
+		       "tcp");
 
-	strcat(local_addr, ":");
-	strcat(local_addr, buffer);
-	snprintf(buffer, sizeof(buffer), "%s",
-		 get_sname(htons(rem_port), "tcp", flag_not & FLAG_NUM_PORT));
-
-	if ((strlen(rem_addr) + strlen(buffer)) > 22)
-	    rem_addr[22 - strlen(buffer)] = '\0';
-
-	strcat(rem_addr, ":");
-	strcat(rem_addr, buffer);
 	timers[0] = '\0';
-
 	if (flag_opt)
 	    switch (timer_run) {
 	    case 0:
@@ -820,8 +871,11 @@
 			 timer_run, (double) time_len / HZ, retr, timeout);
 		break;
 	    }
-	printf("%-4s  %6ld %6ld %-*s %-*s %-11s",
-	       protname, rxq, txq, netmax(23,strlen(local_addr)), local_addr, netmax(23,strlen(rem_addr)), rem_addr, _(tcp_state[state]));
+
+	tcpudp_write(protname,
+		     rxq, txq,
+		     local_addr, rem_addr,
+		     _(tcp_state[state]));
 
 	finish_this_one(uid,inode,timers);
     }
@@ -835,7 +889,7 @@
 
 static void udp_do_one(int lnr, const char *line)
 {
-    char buffer[8192], local_addr[64], rem_addr[64];
+    char local_addr[64], rem_addr[64];
     char *udp_state, timers[64], more[512];
     int num, local_port, rem_port, d, state, timer_run, uid, timeout;
     char *protname;
@@ -927,24 +981,13 @@
 
     if (flag_all || (notnull(remaddr) && !flag_lst) || (!notnull(remaddr) && flag_lst))
     {
-        safe_strncpy(local_addr, ap->sprint((struct sockaddr *) &localaddr, 
-					    flag_not), sizeof(local_addr));
-	snprintf(buffer, sizeof(buffer), "%s",
-		 get_sname(htons(local_port), "udp",
-			   flag_not & FLAG_NUM_PORT));
-	if ((strlen(local_addr) + strlen(buffer)) > 22)
-	    local_addr[22 - strlen(buffer)] = '\0';
-	strcat(local_addr, ":");
-	strcat(local_addr, buffer);
-
-	snprintf(buffer, sizeof(buffer), "%s",
-		 get_sname(htons(rem_port), "udp", flag_not & FLAG_NUM_PORT));
-        safe_strncpy(rem_addr, ap->sprint((struct sockaddr *) &remaddr, 
-					  flag_not), sizeof(rem_addr));
-	if ((strlen(rem_addr) + strlen(buffer)) > 22)
-	    rem_addr[22 - strlen(buffer)] = '\0';
-	strcat(rem_addr, ":");
-	strcat(rem_addr, buffer);
+	tcpudp_one_end(local_addr, sizeof(local_addr),
+		       ap, local_port, (struct sockaddr *) &localaddr,
+		       "udp");
+
+	tcpudp_one_end(rem_addr, sizeof(rem_addr),
+		       ap, rem_port, (struct sockaddr *) &remaddr,
+		       "udp");
 
 	timers[0] = '\0';
 	if (flag_opt)
@@ -963,8 +1006,11 @@
 			 retr, timeout);
 		break;
 	    }
-	printf("%-4s  %6ld %6ld %-23s %-23s %-11s",
-	       protname, rxq, txq, local_addr, rem_addr, udp_state);
+
+	tcpudp_write(protname,
+		     rxq, txq,
+		     local_addr, rem_addr,
+		     udp_state);
 
 	finish_this_one(uid,inode,timers);
     }

Message #87 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Olaf van der Spek <olafvdspek@gmail.com>

To: 254243@bugs.debian.org

Subject: Bug Status?

Date: Mon, 12 Dec 2005 18:36:38 +0100

Hi Bernd,

What is the status of this bug?
Did you decide about the offered patches already?

Greetings,

Olaf

Message #92 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Bernd Eckenfels <be-mail2005@lina.inka.de>

To: Olaf van der Spek <olafvdspek@gmail.com>

Cc: debian-devel@lists.debian.org, 222324@bugs.debian.org, 254243@bugs.debian.org

Subject: Re: net-tools maintenance status

Date: Fri, 16 Dec 2005 20:42:37 +0100

On Fri, Dec 16, 2005 at 08:03:47PM +0100, Olaf van der Spek wrote:
> Bernd?

I dont like the prposed solution, i am looking for a more generic one. The
pressing problem is to display IPV6 netstat on 80char widt correctly without
truncation (#254243). 

I guess this will require some changes to the output formatter, anyway.

I think i will add a non-formatting/non-truncating scritable output instead
of the wide switch, and make netstat on tty observ cols. The question is,
what size to use on non-tty output.

Gruss
Bernd
-- 
  (OO)     -- Bernd_Eckenfels@Mörscher_Strasse_8.76185Karlsruhe.de --
 ( .. )    ecki@{inka.de,linux.de,debian.org}  http://www.eckes.org/
  o--o   1024D/E383CD7E  eckes@IRCNet  v:+497211603874  f:+49721151516129
(O____O)  When cryptography is outlawed, bayl bhgynjf jvyy unir cevinpl!

Message #97 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Olaf van der Spek <olafvdspek@gmail.com>

To: Bernd Eckenfels <be-mail2005@lina.inka.de>

Cc: 222324@bugs.debian.org, 254243@bugs.debian.org

Subject: Re: net-tools maintenance status

Date: Fri, 16 Dec 2005 22:52:12 +0100

On 12/16/05, Bernd Eckenfels <be-mail2005@lina.inka.de> wrote:
> On Fri, Dec 16, 2005 at 08:03:47PM +0100, Olaf van der Spek wrote:
> > Bernd?
>
> I dont like the prposed solution, i am looking for a more generic one. The

Thanks for the response.
In what way would you like it to be more generic?

> pressing problem is to display IPV6 netstat on 80char widt correctly without
> truncation (#254243).

That's not possible without dropping entire columns.

> I guess this will require some changes to the output formatter, anyway.
>
> I think i will add a non-formatting/non-truncating scritable output instead
> of the wide switch, and make netstat on tty observ cols. The question is,
> what size to use on non-tty output.

Message #102 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Olaf van der Spek <olafvdspek@gmail.com>

To: Bernd Eckenfels <be-mail2005@lina.inka.de>

Cc: 222324@bugs.debian.org, 254243@bugs.debian.org

Subject: Re: net-tools maintenance status

Date: Thu, 22 Dec 2005 16:58:25 +0100

On 12/16/05, Olaf van der Spek <olafvdspek@gmail.com> wrote:
> On 12/16/05, Bernd Eckenfels <be-mail2005@lina.inka.de> wrote:
> > On Fri, Dec 16, 2005 at 08:03:47PM +0100, Olaf van der Spek wrote:
> > > Bernd?
> >
> > I dont like the prposed solution, i am looking for a more generic one. The
>
> Thanks for the response.
> In what way would you like it to be more generic?
>
> > pressing problem is to display IPV6 netstat on 80char widt correctly without
> > truncation (#254243).
>
> That's not possible without dropping entire columns.
>
> > I guess this will require some changes to the output formatter, anyway.
> >
> > I think i will add a non-formatting/non-truncating scritable output instead
> > of the wide switch, and make netstat on tty observ cols. The question is,
> > what size to use on non-tty output.

Bernd?

Message #107 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Olaf van der Spek <OvdSpek@LIACS.NL>

To: 254243@bugs.debian.org

Subject: NMU?

Date: Sat, 11 Mar 2006 22:31:58 +0100

Hi Bernd,

Do you mind if an NMU is done to fix this issue?

Olaf

Message #112 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Bernd Eckenfels <be-mail2006@lina.inka.de>

To: Olaf van der Spek <OvdSpek@LIACS.NL>

Cc: 254243@bugs.debian.org

Subject: Re: Bug#254243: NMU?

Date: Sat, 11 Mar 2006 23:29:37 +0100

On Sat, Mar 11, 2006 at 10:31:58PM +0100, Olaf van der Spek wrote:
> Do you mind if an NMU is done to fix this issue?

An NMU for a Wishlist Bug is IMHO not aproperiate, especially not since the
fixing patch may break existing scripts.

Gruss
Bernd
-- 
  (OO)     -- Bernd_Eckenfels@Mörscher_Strasse_8.76185Karlsruhe.de --
 ( .. )    ecki@{inka.de,linux.de,debian.org}  http://www.eckes.org/
  o--o   1024D/E383CD7E  eckes@IRCNet  v:+497211603874  f:+49721151516129
(O____O)  When cryptography is outlawed, bayl bhgynjf jvyy unir cevinpl!

Message #117 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Olaf van der Spek <OvdSpek@LIACS.NL>

To: Bernd Eckenfels <be-mail2006@lina.inka.de>

Cc: 254243@bugs.debian.org

Subject: Re: Bug#254243: NMU?

Date: Sat, 11 Mar 2006 23:33:36 +0100

Bernd Eckenfels wrote:
> On Sat, Mar 11, 2006 at 10:31:58PM +0100, Olaf van der Spek wrote:
>> Do you mind if an NMU is done to fix this issue?
> 
> An NMU for a Wishlist Bug is IMHO not aproperiate, especially not since the
> fixing patch may break existing scripts.

The patch I submitted does not change the output if stdout is not a 
terminal.
Could you please explain how it still breaks existing scripts?

Message #122 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Olaf van der Spek <OvdSpek@LIACS.NL>

To: 222324@bugs.debian.org, 254243@bugs.debian.org

Subject: Severity: normal

Date: Mon, 13 Mar 2006 12:06:32 +0100

Hi,

I've increased the severity to normal as these bugs do affect the 
usefulness of the netstat app.

Olaf

Message #129 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Joel Becker <jlbec@evilplan.org>

To: 254243@bugs.debian.org

Subject: Re: Severity: normal

Date: Fri, 17 Mar 2006 18:23:56 +0000

Olaf van der Spek <OvdSpek@LIACS.NL> wrote:
> I've increased the severity to normal as these bugs do affect the 
> usefulness of the netstat app.

	Can we increase it further?  This is not cosmetic.  I just
suffered an attack on my httpd from a runaway host, and I couldn't
determine which host it was because of the output truncation.  Sure, I
was able to find out with deeper digging, but this sort of thing is
precisely what people use "netstat -an" for!

Joel

-- 

"People with narrow minds usually have broad tongues."

			http://www.jlbec.org/
			jlbec@evilplan.org

Message #134 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Bernd Eckenfels <be-mail2006@lina.inka.de>

To: Joel Becker <jlbec@evilplan.org>

Cc: 254243@bugs.debian.org, control@bugs.debian.org

Subject: Re: Bug#254243: Severity: normal

Date: Fri, 17 Mar 2006 21:41:46 +0100

severity 254243 important
thanks

yes it is possible.

The new version will feature a -X option which enforce script-parseable,
untruncated netstat output. Then the bug can be a wishlist to better use the
available interactive screen space.

Gruss
Bernd


On Fri, Mar 17, 2006 at 06:23:56PM +0000, Joel Becker wrote:
> Olaf van der Spek <OvdSpek@LIACS.NL> wrote:
> > I've increased the severity to normal as these bugs do affect the 
> > usefulness of the netstat app.
> 
> 	Can we increase it further?  This is not cosmetic.  I just
> suffered an attack on my httpd from a runaway host, and I couldn't
> determine which host it was because of the output truncation.  Sure, I
> was able to find out with deeper digging, but this sort of thing is
> precisely what people use "netstat -an" for!
> 
> Joel
> 
> -- 
> 
> "People with narrow minds usually have broad tongues."
> 
> 			http://www.jlbec.org/
> 			jlbec@evilplan.org
> 

-- 
  (OO)     -- Bernd_Eckenfels@Mörscher_Strasse_8.76185Karlsruhe.de --
 ( .. )    ecki@{inka.de,linux.de,debian.org}  http://www.eckes.org/
  o--o   1024D/E383CD7E  eckes@IRCNet  v:+497211603874  f:+49721151516129
(O____O)  When cryptography is outlawed, bayl bhgynjf jvyy unir cevinpl!

Message #143 received at 254243-quiet@bugs.debian.org (full text, mbox, reply):

From: Bernd Eckenfels <be-mail2006@lina.inka.de>

To: nh <nhwith@bonbon.net>

Cc: ecki@debian.org, 254243-quiet@bugs.debian.org

Subject: Re: net-tools-1.60_netstat_verbose_ipv6.patch

Date: Tue, 10 Oct 2006 07:53:16 +0200

Thanks for the patch, it is a known bug, with some possible solutions. Will
most likely add your patch as a "unformatted delimited" version.

See http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=254243 for alternatives

Gruss
Bernd

On Tue, Oct 10, 2006 at 06:28:49AM +0200, nh wrote:
> Hi,
> 
> debian net-tools netstat strips ipv6 raw6/tcp6/udp6 remote/local address 
> fields to length of ~18 (without port).
> I didn't find way how to make it print full address (since it's 
> hardcoded to keep formatting).
> 
> Patch makes it to print full addresses when '-v' switch is used.
> It's applicable on net-tools_1.60-10 and testing/unstable net-tools_1.60-17.
> 
> I'm not sure if there is any special ipv6 version of net-tools, other 
> tool which I missed or particular reason why it's limited (I do prefer 
> to see full address instead of nicely formatted and stripped stuff in 
> some cases :>), so sorry for bothering if it's some of those variants.
> 
> Regards,
> Nathaniel Hier
> 
> 
> 
> 
> 
> 

> --- old/net-tools-1.60/netstat.c	2006-10-10 05:45:48.000000000 +0200
> +++ net-tools-1.60/netstat.c	2006-10-10 05:46:19.000000000 +0200
> @@ -779,16 +779,26 @@
>  		 get_sname(htons(local_port), "tcp",
>  			   flag_not & FLAG_NUM_PORT));
>  
> -	if ((strlen(local_addr) + strlen(buffer)) > 22)
> +	if ((strlen(local_addr) + strlen(buffer)) > 22 && !protname[3])
>  	    local_addr[22 - strlen(buffer)] = '\0';
> +	else if ((strlen(local_addr) + strlen(buffer)) > 22 && 
> +			 !(flag_ver & FLAG_VERBOSE) && protname[3])
> +		local_addr[22 - strlen(buffer)] = '\0';
> +	else if ((strlen(local_addr) + strlen(buffer)) > 46)
> +		local_addr[46 - strlen(buffer)] = '\0';
>  
>  	strcat(local_addr, ":");
>  	strcat(local_addr, buffer);
>  	snprintf(buffer, sizeof(buffer), "%s",
>  		 get_sname(htons(rem_port), "tcp", flag_not & FLAG_NUM_PORT));
>  
> -	if ((strlen(rem_addr) + strlen(buffer)) > 22)
> +	if ((strlen(rem_addr) + strlen(buffer)) > 22 && !protname[3])
> +		rem_addr[22 - strlen(buffer)] = '\0';
> +	else if ((strlen(rem_addr) + strlen(buffer)) > 22 && 
> +			 !(flag_ver & FLAG_VERBOSE) && protname[3])
>  	    rem_addr[22 - strlen(buffer)] = '\0';
> +	else if ((strlen(rem_addr) + strlen(buffer)) > 46)
> +		rem_addr[46 - strlen(buffer)] = '\0';
>  
>  	strcat(rem_addr, ":");
>  	strcat(rem_addr, buffer);
> @@ -820,6 +830,8 @@
>  			 timer_run, (double) time_len / HZ, retr, timeout);
>  		break;
>  	    }
> +
> + 
>  	printf("%-4s  %6ld %6ld %-*s %-*s %-11s",
>  	       protname, rxq, txq, netmax(23,strlen(local_addr)), local_addr, netmax(23,strlen(rem_addr)), rem_addr, _(tcp_state[state]));
>  
> @@ -932,8 +944,15 @@
>  	snprintf(buffer, sizeof(buffer), "%s",
>  		 get_sname(htons(local_port), "udp",
>  			   flag_not & FLAG_NUM_PORT));
> -	if ((strlen(local_addr) + strlen(buffer)) > 22)
> +
> +	if ((strlen(local_addr) + strlen(buffer)) > 22 && !protname[3])
>  	    local_addr[22 - strlen(buffer)] = '\0';
> +	else if ((strlen(local_addr) + strlen(buffer)) > 22 && 
> +			 !(flag_ver & FLAG_VERBOSE) && protname[3])
> +		local_addr[22 - strlen(buffer)] = '\0';
> +	else if ((strlen(local_addr) + strlen(buffer)) > 46)
> +		local_addr[46 - strlen(buffer)] = '\0';
> +
>  	strcat(local_addr, ":");
>  	strcat(local_addr, buffer);
>  
> @@ -941,8 +960,14 @@
>  		 get_sname(htons(rem_port), "udp", flag_not & FLAG_NUM_PORT));
>          safe_strncpy(rem_addr, ap->sprint((struct sockaddr *) &remaddr, 
>  					  flag_not), sizeof(rem_addr));
> -	if ((strlen(rem_addr) + strlen(buffer)) > 22)
> +	if ((strlen(rem_addr) + strlen(buffer)) > 22 && !protname[3])
>  	    rem_addr[22 - strlen(buffer)] = '\0';
> +	else if ((strlen(rem_addr) + strlen(buffer)) > 22 && 
> +			 !(flag_ver & FLAG_VERBOSE) && protname[3])
> +		rem_addr[22 - strlen(buffer)] = '\0';
> +	else if ((strlen(rem_addr) + strlen(buffer)) > 46)
> +		rem_addr[46 - strlen(buffer)] = '\0';
> +
>  	strcat(rem_addr, ":");
>  	strcat(rem_addr, buffer);
>  
> @@ -963,8 +988,9 @@
>  			 retr, timeout);
>  		break;
>  	    }
> -	printf("%-4s  %6ld %6ld %-23s %-23s %-11s",
> -	       protname, rxq, txq, local_addr, rem_addr, udp_state);
> +
> +	printf("%-4s  %6ld %6ld %-*s %-*s %-11s",
> +		   protname, rxq, txq, netmax(23,strlen(local_addr)), local_addr, netmax(23,strlen(rem_addr)), rem_addr, udp_state);	
>  
>  	finish_this_one(uid,inode,timers);
>      }
> @@ -1054,8 +1080,15 @@
>  			   flag_not & FLAG_NUM_PORT));
>          safe_strncpy(local_addr, ap->sprint((struct sockaddr *) &localaddr, 
>  					    flag_not), sizeof(local_addr));
> -	if ((strlen(local_addr) + strlen(buffer)) > 22)
> +
> +	if ((strlen(local_addr) + strlen(buffer)) > 22 && !protname[3])
> +		local_addr[22 - strlen(buffer)] = '\0';
> +	else if ((strlen(local_addr) + strlen(buffer)) > 22 && 
> +			 !(flag_ver & FLAG_VERBOSE) && protname[3])
>  	    local_addr[22 - strlen(buffer)] = '\0';
> +	else if ((strlen(local_addr) + strlen(buffer)) > 46)
> +		local_addr[46 - strlen(buffer)] = '\0';
> +
>  	strcat(local_addr, ":");
>  	strcat(local_addr, buffer);
>  
> @@ -1063,8 +1096,15 @@
>  		 get_sname(htons(rem_port), "raw", flag_not & FLAG_NUM_PORT));
>          safe_strncpy(rem_addr, ap->sprint((struct sockaddr *) &remaddr, 
>  					  flag_not), sizeof(rem_addr));
> -	if ((strlen(rem_addr) + strlen(buffer)) > 22)
> +
> +	if ((strlen(rem_addr) + strlen(buffer)) > 22 && !protname[3])
>  	    rem_addr[22 - strlen(buffer)] = '\0';
> +	else if ((strlen(rem_addr) + strlen(buffer)) > 22 && 
> +			 !(flag_ver & FLAG_VERBOSE) && protname[3])
> +		rem_addr[22 - strlen(buffer)] = '\0';
> +	else if ((strlen(rem_addr) + strlen(buffer)) > 46)
> +		rem_addr[46 - strlen(buffer)] = '\0';
> +
>  	strcat(rem_addr, ":");
>  	strcat(rem_addr, buffer);
>  


-- 
  (OO)     -- Bernd_Eckenfels@Mörscher_Strasse_8.76185Karlsruhe.de --
 ( .. )    ecki@{inka.de,linux.de,debian.org}  http://www.eckes.org/
  o--o   1024D/E383CD7E  eckes@IRCNet  v:+497211603874  f:+49721151516129
(O____O)  When cryptography is outlawed, bayl bhgynjf jvyy unir cevinpl!

Message #148 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Brian Haley <brian.haley@hp.com>

To: 254243@bugs.debian.org

Subject: Bug#254243

Date: Mon, 13 Nov 2006 16:39:34 -0500

Hi,

I also ran into this problem and have tried both patches here.  I don't 
really like the second one since in addition to having to supply a -v 
flag you get spurious warnings depending on what address families are 
installed:

netstat: no support for `AF IPX' on this system.
netstat: no support for `AF AX25' on this system.
netstat: no support for `AF X25' on this system.
netstat: no support for `AF NETROM' on this system.

I know it's the easy solution since USAGI did the same thing in 2002 
though... but not having the default be the full output can lead to this:

# netstat -an
tcp6  0   0 ::ffff:16.123.1.1:22    ::ffff:16.123.1.2:57347 ESTABLISHED

# netstat -anv
tcp6  0   0 ::ffff:16.123.1.1:22    ::ffff:16.123.1.29:57347 ESTABLISHED

If I didn't know to add the -v I would have missed that final 9 in 29 
since it at least looked like a valid address.

Thanks,

-Brian

Message #155 received at 254243@bugs.debian.org (full text, mbox, reply):

From: martin f krafft <madduck@debian.org>

To: 254243@bugs.debian.org

Subject: Status update?

Date: Fri, 5 Sep 2008 15:47:17 +0100

[Message part 1 (text/plain, inline)]

What's the status of this bug report/patch?

-- 
 .''`.   martin f. krafft <madduck@debian.org>
: :'  :  proud Debian developer, author, administrator, and user
`. `'`   http://people.debian.org/~madduck - http://debiansystem.info
  `-  Debian - when you have better things to do than fixing systems

[digital_signature_gpg.asc (application/pgp-signature, inline)]

Message #160 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Olaf van der Spek <Olaf@XWIS.Net>

To: 254243@bugs.debian.org

Subject: (no subject)

Date: Sun, 19 Oct 2008 11:25:22 +0200

I think it should be merged with #222324
I'm willing to continue work on my patch from 2005 if Luk and Martin agree.

Message #165 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Luk Claes <luk@debian.org>

To: Olaf van der Spek <Olaf@XWIS.Net>, 254243@bugs.debian.org

Cc: Bernd Eckenfels <ecki@lina.inka.de>

Subject: Re: Bug#254243: (no subject)

Date: Sun, 19 Oct 2008 12:35:59 +0200

Olaf van der Spek wrote:
> I think it should be merged with #222324
> I'm willing to continue work on my patch from 2005 if Luk and Martin agree.

It's a bit unclear to me what Bernd's objections are and what he
intended to incorporate, so I kind of decided to delay this till after
lenny. But I don't see a reason to not go forward with this once it's
clear to me what Bernd's opinion is...

Cheers

Luk

Message #170 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Olaf van der Spek <Olaf@XWIS.Net>

To: Luk Claes <luk@debian.org>

Cc: 254243@bugs.debian.org, Bernd Eckenfels <ecki@lina.inka.de>

Subject: Re: Bug#254243: (no subject)

Date: Sun, 19 Oct 2008 20:39:06 +0200

Luk Claes wrote:
> Olaf van der Spek wrote:
>> I think it should be merged with #222324
>> I'm willing to continue work on my patch from 2005 if Luk and Martin agree.
> 
> It's a bit unclear to me what Bernd's objections are and what he

Yeah, it's unclear to me too. ;)
That was the core of the issue.

> intended to incorporate, so I kind of decided to delay this till after
> lenny. But I don't see a reason to not go forward with this once it's
> clear to me what Bernd's opinion is...

Message #175 received at 254243@bugs.debian.org (full text, mbox, reply):

From: martin f krafft <madduck@debian.org>

To: Olaf van der Spek <Olaf@XWIS.Net>, 254243@bugs.debian.org

Subject: Re: Bug#254243: (no subject)

Date: Mon, 20 Oct 2008 08:43:44 +0200

[Message part 1 (text/plain, inline)]

also sprach Olaf van der Spek <Olaf@XWIS.Net> [2008.10.19.1125 +0200]:
> I think it should be merged with #222324
> I'm willing to continue work on my patch from 2005 if Luk and Martin agree.

I don't want to have to rely on a width switch. I'd say that netstat
should output line-oriented and stop worrying about terminal size.
First of all, IPv6-connected hosts are unlikely to be so old that
they can only do width-80 or non-tty terminals, and second, the days
of tools like route/arp/netstat which present formatted output
(which is harder to process) are hopefully over in favour of tools
that print the maximum of information and leave it to other tools to
do the formatting. I really like /bin/ip for that reason, and
because its output is the same as its input. Tabular output, like
netstat, is IMHO not worth the trouble, especially if it means that
the output cannot be complete.

-- 
 .''`.   martin f. krafft <madduck@debian.org>
: :'  :  proud Debian developer, author, administrator, and user
`. `'`   http://people.debian.org/~madduck - http://debiansystem.info
  `-  Debian - when you have better things to do than fixing systems

[digital_signature_gpg.asc (application/pgp-signature, inline)]

Message #180 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Olaf van der Spek <Olaf@XWIS.Net>

To: martin f krafft <madduck@debian.org>

Cc: 254243@bugs.debian.org

Subject: Re: Bug#254243: (no subject)

Date: Mon, 20 Oct 2008 12:44:57 +0200

martin f krafft wrote:
> also sprach Olaf van der Spek <Olaf@XWIS.Net> [2008.10.19.1125 +0200]:
>> I think it should be merged with #222324
>> I'm willing to continue work on my patch from 2005 if Luk and Martin agree.
> 
> I don't want to have to rely on a width switch. I'd say that netstat
> should output line-oriented and stop worrying about terminal size.
> First of all, IPv6-connected hosts are unlikely to be so old that
> they can only do width-80 or non-tty terminals, and second, the days
> of tools like route/arp/netstat which present formatted output
> (which is harder to process) are hopefully over in favour of tools
> that print the maximum of information and leave it to other tools to
> do the formatting. I really like /bin/ip for that reason, and
> because its output is the same as its input. Tabular output, like
> netstat, is IMHO not worth the trouble, especially if it means that
> the output cannot be complete.

That's fine for machine readable output, but what about human readable 
output?
And how to handle existing scripts that rely on this specific tabular 
output?

Message #185 received at 254243@bugs.debian.org (full text, mbox, reply):

From: martin f krafft <madduck@debian.org>

To: Olaf van der Spek <Olaf@XWIS.Net>

Cc: 254243@bugs.debian.org

Subject: Re: Bug#254243: (no subject)

Date: Mon, 20 Oct 2008 12:53:28 +0200

[Message part 1 (text/plain, inline)]

also sprach Olaf van der Spek <Olaf@XWIS.Net> [2008.10.20.1244 +0200]:
> That's fine for machine readable output, but what about human readable  
> output?

That's debateable. I find e.g. 'ip route list' output far better
than /sbin/route by now. The same applies to 'ip address list' vs.
/sbin/ifconfig. /usr/sbin/arp was always non-tabular.

> And how to handle existing scripts that rely on this specific tabular  
> output?

those should hardly depend on column or terminal width.

-- 
 .''`.   martin f. krafft <madduck@debian.org>
: :'  :  proud Debian developer, author, administrator, and user
`. `'`   http://people.debian.org/~madduck - http://debiansystem.info
  `-  Debian - when you have better things to do than fixing systems
 
"there was no difference between the behavior of a god
 and the operations of pure chance..."
                              -- thomas pynchon, "gravity's rainbow"

[digital_signature_gpg.asc (application/pgp-signature, inline)]

Message #190 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Olaf van der Spek <Olaf@XWIS.Net>

To: martin f krafft <madduck@debian.org>

Cc: 254243@bugs.debian.org

Subject: Re: Bug#254243: (no subject)

Date: Mon, 20 Oct 2008 12:59:02 +0200

martin f krafft wrote:
> also sprach Olaf van der Spek <Olaf@XWIS.Net> [2008.10.20.1244 +0200]:
>> That's fine for machine readable output, but what about human readable  
>> output?
> 
> That's debateable. I find e.g. 'ip route list' output far better
> than /sbin/route by now. The same applies to 'ip address list' vs.
> /sbin/ifconfig. /usr/sbin/arp was always non-tabular.

# arp
Address                  HWtype  HWaddress           Flags Mask 
   Iface
192.168.1.34             ether   00:17:31:64:23:c5   C 
   eth0
192.168.1.254            ether   00:19:cb:43:af:b7   C 
   eth0

This looks pretty tabular to me.

>> And how to handle existing scripts that rely on this specific tabular  
>> output?
> 
> those should hardly depend on column or terminal width.

So that doesn't have to be taken into account? Good.

Message #195 received at 254243@bugs.debian.org (full text, mbox, reply):

From: martin f krafft <madduck@debian.org>

To: Olaf van der Spek <Olaf@XWIS.Net>

Cc: 254243@bugs.debian.org

Subject: Re: Bug#254243: (no subject)

Date: Mon, 20 Oct 2008 13:09:48 +0200

[Message part 1 (text/plain, inline)]

also sprach Olaf van der Spek <Olaf@XWIS.Net> [2008.10.20.1259 +0200]:
> Address                  HWtype  HWaddress           Flags Mask    Iface
> 192.168.1.34             ether   00:17:31:64:23:c5   C    eth0
> 192.168.1.254            ether   00:19:cb:43:af:b7   C    eth0
>
> This looks pretty tabular to me.

Ha! I forgot! I have arp aliased to arp -a:

  -a     Use alternate BSD style output format (with no fixed columns).

-- 
 .''`.   martin f. krafft <madduck@debian.org>
: :'  :  proud Debian developer, author, administrator, and user
`. `'`   http://people.debian.org/~madduck - http://debiansystem.info
  `-  Debian - when you have better things to do than fixing systems
 
"you don't sew with a fork, so i see no reason
 to eat with knitting needles."
                               -- miss piggy, on eating chinese food

[digital_signature_gpg.asc (application/pgp-signature, inline)]

Message #200 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Olaf van der Spek <Olaf@XWIS.Net>

To: martin f krafft <madduck@debian.org>

Cc: 254243@bugs.debian.org

Subject: Re: Bug#254243: (no subject)

Date: Mon, 20 Oct 2008 13:12:33 +0200

martin f krafft wrote:
> also sprach Olaf van der Spek <Olaf@XWIS.Net> [2008.10.20.1259 +0200]:
>> Address                  HWtype  HWaddress           Flags Mask    Iface
>> 192.168.1.34             ether   00:17:31:64:23:c5   C    eth0
>> 192.168.1.254            ether   00:19:cb:43:af:b7   C    eth0
>>
>> This looks pretty tabular to me.
> 
> Ha! I forgot! I have arp aliased to arp -a:
> 
>   -a     Use alternate BSD style output format (with no fixed columns).

Hehe. I really don't think that's good for netstat. Too many columns and 
too many rows.

But as machine readable format, it's the best way I think.

Message #205 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Luar Roji <cyberplant@roji.net>

To: Olaf van der Spek <Olaf@XWIS.Net>, 254243@bugs.debian.org

Cc: martin f krafft <madduck@debian.org>

Subject: Re: Bug#254243: (no subject)

Date: Mon, 20 Oct 2008 17:46:44 -0200

I think we have to fix it soon, some way or another.. 

I would preffer to keep the actual behaviour so no scripts are harmed. But
it's real that we don't want to have to add a new command line flag always
from now! 

Now, another fact is that netstat was always tabular, and I think we are used
to see it this way.. 

Maybe (only maybe!) we can make a NEW version of netstat, and show a warning
when chopping is happening, so we don't break actual scripts, and in the new
version we can do what we want without being afraid, what do you think? In
the future we can rename the old netstat, and use the new as the original..

netstat-ng? 

On Mon, Oct 20, 2008 at 01:12:33PM +0200, Olaf van der Spek wrote:
> martin f krafft wrote:
>> also sprach Olaf van der Spek <Olaf@XWIS.Net> [2008.10.20.1259 +0200]:
>>> Address                  HWtype  HWaddress           Flags Mask    Iface
>>> 192.168.1.34             ether   00:17:31:64:23:c5   C    eth0
>>> 192.168.1.254            ether   00:19:cb:43:af:b7   C    eth0
>>>
>>> This looks pretty tabular to me.
>>
>> Ha! I forgot! I have arp aliased to arp -a:
>>
>>   -a     Use alternate BSD style output format (with no fixed columns).
>
> Hehe. I really don't think that's good for netstat. Too many columns and  
> too many rows.
>
> But as machine readable format, it's the best way I think.

Message #210 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Olaf van der Spek <Olaf@XWIS.Net>

To: Luar Roji <cyberplant@roji.net>

Cc: 254243@bugs.debian.org, martin f krafft <madduck@debian.org>

Subject: Re: Bug#254243: (no subject)

Date: Mon, 20 Oct 2008 21:50:59 +0200

Luar Roji wrote:
> I think we have to fix it soon, some way or another.. 

Before or after Lenny?

Message #215 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Luar Roji <cyberplant@roji.net>

To: Olaf van der Spek <Olaf@XWIS.Net>

Cc: 254243@bugs.debian.org, martin f krafft <madduck@debian.org>

Subject: Re: Bug#254243: (no subject)

Date: Mon, 20 Oct 2008 19:26:30 -0200

I think it's important enough to have it fixed in lenny, but don't know what
other people think about.

On Mon, Oct 20, 2008 at 09:50:59PM +0200, Olaf van der Spek wrote:
> Luar Roji wrote:
>> I think we have to fix it soon, some way or another.. 
>
> Before or after Lenny?
>

Message #220 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Olaf van der Spek <Olaf@XWIS.Net>

To: Luar Roji <cyberplant@roji.net>

Cc: 254243@bugs.debian.org, martin f krafft <madduck@debian.org>

Subject: Re: Bug#254243: (no subject)

Date: Mon, 20 Oct 2008 23:34:54 +0200

Luar Roji wrote:
> I think it's important enough to have it fixed in lenny, but don't know what
> other people think about.

I think it should've been fixed before Etch. ;)

Message #225 received at 254243@bugs.debian.org (full text, mbox, reply):

From: martin f krafft <madduck@debian.org>

To: Luar Roji <cyberplant@roji.net>, 254243@bugs.debian.org

Subject: Re: Bug#254243: (no subject)

Date: Tue, 21 Oct 2008 10:14:34 +0200

[Message part 1 (text/plain, inline)]

also sprach Luar Roji <cyberplant@roji.net> [2008.10.20.2146 +0200]:
> Maybe (only maybe!) we can make a NEW version of netstat, and show a warning
> when chopping is happening, so we don't break actual scripts, and in the new
> version we can do what we want without being afraid, what do you think? In
> the future we can rename the old netstat, and use the new as the original..

This, in addition to the -w flag, sounds like a good idea! The
question now is whether it's possible to patch the source in such
a way that the release team will let us bring it into lenny. I think
it's worth a try and I am happy to help you make it happen.

-- 
 .''`.   martin f. krafft <madduck@debian.org>
: :'  :  proud Debian developer, author, administrator, and user
`. `'`   http://people.debian.org/~madduck - http://debiansystem.info
  `-  Debian - when you have better things to do than fixing systems

[digital_signature_gpg.asc (application/pgp-signature, inline)]

Message #230 received at 254243@bugs.debian.org (full text, mbox, reply):

From: martin f krafft <madduck@debian.org>

To: Luar Roji <cyberplant@roji.net>

Cc: Olaf van der Spek <olaf@xwis.net>, 254243@bugs.debian.org, philb@gnu.org

Subject: Patch for netstat to prevent chopping IPv6 addresses

Date: Tue, 11 Nov 2008 08:54:01 +0100

[Message part 1 (text/plain, inline)]

tags 254243 patch
thanks

@Phil, net-tools upstream: here is a patch proposal for netstat, you
can find information and the discussion leading up to it here:
  http://bugs.debian.org/254243
If you could acknowledge that you like the solution and want to
integrate it into netstat, then we could try to push it into our
forthcoming stable release, which has IPv6 as a release goal.

@Luk, Debian release team: I'd appreciate a short note whether this
is NMU-able and potentially could become part of lenny.



also sprach Luar Roji <cyberplant@roji.net> [2008.11.11.0837 +0100]:
> Well, I've finally done it!
[...]
> The URL is: http://roji.net/net-tools-files.tar.gz

Hi Luar, this is *really good* work. The patch is minimal and
non-intrusive, I wish more people would work like you. Don't worry
about taking a bit longer, we all have our lives to live. It's the
result in the end which counts. And yours is great.

I've tried the new package and I've created a patch as follows, and
also added the patch tag to the bug report (this email is BCC'd to
control@bugs.debian.org).

Let's hope upstream/Phil likes it as much as I do. I'll wait for his
word before I upload an NMU.


$ debdiff net-tools_1.60-21.dsc net-tools_1.60-21.1.dsc
diffstat for net-tools_1.60-21 net-tools_1.60-21.1

 debian/changelog |    7 +++++++
 netstat.c        |   18 +++++++++++++-----
 2 files changed, 20 insertions(+), 5 deletions(-)

diff -u net-tools-1.60/netstat.c net-tools-1.60/netstat.c
--- net-tools-1.60/netstat.c
+++ net-tools-1.60/netstat.c
@@ -149,6 +149,7 @@
 int flag_igmp= 0;
 int flag_rom = 0;
 int flag_exp = 1;
+int flag_wide= 0;
 int flag_prg = 0;
 int flag_arg = 0;
 int flag_ver = 0;
@@ -786,16 +787,20 @@
     get_sname(htons(local_port), "tcp",
         flag_not & FLAG_NUM_PORT));
 
-       if ((strlen(local_addr) + strlen(buffer)) > 22)
-           local_addr[22 - strlen(buffer)] = '\0';
+       if (!flag_wide) {
+           if ((strlen(local_addr) + strlen(buffer)) > 22)
+               local_addr[22 - strlen(buffer)] = '\0';
+       }
 
  strcat(local_addr, ":");
  strcat(local_addr, buffer);
  snprintf(buffer, sizeof(buffer), "%s",
     get_sname(htons(rem_port), "tcp", flag_not & FLAG_NUM_PORT));
 
-       if ((strlen(rem_addr) + strlen(buffer)) > 22)
-           rem_addr[22 - strlen(buffer)] = '\0';
+       if (!flag_wide) {
+           if ((strlen(rem_addr) + strlen(buffer)) > 22)
+               rem_addr[22 - strlen(buffer)] = '\0';
+       }
 
  strcat(rem_addr, ":");
  strcat(rem_addr, buffer);
@@ -1595,7 +1600,7 @@
     getroute_init();           /* Set up AF routing support */
 
     afname[0] = '\0';
-    while ((i = getopt_long(argc, argv, "MCFA:acdegphinNorstuVv?wxl64", longopts, &lop)) != EOF)
+    while ((i = getopt_long(argc, argv, "MCFA:acdegphinNorstuWVv?wxl64", longopts, &lop)) != EOF)
  switch (i) {
  case -1:
      break;
@@ -1639,6 +1644,9 @@
  case 'i':
      flag_int++;
      break;
+       case 'W':
+           flag_wide++;
+           break;
  case 'n':
      flag_not |= FLAG_NUM;
      break;
diff -u net-tools-1.60/debian/changelog net-tools-1.60/debian/changelog
--- net-tools-1.60/debian/changelog
+++ net-tools-1.60/debian/changelog
@@ -1,3 +1,10 @@
+net-tools (1.60-21.1) unstable; urgency=low
+
+  * Non-maintainer upload.
+  * Added new parameter -W to not truncate IP addresses (Closes: #254243)
+
+ -- Luar Roji <cyberplant@roji.net>  Tue, 11 Nov 2008 05:15:42 -0200
+
 net-tools (1.60-21) unstable; urgency=low
 
   * Update maintainer to net-tools Team.

-- 
 .''`.   martin f. krafft <madduck@debian.org>
: :'  :  proud Debian developer, author, administrator, and user
`. `'`   http://people.debian.org/~madduck - http://debiansystem.info
  `-  Debian - when you have better things to do than fixing systems
 
"aus der kriegsschule des lebens -
 was mich nicht umbringt, macht mich härter."
                                                 - friedrich nietzsche

[digital_signature_gpg.asc (application/pgp-signature, inline)]

Message #237 received at 254243@bugs.debian.org (full text, mbox, reply):

From: martin f krafft <madduck@debian.org>

To: Luar Roji <cyberplant@roji.net>

Cc: Olaf van der Spek <olaf@xwis.net>, 254243@bugs.debian.org

Subject: Re: Patch for netstat to prevent chopping IPv6 addresses

Date: Tue, 11 Nov 2008 10:00:19 +0100

[Message part 1 (text/plain, inline)]

also sprach Luar Roji <cyberplant@roji.net> [2008.11.11.0909 +0100]:
> I forgot to add a description of the flag in the help page, do you 
> want I add it?
> 
> What about the manpage?

Yes, that would make sense. Also, it has been suggested to add
--wide as a synonym.

You can just recreate the diff, no need to create a new changelog
entry.

-- 
 .''`.   martin f. krafft <madduck@debian.org>
: :'  :  proud Debian developer, author, administrator, and user
`. `'`   http://people.debian.org/~madduck - http://debiansystem.info
  `-  Debian - when you have better things to do than fixing systems
 
"if confronted with a choice between all the truth in god's right hand
 and the ever live struggle for truth, coupled with eternal error, in
 god's left, i would choose the left."
                                                   -- gotthold lessing

[digital_signature_gpg.asc (application/pgp-signature, inline)]

Message #242 received at 254243@bugs.debian.org (full text, mbox, reply):

From: martin f krafft <madduck@debian.org>

To: Luar Roji <cyberplant@roji.net>

Cc: Olaf van der Spek <olaf@xwis.net>, 254243@bugs.debian.org, philb@gnu.org

Subject: [updated] Patch for netstat to prevent chopping IPv6 addresses

Date: Tue, 11 Nov 2008 14:53:21 +0100

[Message part 1 (text/plain, inline)]

also sprach martin f krafft <madduck@debian.org> [2008.11.11.1000 +0100]:
> Yes, that would make sense. Also, it has been suggested to add
> --wide as a synonym.
> 
> You can just recreate the diff, no need to create a new changelog
> entry.

Attached is the updated patch and upstream is again on Cc. I changed
the changelog entry a bit from what Luar sent, and verified that the
patch is sane and obeys style rules etc. I also rephrased the
manpage description.

Still waiting for upstream to give an okay. This is not to sound
impatient, just telling the world that I haven't received one yet.
:)

-- 
 .''`.   martin f. krafft <madduck@debian.org>
: :'  :  proud Debian developer, author, administrator, and user
`. `'`   http://people.debian.org/~madduck - http://debiansystem.info
  `-  Debian - when you have better things to do than fixing systems
 
quantum mechanics: the dreams stuff is made of.

[net-tools_1.60-21.1.debdiff (text/plain, attachment)]

[digital_signature_gpg.asc (application/pgp-signature, inline)]

Message #247 received at 254243-close@bugs.debian.org (full text, mbox, reply):

From: Luk Claes <luk@debian.org>

To: 254243-close@bugs.debian.org

Subject: Bug#254243: fixed in net-tools 1.60-22

Date: Sun, 16 Nov 2008 17:02:03 +0000

Source: net-tools
Source-Version: 1.60-22

We believe that the bug you reported is fixed in the latest version of
net-tools, which is due to be installed in the Debian FTP archive:

net-tools_1.60-22.diff.gz
  to pool/main/n/net-tools/net-tools_1.60-22.diff.gz
net-tools_1.60-22.dsc
  to pool/main/n/net-tools/net-tools_1.60-22.dsc
net-tools_1.60-22_i386.deb
  to pool/main/n/net-tools/net-tools_1.60-22_i386.deb



A summary of the changes between this version and the previous one is
attached.

Thank you for reporting the bug, which will now be closed.  If you
have further comments please address them to 254243@bugs.debian.org,
and the maintainer will reopen the bug report if appropriate.

Debian distribution maintenance software
pp.
Luk Claes <luk@debian.org> (supplier of updated net-tools package)

(This message was generated automatically at their request; if you
believe that there is a problem with it please contact the archive
administrators by mailing ftpmaster@debian.org)


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Format: 1.8
Date: Sun, 16 Nov 2008 17:53:05 +0100
Source: net-tools
Binary: net-tools
Architecture: source i386
Version: 1.60-22
Distribution: unstable
Urgency: low
Maintainer: net-tools Team <pkg-net-tools-maintainers@lists.alioth.debian.org>
Changed-By: Luk Claes <luk@debian.org>
Description: 
 net-tools  - The NET-3 networking toolkit
Closes: 254243
Changes: 
 net-tools (1.60-22) unstable; urgency=low
 .
   * Add new command line option -W/--wide to netstat which prevents
     truncation of IPv6 addresses (Closes: #254243). Thanks Luar Roji.
Checksums-Sha1: 
 de5608534270fd0d2b4621dd7a50613ceacdf291 1070 net-tools_1.60-22.dsc
 0e4502c63ebc4201b4a6d458da7e0663dcb345b2 191308 net-tools_1.60-22.diff.gz
 e974963aeab15af592baa08d9a8132a32f55d175 264606 net-tools_1.60-22_i386.deb
Checksums-Sha256: 
 b2018137415aa14f4fe0c38249650ab0801cdb37dee629c6e5c2123e6bef386f 1070 net-tools_1.60-22.dsc
 13b422ab816e96f77c1f29cfad5885d9c12a402c6c908314d7de3a681ee75138 191308 net-tools_1.60-22.diff.gz
 c66b55778c6d6db78da3303f81e2f5f35e18b32d593e5af7ec2d9c3f5db9ccba 264606 net-tools_1.60-22_i386.deb
Files: 
 4363de6eca4bbcdb9c718fc9f4caeda6 1070 net important net-tools_1.60-22.dsc
 2b9b816167b67d2714de52c845828644 191308 net important net-tools_1.60-22.diff.gz
 bba2a9db8105ed8da392150485f24f1b 264606 net important net-tools_1.60-22_i386.deb

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)

iEYEARECAAYFAkkgUNIACgkQ5UTeB5t8Mo3RPwCeN9nL/K/Z3wqiu/HEgc6dKzxz
tTMAoIjV+yyIfdbWFogmcRuyl2FLsMc7
=1+6N
-----END PGP SIGNATURE-----

Message #252 received at 254243@bugs.debian.org (full text, mbox, reply):

From: martin f krafft <madduck@debian.org>

To: Luar Roji <cyberplant@roji.net>

Cc: Olaf van der Spek <olaf@xwis.net>, 254243@bugs.debian.org, philb@gnu.org

Subject: Re: Bug#254243: [updated] Patch for netstat to prevent chopping IPv6 addresses

Date: Sun, 16 Nov 2008 18:36:04 +0100

[Message part 1 (text/plain, inline)]

also sprach martin f krafft <madduck@debian.org> [2008.11.11.1453 +0100]:
> Still waiting for upstream to give an okay. This is not to sound
> impatient, just telling the world that I haven't received one yet.

Luk informed me that the patch has been committed upstream and he
pushed 1.60-22 to the archive just now.

Thanks Luar. You've made the (IPv6) world a better place. :)

-- 
 .''`.   martin f. krafft <madduck@debian.org>
: :'  :  proud Debian developer, author, administrator, and user
`. `'`   http://people.debian.org/~madduck - http://debiansystem.info
  `-  Debian - when you have better things to do than fixing systems
 
if loving linux is wrong, i don't want to be right.

[digital_signature_gpg.asc (application/pgp-signature, inline)]

Message #257 received at 254243@bugs.debian.org (full text, mbox, reply):

From: Ferenc Wagner <wferi@niif.hu>

To: 254243@bugs.debian.org

Cc: Luk Claes <luk@debian.org>

Subject: will it get a freeze exception?

Date: Tue, 02 Dec 2008 15:10:50 +0100

Hi,

Thanks for finally fixing this long standing and disturbing bug!
Do you think it will get a Lenny freeze exception? :)
-- 
Cheers,
Feri.

Send a report that this bug log contains spam.