Hi, On Sun, Feb 05, 2012 at 03:02:40PM +0100, David Sommerseth wrote: > >as promised :-) - it now passes all my t_client.rc tests, and smells > >a lot less funny. We now might even throw out all these weird > >#ifdefs from tun.h... > > Maybe an attachment would be nice? ;-)
And I thought you could just fetch this from the git repo on my laptop
(suspended to disk and turned off, of course!). Well, that's what you
get when debugging assert() crashes instead concentrating on your
e-mail :-)
gert
--
USENET is *not* the non-clickable part of WWW!
//www.muc.de/~gert/
Gert Doering - Munich, Germany g...@greenie.muc.de
fax: +49-89-35655025 g...@net.informatik.tu-muenchen.de
From df533746e6bf6b66020b6da4720b20e26878aedf Mon Sep 17 00:00:00 2001
From: Gert Doering <g...@obsd49.ov.greenie.net>
List-Post: openvpn-devel@lists.sourceforge.net
Date: Sun, 5 Feb 2012 13:35:03 +0100
Subject: [PATCH] Platform cleanup for OpenBSD
Turn around initialization order (IFCONFIG_AFTER_TUN_OPEN) to make it
"the same as all other platforms besides Windows" (tun.h).
Remove "ifconfig destroy / ifconfig create" from open_tun() and change
to generic "create tun device by opening /dev/tunN" approach, thus
cleaning up the IFCONFIG_BEFORE_TUN_OPEN bit.
Add "-link0" for ifconfig calls in tun mode, to make sure that even if
we happen to re-use a not-cleaned-up tun interface in tap mode, it will
then be setup correctly (-link0 -> tun, link0 -> tap).
Add correct ifconfig calls for "topology subnet".
On tunnel close, only call "ifconfig destroy" if it was a tap interface
(tun + link0), because those do not auto-disappear (OpenBSD bug?)
Get rid of READV/WRITEV #ifdef's - as per the man page, these calls have
been added to 4.2BSD, and there never was an OpenBSD version without.
Tested on OpenBSD 4.9 with tun+tap, ipv4+ipv6, topology net30+subnet
Signed-off-by: Gert Doering <g...@greenie.muc.de>
---
tun.c | 60 +++++++++++++++++++++++++++++++++---------------------------
tun.h | 2 +-
2 files changed, 34 insertions(+), 28 deletions(-)
diff --git a/tun.c b/tun.c
index 1527ac8..1a4a16f 100644
--- a/tun.c
+++ b/tun.c
@@ -880,29 +880,15 @@ do_ifconfig (struct tuntap *tt,
#elif defined(TARGET_OPENBSD)
/*
- * OpenBSD tun devices appear to be persistent by default. It seems in
order
- * to make this work correctly, we need to delete the previous instance
- * (if it exists), and re-ifconfig. Let me know if you know a better
way.
+ * On OpenBSD, tun interfaces are persistant if created with
+ * "ifconfig tunX create", and auto-destroyed if created by
+ * opening "/dev/tunX" (so we just use the /dev/tunX)
*/
- argv_printf (&argv,
- "%s %s destroy",
- IFCONFIG_PATH,
- actual);
- argv_msg (M_INFO, &argv);
- openvpn_execve_check (&argv, es, 0, NULL);
- argv_printf (&argv,
- "%s %s create",
- IFCONFIG_PATH,
- actual);
- argv_msg (M_INFO, &argv);
- openvpn_execve_check (&argv, es, 0, NULL);
- msg (M_INFO, "NOTE: Tried to delete pre-existing tun/tap instance -- No
Problem if failure");
-
/* example: ifconfig tun2 10.2.0.2 10.2.0.1 mtu 1450 netmask
255.255.255.255 up */
if (tun)
argv_printf (&argv,
- "%s %s %s %s mtu %d netmask 255.255.255.255 up",
+ "%s %s %s %s mtu %d netmask 255.255.255.255 up
-link0",
IFCONFIG_PATH,
actual,
ifconfig_local,
@@ -910,6 +896,19 @@ do_ifconfig (struct tuntap *tt,
tun_mtu
);
else
+ if ( tt->topology == TOP_SUBNET )
+ {
+ argv_printf (&argv,
+ "%s %s %s %s mtu %d netmask %s up -link0",
+ IFCONFIG_PATH,
+ actual,
+ ifconfig_local,
+ ifconfig_local,
+ tun_mtu,
+ ifconfig_remote_netmask
+ );
+ }
+ else
argv_printf (&argv,
"%s %s %s netmask %s mtu %d broadcast %s link0",
IFCONFIG_PATH,
@@ -1959,10 +1958,6 @@ read_tun (struct tuntap* tt, uint8_t *buf, int len)
#elif defined(TARGET_OPENBSD)
-#if !defined(HAVE_READV) || !defined(HAVE_WRITEV)
-#error openbsd build requires readv & writev library functions
-#endif
-
/*
* OpenBSD has a slightly incompatible TUN device from
* the rest of the world, in that it prepends a
@@ -2006,15 +2001,26 @@ open_tun (const char *dev, const char *dev_type, const
char *dev_node, struct tu
}
}
-/* the current way OpenVPN handles tun devices on OpenBSD leads to
- * lingering tunX interfaces after close -> for a full cleanup, they
- * need to be explicitely destroyed
+/* tun(4): "If the device was created by opening /dev/tunN, it will be
+ * automatically destroyed. Devices created via ifconfig(8) are
+ * only marked as not running and traffic will be dropped
+ * returning EHOSTDOWN."
+ * --> no special handling should be needed - *but* OpenBSD is misbehaving
+ * here: if the interface was put in tap mode ("ifconfig tunN link0"), it
+ * *will* stay around, and needs to be cleaned up manually
*/
-
+
void
close_tun (struct tuntap* tt)
{
- if (tt)
+ /* only *TAP* devices need destroying, tun devices auto-self-destruct
+ */
+ if (tt && tt->type == DEV_TYPE_TUN )
+ {
+ close_tun_generic (tt);
+ free(tt);
+ }
+ else if (tt)
{
struct gc_arena gc = gc_new ();
struct argv argv;
diff --git a/tun.h b/tun.h
index c0785dd..513257f 100644
--- a/tun.h
+++ b/tun.h
@@ -283,7 +283,7 @@ ifconfig_order(void)
#elif defined(TARGET_SOLARIS)
return IFCONFIG_AFTER_TUN_OPEN;
#elif defined(TARGET_OPENBSD)
- return IFCONFIG_BEFORE_TUN_OPEN;
+ return IFCONFIG_AFTER_TUN_OPEN;
#elif defined(TARGET_DARWIN)
return IFCONFIG_AFTER_TUN_OPEN;
#elif defined(TARGET_NETBSD)
--
1.7.3.5
pgpyAXycqMF_p.pgp
Description: PGP signature
