From mboxrd@z Thu Jan 1 00:00:00 1970 From: Marian Stagarescu Subject: PPTP conntrack w/out DEBUG is broken ??? Date: Wed, 25 Sep 2002 16:48:42 -0400 Sender: netfilter-devel-admin@lists.netfilter.org Message-ID: <200209251648.42936.marian@ti.com> References: <87d6r9jaa4.fsf@sensi.yemanja.net> <1032690651.9594.13.camel@tux> Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Return-path: In-Reply-To: <1032690651.9594.13.camel@tux> Errors-To: netfilter-devel-admin@lists.netfilter.org List-Help: List-Post: List-Subscribe: , List-Id: List-Unsubscribe: , List-Archive: Content-Type: text/plain; charset="iso-8859-1" To: Martin Josefsson , =?iso-8859-15?q?C=E9dric=20Rivard?= Cc: Netfilter-devel , Netfilter i run into one problem with PPTP conntrack: run it with debug on: works run it w/out debug on: does not. i also turned on the patch below and configure the CONFIG_NETFILTER_DEBUG= : compile ip_conntrack_pptp with debug on: all ok compile ip_conntrack_pptp w/out debug: does not even print the ASSERT rea= dlock=20 statements from CONFIG_NETFILTER_DEBUG, PPTP connection is not establishe= d. this debug issue was tracked down to a single printk statement in=20 ip_conntrack_pptp.c: line 244,245: DEBUGP("%s, CID=3D%X, PCID=3D%X\n", strMName[msg] =20 ntohs(*cid), ntohs(*pcid)); i run this on a 2.4.17 mips little endian, single proc, latest pptp ALG f= iles=20 from netfilter, newnat13. how this can be ?=20 thanks for any input, marian On Sunday 22 September 2002 06:30 am, Martin Josefsson wrote: > On Fri, 2002-09-20 at 11:11, C=E9dric Rivard wrote: > > Hi all, > > > > As I've already reported, there is a problem with the PPTP conntrack > > module of pom. It can't be built with the option CONFIG_NETFILTER_DEB= UG. > > And I still receive questions about the first thread I started on thi= s > > problem but without pointing the cause. I hope that everyone is now a= ware > > of this issue. > > This issue came to my attention in the end of August, I produced a patc= h > and I never heard from that person again. The patch is quite ugly but i= t > does solve this specific problem. I was waiting for some more input fro= m > other developers before submitting it to cvs but now I've given up on > that and I've submitted it. patch-o-matic in cvs now works with > debugging enabled. > > or you can apply this patch on your kernel after patching with the > patch-o-matic you are currently using: > > http://www.netfilter.org/~gandalf/patches/ip_conntrack_proto_gre.c-lock= fix