I've been running tincd on several machines without a problem. I'm now setting up a new intel-type dual-processor machine, with a custom compiled 2.4.9 linux kernel. When I start the daemon: tincd -n vpn -dddddddddd I get the following debug information in in /var/log/messages Aug 27 17:26:49 q tinc.vpn[971]: Process 972 (tinc-up) exited with non-zero status 255 Aug 27 17:26:49 q tinc.vpn[971]: Ready: listening on port 30001 Aug 27 17:26:49 q tinc.vpn[971]: Trying to connect to 55.55.55.55 Aug 27 17:26:49 q tinc.vpn[971]: Connected to 55.55.55.55 port 30001 Aug 27 17:26:49 q tinc.vpn[971]: Error while reading from ethertap device: File descriptor in bad state Aug 27 17:26:50 q last message repeated 19806 times If I start tinc with the no detach option: tincd -n vpn -D -dddddddddd I get this: SIOCSIFHWADDR: No such device SIOCSIFADDR: No such device vpn: unknown interface: No such device SIOCSIFNETMASK: No such device vpn: unknown interface: No such device I have tinc configured to use the Universal Tun functionality; it is installed and running as a module. (I've also tried with tun compiled into the kernel) The tinc-up is as follows: ifconfig $NETNAME hw ether fe:fd:0:0:0:0 ifconfig $NETNAME 10.10.13.1 netmask 255.255.0.0 -arp Any ideas?? - Tinc: Discussion list about the tinc VPN daemon Archive: http://mail.nl.linux.org/lists/ Tinc site: http://tinc.nl.linux.org/
On Mon, Aug 27, 2001 at 05:46:59PM -0400, Glen wrote:> I've been running tincd on several machines without a problem. > I'm now setting up a new intel-type dual-processor machine, with a custom > compiled 2.4.9 linux kernel.[...]> Aug 27 17:26:49 q tinc.vpn[971]: Error while reading from ethertap device: > File descriptor in bad state > Aug 27 17:26:50 q last message repeated 19806 timesThere have been some changes in the header files of the tun/tap driver in the latest kernels, you might have to recompile tinc to match your kernel version. -- Met vriendelijke groet / with kind regards, Guus Sliepen <guus@sliepen.warande.net> -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 232 bytes Desc: not available Url : http://brouwer.uvt.nl/pipermail/tinc/attachments/20010828/18379dec/attachment.pgp
Glen wrote:> > I've been running tincd on several machines without a problem. > I'm now setting up a new intel-type dual-processor machine, with a custom > compiled 2.4.9 linux kernel.if you are running rh7.1 (other linux's may have done this as well). your /usr/include/asm and /usr/include/linux are real directories with 2.4.2 headers. I came across this exact problem while working with the vtun package. The problem is the ioctl's for the tun/tap interface changed from a shift to an _IOW which comes up with very different values. /usr/include/asm should be a symbolic link to /usr/src/linux/include/asm or /usr/src/linux/include/asm-i386 /usr/include/linux should be a symbolic link to /usr/src/linux/include/linux Hope this helps, Eric - Tinc: Discussion list about the tinc VPN daemon Archive: http://mail.nl.linux.org/lists/ Tinc site: http://tinc.nl.linux.org/
Reasonably Related Threads
- Rsyslog and "Error while reading from Linux ethertap device /dev/net/tun: File descriptor in bad state"
- ethertap dropping packets?
- No response to our critical packet problem
- [Bridge] Bridging with ethertap causing kernel oops
- Problem withs hosts/ -files & Ethertap with 2.4.x