From: Josef S. <js1...@in...> - 2006-07-05 18:29:36
|
Hello, after finding a combination of kernel and root fs which don't throw=20 NPTL-related errors on me (like the ones linked on the "new" webpages do),= =20 working inside UML is all fine but I cannot get the eth0 interface up. This is on a standard Debian/sarge host system (kernel 2.4.27 without any=20 patches like SKAS) No matter which kernel I use, including various flavours of 2.4.27, 2.6.14 = and=20 2.6.16, whenever "ifconfig eth0 up" is executed, the kernel will hang. Even= =20 if written exactly like that, i.e. with the IP address missing, and error=20 message will come up but then the system will still hang. (none):~# ifconfig eth0 up dev_ip_addr - device not assigned an IP address Networking uses tuntap and seems to be detected ok: Netdevice 0 : TUN/TAP backend - IP =3D xxx.xxx.xxx.xxx Unfortunately, gdb won't tell me much about this, but at least something: remove_umid_dir - actually_do_remove failed with err =3D -2 BUG: warning=20 at /home/jdike/linux/2.6/linux-2.6.16/kernel/irq/manage.c:274/free_irq() 08962c2c: [<0805857c>] dump_stack+0x1c/0x20 08962c44: [<0808a440>] free_irq+0x3e/0xe3 08962c70: [<0805ca4b>] free_winch+0x47/0x55 08962c84: [<0805ca94>] winch_cleanup+0x13/0x24 08962c94: [<08057781>] do_uml_exitcalls+0xd/0x1b 08962ca0: [<08057d10>] uml_cleanup+0x12/0x19 08962ca8: [<08065a10>] last_ditch_exit+0x23/0x2b 08962cc8: [<400446f8>] 0x400446f8 08962fb4: [<08066420>] write_sigio_thread+0x27/0x139 08962fe0: [<400f292a>] 0x400f292a NB: would it be a good idea to add "handle SIGUSR1 noprint nostop" to the=20 debugging page on UML? Josef =2D-=20 Tous ces gens qui confondent informatique et bureautique, carte et territoi= re, interface et programme, boutons color=C3=A9s et facilit=C3=A9... =2D S. Blondeel |