On Fri, Sep 16, 2005 at 07:17:54AM +0200, Tobias Walkowiak wrote: > On Thu, Sep 15, 2005 at 05:13:07PM +0200, Max Vozeler wrote: > > Tobias, does it make a difference if you add loop to /etc/modules? > > no. it was one of my first attempts to get it running. didn't help OK, too bad :) That helped on my udev test system. Thinking about it, perhaps there is some race condition involved when udev creates the device nodes. ISTR that this happens asynchronously to the modprobe command, so when the modprobe returns it is not certain that all loop device nodes have been created. To further debug this it would help to remove the stderr redirection from swapon calls in S10checkroot.sh and S35mountall.sh. Perhaps you could try to add a sleep 5 before the swapon in mountall.sh and see if that makes a difference. Another idea would be" ls -l /dev/loop*; lsmod| grep loop" before and after the swapons. > it really seems that the whole thing is due to udev and the start- / > mount-scripts. will have to (and can ;) live with that That's OK, but if we can solve it we should try. Discussion goes towards making it the default in future Debian releases, so this will probably be affecting more people in the not-too-distant future. cheers, Max - Linux-crypto: cryptography in and on the Linux system Archive: http://mail.nl.linux.org/linux-crypto/
