Ken Moffat wrote: Hi Ken,
It worked like a charm :-D Thanks for your help - it was just what I needed :-D Best Regards Michael > 2009/10/16 Michael L. <[email protected]>: > > Ah. I think I've had the same problem (it fails during the netfs script). > In my case, that was on a CLFS build, and I assumed something in > that toolchain was causing the problem. Commenting the following > chunk of code in the netfs script worked around it for me.: > > > # Try and stop processes the nice way > # (probably won't work in most cases) > /bin/fuser -SIGTERM -km $NETMOUNTS > /dev/null > > # Check and see if it found anything. If it > # did, then give 3 seconds for things to exit > # the nice way before killing them off. > # This one will work all of the time! > if [ $? = 0 ] > then > /bin/sleep 3 > /bin/fuser -km $NETMOUNTS > /dev/null > fi > > > After that, it continues to umount _netdev, and then to umount by > fstype, so commenting the code on a desktop doesn't seem to > cause any problems. > > All I've established so far is that the call to fuser with -SIGTERM > is described as 'Terminated' (status 143) if run manually, and the > second call is described as 'Killed' (status 137). I've been > assuming something in the clfs toolchain might have caused fuser > to be miscompiled [ mid-September book, with 2.6.31 headers. > If this is LFS, and your problem turns out to match, then it might > be a general problem with psmisc-22.8. > > FWIW, trying to run strace didn't do a lot of good (strace > appeared to be killed by the SIGTERM). > > ĸen > -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
