Hide Forgot
Description of problem: When issuing a umount command you can get the following message "mount disagrees with the fstab". This message doesn't give any usefull information about what really happens. Here's what was the trigger of it in my install and how I fixed it : I had this line in my fstab file * server:/share /share nfs noauto,user,defaults 0 0 but then umount gave me the not very helpfull message... I found this doing a 'cat /etc/mtab | grep share' server:/share/ /share nfs rw,nosuid,nodev,noexec,relatime,vers=3,rsize=32768,wsize=32768,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,mountaddr=192.168.0.1,mountvers=3,mountport=892,mountproto=udp,local_lock=none,addr=192.168.0.1 0 0 I finally understood that the PRECISE value in mtab and fstab syntaxicaly didn't match, then I changed my fstab line to this : * server:/share/ /share nfs noauto,user,defaults 0 0 This time no more problem. The error messages needs to be more verbose and explicative about what the problem is, or the mount/umount executables needs to be modified to be more aware about a simple "/" difference. The Libc logic "strcmp(str1, str2)≠0 means different" is a bit overkill in this context and needs a better approach. Version-Release number of selected component (if applicable): How reproducible: always Steps to Reproduce: 1.do not include a / at the end of a nfs mount source in fstab 2.mount 3.umount Actual results: says that the umount is not possible because "mount disagrees with the fstab" Expected results: understands that /share/ and /share means the same and do the umount Additional info:
This is known (and fixed) issue in F17... well, it seems that we have to backport the umount part of the upstream commit b106d052383083b80c0dc41f1555d2661db00374 to F16.
It's not so fatal problem to backport the patch to Fedora16. Closing.
just for the record, I disagree totally your point-of-view on this matter. The devil is in the details and lots of those "minor" concerns spoils every Fedora distribution I worked on since many years. Having a "killer feature" does not make a release the better ever, it's about being bug, regression and issue free.