Couldn't find mount point

David Eckelkamp davide%cadillac.cad.mcc.com at mcc.com
Sat Nov 4 01:48:13 AEST 1989


++++ In Sun-Spots Digest, v8n178 [Miscellaneous]; 
++++ stanonik at nprdc.navy.mil (Ron Stanonic) [RS>] says.
RS> We're getting the following complaint from cron
RS> Your "cron" job
RS> find / -name .nfs\* -mtime +7 -exec rm -f {} \; -o -fstype nfs -prune
RS> produced the following output:
RS> Couldn't find mount point for home
RS> We suspect the complaint is related to automount unmounting something.
RS> Any ideas why this complaint?  Is it indicative of a problem?
RS> Thanks,

Your suspicions are partially correct. The problem is related to the
automouter but is really a bug in the 'find' command. The bug is exercised
only when you use the '-fstype' option to find. We have source code for
4.0 and I was able to find and correct the problem for us. I also
submitted a bug report to Sun with a partial fix.

The problem is in the routine that returns the file system type for the
current filename being processed (gettype in find.c). The routine reads
the mount table looking for the mount trying to find the mount point for
the file system of the current filename. If the file system type is either
IGNORE or SWAP the routine will not find the mount point and print the
error. The automounter mounts itself as type "ignore", at least it did in
4.0 and/or 4.0.1. In 4.0.3 the automounter now mounts itself as type "nfs"
which should not tickle the bug. 

The work-around is to not use the '-fstype' option. The way I did that for
our updatedb script was to run a seperate find command for each physical
device that is mounted on a machine. That is not really very elegant, but
it does work :-)

If you want more info, or the patch that I used to fix the problem, just
ask.

David Eckelkamp         
ARPA: eckelkamp at mcc.com
uucp: {uunet|harvard|gatech|pyramid}!cs.utexas.edu!milano!cadillac!davide
Microelectronics and Computer Technology Corp. (MCC) VLSI CAD Program
3500 W.Balcones Center Dr., Austin,TX 78759            [512] 338-3796



More information about the Comp.sys.sun mailing list