Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 2 Jul 2009 10:52:35 -0400 (EDT)
From:      Rick Macklem <rmacklem@uoguelph.ca>
To:        "Bjoern A. Zeeb" <bzeeb-lists@lists.zabbadoz.net>
Cc:        FreeBSD current mailing list <current@freebsd.org>
Subject:   Re: Install from NFS onto NFS fails on amd64
Message-ID:  <Pine.GSO.4.63.0907021044350.16757@muncher.cs.uoguelph.ca>
In-Reply-To: <20090702121640.D245@maildrop.int.zabbadoz.net>
References:  <20090702121640.D245@maildrop.int.zabbadoz.net>

next in thread | previous in thread | raw e-mail | index | archive | help


On Thu, 2 Jul 2009, Bjoern A. Zeeb wrote:

> Hi,
>
> trying to installworld from NFS to NFS (which also is the NFS Root
> of the installing and to be updated machien) on amd64 always fails
> with this error.  Anyone knows why that is?
>
[stuff snipped]
> install -s -o root -g wheel -m 555   ldd32 /usr/bin
> rm: /tmp/install.uMl44I7W: Directory not empty
> *** Error code 1
>
> Stop in /zoo/bz/HEAD.svn.
> *** Error code 1
>
> Stop in /zoo/bz/HEAD.svn.
> lion1# 
Well, at a quick glance, I can't see anything in either the NFS client nor
server that generates ENOTEMPTY for rmdir. It would seem to me that it
was generated by the underlying file system on the NFS server for some
reason. (Same goes for the syscall above the client side NFS.)

You could confirm this with a printf() right after VOP_RMDIR() in
sys/nfsserver/nfs_serv.c. (A tcpdump/wireshark trace would tell you
if the ENOTEMPTY was reported by the server, which looks like it is
the case.)

Not much help, but...rick



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.GSO.4.63.0907021044350.16757>