Machine A.
mount -t iso9660 -o ro,loop /foo/iso /mnt/iso
/etc/exports contains /mnt *(rw,no_root_squash,no_all_squash)
Machine B.
mount -t nfs A:/mnt/iso /local/iso
B says
mount: A:/mnt/iso failed, reason given by server: Permission denied
A says
rpc.mountd: authenticated mount request from ... for /mnt/iso(/mnt)
rpc.mountd: getfh failed: Operation not permitted
strace rpc.mountd on A
stat64("/mnt/iso", {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0
nfsservctl(0x8, 0xbfffda80, 0x80553a0) = -1 EPERM (Operation not permitted)
If A does not have the loopback mounted then B can mount A:/mnt/iso
over NFGS, getting an empty directrory. Not a permission problem.
A and B are running 2.4.18-XFS, nfs-utils-0.3.1-13, mount-2.11g-5.
A is built with
CONFIG_NFS_FS=m
CONFIG_NFS_V3=y
# CONFIG_ROOT_NFS is not set
CONFIG_NFSD=m
CONFIG_NFSD_V3=y
CONFIG_SUNRPC=m
CONFIG_LOCKD=m
CONFIG_LOCKD_V4=y
I have a note from April 2000 where NFS access to a loopback mount used
to work. Before I dig through two years of kernels to find out where
it stopped working, is it valid to access loopback via NFS?
AFAIK doing NFS first then loopback on the local NFS directory has
never worked.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/