NFS: Fix inode corruption in nfs_prime_dcache()

Due to inode number reuse in filesystems, we can end up corrupting the
inode on our client if we apply the file attributes without ensuring that
the filehandle matches.
Typical symptoms include spurious "mode changed" reports in the syslog.

We still do want to ensure that we don't invalidate the dentry if the
inode number matches, but we don't have a filehandle.

Fixes: fa9233699c ("NFS: Don't require a filehandle to refresh...")
Signed-off-by: Trond Myklebust <[email protected]>
Cc: [email protected] # v4.0+
Tested-by: Oleg Drokin <[email protected]>
Signed-off-by: Anna Schumaker <[email protected]>
This commit is contained in:
Trond Myklebust 2016-09-22 13:38:52 -04:00 • committed by Anna Schumaker
commit 7dc72d5f7a

View file

@ -435,11 +435,11 @@ int nfs_same_file(struct dentry *dentry, struct nfs_entry *entry)
return 0;
nfsi = NFS_I(inode);
if (entry->fattr->fileid == nfsi->fileid)
return 1;
if (nfs_compare_fh(entry->fh, &nfsi->fh) == 0)
return 1;
return 0;
if (entry->fattr->fileid != nfsi->fileid)
return 0;
if (entry->fh->size && nfs_compare_fh(entry->fh, &nfsi->fh) != 0)
return 0;
return 1;
}
static
@ -525,6 +525,8 @@ again:
&entry->fattr->fsid))
goto out;
if (nfs_same_file(dentry, entry)) {
if (!entry->fh->size)
goto out;
nfs_set_verifier(dentry, nfs_save_change_attribute(dir));
status = nfs_refresh_inode(d_inode(dentry), entry->fattr);
if (!status)
@ -537,6 +539,10 @@ again:
goto again;
}
}
if (!entry->fh->size) {
d_lookup_done(dentry);
goto out;
}
inode = nfs_fhget(dentry->d_sb, entry->fh, entry->fattr, entry->label);
alias = d_splice_alias(inode, dentry);