Re: [PATCH 22/26] sys_mknodat(): elevate write count for vfs_mknod/create()

[Dave Hansen <hansendc@us.ibm.com>]

On Sat, 2007-07-07 at 20:25 +0200, Jan Engelhardt wrote:
> On Jul 5 2007 15:43, Dave Hansen wrote:
> 
> >@@ -1911,13 +1911,27 @@ asmlinkage long sys_mknodat(int dfd, con
> > 	error = do_path_lookup(dfd, tmp, LOOKUP_PARENT, &nd);
> > 	if (error)
> > 		goto out;
> >+
> > 	dentry = lookup_create(&nd, 0);
> > 	error = PTR_ERR(dentry);
> >+	if (error)
> >+		goto out_unlock;
> > 
> > 	if (!IS_POSIXACL(nd.dentry->d_inode))
> > 		mode &= ~current->fs->umask;
> >-	if (!IS_ERR(dentry)) {
> >-		switch (mode & S_IFMT) {
> >+	if (S_ISDIR(mode)) {
> >+		error = -EPERM;
> >+		goto out_dput;
> 
> Hm, I wonder why mknod(,S_IFDIR,) returns -EPERM rather than -EINVAL like the
> others? It is a bit misleading; EPERM usually means root could do it
> somehow (unless the kernel is really pedantic).
> IMHO this if(S_ISDIR) should go.

I'd prefer to leave that fight for another day. :)  It'd be easy to fix
up later.

> >+	if (!S_ISREG(mode)  && !S_ISCHR(mode) && !S_ISBLK(mode) &&
> >+	    !S_ISFIFO(mode) && !S_ISSOCK(mode) && (mode != 0)) {
> 
> If you align the two &&s  (^) you can also align ^ the others :)

Good point.  I'll fix it up.