Hi!
On Tue 11-04-23 09:47:08, Carlos Maiolino wrote:
> My apologies it took a while to get back to you on this one, I was
> accumulating reviews before running through all of them.
Sure, that is fine.
> > > inode = shmem_get_inode(idmap, dir->i_sb, dir, mode, 0, VM_NORESERVE);
> > > - if (inode) {
> > > - error = security_inode_init_security(inode, dir,
> > > - NULL,
> > > - shmem_initxattrs, NULL);
> > > - if (error && error != -EOPNOTSUPP)
> > > - goto out_iput;
> > > - error = simple_acl_create(dir, inode);
> > > - if (error)
> > > - goto out_iput;
> > > - d_tmpfile(file, inode);
> > > - }
> > > +
> > > + if (IS_ERR(inode))
> > > + return PTR_ERR(inode);
> >
> > This doesn't look correct. Previously, we've called
> > finish_open_simple(file, error), now you just return error... Otherwise the
> > patch looks good to me.
>
> I see what you mean. But, finish_open_simple() simply does:
>
> if (error)
> return error;
>
> So, calling it with a non-zero value for error at most will just add another
> function call into the stack.
I see, I didn't look inside finish_open_simple() :). Well, it is at least
inline function so actually no function call.
> I'm not opposed to still call finish_open_simple() but I don't think it adds
> anything.
Yeah, the value of the call is questionable but given it takes 'error'
argument only so that it could do that if (error) return error; dance, it
seems to be very much expected it is called in error cases as well.
> If you prefer it being called. I thought about adding a new label, something
> like:
>
> if (IS_ERR(inode))
> goto err_out;
> .
> .
> .
> d_tmpfile(file, inode)
> err_out:
> return finish_open_simple(file, error)
> .
> .
> .
>
> Would it work for you?
Yeah, I'd prefer we keep calling finish_open_simple(). Adding a label for
it is fine by me.
Honza
--
Jan Kara <jack@xxxxxxxx>
SUSE Labs, CR
