From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-fsdevel-owner@vger.kernel.org>
Received: from zeniv.linux.org.uk ([195.92.253.2]:53120 "EHLO
	ZenIV.linux.org.uk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1753583AbcCJT76 (ORCPT
	<rfc822;linux-fsdevel@vger.kernel.org>);
	Thu, 10 Mar 2016 14:59:58 -0500
Date: Thu, 10 Mar 2016 19:59:51 +0000
From: Al Viro <viro@ZenIV.linux.org.uk>
To: "Drokin, Oleg" <oleg.drokin@intel.com>
Cc: "Dilger, Andreas" <andreas.dilger@intel.com>,
	Linus Torvalds <torvalds@linux-foundation.org>,
	"<linux-fsdevel@vger.kernel.org>" <linux-fsdevel@vger.kernel.org>,
	Theodore Ts'o <tytso@mit.edu>, Mark Fasheh <mfasheh@suse.com>
Subject: Re: races in ll_splice_alias() and elsewhere (ext4, ocfs2)
Message-ID: <20160310195951.GL17997@ZenIV.linux.org.uk>
References: <20160308160537.GV17997@ZenIV.linux.org.uk>
 <498D5A19-9E55-48D7-B5CF-34CA5769FF7F@intel.com>
 <20160308211148.GX17997@ZenIV.linux.org.uk>
 <A5CA4682-77B1-4DF4-A784-7C7C2FAFC9AF@intel.com>
 <20160309003416.GY17997@ZenIV.linux.org.uk>
 <7C3EBB6F-54AC-4744-BEC1-33EA82216F85@intel.com>
 <20160309012658.GZ17997@ZenIV.linux.org.uk>
 <34C2B1C3-2B7F-490B-A03A-3BCDDFC8BE48@intel.com>
 <20160310022041.GF17997@ZenIV.linux.org.uk>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20160310022041.GF17997@ZenIV.linux.org.uk>
Sender: linux-fsdevel-owner@vger.kernel.org
List-ID: <linux-fsdevel.vger.kernel.org>

On Thu, Mar 10, 2016 at 02:20:42AM +0000, Al Viro wrote:

> I'll pick it.  Mind if I fold it into the one I'd posted (with credits,
> obviously)?

Actually, it's still racy - there's a window between d_obtain_alias() and
ll_d_init() where it can be picked by d_splice_alias().

Sigh...  I'm really tempted to just add ->d_init() and let d_alloc() call it
if non-NULL.  

The question is what should it get and what should it be the method of...

How about
	int (*d_init)(struct dentry *);
in dentry_operations, with __d_alloc() ending with
        d_set_d_op(dentry, dentry->d_sb->s_d_op);
	if (unlikely(dentry->d_op && dentry->d_op->d_init)) {
		int err = dentry->d_op->d_init(dentry);
		if (unlikely(err)) {
			dentry_free(dentry);
			return ERR_PTR(err);
		}
	}
        this_cpu_inc(nr_dentry);

        return dentry;
}

Then lustre would simply have

int ll_d_init(struct dentry *de)
{
	struct ll_dentry_data *lld = kzalloc(sizeof(*lld), GFP_NOFS);
	if (unlikely(!lld))
		return -ENOMEM;
	lld->lld_invalid = 1;
	smp_wmb();	/* read barrier in whatever will find us */
	de->d_fsdata = lld;
        return 0;
}

as its ->d_init() and forget about all that mess.

Objections, better ideas?