Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Fri, 2 Apr 2021 14:27:08 +0100
From:   Matthew Wilcox <willy@infradead.org>
To:     Hugh Dickins <hughd@google.com>
Cc:     Andrew Morton <akpm@linux-foundation.org>, linux-mm@kvack.org,
        linux-fsdevel@vger.kernel.org, linux-nvdimm@lists.01.org,
        linux-kernel@vger.kernel.org
Subject: Re: BUG_ON(!mapping_empty(&inode->i_data))
Message-ID: <20210402132708.GM351017@casper.infradead.org>
References: <alpine.LSU.2.11.2103301654520.2648@eggly.anvils>
 <20210331024913.GS351017@casper.infradead.org>
 <alpine.LSU.2.11.2103311413560.1201@eggly.anvils>
 <20210401170615.GH351017@casper.infradead.org>
 <20210402031305.GK351017@casper.infradead.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20210402031305.GK351017@casper.infradead.org>
Precedence: bulk

On Fri, Apr 02, 2021 at 04:13:05AM +0100, Matthew Wilcox wrote:
> +	for (;;) {
> +		xas_load(xas);
> +		if (!xas_is_node(xas))
> +			break;
> +		xas_delete_node(xas);
> +		xas->xa_index -= XA_CHUNK_SIZE;
> +		if (xas->xa_index < index)
> +			break;

That's a bug.  index can be 0, so the condition would never be true.
It should be:

		if (xas->xa_index <= (index | XA_CHUNK_MASK))
			break;
		xas->xa_index -= XA_CHUNK_SIZE;

The test doesn't notice this bug because the tree is otherwise empty,
and the !xas_is_node(xas) condition is hit first.  The next test will
notice this.