From mboxrd@z Thu Jan  1 00:00:00 1970
From: Christoph Hellwig <hch@infradead.org>
Subject: Re: inode_dio_wait and hole-punch?
Date: Tue, 15 May 2012 02:59:44 -0400
Message-ID: <20120515065944.GC7373@infradead.org>
References: <alpine.LSU.2.00.1205131414270.1547@eggly.anvils>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Cc: Christoph Hellwig <hch@infradead.org>,
	linux-fsdevel@vger.kernel.org
To: Hugh Dickins <hughd@google.com>
Return-path: <linux-fsdevel-owner@vger.kernel.org>
Received: from 173-166-109-252-newengland.hfc.comcastbusiness.net ([173.166.109.252]:36794
	"EHLO bombadil.infradead.org" rhost-flags-OK-OK-OK-OK)
	by vger.kernel.org with ESMTP id S1755718Ab2EOG7o (ORCPT
	<rfc822;linux-fsdevel@vger.kernel.org>);
	Tue, 15 May 2012 02:59:44 -0400
Content-Disposition: inline
In-Reply-To: <alpine.LSU.2.00.1205131414270.1547@eggly.anvils>
Sender: linux-fsdevel-owner@vger.kernel.org
List-ID: <linux-fsdevel.vger.kernel.org>

On Sun, May 13, 2012 at 02:26:40PM -0700, Hugh Dickins wrote:
> In looking through hole-punching implementations, I got the impression
> that perhaps ext4 and ocfs2 might need to inode_dio_wait() before they
> truncate pagecache.  And that might require ext4 to hold i_mutex there?

Anything freeing blocks should be protected by inode_dio_wait, and any
call to it needs i_mutex or equivalent protection against starting new
direct I/O.