Date: Wed, 23 Oct 2013 03:03:03 -0700 From: Christoph Hellwig To: Richard Weinberger Subject: Re: [PATCH, RFC] what's the point of mtd_inodefs? Message-ID: <20131023100303.GA20373@infradead.org> References: <20131022153313.GA11923@infradead.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: Cc: Christoph Hellwig , "linux-mtd@lists.infradead.org" , David Woodhouse , LKML , kirill@shutemov.name List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Wed, Oct 23, 2013 at 07:50:17AM +0200, Richard Weinberger wrote: > On Tue, Oct 22, 2013 at 5:33 PM, Christoph Hellwig wrote: > > I've been looking at mtdchar as part of a bigger series touching all > > kidns of places and really wonder what the point of mtd_inodefs is. > > As far I can tell it was introduced because of that issue: > http://lists.infradead.org/pipermail/linux-mtd/2010-April/029593.html Not sure how that issue came up as writeback only ever uses mapping->backing_dev_info for block devices, else it only uses s_bdi to avoid this issue. As pointed out somewhere in that thread the mem driver uses a similar scheme.