From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from bombadil.infradead.org ([65.50.211.133]:53337 "EHLO bombadil.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751673AbdERGbK (ORCPT ); Thu, 18 May 2017 02:31:10 -0400 Date: Wed, 17 May 2017 23:31:09 -0700 From: Christoph Hellwig Subject: Re: [PATCH 1/3] xfs: remove double-underscore integer types Message-ID: <20170518063109.GB722@infradead.org> References: <20170518012618.GT4519@birch.djwong.org> <20170518013043.GU4519@birch.djwong.org> <20170518060158.GP17542@dastard> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20170518060158.GP17542@dastard> Sender: linux-xfs-owner@vger.kernel.org List-ID: List-Id: xfs To: Dave Chinner Cc: "Darrick J. Wong" , xfs , Eric Sandeen > I'm not sure that this is entirely correct when it comes to sparse > endian notations or the way the __ types were intended to be used. > ISTR we used the __ types were originally for the in-memory endian > converted variable definitions that shadowed the on-disk structures. > The cleanup plan I was planning to do was to convert these all to > the linux kernel definitions of __[s,u][8,16,32,64] so it was clear > they shadow on disk structures of specific sizes. > > Once that was done, everything else could be converted to c99 types > (like you've done above) and then we'd be free of all the old > __[u]int*_t types.... Sparse doesn't care - what matters there are the __be* and __le* types. Otherwise it's just a decision if we want to use C99 or Linux types, and if we want to use Linux types if we want to use the __-prefixed ones to avoid name collisions in userspace. I'm fine either way - I'm used to typing uXX so that'd be my first preference, but the uintX_t is a close second, so I'd be perfectly fine with this patch: Reviewed-by: Christoph Hellwig