From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoph Hellwig Subject: Re: [PATCH 1/1] kpartx: Use __kernel_daddr_t for solaris_x86_slice.s_start Date: Tue, 9 Jul 2019 06:34:36 -0700 Message-ID: <20190709133436.GA2231@infradead.org> References: <20190708213551.26349-1-petr.vorel@gmail.com> <20190709024742.GA22293@infradead.org> <20190709080204.GA12875@dell5510> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Content-Disposition: inline In-Reply-To: <20190709080204.GA12875@dell5510> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: dm-devel-bounces@redhat.com Errors-To: dm-devel-bounces@redhat.com To: Petr Vorel Cc: Christoph Hellwig , dm-devel@redhat.com List-Id: dm-devel.ids On Tue, Jul 09, 2019 at 10:02:05AM +0200, Petr Vorel wrote: > > > What this really should use is fixed size types. > If it's not specific to __kernel_daddr_t nor daddr_t ("The type of a disk > address") and long is sufficient for all platforms, that's even better. > > I'd be just for removing typedef int daddr_t comment. > > BTW gpart also uses struct solaris_x86_slice, with daddr_t [1]. > I've filed a PR [2], but I guess I'll change it to long. solaris_x86_slice is an on-disk format, defined for good old 32-bit x86 Solaris. So the question is not if it is enough, but if it matches what Solaris does. I don't have the Solaris source at the moment, but here is what the Linux kernel uses: struct solaris_x86_slice { __le16 s_tag; /* ID tag of partition */ __le16 s_flag; /* permission flags */ __le32 s_start; /* start sector no of partition */ __le32 s_size; /* # of blocks in partition */ };