Date: Sun, 26 Jan 2003 23:09:58 -0800 From: Marcel Moolenaar <marcel@xcllnt.net> To: phk@FreeBSD.org Cc: Nate Lawson <nate@root.org>, cvs-committers@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: src/sbin/disklabel disklabel.c Message-ID: <20030127070958.GA2431@dhcp01.pn.xcllnt.net> In-Reply-To: <15853.1043649348@critter.freebsd.dk> References: <Pine.BSF.4.21.0301262230140.84307-100000@root.org> <15853.1043649348@critter.freebsd.dk>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Jan 27, 2003 at 07:35:48AM +0100, phk@FreeBSD.org wrote: > In message <Pine.BSF.4.21.0301262230140.84307-100000@root.org>, Nate Lawson wri > tes: > >On Sun, 26 Jan 2003 phk@freebsd.org wrote: > >> Robert Watson writes: > >> >We've always had a problem with ioctl's operating on storage devices > >> >regardless of the open mode (and permitted access modes) for the devidce > >> >nodes. > >> > >> And it ain't going to get any better as we get more weird "disks" > >> in the GEOM framwork. > > > >I'm not sure why boot block/disk label writes can't be done as ordinary > >writes and semantic checking performed by whichever GEOM layer thinks it > >knows best about that portion of a disk. What is wrong with that > >approach? > > It doesn't work without running a lot of special-case testing code, > possibly several times, on each and every disk I/O request. > > Considering how often people update their bootblocks, this would be > a totally unacceptable tradeoff. One of the thoughts I have on the issue is to have the geom classes have device nodes as side-entrances, independent of the producer- consumer interfaces. Direct access to the device nodes can then easily be checked with each read or write without pessimizing the P/C layering. The ioctl interface not not suited for GPT as the amount of data that has to be read or written is not fixed, as are the LBAs at which the data has to be written. -- Marcel Moolenaar USPA: A-39004 marcel@xcllnt.net To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030127070958.GA2431>