Date: Mon, 1 Jun 1998 09:48:25 +0930 From: Greg Lehey <grog@lemis.com> To: walton@nordicdms.com, freebsd-questions@FreeBSD.ORG Subject: Re: fdisk slice oddities Message-ID: <19980601094825.S22406@freebie.lemis.com> In-Reply-To: <19980531235417422.AAA277@mail.nordicdms.com>; from Dave Walton on Sun, May 31, 1998 at 04:54:17PM %2B0000 References: <19980530073409264.AAA277@mail.nordicdms.com>; <19980530185741.K20360@freebie.lemis.com> <19980531235417422.AAA277@mail.nordicdms.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, 31 May 1998 at 16:54:17 +0000, Dave Walton wrote: > On 30 May 98 at 18:57, Greg Lehey wrote: > >> Dedicated mode. It's not only not dangerous, it's safer than >> non-dedicated mode. > > Really? Why is it safer? It doesn't rely on the BIOS as much. It doesn't rely on Microsoft partition tables. I think, but am not sure, that once the bootstrap has been loaded from the first track (which is the first track no matter how the geometry is defined), it can always find the rest of the system. > Sysinstall calls it "dangerously dedicated", BTW. A lot of people use this term. I haven't seen a valid reasoning for it yet. >>> Is there some particular reason why dangerously dedicated mode >>> changes the drive geometry? >> >> It might come closer to getting it right. > > That's just it... It's not changing to a more correct geometry, it's > changing to a geometry with < 1024 cylinders. It doesn't really matter what geometry you select as long as you can address the data on the disk. The only reason for these fake cylinders, heads and sector counts is that some controllers require them. In fact, it's not always < 1024 cylinders. > For example, on this 6.5GB drive that I'm working with right now, > dmesg and fdisk report the geometry as 13328/15/63. If I use the > whole disk in "dedicated mode", the geometry changes to 784/255/63. I have an IBM DHEA drive here, 6.4 GB. dmesg tells me: wdc1: unit 0 (wd2): <IBM-DHEA-36480>, DMA, 32-bit, multi-block-16 wd2: 6197MB (12692736 sectors), 12592 cyls, 16 heads, 63 S/T, 512 B/S Like all my disks, it's running in dedicated mode. disklabel tells me: sectors/track: 63 tracks/cylinder: 16 sectors/cylinder: 1008 cylinders: 12591 In other words, with the exception of the number of cylinders, the same geometry. I suspect that a cylinder has been reserved for something else, such as bad track tables. >> I suppose the most important difference is that no modern disks have a >> rigid geometry any more (in other words, the terms cylinders, heads >> and sectors are all faked). All we care about is the LBA (logical >> block addressing) mode. > > Which is why I'm surprised that fdisk changes to something other than > LBA mode for "dedicated mode". I think this is historical. We should change it. >> Apart from the observation, did you have any problems? > > No, it just strikes me as a peculiar behavior, so I thought I'd > mention it. > > By the way, given the choice between an LBA geometry and a <1024 > geometry, is there any particular reason to go with one over the > other? I suppose there could be a minimal difference in efficiency. I haven't looked at the code, though, so I can't say for sure. Greg -- See complete headers for address and phone numbers finger grog@lemis.com for PGP public key To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19980601094825.S22406>