Date: Mon, 04 Nov 2002 01:07:22 -0700 (MST) From: "M. Warner Losh" <imp@bsdimp.com> To: jhay@icomtek.csir.co.za Cc: n_hibma@van-laarhoven.org, current@FreeBSD.ORG, phk@critter.freebsd.dk Subject: Re: umass CF geometry problems, was Re: fdisk -BI ob clean disk broken Message-ID: <20021104.010722.110976721.imp@bsdimp.com> In-Reply-To: <200211040802.gA482oJj093121@zibbi.icomtek.csir.co.za> References: <20021104084214.D17734-100000@uitsmijter.van-laarhoven.org> <200211040802.gA482oJj093121@zibbi.icomtek.csir.co.za>
next in thread | previous in thread | raw e-mail | index | archive | help
In message: <200211040802.gA482oJj093121@zibbi.icomtek.csir.co.za> John Hay <jhay@icomtek.csir.co.za> writes: : > What is the GET_GEOMETRY used for anyway? : : Well the short version of the problem is that "fdisk -BI <disk>" works : on -stable to get a FreeBSD partition on the Compact Flash. This does : not work on -current anymore. I have traced that back to the commit : in umass.c rev 1.61 that removed the fake geometry setting and just : leave the cylinders, heads and sectors_per_track zero. This cause : fdisk to coredump with a floating point error. fdisk is using them, btw, to create a MBR which needs these fields to be somewhat sane. The floating point error likely is because we're dividing by zero on this case: #define RoundCyl(x) ((((x) + cylsecs - 1) / cylsecs) * cylsecs) if cylsecs is 0, guess what happens. We do similar things with dos_cylsecs in init_sector0. There's also code in get_params() that devides by dos_heads * 512 * dos_sectors: static int get_params() { int error; u_int u; off_t o; error = ioctl(fd, DIOCGFWSECTORS, &u); if (error == 0) sectors = dos_sectors = u; error = ioctl(fd, DIOCGFWHEADS, &u); if (error == 0) heads = dos_heads = u; dos_cylsecs = cylsecs = heads * sectors; disksecs = cyls * heads * sectors; error = ioctl(fd, DIOCGSECTORSIZE, &u); if (error != 0) u = 512; error = ioctl(fd, DIOCGMEDIASIZE, &o); if (error == 0) { disksecs = o / u; cyls = dos_cyls = o / (u * dos_heads * dos_sectors); } return (disksecs); } fdisk likely should do something sane in the face of such insanity, but it is unclear what and fdisk is a royal pita to work on anyway :-( Warner To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20021104.010722.110976721.imp>