Date: Fri, 6 Feb 1998 18:36:49 +0000 (GMT) From: Terry Lambert <tlambert@primenet.com> To: cgull+usenet-886763413@smoke.marlboro.vt.us (john hood) Cc: alk@pobox.com, hackers@FreeBSD.ORG Subject: Re: wd0s1e hard errors Message-ID: <199802061836.LAA12807@usr01.primenet.com> In-Reply-To: <199802061118.GAA08425@smoke.marlboro.vt.us> from "john hood" at Feb 6, 98 06:18:59 am
next in thread | previous in thread | raw e-mail | index | archive | help
> > Should bad144 be retired? > > Yup, from current, anyway. > > It adds complexity to the wd driver, various bits of the code have > often been broken, and I'd guess that support for > MFM/RLL/ESDI/antiqueIDE controllers has suffered enough rot that it > doesn't work anyway. The code isn't broken. You just can't have replacement sectors over the 1024 boundary. My persional opinion is that it should go "slice a", "replacement sectors", "slice b", and so on. This would better guarantee that sectors will not be over the 1024 boundry, as well as allowing for expansion of the table by reducing the amount of swap, if necessary. Even without this, the problems people see are attributable to pilot error; you can perform the same workaround (keeping the replacement sectors under 1024) by setting up two DOS partitions, and putting "slice a" in one by itself. The bad144 code doesn't belong in the disk driver, and more than the partition/disklabel/slice stuff belongs in the disk driver. Just because it's in the wrong place doesn't mean you should get rid of it -- if that were true, unless you used Julian's slice code on top of devfs, DOS partitioning and BSD disklabel management would be diked out tomorrow. The standard IDE driver works with wd drives (that's why it'still called "wd" instead of "ide"). Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199802061836.LAA12807>