Date: Sun, 9 Aug 1998 04:32:57 +1000 From: Bruce Evans <bde@zeta.org.au> To: grog@lemis.com, mike@smith.net.au Cc: bde@zeta.org.au, hardware@FreeBSD.ORG, ks@itp.ac.ru Subject: Re: IBM 16 GB IDE HDD Message-ID: <199808081832.EAA01574@godzilla.zeta.org.au>
next in thread | raw e-mail | index | archive | help
>> >> 0x1000 Use LBA addressing instead of the default CHS addressing. >> > >> > This is seriously incomplete and out of date. LBA addressing should almost >> > never be used, since no cases are known where it is necessary, and cases >> > are known where it causes trashed disks (dumping to any drive where the >> > default geometry is not the same as the fake LBA geometry). >> >> What am I missing here? I thought LBA meant "logical block >> addressing". Where does "geometry" (by which I understand bogus >> mapping to cylinder/head/sector values) come into it? > >It has to do with the fact that for legacy interoperability reasons the >LBA and CHS numbers in the MBR have to match up, and the 'wd' driver >still uses the CHS interface to the disk registers (lowest common >denominator). No, it has to do with bugs in the driver. Due to lack of personhours, parts of the driver still use CHS while other parts of the driver use LBA. >The drive itself changes the mapping between the CHS register values >and the LBA address based on what it thinks it's geometry is. Unless >you know what it thinks it is, you can't second-guess it to get your >data where you think it's going. No, the driver decides the mapping. It can choose almost any mapping (except for pre-IDE drives which don't support setting the mapping and yet-to-be-made drives which don't support any mapping). It chooses the drive's default one for simplicity. Bruce To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hardware" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199808081832.EAA01574>