Date: Sun, 19 Mar 1995 18:45:26 +1000 From: Bruce Evans <bde@zeta.org.au> To: phk@ref.tfs.com, rgrimes@gndrsh.aac.dev.com Cc: bde@zeta.org.au, dufault@hda.com, freebsd-hackers@freefall.cdrom.com, pst@Shockwave.COM Subject: Re: kern/248: scbus attach/probe printf inconsistency Message-ID: <199503190845.SAA12153@godzilla.zeta.org.au>
next in thread | raw e-mail | index | archive | help
>> Ahh, we misunderstand each other! >> >> I'm only talking about sd.c, not wd.c. >> >> For wd.c disks the information has a very real value, for scsi: absolutely >> none. >For scsi the only unreal value is sectors/track. All other values are >as reported by the drive. I think this is why the wd info is more valuable. The IDE drives that I'm familiar with (not many) report translated values in wdgetctlr(). For drives smaller than 504MB, it is possible for the translated values to be usable both in the partition table and for accessing the drive, so there is little possiblity of confusion. I think the translated values become unsuitable for accessing the drive above 504MB, so wd.c breaks. Some drives report more-physical values in a place that `struct wdparams' doesn't support. The driver doesn't really need to know about these - it can invent a suiatble geomtry. Bruce
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199503190845.SAA12153>