> On Sep 29, 2017, at 10:35 AM, Jules Richardson via cctalk > <cctalk@classiccmp.org> wrote: > > On 09/29/2017 11:20 AM, Jon Elson via cctalk wrote: >> I can only imagine a real AT would be even less likely to handle a drive >> over about 40 MB. > > I think the limit was normally 512MB in the old c/h/s addressing days, wasn't > it? For a connected drive, the BIOS set aside 4 bits for the number of heads > per cylinder, 6 bits for the sectors per track, and 10 bits for the number of > cylinders - i.e. maximums of 16, 64 and 1024. At 512 bytes per sector, that > came out as 512MB.
Yes, and I can say that I’m the one responsible for that restriction. :-o At the time we needed to support our LBA controllers and we figured that the restriction was OK because we’d have a BIOS replacement long before disk drives got that big. We were wrong on both counts: drives got bigger a whole lot faster than we expected and that a BIOS replacement took a whole lot longer to make it into systems. *sigh* > > Older BIOS firmware provided no means for the user to define the geometry of > a connected drive - just a list of predefined types, and those often maxed > out at far less than any 512MB limit. There were various software solutions > to get around it, though. The above (512MB max disk size) came about with the PS/2. Prior to that BIOS just had predefined types. TTFN - Guy
