From: BillW50 on
In news:hs4n0p$33b$1(a)news.eternal-september.org,
Barry Watzman typed on Sat, 08 May 2010 18:01:58 -0400:
> Re: "But all of the ones I have seen was limited by the BIOS alone."
>
> No, the actual IDE port hardware itself can make bypassing the 137 GB
> barrier impossible. There are address registers in the hardware that
> may not support 48-bit LBA addressing. [BTW, the scheme which
> preceded 48-bit LBA, which limits you to 137GB, was 28-bit LBA]
>
> Partitioning is not a solution in the sense that you were proposing,
> although it can let you use only the 1st 137GB of a larger drive (the
> rest being entirely and completely unused). This has to be done with
> care, because if the OS attempts a write beyond 137GB, you can be
> hosed.
> Of course, there are systems in which the only limitations are the OS
> or the BIOS.

Yes like I said, I am sure you can dig up a controller old enough that
can't handle 48-bit LBA addressing. I don't doubt this for a second.
Although since 2000 (that is 10 years old now) that I haven't seen been
able to handle it. If you or anybody else have seen a controller newer
than 2000 that can't, I sure would love to hear about it.

And if you go 20 years back, I don't think there was any PATA
controllers yet. Well they started about this time from what I remember.
So any controller built between 1990 to 2000 would be in question IMHO.
There was some BIOS that couldn't handle 48-bit LBA up to 2002 or 2003
(even though the controller was 2000 or newer). But that isn't a problem
since you can always use partitions to get around that problem.

Anything older than 2000, well most of them couldn't run an OS that
supports 48-bit LBA (well) anyway. So forget anything larger than 137GB
anyway. At least on the ATA bus anyway.

--
Bill
Gateway M465e ('06 era) - Windows XP SP3