Re: 2.2.9 -> 2.2.10 lilo err

Anthony Barbachan (barbacha@Hinako.AMBusiness.com)
Wed, 16 Jun 1999 18:38:28 -0400


> In <003501beb6f5$f07fc260$04c809c0@hotmail.com> Anthony Barbachan
(barbacha@Hinako.AMBusiness.com) wrote:
> > Does anybody know if this LILO problem is going to be fixed anytime
> > soon?
>
> Not, AFAIK. Can of worms there: you can access disk beyond 1024 cylinder
only
> with new extended BIOS calls and quite a few BIOSes has subtle errors
there :-((
>
> > Is there some reason why LILO can't be fixed to accomodate the > 1023
> > cylinder situations?
>
> What for ? It will need TONS of debugging to catch all possible weird
> implementations and buy you [almost] nothing.
>
> > NT and Win9x do not seam to have this problem about booting after the
> > 1023rd cylinder.
>
> Windows98 does not boots from partitions started after 1023 cylinder here.
> And since msdos.sys/ntkrnl usually (but not always) sit in the very
beginning
> of disk (installer will try to do this) this problem usually is just
irrelevant.
> But Win9x and NT HAVE this problem as well. At least on some systems...
>

Win98 using FAT32 and NT 5 Beta 3 on NTFS seam fine here on large hard
drives and do not complain when being installed on partitions that span the
1023 cylinder boundery, however I do not know about all systems.
Considering the wide use of defragmenters I would think the installer would
at least warn of possible errors, unless the Windows filesystems ensure that
their kernels are kept below 1024. If thats the case it may not be a bad
idea to have similar functionality in the next Linux filesystem.

> > Drives are already huge and just getting bigger every day so I suspect
more
> > and more people are going to start running into this problem; especially
those
> > who may already have the first part of their drive in use by another OS
or
> > those who do not want the complication of multiple partitions on their
drive
> > as well as those less knowledgable users for whom this adds a
complication to
> > their system setup.
>
> So you volunteer to do it, right ?

Anybody know where I can get information on the extended BIOS calls?
While LILO may not be easily modifiable to boot after the 1023rd cylinder in
all cases, I suspect that most cases in which the problem may appear can be
covered by a LILO that support the extended BIOS call. Since this problem
really only affects > 8GB drives (as LBA no longer can adjust the number of
cylinders below 1024), which themselves are only supported by the latest
BIOSes who probably have resolved any bugs or problems with their extended
BIOS calls.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/