On 7/31/08, Jan Henrik Sylvester <[EMAIL PROTECTED]> wrote:
> I guess I should have put it more concise as some important details were
> lost in the middle of the background story.
> --- The important part: --->
> If I boot the FreeBSD 7.0-RELEASE livefs CD, I have my 232GB hd limited
> to 128GB. (According to dmesg and some dd writing tests.)
> If I boot the dynamic drive overlay MBR that changes the INT13 bios
> routines and let it boot the livefs CD, FreeBSD can access all 232GB.
> These two cases happen with the same MBR and the same content on the drive.
> <--- (Important part ending) ---
>  From my limited understanding: This cannot be about the FreeBSD
> installation transfered by firewire, since it is long gone. This cannot
> be about the FreeBSD MBR, since it is not involved. This is not about
> the partition table, since it is the same.

It CAN be about the partition table, at least in principle, although
I'd think that if the dmesg says the drive is 128 GB that's probably
not what is happening. An invalid partition table can be interpreted
differently by different systems (I know that from painful
experience), and of course, by different releases of FreeBSD. And
Vista in particular seems to like to do strange things to partition
tables, FWIW.

> Doing the same with a Linux live CD (Knoppix), I can access the whole
> drive in both cases.
> Windows reports 232GB, too, but according to this (German) posting
> http://www.acer-userforum.de/thread.php?postid=40207 writing above 128GB
> will wrap around for the 855GME chipset driver (I have 855PM). The
> posting claims ddo would solve it, which I cannot understand, if the
> Windows driver ignores the bios information. (The posting is about Acer
> Travelmate 661. I got the 800 from the same time.)
> My laptop is more than five years old and I have got the latest bios for
> years, apparently without 48bit LBA.
> I am kind of lost. I cannot understand the disassembled MBR and even if
> I could, I do not think I would want to create my own boot manager /
> INT13 handler. I guess reading FreeBSD source code would be next... but
> I am not very confident there, either.
> The alternative is that my understanding of the problem is totally wrong.
> Thanks for more helping to think through this mess of information I got
> during the last day.

FreeBSD has to make some educated guesses about what disk geometry to
assume: it has to try to make assumptions that will be compatible with
what other operating systems are likely to do, so at times it is
possible that it will opt for something safe but not "right" in order
to balance conflicting information it gets from the BIOS, the
partition table, and the drive itself. That's why you have the option
of changing its assumptions when you do a sysinstall. What the livefs
cd does, I do not know, but the only way you will know what would
happen if you installed FreeBSD on, and boot from, that hard drive
will be to actually install FreeBSD on that hard drive and see what
happens. The one thing I am certain of is that it should have no
problem booting from and using a 232 GB drive.

- Bob
freebsd-questions@freebsd.org mailing list
To unsubscribe, send any mail to "[EMAIL PROTECTED]"

Reply via email to