Some notes on RootOnZFS article in wiki
jhb at freebsd.org
Wed Dec 23 14:38:35 UTC 2009
On Tuesday 22 December 2009 9:11:43 am Dag-Erling Smørgrav wrote:
> "James R. Van Artsdalen" <james-freebsd-current at jrv.org> writes:
> > It definitely breaks things *when booting* to depend in any way on a
> > partition table since there may not be one. By the mid 90's nearly
> > every OS was putting in at least dummy partition tables for the same
> > reason GPT does - to lessen the risk of accidental clobbering of the
> > disk - but that's just a convention. I'm sure there are still a few
> > customized VAR-things out there that don't bother with a partition table.
> I can assure you that Windows does not put in a dummy partition table,
> and will not boot if the partition is not active.
That is due to Windows' MBR bootcode, not the BIOS.
> I can also assure you that the BIOS on my current laptop (ThinkPad T60)
> *does* care about the partition table, because the BIOS boot menu has an
> option to launch the rescue & recovery utility, which is located on a
> DOS partition at the end of the disk (although the BIOS works fine if
> the R&R partition is missing)
> > A number of vendors have taken to putting "hidden" system partitions on
> > the disk with various utilities that can be run via a hotkey press
> > during POST. These schemes have to use MBR-like code from the BIOS ROM
> > to boot their system partition and that pseudo-MBR must read and
> > interpret the partition table to find the system partition. But during
> > system boot itself the MBR sector is read and if the last word in that
> > sector is 0xAA55 then the BIOS executes the MBR code blind as to what is
> > on the disk. It's the MBR code that's read from the disk that scans the
> > partition table, if there is one.
> I can't quite parse that.
> The R&R partition on my T60 is not hidden in any way.
Umm, you just said you have a recovery partition on your T60 which is what he
is talking about here! Geez. Presumably by "hidden" he means that it doesn't
show up as a volume under Windows or OS X, so for the vast majority of folks
it is hidden.
> > There were attempts for a time to check for boot sector virii before
> > booting but those were always so problematic that I never did that, and
> > I don't the the other main BIOS teams did it either.
> I've had machines that had a BIOS option to check if the boot sector had
> been modified and warn the user before booting. It worked just fine.
That is not what he is talking about. He is talking about trying to analyze
the boot sector for known signatures, etc., not the simple write-protection
Now, there are some examples I know of that involve BIOS's having too-intimate
knowledge of disk partitioning schemes. One is that the floppy emulation mode
in the "El Torito" CD booting code of certain IBM BIOSes assume that the
floppy disk is actually a DOS floppy disk complete with a BPB that the IBM
BIOS writes to after it has been loaded. We have a dummy BPB in boot1.S as a
workaround for that brain damage. Another instance is that some BIOS's
attempted to help with the C/H/S geometry disaster by examining the C/H/S and
LBA addresses in the MBR to determine what C/H/S geometry the MBR was
expecting. They would then compute that C/H/S geometry and honor that for
C/H/S to LBA conversions. It generally worked well. However, if you used a
heads value of 256 in a C/H/S value in the MBR it would end up causing a
divide by zero that caused the boot to hang. The original fake MBR in boot1.S
for "dangerously dedicated" mode had a partition that tickled this particular
bug and was changed to use 255 heads. There are more details on these in the
commit logs of boot1.S.
Personally, most of my headaches dealing with BIOS's these days are dealing
with SMM rather than disks. As an OS developer I loathe SMM. :)
More information about the freebsd-current