Re: RFC: from FIBMAP to FIONDEV

Jeff Garzik (jgarzik@pobox.com)
Tue, 15 Jun 1999 13:59:20 -0400 (EDT)

On Tue, 15 Jun 1999, Werner Almesberger wrote:
> > With RAID 1, lilo can install multiple copies of LILO, one on each disc,
> > each one told about the blocks which exist on its own disc.
>
> Well, technically it could, because it happens to know where the data
> goes. However, this would involve allocating space in the RAID, via
> the file system, and then filling it via direct writes to the underlying
> devices. Pretty horrible, if you ask me.
>
> No, as far as LILO is concerned, I consider this a case of policy at a
> higher level. LILO just writes a single boot loader in a well-defined,
> clean way. If there's replication in the RAID, you install multiple
> boot loaders, and you tell LILO explicitly where they go.

When this was discussed a while ago (maybe on another list?), I
suggested a reserved space at the beginning of each RAID partition.
That seemed to encompass the least amount of work on the part of
everybody.

When installing a kernel, a special ioctl (before any writes occur)
would tell the RAID software that the file being written should be
stored in the special reserved area. When the file is actually written,
whether it be RAID0/1/5/whatever, the file output is multiplexed to a
reserved space at the beginning of each partition in the md set.

At that point, LILO need only be extended to output to multiple MBRs/boot
sectors, with each MBR pointing to the md reserved space on that
particular drive. It might need to pass a magic cookie or two to the
kernel to tell md it is booting off a secondary boot sector.

Jeff

-
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/