[RFC] Remove requirement of alignment to track from MBR scheme

[Andrey V. Elsukov]

On 23.05.2011 20:14, Warner Losh wrote:
> Since we don't have a good review tool for the project, this may seem a
> little fragmented, but here goes:
> 
> +	if (pp->sectorsize > 4096)
> +		return (EOPNOTSUPP);
> 
> What's the motivation for this?  This seems unrelated to the problem at
> hand.  It may be good, but 4096-byte sector size is very new and I'm not
> sure MS-DOS can cope...  :)  There has been a long tradition in the MBR,
> however, of supporting different sector sizes to allow old-old-old
> school floppies that had 128 or 256 byte sectors to work, as well as
> allowing the supper-compressed 1.77MB floppies to work with their
> 1024-byte (or was that 2048-byte?) sectors.  Not sure this really
> matters for reading, and for writing this case is so far beyond the edge
> I'm not even sure why I bring it up[*]

We have similar check in the g_part_mbr_probe function, and since we do
not support sectorsize > 4096 in the probe method, i do not think we
should try to create MBR for these providers.

> In any event, I'd be tempted to use a #define for 4096 like
> MBR_MAX_SECTOR_SIZE.
> 
> -	msize = MIN(pp->mediasize / pp->sectorsize, UINT32_MAX);
> +	msize = MIN(pp->mediasize / pp->sectorsize, 2 * UINT32_MAX);
> 
> Why this change?  I think that it is in two places.

Currently we have limit to msize = UINT32_MAX, but partition in MBR has
start offset and size (not end offset). Theoretically it can have size
that is up to UINT32_MAX sectors, also start offset can be UINT32_MAX.
And for example, for 4T disk we can have 2 partitions with 2TB size.

-- 
WBR, Andrey V. Elsukov

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 488 bytes
Desc: OpenPGP digital signature
Url : http://lists.freebsd.org/pipermail/freebsd-geom/attachments/20110523/dc9972d3/signature.pgp