On Mon, 25 Aug 2014, Oliver Neukum wrote:
> On Mon, 2014-08-25 at 10:58 +0000, Alfredo Dal Ava Junior wrote:
>
> > - 1TB and 2TB: READ_CAPACITY_10 returns correct size value
> > - 3TB and 4TB: READ_CAPACITY_10 returns size in a 2TB modulus
> >
> > If we fix capacity size by reporting (READ_CAPACITY_10 + MODULO_2TB), the result
> > will be invalid when user plugs a <2TB HDD. An idea (bring by Oliver) is:
>
> It is worse than that. Pretty soon a 4.7TB disk will also be plausible.
>
> > first guess reading last sector using modulus result to check if size is valid.
>
> It is necessary that a virgin disk also be handled correctly.
> We cannot use the partition table (besides it being a layering
> violation)
>
> It would propose (in pseudo code)
>
> if (read_capacity_16(device) < 0) {
> lower_limit = read_capacity_10(device);
> for (i = 1; i++; i < SANITY_LIMIT) {
> err = read_sector(device, lower_limit + i * 2TB-SAFETY);
> if (err == OUT_OF_RANGE)
> break;
> }
> if (i < SANITY_LIMIT)
> return (i - 1) * 2TB + lower_limit;
> else
> return ERROR;
Don't forget that lots of disks go crazy if you try to read from a
nonexistent block, that is, one beyond the end of the disk.
IMO, this bug cannot be worked around in any reasonable manner. The
device simply cannot handle disks larger than 2 TB.
Alan Stern
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
