re: partitionSizeHi in raidframe component label, take 2 (wasRe: partitionSizeHi in raidframe component label)

[matthew green <mrg%eterna.com.au@localhost>]

> - Userland fix is added.
> - rf_reasonable_label() is also changed to use accessor.  Initially I
>   thought it is a special case since the comment in
>   rf_fix_old_label_size implies that the numBlocksHi may be garbarge
>   when the function is called, but even if it is garbage the result of
>   comparison is same.

looks good, and the above reasoning seems valid to me.

thanks,


.mrg.