40
/*
* Interleave description table.
* Computed at boot time to speed irregular-interleave lookups.
* The idea is that we interleave in "groups". First we interleave
* evenly over all component disks up to the size of the smallest
* component (the first group), then we interleave evenly over all
* remaining disks up to the size of the next-smallest (second group),
* and so on.
*
* Each table entry describes the interleave characteristics of one
* of these groups. For example if a concatenated disk consisted of
* three components of 5, 3, and 7 DEV_BSIZE blocks interleaved at
* DEV_BSIZE (1), the table would have three entries:
*
* ndisk startblk startoff dev
* 3 0 0 0, 1, 2
* 2 9 3 0, 2
* 1 13 5 2
* 0 - - -
*
* which says that the first nine blocks (0-8) are interleaved over
* 3 disks (0, 1, 2) starting at block offset 0 on any component disk,
* the next 4 blocks (9-12) are interleaved over 2 disks (0, 2) starting
* at component block 3, and the remaining blocks (13-14) are on disk
* 2 starting at offset 5.
*/