Re: Big files in ext2fs (but not i_osync)

Chip Salzenberg (
Sat, 28 Feb 1998 12:55:54 -0500

Color me inattentive -- I completely missed the fact that i_osync is
*already* use only in memory and is not written out. Please forget I
ever mentioned it. But please *do* remember the correct point I made,
that we only need 16 more bits right now -- well, really, only 12.

Since Remy explains that i_{file,dir}_acl are both unavailable, it
seems that we can do one of these:

(1) Decide that fragments are the wave of the past and re-use the
fragment fields for other purposes, including more size bits;

(2) Steal one of the block pointers for big files.

Option (2) would increase the use of double- and triple-indirects for
exactly the files that can least afford them, but it would be 100%
backward-compatible and would not impact the use of fragments if we
ever decide they're a Good Thing.

For that matter, we could decide that Really Big Files should not have
any direct pointers, only indirects. That would actually be an
optimization that would more than make up for stealing one direct
pointer for extra size bits.

Chip Salzenberg                - a.k.a. -               <>
"I brought the atom bomb.  I think it's a good time to use it."  //MST3K

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to