On Mon, Feb 18, 2002 at 11:43:38PM +0100, Daniel Phillips wrote:
> We want to stay with the shift counts. They should be the primary currency
> of size measurement. You can add shift counts together and get nice, compact
> code, whereas with absolute size you often have to ugly things - e.g., it's a
> pain to divide by blocksize when you have it as an absolute number, it's easy
> when you have it as a shift.
>
> If you are going to the trouble of fixing this, please don't use absolute
> size as the primary measure, use a shift count.
Most of this is targetted at userland which needs byte counts (size in
sectors was a bug introduced after the original BLKGETSIZE64 went in).
Using the number of sectors in kernel is perhaps more efficient, but it
is a microoptimization that won't show up on any benchmarks.
-ben
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Sat Feb 23 2002 - 21:00:29 EST