From owner-freebsd-hackers Wed Mar 22 22:34:06 1995 Return-Path: hackers-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id WAA01046 for hackers-outgoing; Wed, 22 Mar 1995 22:34:06 -0800 Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.34]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id WAA01035 for ; Wed, 22 Mar 1995 22:33:53 -0800 Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.9/8.6.9) id QAA04092; Thu, 23 Mar 1995 16:28:32 +1000 Date: Thu, 23 Mar 1995 16:28:32 +1000 From: Bruce Evans Message-Id: <199503230628.QAA04092@godzilla.zeta.org.au> To: PVinci@ix.netcom.com, hackers@FreeBSD.org Subject: Re: ata/ide sector translation modes c/h/s vs. LBA? Sender: hackers-owner@FreeBSD.org Precedence: bulk >Why does the ATA standard make such a big stink about LBA mode?? It doesn't. ATA-2 has only 42 lines of normal standardese about it. >As I see it, either way, you have to translate either mode into: > Sector Number > Cylinder Low > Cylinder High > Dev/Head No. In LBA mode, you only have to "translate" a 24 bit LBA into 3 bytes by writing it a byte at a time to the controller registers. The drive may or may not have to do more work to convert an LBA to an internal form than to convert from the currently specified geometry to an internal form. >So, how does LBA mode improve performance -- isn't it just a different >set of calculations that need be done before *Each* command?? Am I >missing something? It solves a whole microsecond or two, depending on how fast the cpu can do one div-mod, one div and one mod. Bruce