Re: Problems with IDE on linux 2.6.22.X

[Rene Herman <rene.herman@gmail.com>]

On 08/30/2007 11:16 PM, Greg Freemyer wrote:

> On 8/30/07, Rene Herman <rene.herman@gmail.com> wrote:

>> Well -- the world where ATA, SCSI, USB, Firewire and what have you are
>> low-level drivers to a unifying storage layer is under non too obscure
>> definitions sort of not non-wonderful...
>>
> 
> USB / Firewire / FC / iSCSI are all SCSI transports and fit within the
> SCSI subsystem by design.
> 
> ie. Just like ethernet, DSL, T-1, etc can all carry IP traffic with no
> conceptual conflict, many media by design carry SCSI traffic.
> 
> The PATA and SATA physical layer typically carry ATA commands and
> having them tied into the SCSI stack is an aberration that I hope will
> be eliminated some day.
> 
> ATAPI is an exception.  Not sure where that would end up in a perfect world.

As said, if you make a bit of an effort to view the former SCSI stack as a 
unified storage midlayer the abberation becomes less abberational (if that's 
a word).

Real SCSI, the other SCSI transports and ATAPI would just use more of the 
common mid-layer than P/SATA would. I'd expect the way forward would be to 
just refactor things until someone notices that drivers/scsi is the wrong 
place for sd.c and sr.c and moves them to drivers/block or whereever.

Practically, the PATA driver gives me (almost) the same throughput as the 
old IDE driver does, and given that I need the former SCSI stack _anyway_ 
for my external USB harddrive, I don't see a pressing need to carry along 
yet another storage stack for my harddrive.

Rene.