MP3 player, IDE interface

Chris Caudle chris at
Wed May 8 21:53:52 UTC 2002

On Tuesday 07 May 2002 03:44 pm, Peter Mueller wrote:
> an IDE Disk to store music files.

What will be driving the IDE interface (e.g. custom interface, integrated 
circuit from CMD, etc.).
In the PC world, there is an historical standard for the control registers of 
an ATA interface, which was formalized by Intel for the PCI to IDE 
controllers.  If not using that interface, it may be beneficial to make your 
interface conform as closely to that standard as possible just to make it 
easier to port existing code.

I am not familiar enough with the ATA programming model yet to know whether it 
would be beneficial to split such an interface into multiple layers, similar 
to the way SCSI devices are handled on NT or Linux, with a SCSI layer which 
builds the SCSI command packets, and a hardware interface layer which handles 
the details of interfacing with the hardware to get the command packet driven 
onto the bus.  If anyone else has an opinion one way or the other I would be 
interested to hear it.

I downloaded an ATA driver placed in the public domain  by Hal Landis, but 
haven't had a chanced to study what modifications will be needed to make it 
usable under RTEMS.  It was written as a DOS driver, so there are probably a 
lot of memory model assumptions which don't hold in RTEMS.
You can find that at:

Lots of good ATA info in general at that site.

-- Chris Caudle

More information about the users mailing list