ioctl's and USB drives (and just plain ATA)

[Leonard den Ottolander]

Hi Jeff,

On Tue, 2005-08-09 at 20:38, Jeff Layton wrote:
> This may or may not work, depending on your hardware. You have a
> controller that does USB storage -> IDE translation. There's no
> guarantee that this translator will pass the appropriate ioctl() stuff
> to the actual IDE controller/drive.
> 
> Then again, it might...  :-)

Back to square one :-s . Anyway, if the SCSI emulation layer is dropping
these commands it doesn't really matter if the USB device gets the
translation right. Is there a way to work around the SCSI emulation
layer so these commands at least get passed to the device?

Related question: If I want to use TASKFILE ioctls with a 2.4 kernel do
I need to configure the kernel with CONFIG_IDE_TASK_IOCTL or
CONFIG_IDE_TASKFILE_IO? I saw some comment that suggest the latter, but
I somehow get the impression that is about a different IDE interface and
not about just sending raw TASKFILE ioctls which I need to
implement/improve the ATA security features in hdparm.

Leonard.

-- 
mount -t life -o ro /dev/dna /genetic/research