ATA Passthru unimplemented protocols

ATA Passthru unimplemented protocols

[Timothy Thelin]

Are there plans to eventually do them or is anyone actively working on them?

Thanks,
Tim Thelin

Re: ATA Passthru unimplemented protocols

[Jeff Garzik]

Timothy Thelin wrote:
> Are there plans to eventually do them or is anyone actively working on them?

Such as?

	Jeff

RE: ATA Passthru unimplemented protocols

[Timothy Thelin]

I'm interested in soft reset (1) and return response information (15).  But
I didn't know who was doing what or had plans for what, which is why the
question was so general.  If I end up working on these two myself, I'd be
nice to  coordinate with existing efforts rather than collide with them =)

 

-----Original Message-----
From: Jeff Garzik [mailto:jgarzik@pobox.com]
Sent: Wednesday, July 27, 2005 1:31 PM
To: Timothy Thelin
Cc: linux-ide@vger.kernel.org
Subject: Re: ATA Passthru unimplemented protocols


Timothy Thelin wrote:
> Are there plans to eventually do them or is anyone actively working on
them?

Such as?

	Jeff

Re: ATA Passthru unimplemented protocols

[Jeff Garzik]

Timothy Thelin wrote:
> I'm interested in soft reset (1) and return response information (15).  But
> I didn't know who was doing what or had plans for what, which is why the
> question was so general.  If I end up working on these two myself, I'd be
> nice to  coordinate with existing efforts rather than collide with them =)

Certainly!  Your question was just a bit vague :)

The soft reset would be especially nice to have, but would require a bit 
of work to synchronize with other devices on the bus (needed for PATA 
only, really).

Another line item that I would like to see happen is "flagged 
taskfiles", which allows the user to specify which shadow registers to 
touch and -- more importantly -- which not to touch.  This is important 
for vendor-specific commands (particularly older PATA devices).

	Jeff

RE: ATA Passthru unimplemented protocols

[Timothy Thelin]

Well it doesn't seem that nasty to adopt flagged access.  From a userland
API point of view how about adopting the IDE driver's taskfile mechanism,
and simply translating that into the yet-to-be internal flagged mechanism?

Tim



-----Original Message-----
From: Jeff Garzik [mailto:jgarzik@pobox.com]
Sent: Wednesday, July 27, 2005 2:09 PM
To: Timothy Thelin
Cc: linux-ide@vger.kernel.org
Subject: Re: ATA Passthru unimplemented protocols


Timothy Thelin wrote:
> I'm interested in soft reset (1) and return response information (15).
But
> I didn't know who was doing what or had plans for what, which is why the
> question was so general.  If I end up working on these two myself, I'd be
> nice to  coordinate with existing efforts rather than collide with them =)

Certainly!  Your question was just a bit vague :)

The soft reset would be especially nice to have, but would require a bit 
of work to synchronize with other devices on the bus (needed for PATA 
only, really).

Another line item that I would like to see happen is "flagged 
taskfiles", which allows the user to specify which shadow registers to 
touch and -- more importantly -- which not to touch.  This is important 
for vendor-specific commands (particularly older PATA devices).

	Jeff

Re: ATA Passthru unimplemented protocols

[Jeff Garzik]

Timothy Thelin wrote:
> Well it doesn't seem that nasty to adopt flagged access.  From a userland
> API point of view how about adopting the IDE driver's taskfile mechanism,
> and simply translating that into the yet-to-be internal flagged mechanism?

There is actually already an internal flagged mechanism, but we only 
have two flags at present: ATA_TFLAG_ISADDR and ATA_TFLAG_DEVICE.

	Jeff