Re: [PATCH 2/3] WIN_* -> ATA_CMD_* conversion: update WIN_* users to use ATA_CMD_*

[Chris Wedgwood]

On Fri, Nov 05, 2004 at 10:45:04PM -0500, Ross Biro wrote:

> Just a reminder, this error recovery doesn't work on many modern
> hard drives, and is a violation of all ATA specs after ATA-2*.

I think this code is the only user of that token and I wonder how well
tested this is?  Bart?  Andre?

> In particular, most Maxtor and Western Digital Drives will not
> recover from errors with this command sequence.  The preferred error
> recovery is to do a reset followed by a set features, because that
> is what Windows does (as told to me by a drive vendor).

I assume for Windows they do that for all drives?  Even older ones?

I also wonder what happens with TCQ when you get an error?  Do you
just retry everything outstanding?

> I've tested the reset/set features method of error recovery and it
> works on all the drives I've tried.  I have not tried it on any
> older drives, or any other types of ATAPI devices.

It probably should get fixed, there just doesn't seem to be much
incentive to beat on the old IDE code though :( Minor cleanups seem
worthwhile but anything intrusive I worry will break some hard-to-test
platform for someone.

That said, if libata gets PATA support merged then know this is
wonderful there.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/