atapicam cd error

[Scott Long]

Joe Altman wrote:
> On Fri, May 25, 2007 at 06:30:00PM +0200, Oliver Fromme wrote:
>> Joe Altman wrote:
>>  > You have atapicam as a kernel module; and what version of atapi-cam.c?
>>
>> No, I've always compiled it statically in my kernel.
> 
> That's what I meant to say; that it was a device in the kernel:
> 
> device atapicam
> 
>>  > It looks like this one is current: Revision 1.54 
>>
>> That's the HEAD version (in 7-current).
> 
> Yes, I realized it was something like that after I reviewed the
> handbook.
> 
>> By the way, I am _not_ using k3b or any other krap.  ;-)
> 
> k3b sounds like something under KDE; but ISTM that neither Grip nor
> XMMS are crap. Also in my experience, if atapicam is a kernel device,
> the Plextor I have is completely useless. Without atapicam as a
> device, the Plextor works for a variety of tasks.
> 
> I wonder if atapi-cam.c interacts, in my setup and in some way, with
> some other kernel device and so makes it appear that atapicam is the
> source of the problem?
> 

atapi-cam is a hack on top of the ata driver.  I don't say that out of
disrespect for the authors of it, I think that they've done a very fine
job with a very difficult task.  What I mean is that the ata driver
doesn't provide the facilities that atapicam needs in order to function
reliably.  So there is a bit of a tug-of-war between the ata core and
acd drivers, and the atapicam driver; both think that they know best
what's going on and what needs to happen, and both wind up accidentally
stomping on each other as a result.  In addition, the acd driver does
know a little more about quirky hardware and is able to cope better with
the special needs of those devices.

So given all of that, it's great that atapicam works as well as it does.
I have a project in the works to provide better integration between
these two opposing sides (and please don't worry about offering to be an
eager early tester right now), so we'll see how that works out over the
next few months.

Scott