On Tue, Sep 12, 2006 at 06:22:19AM -0700, Thomas Carlson wrote:
> First of all, thanks for all your help, Allen.  You are knowledgeable, 
> yet patient with those of us who aren't, a real credit to the list.

Thanks.

> The manual entry for the se driver indicates that:
> 
> "In addition, to facilitate SCSI commands issued by userland programs, 
> there are open(), close(), and ioctl() entry points.  This allows a 
> user program to, for example, display the EA41x statistic and download 
> new code into the adaptor - functions which can't be performed through 
> the ifconfig(8) interface."

Yes, it does.  But it doesn't do anything special with the calls.
They're really just passed to the scsipi layer, and I don't know
of any tools that use them at the moment.  I read that a little
differently from how you read it (and I could be wrong).  I think
the intent is that you could write an application (adding code in
the driver, too) to download statistics from the device during
operation (# of collisions, overruns, underruns, whatever) or to
update the firmware in the device.  ifconfig(8) can't do those
things.  The driver clearly has code in it to set the address,
etc., in response to the calls that ifconfig(8) make.

So the question I have is whether or not the device needs new code
loaded into it after a bus or device reset.

I think the next steps are to:

	1) Try booting both with and without loading the MacOS driver
	   for the Cabletron.  See if the behavior is different.

	2) Start building kernels to try to see better what's happening
	   (or not happening) in the driver.

-allen

-- 
Allen Briggs  |  http://www.ninthwonder.com/~briggs/  |  briggs@ninthwonder.com