> 
> 
> >From Marc Wandschneider:
> >| From: Alistair G. Crooks  <netmail!agc@uts.amdahl.com>
> >|
> >|[...]
> >| I like the new messages from the wd driver.
> >
> >        on a related note:  is any work being done for this driver in 
> either
> >        the current kernel or the magnum branch?  it's bad and getting 
> worse,
> >        almost to the point where i'm recommending that people with ide
> >        drives should check out linux or freebsd instead for stability's
> >        sake.
> 
> I am using the wd driver as it exists in -current, and have not had nearly 
> as many of the reported problem that I have seen lately.   I HAVE noticed 
> that when the controller hicks up, I will get "soft fsbn-type" errors when 
> the controller hicks up, and then the system will proceed merrily along from 
> there, unless it is during the fsck.  If it reports errors during the 
> reboot, I will get unfixable errors.  It is no big deal, and has only 
> happened once.
> 
[...]

Just for another data point (no intent to cause a war)

My system has 2 IDE drives,  a Conner 340M and a Quantum 240M.
Also, a fast VESA-bus IDE controller, 486-DX 33, 16MB ram.

with NetBSD-current (as of Jan 9), I could not access the 2nd drive at all
(I tried both as the master - no dice either way).  As soon as I try
to do something simple like 'disklabel -r' or 'fdisk' of wd1,
I get 3 timeout messages from the wd driver, and then a complete hang.

I just loaded FreeBSD-current (as of Jan 17) and it works like a champ.
Both drives are accessible, didn't even have to re-disklabel the 2nd
drive - just mount it and works great.  
The latest FreeBSD has a much improved wd driver,  and some of the 
timeouts have been increased, as well as some delays added to
"conform to the ATA spec".  

I recommend that someone look at porting this new driver to NetBSD.
'Works for me!'

Jim


------------------------------------------------------------------------------