From the message:

/* TODO: find and fix the timing bugs apparent on some controllers */

found in wd.c (1.30 1993/12/24), I assumed someone might be interested in this. 

I have a Dell 450DE with an IDE drive. I installed from the NetBSD-current 
tar files that cgd announced shortly after Christmas (no problems doing that
by the way) and went about building a new kernel. The new wd.c causes my
system to hang at the point shown below. I should mention that this has
happened before, but I was a bad citizen and just replaced wd.c with the 
one that came with 0.9 without reporting the error. This time, I at least 
rebuilt a kernel with WDCNDELAY_DEBUG which provided the information below:

[...]
lpa0 at 0x378-0x37f on isa
wdc0: timeout took 850us
wdc0: timeout took 1612275us
wdc0 at 0x1f0-0x1f7 irq 14 on isa
wdc0: timeout took 16100us
wd0 at wdc0 targ 0: 234MB 723 cyl, 13 head, 51 sec <QUANTUM LP240A GM240A01X>
fdc0 at 
[...]
biomask 4060 ttymask 21a netmask 200
wdc0: timeout took 16325us
wdc0: timeout took 394350us
wdc0: timeout took 575us
wdc0: timeout in wdintr WDCS_BUSY
wdc0: timeout in wdintr WDCS_DRQ
wdc0: timeout in wdintr WDCS_DRQ
[HANG]

===
Jason Brazile 					brazile@cs.utexas.edu
Graduate Student				Dept of Computer Science 
"People say I'm apathetic but I don't care" 	University of Texas

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