> The kernel prints some message which I cannot repeat as-is because the
> kernel doesn't seem to use XON/XOFF handshaking so that the output on the
> serial console gets corrupted. However, it's the error message from
> sys/dev/wdc/ata/ata_wdc.c in wdc_ata_bio_intr():
> 	"device timeout, c_bcount=%s, c_skip%d".
> 
> This repeats forever and kernel can't boot any further. I've tried to
> revert a few of the changes WRT ata*, wdc* and pciiide* but didn't have
> any success. Either the problem persisted or the harddisk wasn't detected
> at all.
> 
> 
> This is from the kernel and configuration which still works:
> 
> NetBSD shockwave 2.0E NetBSD 2.0E (SHOCKWAVE) #2: Fri Apr 30 01:29:53 CEST 2004  bin@cyclonus:/usr/obj/sys/arch/i386/compile/SHOCKWAVE i386
> 
> [...]
> pciide0 at pci0 dev 10 function 0
> pciide0: vendor 0x10ad product 0x0001 (rev. 0x02)
> pciide0: hardware does not support DMA
> pciide0: primary channel wired to compatibility mode
> pciide0: primary channel interrupting at irq 14
> atabus0 at pciide0 channel 0
> pciide0: primary channel ignored (other hardware responding at addresses)
> pciide0: secondary channel wired to compatibility mode
> pciide0: secondary channel interrupting at irq 15
> atabus1 at pciide0 channel 1
> pciide0: secondary channel ignored (not responding; disabled or no drives?)

hi,
can you try a kernel without the pciide devices, only wdc at isa ?
Also I'd be interested in seeing the pciide-related lines from the newer
kernel. You should be able to enter ddb and use the dmesg command.
If this doesn't work, you can use boot -d, set a breakpoint on atabusconfig():
break atabusconfig
continue

Then it should break to ddb before probing the IDE bus and you can use
the dmesg command.

-- 
Manuel Bouyer <bouyer@antioche.eu.org>
     NetBSD: 26 ans d'experience feront toujours la difference
--