> I finally found that the code in sl82c105c was used with this chip and
> that part of this code is under compile option "CONFIG_ARCH_NETWINDER".
> This WEIRD, since this option only means that netwinder support is
> enabled, not that we are running on a netwinder. If I add support for
> both our board _and_ netwinder in the kernel compile options, then
> magically, my IDE began to work.
The problem is that sl82c105 has never been properly ported to Linux,
until I added the support. Unfortunately, the original file was for
another architecture, and so I wasn't going to put the support in
for every architecture.
> I've not yet figured out what this added IDE code does (and why it's only
> defined for the netwinder), but I'm still having a problem with the
> second IDE controller of the winbond, which is still giving me
> lost_interrupts.
The second channel has never been tested, so it's quite possible that there's
some bugs there. It may even be that channel 1 is updated with the channel
2 info, so I think this is the place to look.
unsubscribe: body of `unsubscribe linux-arm' to [EMAIL PROTECTED]
