Ho hum..

About a year ago I started moving drivers used by the sparc port
towards the bus_space(9) and bus_dma(9) model. This was done in
a number of steps. It was clear that several drivers would be needing
actual testing after the surgery applied to them. I'm relying
on "-current" users who occasionally build new kernels and see if
those still function with their hardware.

Obviously, this has not been the case for several drivers (mostly for
aging devices on sun4 systems), as evidenced by bugs that I found
in the last few days in the `si' and `ie' drivers which have made
those devices non-functional in -current for nearly a year.

This raises the question whether we can sanely continue to claim to
have support for this hardware in the upcoming 1.4 release. Unless
it can be demonstrated that a device on the `supported list' can
be succesfully driven by a -current/1.4-branch kernel, the answer
is going to be "no", I guess...

-pk