[RFC PATCH 0/5] CELL bogus_console port to hvc_console backend driver

[Ryan S. Arnold]

The following patch-set was created against the 2.6.14-rc5 CBE (cell
broadband environment) patch-set provided by Arnd Bergman.

The purpose of this patch-set is to port the CELL IBM Full System Simulator
bogus_console.c driver to an hvc_console back-end driver, namely hvc_fss.c.

Our intention is to support binary-compatibility of all hvc_console
back-end drivers such that all drivers can be built into the kernel at
configuration time but only one back-end (the one that detects that it is
running on the right hardware) actually registers the front-end
driver at console init.

This is a request-for-comments.  Please contribute any suggestions
especially in-regards-to the Makefile and Kconfig changes.  I'm not very
experienced with them.

I do realize that the current mainline kernel has some significant
differences in the hvc_console driver versus the 2.6.14-rc5 kernel.  We'll
address this when the cell patches go upstream.

I've tested these patches on the CELL IBM Full System Simulator and
the console works fine.  I've not had a chance to test the hvc_vio
back-end on ppc64 hardware but I'll do this once I have a chance.

I suspect that there will be some udev magic required to get the /dev/hvc0
device to appear on the CELL simulator since there isn't an actual serial
device.

Using this driver one can actually execute agetty on the console as well.

Thanks,
Ryan S. Arnold
IBM Linux Technology Center