[kbd] unicode_start and hvc consoles

[Josh Boyer]

Hi,

While trying to run unicode_start on a POWER machine, we ran into the
following error:

	[root at elm17d211 ~]# tty
	/dev/hvc0

	[root at elm17d211 ~]# unicode_start
	unicode_start skipped on /dev/hvc0
	[root at elm17d211 ~]#

It seems the unicode_start script checks for a valid tty device by
looking for /dev/console, /dev/vc*, or /dev/tty*.  On POWER machines,
the console device is often provided by /dev/hvc0.  I believe that can
also be true on Xen guests and in some cases ARM installations today as
well.

Do you know of a valid reason the /dev/hvc0 would be excluded, or is
this simple a case of someone not reporting the issue as of yet.  I
would appreciate any thoughts you have.  Please CC me on reply.

josh