>> In particular, I mostly know how to use the blitter [on a cg6].

> Yeaaaa!

It's really not that bad; read the sunGX.[ch] files in X11R6.4's
programs/Xserver/hw/sun directory (if I've got that path right - it's
from memory).

> Can you do enough with the colormapping that we can do reliable ansi
> color sequences on the console, eventually, do ya think?

I believe I know enough about the harwdare.  I am not convinced I know
enough about hooking it into the rcons/wscons code - it took chasing
through at least half a dozen files, structures, and
pointers-to-functions to figure out what I had to do to drop in a
cg6-specific copyrows routine.  I can probably make it work, but am not
sure how fragile the result will end up being.

Incidentally, the kernel colormap code is way, way over-complicated for
the cg6, probably due to attempts to share it with other framebuffers
using Brooktree RAMDACs.  (Most briefly, on the cg6 there is no need to
load multiples of four colormap entries; you can set bt->addr to
anything from 0 to 255 and then do three stores to bt->cmap, and
thereby load just one RGB triple in the colormap, ignoring the
four-byte alignment goop the bt_subr.c code is full of.)

>> This has finally allowed me to get rid of the execrably slow
>> scrolling when using a cg6 under RASTERCONSOLE.  (Indeed, on at
>> least one of my machines, the console scrolling is equally slow -
>> perhaps due to an old ROM rev.)
> What rev?

Old enough that it boots from sd(0,0,0) instead of /sbus/esp/sd@0. :)
Fuzzy memory says 1.3, but I'm not sure that wasn't another machine.

> I have the fastcons eeprom code if you want it.

I've got some fast console code too; I haven't bothered installing it
on this machine.  After all, the console isn't used for much, only for
booting before it gets to discovering the cg6 console, at which point
the (now highly usable) cg6 scrolling code kicks in.

>> Would there be any interest in getting this code into the tree?
> As the proud owner of an IPX (soon to be the proud owner of a FREE
> ss5 with a TGX (cg6 rev 11)),

*drool*

> of course I'd like it in there.

Thanks for the support.  However (as recent events exemplify) just
because you and I would like something doesn't mean it'll go in. :-/
While this is surely less controversial than the "recent events" code,
I *have* been surprised before....

Want a copy separately and immediately? :-)  Four files need patching,
all in sys/arch/sparc/dev: cgsix.c, cvsixvar.h, and cgsix_{obio,sbus}.c
(haven't been able to test the obio variant).

					der Mouse

			       mouse@rodents.montreal.qc.ca
		     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B