Re: machfb (sort of) working on Mach64 GX

[David Riley <fraveydank%gmail.com@localhost>]

This patch against -current does appear to work (ugh, I had forgotten how big 
the default font was). Thanks!  Best of luck.

I wouldn't be too optimistic about this fixing the iMac, though I wouldn't say 
there's *no* chance. By and large, the only changes I made were to the PCI and 
internal device matching and handling for really early parts (the GX/CX 
(pre-RAGE) were made during the PCI/ISA transition, and the internal device ID 
registers didn't match the PCI product IDs like all the later ones, and there 
were a handful of registers which read differently in the early chips).  I 
still haven't added in any of the handling for the motley crew of external PLLs 
used by the GX/CX cards, so actual modeswitches are probably out for now.

Still, adding a register-only mode-switch to the default mode made the big 1k 
pixel offset go away on mine, so maybe it'll help yours.  Essentially, I found 
that OpenFirmware initializes the screen with an offset for some reason 
(perhaps there's display-specific data there; the hardware cursor data is at 
the end of the memory range, not the beginning), so perhaps something is off 
there.

The other possibility is (as mentioned) the CRT handling being buggy, since the 
iMac had some interesting restrictions on modes.  I wish my parents' old Rev B 
iMac was still alive and well... maybe they still have it.  I'll check.

Can you send a photo/video of the screen?  Might help in diagnosing.  I'd 
suggest playing with registers in ddb as well to see if you can get it scanning 
right (that's how I figured out where the offset was happening on mine), but 
the iMac doesn't have a serial port unless you got an iPort, so you'd be flying 
blind on the console end of things.

- Dave

On Apr 28, 2011, at 11:32 AM, John D. Baker wrote:

> I've adapted your patch to -current (machfb.c r1.63) in case your
> findings might get the RagePro in the original iMac 233Mhz G3 to
> work.
> 
> There was some rearrangement so the "last" patch actually came earlier.
> 
> The patched "machfb.c" compiles.  I hope I can test it this afternoon,
> otherwise it will be next week before I can test it.
> 
> -- 
> |/"\ John D. Baker, KN5UKS               NetBSD     Darwin/MacOS X
> |\ / jdbaker[snail]mylinuxisp[flyspeck]com    OpenBSD            FreeBSD
> | X  No HTML/proprietary data in email.   BSD just sits there and works!
> |/ \ GPGkeyID:  D703 4A7E 479F 63F8 D3F4  BD99 9572 8F23 E4AD 
> 1645<machfb-r1.63_patch.diff>