From: Michael Joosten [joost@c-lab.de]
Sent: Friday, October 26, 2001 12:53 PM
To: Info-VAX@Mvb.Saic.Com
Subject: Re: PC164, OVMS7.2, isacfg and graphic cards (MACH64, S3 968)

Fred Kleinsorge wrote:
> 

> >I don't think you can use the later versions of the Trio64 that had a
> >collection of extra letters on the end V+ V2+ or something similar.
> 
> You can use the V+, but not the V2.  I actually *have* a server for the V2
> tucked away someplace, but it was never released, and we never sold boards
> with a V2 on it.

Indeed, the DECW$SERVER_DDX_S3.EXE is somewhat picky... With a friend I
checked the options, so most of what follows is just for the records,
err, Google:

Number Nine GXE64 (S3 Vision868) : Mentioned in
DECW$DEVICE_CONFIG_GQ.COM, but just does NOT work (this
DECW$SERVER_0_ERROR.LOG version is already purged, so citing a DUMP):

 3A74696E 496E6565 72635373 6D763373 s3vmsScreenInit: 0000F0
 69562033 53206570 79742070 69684320  Chip type S3 Vi 000100
 70707573 20746F6E 20343638 6E6F6973 sion864 not supp 000110
 72635373 6D763373 00000A64 6574726F orted...s3vmsScr 000120
 4E574F4E 4B4E5520 3A74696E 496E6565 eenInit: UNKNOWN 000130
 61666564 202C6570 79742070 69686320  chip type, defa 000140
 34366F69 72542033 53206F74 20746C75 ult to S3 Trio64 000150
 25783020 3A766552 00000000 0000000A ........Rev: 0x% 000160
 0A642520 3D3D2050 4E50202C 78322E32 2.2x, PNP == %d. 000170
 72635373 6D763373 00000000 00000000 ........s3vmsScr 000180
 79742070 69684320 3A74696E 496E6565 eenInit: Chip ty 000190
 0000202C 34366F69 72542033 53206570 pe S3 Trio64, .. 0001A0
 3A74696E 496E6565 72635373 6D763373 s3vmsScreenInit: 0001B0
 72542033 53206570 79742070 69684320  Chip type S3 Tr 0001C0
 73696E69 46202D20 0000202C 32336F69 io32, 

Still, the plain SYS$GQSDRIVER starts and shows the 'OpenVMS Alpha
Operator Console Label'. But no X11 server, sorry.

Then, the ELSA GLoria-8 (NO, this is not a Synergy!) with S3 968:
Needed the usual PCI-Id tweaks in DECW$USER_CONFIG.DAT, then I got the
driver started and the console label.

The XServer, though, identified this one as Trio32 with 1MB max. It did
produce an image, but alas, with a VSync of 23Hz (early TV sets of the
'30 ??) and HSync of 18kHz. So, forget it.
I'm somewhat surprised that it did NOT default to Trio64, as the
messages in the DUMP seem to indicate. PCI-Id of 968 from Xfree86:
#define PCI_CHIP_968            0x88F0 vs.
#define PCI_CHIP_TRIO           0x8811  for the Trio32's
#define PCI_CHIP_AURORA64VP     0x8812  and 64/64V
#define PCI_CHIP_TRIO64UVP      0x8814

While asking some colleagues for their old cards, I went back to the ATI
Mach64, which I finally got working by using a 'brute force' SYSMAN IO
CONNECT: Assuming the card in ISA slot 1, ANA/SYS CLUE CONF showing that 
the ISA bus is at TR 3 (PC164), the following managed to activate the
driver:
mcr sysman io con /vector=0/node=1/adap=3/csr=0 gqa0
/driver=sys$gqddriver
(did I mention that 'isacfg' did not work for me?)

Still, as Fred (or was it Hoff?) mentioned quite some time ago, this is
just an emergency solution, just for a box in the server room, but not
as desktop: With 1280x1024, I only get a stable, non-jittering image at
60Hz, slightly jittering at 66Hz, and completely unacceptable at 70Hz.
Either you use an old monitor or an LCD. 

Taking a look into Xfree86's 'Cards' file:

NAME Number Nine GXE64 with S3 Trio64  # there seems to be also a Trio
version of this card...
SEE S3 Trio64 (generic)

NAME Diamond Stealth 64 DRAM with S3 Trio64
SEE S3 Trio64 (generic)

NAME Diamond Stealth64 Graphics 2xx0 series (Trio64)
SEE S3 Trio64 (generic)

NAME Diamond Stealth 64 DRAM SE
SEE S3 Trio32 (generic)

NAME Diamond Stealth64 Video 2001 series (2121/2201)
SEE S3 Trio64V+ (generic)

NAME Cardex Trio64
SEE S3 Trio64 (generic)

NAME Cardex Trio64Pro
SEE S3 Trio64 (generic)

NAME Miro Crystal 12SD
SEE S3 Trio32 (generic)

NAME Miro Crystal 22SD
SEE S3 Trio64 (generic)

NAME SPEA/V7 Mirage P64 with S3 Trio64
SEE S3 Trio64 (generic)

NAME Number Nine FX Vision 330
SEE S3 Trio64 (generic)

NAME Number Nine FX Motion 331
SEE S3 Trio64V+ (generic)

NAME ASUS Video Magic PCI VT64
SEE S3 Trio64 (generic)

NAME VideoLogic GrafixStar 400
SEE S3 Trio64V+ (generic)

NAME VideoLogic GrafixStar 300
SEE S3 Trio64 (generic)

NAME 2 the Max MAXColor S3 Trio64V+
SEE S3 Trio64V+ (generic)

NAME DataExpert DSV3365
SEE S3 Trio64V+ (generic)

NAME ExpertColor DSV3365
SEE S3 Trio64V+ (generic)

NAME DSV3326
SEE S3 Trio64V+ (generic)

NAME STB Powergraph 64 Video
SEE S3 Trio64V+ (generic)

NAME STB Powergraph 64
SEE S3 Trio64 (generic)

NAME ELSA Winner 1000TRIO
SEE S3 Trio64 (generic)

NAME ELSA Winner 1000TRIO/V
SEE S3 Trio64V+ (generic)

NAME Hercules Terminator 64/Video
SEE S3 Trio64V+ (generic)

NAME Hercules Graphite Terminator 64/DRAM
SEE S3 Trio64 (generic)

Still, that probably does not mean that two cards with the same name
might have two different chips, as usual (think PCish)...

If I'm not mistaken, DEC used the Number Nine FX Motion 331 or Vision
330.

To be continued (perhaps), when the next candidates appear.

-- 
Michael Joosten, SBS C-LAB, joost@c-lab.de
Fuerstenallee 11, 33094 Paderborn, Germany
Phone: +49 5251 606127, Fax: +49 5251 606065
C-LAB is a cooperation of University Paderborn & SIEMENS