[parisc-linux] C160 STI console woes

[parisc-linux] C160 STI console woes

[James P. Kinney III]

[-- Attachment #1: Type: text/plain, Size: 1859 bytes --]

OK, I've enclosed the dmesg output. The card had a sticker which lists
it as a 4017B. dmesg calls it a 4017b-lz. There is an expansion card on
it with a bunch of Motorola chips and a large one with a heat sink. It
is plugged into the GSC socket, not the PCI socket. (Why is the PCI
socket "backwards"? I've got a great Matrox Millineum ...)

grep -i sti dmesg.txt
Kernel command line: root=/dev/sda5 HOME=/ console=tty0 sti=0
sti_font=VGA8x16 TERM=linux
POSIX conformance testing by UNIFIX
STI word mode ROM at f0026000, hpa=fa000000
STI word mode ROM, id 2d08c0a7-9a02587, conforms to spec rev. 8.07
STI device: INTERNAL_EG_1280
STI byte mode ROM at f8000000, hpa=f8000000
STI byte mode ROM, id 2bcb015a-9a02587, conforms to spec rev. 8.04
STI device: HPA4071B_LZ
fb0: stifb 1280x1024-8 frame buffer device, id: 2d08c0a7, mmio:
0xfa100000
fb1: stifb 1280x1024-32 frame buffer device, id: 2bcb015a, mmio:
0xf8100000
sticonsole_init: searching for STI ROMs

grep -i graphics dmesg.txt
18. Gecko GSC Core Graphics (10) at 0xfa000000 [8/24], versions 0x16,
0x0, 0x85
21. Coral SGC Graphics (10) at 0xf8000000 [10/8], versions 0x4, 0x0,
0x77

Before the firmware upgrade, the Coral SGC card was listed as a
"Hyperdrive optional graphics (10) at 0xf8000000 [10/8], versions 0x5,
0x0, 0x77

It seems that sti=1 should turn on console on the graphics monitor.
Unless I need a different graphics terminal other than tty0 or tty1.

-- 
James P. Kinney III   \Changing the mobile computing world/
President and COO      \          one Linux user         /
Local Net Solutions,LLC \           at a time.          /
770-493-8244             \.___________________________./

GPG ID: 829C6CA7 James P. Kinney III (M.S. Physics)
<jkinney@localnetsolutions.com>
Fingerprint = 3C9E 6366 54FC A3FE BA4D 0659 6190 ADC3 829C 6CA7 



[-- Attachment #2: Type: application/pgp-signature, Size: 232 bytes --]

Re: [parisc-linux] C160 STI console woes

[Grant Grundler]

"James P. Kinney III" wrote:
> (Why is the PCI socket "backwards"? I've got a great Matrox Millineum ...)

You mean the keying of the PCI slot?
I don't know if C160 supports 3.3v, 5.5v, or both. I'd need to dig
up a product data sheet to tell you which the C160 (not C160L)
supported. But if your card doesn't fit the keys in the slot,
it's *not* electrically compatible.

Secondly, folks with broken Dino's built-in, can *only* plug
in HP-UPX supported graphics cards once firmware is upgraded
to 5.x or better (on older B and C-class machines).
PDC won't boot with anything else installed - but no harm
in trying (unlike mismatched PCI slot keys).

grant

Re: [parisc-linux] C160 STI console woes

[Martin K. Petersen]

>>>>> "Grant" == Grant Grundler <grundler@puffin.external.hp.com> writes:

Grant> You mean the keying of the PCI slot?  I don't know if C160
Grant> supports 3.3v, 5.5v, or both. I'd need to dig up a product data
Grant> sheet to tell you which the C160 (not C160L) supported. But if
Grant> your card doesn't fit the keys in the slot, it's *not*
Grant> electrically compatible.

C160 has 3.3V only (i.e. it doesn't take most cheapo PC cards).

I ended up buying Symbios branded SCSI controllers for my C160s.  They
support both 3.3 and 5V operation.

And I have yet to see a PCI graphics board with universal/3.3V
connector.  Except for HP ones, that is.

-- 
Martin K. Petersen      Cereal Bowl Engineer, Linuxcare, Inc.
http://mkp.net/         SGI XFS, Linux/PA-RISC, GNOME

Re: [parisc-linux] C160 STI console woes

[Alan Cox]

> And I have yet to see a PCI graphics board with universal/3.3V
> connector.  Except for HP ones, that is.

You want 5v for the fan 8(

Re: [parisc-linux] C160 STI console woes

[Grant Grundler]

Alan Cox wrote:
> > And I have yet to see a PCI graphics board with universal/3.3V
> > connector.  Except for HP ones, that is.
> 
> You want 5v for the fan 8(

I'm told the "3.3v" vs "5v" only indicates which voltage is
used for signalling - not which power rail the card is using.
Some 3.3v cards pull power from the 5v rail and convert it to 3.3v
in order to get a "cleaner" or more accurate 3.3v than what
the PCI Spec (and some vendors) provides.

And for PCI 2.1, 3.3v power rail is optional.
That turned into a requirement for PCI 2.2 compliance.

(Caveat: this is all second hand info - I'm a SW dork - not HW eng.
  but I believe it's correct.)

grant