Re: SparcStation 4 serial console

[Mouse <mouse%Rodents-Montreal.ORG@localhost>]

> Yes, I also have the VRAM card for the SS20 and SS10 (the half-RAM
> version though) but removed it to free up more RAM place since I use
> it without video.

Yes, if I were running a '20 without video I'd pull any video support
it had.  In general, I'd do that since it's less likely to break on the
shelf than when powered up; in the case of the cg14 it also, as you
say, frees up another slot for RAM.

> It was so slow anyay with NetBSD compared to Solaris.

Yes.  We have significantly less documentation on the hardware (though
someone found enough to build useful levels of acceleration into the
wscons code, IIRC).

> I think I have a spare one now if I didn't give it away... should I
> check for you?

Well, I do wish I had spare cg14s, but I can't really justify asking
for any of your spares...especially since, if the implications of your
address's TLD are correct, there's an ocean between us.

> diag-switch is false.  Didn't check on SS10 and SS20, but if i
> remember with diag-switch it would not autoboot...

Well, you need to set diag-file and diag-device instead of boot-file
and boot-device, but otherwise I *thought* auto-boot? was obeyed even
when diag-switch? is true.  Maybe my memory is misleading me, or maybe
it's ROM-version-specific....

>> Also, what are ttya-mode and ttyb-mode?  [...]
> ttyb-rts-dtr-off=false
> ttyb-ignore-cd=true
> ttya-rts-dtr-off=false
> ttya-ignore-cd=true
> ttyb-mode=9600,8,n,1,-
> ttya-mode=9600,8,n,1,-

Okay, nothing noteworthy there.

>> Another possibility is that the serial output driver on the '4 is
>> fried.  If you boot it and write to the serial ports from NetBSD, do
>> you get any output?
> No.  No output... but the idea of Brian with getty gave me an idea.
> I enabled getty on both ttya & ttyb

> I tinkered to connect with that cable and a couple of other
> combinations, one never knows.  I noticed login appeared shortly and
> getty, like: [...]

> With some patience and attempts, I could blindly login! and type a
> command.  So I can send data to the SS4, but not receive it.

That _does_ sound like dead output.  Hmm, I wonder if power to the
level converters is fused or some such, because this sounds a lot like
the output side simply not working.

> I tried both serial port A and B! I can login in both, but get no
> output.

> I think it is unlikely that both output drivers are fried...

Yes, unless it's something common, like the same level shifter chip
being used for both and that chip is totally dead.  That's why I'm
speculating about a power issue.  If the high voltage (+12V and -12V, I
think) to the level shifters are dead, that would account for this very
neatly.

> However minicom reports me always offline, the cables probably don't
> carry DTR over correcly (or CD?)

Possibly; I don't know minicom.

> I wonder if that could disturb the SS4.. but not the SS10 & SS20...

Possible but unlikely, would be my guess.

If this were my machine, I'd probably try to find doc on which chip
does the level shifting and which pins are the high voltage supplies,
to check whether it's getting ±12V.  I'd also check the non-data pins
to see if _anything_ is getting driven.

What are you using for storage?  I'm wondering if perhaps ±12V is
totally broken.  While I'm far from sure, I suspect the only things
that use it are disks and the serial ports, and if you're running
diskless, or using a SCSI<->SATA adapter with a laptop drive, you might
not notice if +12V is dead.  And I think -12V might be used by _only_
the serial ports.  (Anyone out there know what uses ±12 on the '4?)

/~\ The ASCII				  Mouse
\ / Ribbon Campaign
 X  Against HTML		mouse%rodents-montreal.org@localhost
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B