[rescue] Subject: (no subject)
Chase Rayfield
cusbrar2 at yahoo.com
Sun Apr 2 12:35:20 CDT 2017
The TGX / CG6 cards do have a geometry engine or something of that nature for
wireframe acceleration...Michael Lorenz /B maccallan of netbsd I has full
documentation I think but I'm not sure he can release it.
The SX has an integer vector engine... so could potentially be useful for
other things as well. NetBSD actually has an accelerated driver using it...
fast enough to run KDE3
Chase
On Sunday, April 2, 2017 1:00 PM, "rescue-request at sunhelp.org"
<rescue-request at sunhelp.org> wrote:
Send rescue mailing list submissions to
B B B rescue at sunhelp.org
To subscribe or unsubscribe via the World Wide Web, visit
B B B http://www.sunhelp.org/mailman/listinfo/rescue
or, via email, send a message with subject or body 'help' to
B B B rescue-request at sunhelp.org
You can reach the person managing the list at
B B B rescue-owner at sunhelp.org
When replying, please edit your Subject line so it is more specific
than "Re: Contents of rescue digest..."
Today's Topics:
B 1. Re: SPARCstation 5: framebuffer resolution wont change
B B B (Romain Dolbeau)
B 2. Re: SPARCstation 5: framebuffer resolution wont change (Mouse)
B 3. (no subject) (Mouse)
B 4. Re: (no subject) (Romain Dolbeau)
----------------------------------------------------------------------
Message: 1
Date: Sun, 2 Apr 2017 11:35:35 +0200
From: Romain Dolbeau <romain at dolbeau.org>
To: The Rescue List <rescue at sunhelp.org>
Subject: Re: [rescue] SPARCstation 5: framebuffer resolution wont
B B B change
Message-ID:
B B B <CADuzgbq94qHGJbqd4nfMKq=HUkLHPnvoXBR+vjYCwhJjTJaatQ at mail.gmail.com>
Content-Type: text/plain; charset=UTF-8
2017-04-02 10:11 GMT+02:00 Romain Dolbeau <romain at dolbeau.org>:
> 4) "nvrun"
> 5) "s10775" to get 1024x768 VESA
Checking on actual hardware, you need to cd to the device first, i.e.:
4b) cd /sbus/iommu/cgsix at 2
where the words are defined. (the exact name of the cgsix device might
change, use show-devs)
Cordially,
--
Romain Dolbeau
------------------------------
Message: 2
Date: Sun, 2 Apr 2017 10:21:14 -0400 (EDT)
From: Mouse <mouse at Rodents-Montreal.ORG>
To: The Rescue List <rescue at sunhelp.org>
Subject: Re: [rescue] SPARCstation 5: framebuffer resolution wont
B B B change
Message-ID: <201704021421.KAA27117 at Stone.Rodents-Montreal.ORG>
Content-Type: text/plain; charset="iso-8859-1"
>> Modern flatscreen monitors have a nasty tendency to (a) have very
>> limited ability to cope with signals not exactly matching what their
>> EDID says they want and (b) throw up their figurative hands and
>> display either an "I don't like that signal" indication or nothing
>> at all when faced with a signal they don't like.
> This is pretty much a necessity since a flat-panel is, by definition
> a digital display and the older multi-sync CRTs are inherently
> analogue.
I find this argument..unconvincing.
"Digital devices are made with analog components."
> Contemplate for a moment what goes into each.B In the analogue
> multi-sync, the thing picks appropriate scan and frame rates, figures
> out the porches, syncs up and simply channels the analogue video
> signals to the electron-guns.
A flat-screen could do exactly the same thing, except feeding the video
signal to an A->D instead, at a clock rate PLL-synced to the input.
Phase is slightly more interesting, but still not all that difficult.
> Another failing-point for flat-panels is that they look like utter
> cr at p once off their native resolution,
Not always.B The early ones were (almost?) all willing to letterbox
instead of scaling (and I prize the few I have that are); modern ones
less often, and scaling is sometimes done well and sometimes poorly.
(Indeed, what constitutes "well" and "poorly" depends at least in part
on the application.)
/~\ The ASCIIB B B B B B B B B B B B B Mouse
\ / Ribbon Campaign
XB Against HTMLB B B B B B mouse at rodents-montreal.org
/ \ Email!B B B B B 7D C8 61 52 5D E7 2D 39B 4E F1 31 3E E8 B3 27 4B
------------------------------
Message: 3
Date: Sun, 2 Apr 2017 10:29:37 -0400 (EDT)
From: Mouse <mouse at Rodents-Montreal.ORG>
To: rescue at sunhelp.org
Subject: [rescue] (no subject)
Message-ID: <201704021429.KAA06519 at Stone.Rodents-Montreal.ORG>
Content-Type: text/plain; charset="iso-8859-1"
>> I once spent an entire afternon trying to get a 1920x1080 flatscreen
>> to display 1920x1080 from a SPARCstation-20, and failed.
> TGX or SX ?
cg14, which I think means SX.
> TGX+ should be possible in some cases... (you need the 4 MiB of the +
> for the resolution, TGX won't do anything larger than the default
> 1152x900 with just 1 MiB)
It won't?B I know at least one framebuffer I played with was willing to
do 1920x1080 with only 1M, but it did so by repeating the pixels: after
1048576 pixels into the display, it started over.B But I can't recall
whether that was a cg6 or a cg14, or maybe even something else (I know
I have at least one leo...).
> Most flatscreen won't sync to anything with a refresh rate not very
> close to 60 Hz in practice.
Yeah.B "Improved to the point where they're not capable of what
thirty-year-old monitors did routinely."
> TGX(+) were amazing piece of hardware :-)
Indeed.B I just wish they were better documented.B (I've seen reason to
think there's a 3D transform engine in there, albeit not other things
that would today be considered essential for 3D rendering, but haven't
figured out how to do anything but the default of ortho projection
straight down the Z axis.)
/~\ The ASCIIB B B B B B B B B B B B B Mouse
\ / Ribbon Campaign
XB Against HTMLB B B B B B mouse at rodents-montreal.org
/ \ Email!B B B B B 7D C8 61 52 5D E7 2D 39B 4E F1 31 3E E8 B3 27 4B
------------------------------
Message: 4
Date: Sun, 2 Apr 2017 17:07:58 +0200
From: Romain Dolbeau <romain at dolbeau.org>
To: The Rescue List <rescue at sunhelp.org>
Subject: Re: [rescue] (no subject)
Message-ID:
B B B <CADuzgboM7ue54p5G9PwcsrhQo1zg07jXfKcA2vFn-GhLmdJDoA at mail.gmail.com>
Content-Type: text/plain; charset=UTF-8
2017-04-02 16:29 GMT+02:00 Mouse <mouse at rodents-montreal.org>:
> It won't?
Not with any degree of usefulness, you need almost 2 MiB just to store
the pixel values for 1920x1080 ...
> I know at least one framebuffer I played with was willing to
> do 1920x1080 with only 1M, but it did so by repeating the pixels: after
> 1048576 pixels into the display, it started over.
... so 1 MiB just won't be enough :-)
So, yes,B it's possible the regular TGX could be convinced to output
1920x1080 same as a TGX+, but then the framebuffer memory will come
from god-knows-where, possibly repeating the first MiB as you observed
if you're lucky (rather than reading some random bits elsewhere in
sbus memory space)... so I stand by my statement that you need a TGX+
for 1920x1080 :-) (or even for 1280x1024 on a 17" LCD, unfortunately,
since that require 1 1/4 MiB of framebuffer memory).
Cordially,
--
Romain Dolbeau
------------------------------
Subject: Digest Footer
_______________________________________________
rescue maillistB -B rescue at sunhelp.org
http://www.sunhelp.org/mailman/listinfo/rescue
------------------------------
End of rescue Digest, Vol 173, Issue 3
**************************************
More information about the rescue
mailing list