Re: console.c: scrolling bug found

Steven S. Dick (ssd@nevets.oau.org)
Fri, 16 Feb 96 02:36 EST


Ulrich Windl <alfred!ucf-cs!rz.uni-regensburg.de!Ulrich.Windl> wrote:
>On 14 Feb 96 at 23:24, Riku Saikkonen wrote:
>[...]
>> This causes the occasional incorrect display in one frame (it is fixed when
>> the video card does the next display refresh since the outb_p()s have then
>> been completed).
[...]
>I'm afraid that this might be hardware dependent. Maybe some hardware
>has a latch for the new start address. I've never seen the effect on
>a Paradise VGA, or an ATI VGA Wonder XL.

I have an ET4000 card, and I see the scrolling glitch quite frequently.
I've written drivers for cards with hardware scrolling, so first time
I saw it, I knew exactly what it was, and ignored it. It doesn't really
bother me, and I've always ignored it.

As to being a security hole what human can really see information in a
single frame glitch???? (: And don't whine about cameras--if people are
gonna do that kinda thing, physical hardware attack (i.e., just find the
VC it was on previously) anyway.

>Isn't there also a bit to
>disable generation of video signals.

Ick, don't do this...it'll make it flicker.
I'd rather have the glitch.

If this bug gets fixed at all, the fix better not be worse.
Also, one of my very favorite features of linux is the speed the
console scrolls at. I'd be *real* unhappy if anything slowed it down.

However, I'm not sure I'd notice if the scrolling paused updating the
scroll registers a teeny bit while it continued updating the video ram...
(i.e. jumpier scrolling by a few microseconds, but not slower scrolling).

>> So, I guess it's reasonable to leave this slight glitch in, since it's so
>> harmless...

Yes. If you can fix it w/o causing other problems, go ahead,
but otherwise, even if "other problems" are small, leave it alone.

That's my vote anyway.

Steve
ssd@nevets.oau.org