[MPlayer-dev-eng] [PATCH] vo_vesa fixes for some cards

[The Wanderer]

D Richard Felker III wrote:

> On Sun, Oct 03, 2004 at 03:42:09AM -0400, The Wanderer wrote:
> 
>>> use svgalib_helper instead. it works perfectly.
>> 
>> I think I've heard that mentioned, but I don't know where to find
>> it. (Admittedly I also haven't yet Googled - I'd do that before
>> posting, but I'm in a hurry to get to my food before it gets cold.)
> 
> it comes with recent svgalib. but be warned, it contains a very
> stupid vulnerability that lets any user with access to the helper
> access all of kernel memory read/write.. someday i'm gonna submit a
> patch.

I'll pass that on to my siblings, then, and see about giving it a try
myself. If/when you do get around to submitting a patch, I'd be glad if
you could somehow let me know (since I don't follow svgalib development,
at least not yet), because I don't like using things with known security
flaws.

>>> no. anything that uses rgb colorspace is VERY slow. that includes
>>> vesa. even something slow like xv will be faster than vesa...
>>> the only slower vo is x11.
>> 
>> ...which was my only real other option back then, since I think I
>> had some problems with xv (I don't remember what) and I had no 3D 
>> acceleration and so couldn't use gl - and there must have been
>> *some* problem with sdl, but I don't remember what other than its
>> being too slow when doing fullscreen. (VESA had the other
>> disadvantage of using what looked like reverse video in RM files...
>> which is why I was so glad to finally get better hardware.)
> 
> just curious...what card was this? maybe you mentioned it in another 
> email but if so i forgot.

I don't think I mentioned it, no. 'Twas a quite basic PCI video card, I
forget exactly what - something in my memory is saying "ATI RAGE Pro",
but that may be the video card one of my brothers had at the same time
(he now has something much better).

>>> you must be doing something stupid like trying to scale the video
>>> to fullscreen. a 320x240 movie will play just fine even on a
>>> low-end pentium.
>> 
>> Nope, simple 'mplayer -fs filename.avi'. The system in question was
>> a
> 
> depending on how long ago this was, maybe mplayer did
> software-scaling with -fs by default in some vo's...

Possible, but it seems a touch unlikely. This was back in some 0.90
version - lemme see...the earliest MPlayer tarball in my archives is
0.90rc2. (I tend to keep all of my old installers - I'm an inveterate
and probably incurable pack rat.)

>> 475MHz AMD K* (I don't remember the specifics), with IIRC 192MB of
>> RAM - not spectacular (except in that I've never even heard of
>> another machine with such a strange processor speed number), but
>> should certainly have been enough for the purpose.
> 
> you're right, it is a strange speed. afaik it was overclocked. the 
> highest speed available for most k6 processors was 450. one model
> also had a 500mhz version, but that was rare. so i expect whoever
> sold it to you overclocked it (or lied about its speed so you would
> overclock it yourself).

Well, I certainly never intentionally overclocked it, and I thought the
vendor was "respectable" enough that they wouldn't have done it
themselves (much less without telling me) - but it is still about the
only explanation I've been able to think of. ISTR its having been
formally identified as a K6-2, but I don't know for certain.

I think that CPU is still around, in one of the other damn-slow machines
I mentioned - would it be worth the effort for me to find out which and
see what other information I can dig up on it?

-- 
       The Wanderer

Warning: Simply because I argue an issue does not mean I agree with any
side of it.

A government exists to serve its citizens, not to control them.