[MPlayer-dev-eng] [RFD] Something...

[D Richard Felker III]

On Mon, Apr 14, 2003 at 12:43:31AM +0200, Daniel Egger wrote:
> Am Son, 2003-04-13 um 19.43 schrieb D Richard Felker III:
> 
> > But there's no radeon_vid kernel module...
> 
> Whatever you mean by "radeon_vid", there's a framebuffer module,
> a X driver and a quite accelerated DRI module - all available in source.

I mean a way to use the hardware video scaler/colorspace conversion.
XV is not useful since the horrible xf86 code copies the image around
(some drivers even re-pack it!) with super-slow C functions before it
gets to the display, making it about as slow as doing the colorspace
conversion in software with -vo fbdev.

> > Also I was commenting about the ppc.
> 
> PowerBook *IS* ppc.

Right.

> > I'll cite again the root exploit that affects any suid program using
> > DGA then dropping root. They were notified and to my knowledge never
> > fixed it, though I haven't checked the latest sources. It takes
> > incompetence to make such a horrible bug in the first place, though.
> 
> I don't know about this very bug but DGA is deprecated and never worked
> well anyways, so it's quite understandable that they don't care much
> about it. You seem to be pretty confident that mplayer for instance
> is safe in this regard, I wouldn't if I were you.

No, I'm actually pretty confident that MPlayer is NOT safe in this
regard, and that a malformed movie (at least with certain demuxers)
could execute arbitrary code as the user running MPlayer. Hopefully
Arpi's G2 code is better. :)

Rich