[MPlayer-users] VESA stopped working with Radeon.

TJ tianjy at jippii.fi
Tue Jun 24 22:35:02 CEST 2003


Sylvain Petreolle wrote:
> I disagree with you here.
> Im using Radeon 9000 Pro with mplayer -vo vesa, it has the problem.
> Try with mplayerxp -vo vesa => no problem at all.

Crashes in the same way as mplayer (just grabbed the cvs snapshot and 
compiled, had to disable dvdnav and lirc). Do you have glibc version 
>=2.2.5? See the link at the bottom of this post and the posts 
linked/followupped from there. If you do have a new glibc, it'd be 
interesting to know why it works for you but not for me and many/some 
others.

Ruben Garcia wrote:
> What about using "normal, unoptimized pthreads" when compiling glibc?
> anyway, MPlayer is supposed to be single threaded, so the optimized 
> pthreads wouldn't be too useful.

Anyone trying this? I haven't yet compiled glibc, I had hopes I wouldn't 
have to...

and finally, Alex Sisson wrote:
> with the latest main cvs, ive tracked it down to line 
> 740 of osdep/lrmi.c - this chunk
> 
> #else
>         asm volatile (
>          "int $0x80"
>          : "=a" (r)
>          : "0" (113), "b" (vm));
> #endif
> 
> if comment it out i get a vesa failed error rather
> than a segfault )
> 
> i dont really understand that asm, so dont know how to
> fix it though :-(

Basically the asm just does the vm86-call, and should work. I changed it 
to use the newer syscall #166 instead, but it didn't change a thing. See 
http://mplayerhq.hu/pipermail/mplayer-dev-eng/2003-June/019289.html and 
follow-ups.

It gives me (too) vesa failed whenever I comment the section out and use 
the libc's vm86-call instead of the assembler thing. Vm86() returns -1 
(failed), errno 14 'EFAULT/Bad address' (I think).

-TJ

__
Ota itsellesi luotettava kotimainen email http://www.jippii.fi/
Tutustu samalla netin parhaaseen pelipaikkaan JIPPIIGAMESIIN.



More information about the MPlayer-users mailing list