[MPlayer-users] -vo svga segmentation fault

[Ken Acmens]

p_l at gmx.fr wrote:

>[Automatic answer: RTFM (read DOCS, FAQ), also read DOCS/bugreports.html]
>On Wed, Jul 31, 2002 at 09:14:40PM -0400, Ken Acmens wrote:
>
>>GS HUNT wrote:
>>
>
>>>Hmm... Neomagic has had some problem in the past with the Xfree 
>>>server. If you have the money... i'd get a better card. A quick glance 
>>>around the web suggests that there are still problems with neomagic 
>>>cards and Xfree server... Might want to check for lastest patches for 
>>>the Xfree server.
>>>
>
>>>Ken Acmens wrote:
>>>
>
>>>>vo_svga: preinit - maxframes=1
>>>>Opening video decoder: [mpegpes] MPEG 1/2 Video passthrough
>>>>VDec: vo config request - 720 x 480 (preferred csp: Mpeg PES)
>>>>Segmentation fault
>>>>
>>>>This happens with MPlayer-20020731 and MPlayer-0.90pre5.
>>>>
>>>>I've traced (by inserting more printfs) the problem to the vga_init() 
>>>>function call in checksupportedmodes() in libvo/vo_svga.c
>>>>
>>>>Strangely, the demos that come with svgalib-1.4.3 call vga_init() and 
>>>>run just fine.  For some reason mplayer can't call that function...
>>>>
>>>>Does anyone have any ideas?
>>>>
>
>>>>Neomagic Corporation NM2380 [MagicMedia 256XL+] (rev 10)
>>>>
>>>>Video Driver: svgalib-1.4.3 configured as VESA in /etc/vga/libvga.config
>>>>
>>>>Computer: Sony Vaio PCG-F560
>>>>
>
>>My computer is a notebook.  I'm not aware of any way to replace the 
>>video chip.  I wish I could, because it is a crappy chip.  Do you know 
>>of a way?  PCMCIA perhaps?  Docking station?
>>
>
>Sweet dreams :)
>
>>I'm not running X.  I'm trying to play movies without X because I'm not 
>>happy with mplayer's playback under X.  Here's some info about svgalib: 
>>http://www.svgalib.org/vibber/svgalib-dev-faq-2.html
>>
>>I've also tried -vo fbdev and I'm ALMOST satisfied with it.  I'm trying 
>>svgalib just in case it works that tiny bit smoother.
>>
>
>Did you try running mplayer with plain VESA instead of VESA thru svgalib ?
>
Yes. -vo vesa also causes a seg fault.

>Maybe it's a bug in your chip regarding VESA support so neither svgalib
>nor vesa might be able to workaround it :/
>
You could be right.