Robin Redeker wrote:
> 
> On Sat, Mar 23, 2002 at 08:13:33AM -0700, Brian Paul wrote:
> > Robin Redeker wrote:
> > >
> > > Hi,
> > >
> > > i am working on a level editor and when linking
> > > the program with efence, i get a segmentation fault,
> > > here the backtrace:
> > > -----------------------------------------
> [..snip..]
> > > ------------------------------------------
> > >
> > > Its only occuring when linking to efence.
> > > The program works on others people computers with, for example, nVidias
> > > drivers. So i think, this is a bug in dri libGL.
> > >
> [.snip.]
> >
> >
> [.snip.]
> >
> > You should at least try recompiling the GLU library (get it from the Mesa
> > distro) with -g and examine the instruction that's causing the segfault.
> > Is it a read or write problem, for example?
> >
> 
> Hm, i compiled now the libGLU in src-glu/ from Mesa-4.0.1 source.
> After linking it, installing it to /opt/Mesa-4.0.1/lib/ and linking the
> Program to it, the segfault i posted didn't occur.
> 
> from ldd of programm:
>         libgtkgl.so.5 => /opt/lib/libgtkgl.so.5 (0x40066000)
>         libGL.so.1 => /usr/lib/libGL.so.1 (0x4006c000)
>         libGLU.so.1 => /opt/lib/libGLU.so.1 (0x400d8000)
> elmex@enceladus: ~$ ll /opt/lib/libGLU.so.1
> lrwxrwxrwx    1 root     root           31 Feb  7 19:04 /opt/lib/libGLU.so.1 -> 
>/opt/Mesa-4.0.1/lib/libGLU.so.1*
> (I am shure the libs pointing to the compiled one (i removed old libGLU))

Great.  So it doesn't sound like you have a GLU bug after all...
 
> Butanother one:
> ---------------------------------------
> #0  0x425344e2 in radeonInitTextureObjects (ctx=0x49ca7e04) at radeon_tex.c:698
> #1  0x4253475e in radeonInitTextureFuncs (ctx=0x49ca7e04) at radeon_tex.c:747
> #2  0x4252b39f in radeonCreateContext (dpy=0x4088fac8, glVisual=0xbfffebec, 
>driContextPriv=0x49bdefe8,
>     sharedContextPrivate=0x4234c73c) at radeon_context.c:378
> #3  0x424021c4 in driCreateContext (dpy=0x4088fac8, vis=0x49888e70, 
>sharedPrivate=0x4234afe8, pctx=0x49bdcff0)
>     at dri_util.c:849
> #4  0x400bb799 in CreateContext (dpy=0x4088fac8, vis=0x49888e70, 
>shareList=0x42346ae4, allowDirect=1, contextID=0)
>     at glxcmds.c:169
> #5  0x400bb8c1 in glXCreateContext (dpy=0x4088fac8, vis=0x49888e70, 
>shareList=0x42346ae4, allowDirect=1)
>     at glxcmds.c:206
> #6  0x40067adb in gdk_gl_context_share_new (visual=0x408d7f10, sharelist=0x422dcff4, 
>direct=1) at gdkgl.c:149
> #7  0x4006838b in gtk_gl_area_share_new (attrlist=0x8069db4, share=0x471bffc8) at 
>gtkglarea.c:125
> ---------------------------------------
> 
> This segfault occurs _after_ i did, what the first segfault caused.
> This occurs, when opened the 'TexturePicker' one or two times.
> 
> Here is some info, which may helps:
> -------------------------
> (gdb) f 0
> #0  0x425344e2 in radeonInitTextureObjects (ctx=0x49ca7e04) at radeon_tex.c:698
> 698                      (radeonTexObjPtr)texObj->DriverData );
> (gdb) p texObj
> $3 = (struct gl_texture_object *) 0x42357e50
> (gdb) p texObj->DriverData
> $4 = (void *) 0x471b3ed4
> (gdb) p *((radeonTexObjPtr)texObj->DriverData)
> $5 = {next = 0x471b5ed4, prev = 0x49be4b4c, tObj = 0x42357e50,
>     memBlock = 0x0, bufAddr = 0, dirty_images = 0,
>   bound = 0, heap = 0, image = {{x = 0, y = 0, width = 0, height = 0,
>     data = 0x0} <repeats 12 times>},
>   totalSize = 0, pp_txfilter = 13, pp_txformat = 2147483648,
>     pp_txoffset = 0, pp_border_color = 0, firstLevel = 0,
>   lastLevel = 0}
> ------------------
> (gdb) f 1
> #1  0x4253475e in radeonInitTextureFuncs (ctx=0x49ca7e04) at radeon_tex.c:747
> 747        radeonInitTextureObjects( ctx );
> ------------------------

If you think you have a Radeon bug, some additional information would be
helpful.  Top of the list would be a way for developers to reproduce the
problem on their system.  Secondarily, knowing whether the problem
existed in the previous verson of the Radeon driver (the version in
XFree86 4.2 or earlier), or if it exists in other DRI drivers would be
helpful in narrowing this down.

There is also a newer TCL driver under development that you should also
now about.  You can check out source from the tcl-0-0-branch, or just
download the binaries from ftp://ftp.tungstengraphics.com/dri

--                             /\
         Jens Owen            /  \/\ _    
  [EMAIL PROTECTED]  /    \ \ \   Steamboat Springs, Colorado

_______________________________________________
Dri-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/dri-devel

Reply via email to