Bug 1922

Summary:	Weird unresolved symbols when compiling from CVS
Product:	xorg	Reporter:	Michal Kepien <michal.kepien>
Component:	Server/General	Assignee:	Xorg Project Team <xorg-team>
Status:	RESOLVED FIXED	QA Contact:
Severity:	major
Priority:	high	CC:	eich, krh, mat, robin, sndirsch
Version:	git
Hardware:	x86 (IA32)
OS:	Linux (All)
Whiteboard:
i915 platform:		i915 features:

Description Michal Kepien 2004-11-25 15:38:04 UTC

I'm trying to get my Savage IX running with DRI. I'm precisely following the
instructions supplied at http://dri.sourceforge.net/cgi-bin/moin.cgi/Building.
After everything is compiled, X.org starts fine, but when I review Xorg.0.log,
it turns out that something's screwed up. Here's a snippet:

Symbol printf from module /usr/X11R6/lib/modules/linux/libdrm.a is unresolved!
Symbol puts from module /usr/X11R6/lib/modules/fonts/libtype1.a is unresolved!
Symbol printf from module /usr/X11R6/lib/modules/extensions/libdri.a is unresolved!

Each of these errors is printed out a few times. Before it happens, X.org
tells me that DRI was successfully initialized ("DRI is enabled"). When I run
glxinfo (or any other OpenGL application) afterwards, it crashes my Xserver
telling me that "An undefined function has been called". What's even more
frustrating, the same thing happens for me both when I update the X.org version
installed from a Slackware package (6.7.0) and when I compile the whole X.org
source tree from scratch! I'm more than confused by the contents of those
messages - unresolved symbol *printf*? Or puts? Even such a newbie as me knows
these are standard C I/O functions so I sense something weird in here...

Got any clues? Could this be a CVS error? Or maybe something's wrong with my
configuration? I'm using a self-compiled 2.4.27 kernel and Slackware 10.0 as I
mentioned before. gcc version 3.3.4, glibc version 2.3.2. Is anything else
worth mentioning? Dunno. If you'd like to review my whole Xorg.0.log, here it
is:

http://kempniu.no-ip.com/files/Xorg.0.log

I'd be grateful for any clues as I already saw posts about successful DRI
installations using Savage IX and I can't wait to get my very own working :)

Comment 1 Stefan Dirsch 2004-12-02 13:10:20 UTC

Try this one. 
 
--- programs/Xserver/GL/dri/dri.c.~1.6.~        2004-12-02 17:15:14.000000000   
+0100   
+++ programs/Xserver/GL/dri/dri.c       2004-12-02 17:16:08.026064392 +0100   
@@ -1136,8 +1136,6 @@ DRIGetDrawableInfo(ScreenPtr pScreen,   
     WindowPtr          pWin, pOldWin;   
     int                        i;   
    
-    printf("maxDrawableTableEntry = %d\n",   
pDRIPriv->pDriverInfo->maxDrawableTableEntry);   
-   
     if (pDrawable->type == DRAWABLE_WINDOW) {   
        pWin = (WindowPtr)pDrawable;   
        if ((pDRIDrawablePriv = DRI_DRAWABLE_PRIV_FROM_WINDOW(pWin))) {   
 
Does it help?

Comment 2 Stefan Dirsch 2004-12-02 13:26:45 UTC

http://lists.freedesktop.org/pipermail/xorg/2004-December/004895.html 
 
"Yes, the fix is in CVS now." 
[...] 
 
But I still see this problem with current CVS (2004-12-02). 
 
[...] 
Symbol printf from module /usr/X11R6/lib/modules/fonts/libtype1.a is 
unresolved! 
Symbol puts from module /usr/X11R6/lib/modules/fonts/libtype1.a is unresolved! 
[...] 
 
I use the nvidia driver. Therefore the unresolved symbols in libdri.a/libdrm.a 
might still be there as well.

Comment 3 Stefan Dirsch 2004-12-02 13:28:18 UTC

Maybe Kristian can check this again?

Comment 4 Kristian Høgsberg 2004-12-02 14:08:57 UTC

Hmm... this time I actually *did* commit it, so it shoud work now.

Comment 5 Stefan Dirsch 2004-12-02 15:00:10 UTC

Just verified. Works now. Thanks!

Comment 6 Kristian Høgsberg 2004-12-02 18:51:12 UTC

*** Bug 1999 has been marked as a duplicate of this bug. ***

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.