please local # prelink -avfmR prelink: /usr/bin/qtParse: Using /usr/libexec/ld-elf.so.1, not /lib/ld-linux.so.2 as dynamic linker Laying out 488 libraries in virtual address space 41000000-50000000 Assigned virtual address space slots for libraries: /lib/ld-linux.so.2 44a6a000-44a7ecf0 /usr/lib/libGLcore.so.1 44a81000-44f60da8 /lib/libc.so.6 44f63000-45092a04 /lib/libm.so.6 45095000-450b7a90 [snip] Prelinking /usr/lib/libpng.so.3.1.2.5 Prelinking /usr/lib/opengl/nvidia/lib/libGLcore.so.1.0.4363 prelink: Warning: /usr/lib/libGLcore.so.1 has undefined non-weak symbols Segmentation fault unmerging nvidia-glx allows prelinking to successfully complete. This has shown up consistently with 20030217, 20030505, and a hand built 20030522. I also tried compiling prelink and elfutils with just CFLAGS="-pipe", same result. I intend to file a bug report directly with the prelink folks with more debugging info, but I thought you might like to know, too. Reproducible: Always Steps to Reproduce: 1. 2. 3. Portage 2.0.48-r5 (default-x86-1.4, gcc-3.3, glibc-2.3.2-r3) ================================================================= System uname: 2.4.21-ck3 i686 AMD Athlon(tm) MP 2000+ GENTOO_MIRRORS="http://gentoo.oregonstate.edu http://distro.ibiblio.org/pub/Linux/distributions/gentoo" CONFIG_PROTECT="/etc /var/qmail/control /usr/kde/2/share/config /usr/kde/3/share/config /usr/X11R6/lib/X11/xkb /usr/kde/3.1/share/config /usr/share/config" CONFIG_PROTECT_MASK="/etc/gconf /etc/env.d" PORTDIR="/usr/portage" DISTDIR="/usr/local/portage/distfiles" PKGDIR="/usr/portage/packages" PORTAGE_TMPDIR="/usr/local/portage" PORTDIR_OVERLAY="/usr/local/portage/ebuilds" USE="x86 oss apm cups foomaticdb mad mikmod nls xml2 gdbm slang bonobo ruby libwww motif X 3dnow mmx directfb sse dga opengl xv fbcon kde qt qtmt arts tcltk aalib imlib ncurses readline sdl svga lcms gif jpeg png tiff gd avi mpeg quicktime esd gtk gtk2 -gnome alsa ggi dvd xmms oggvorbis encode pam ssl crypt tcpd mozilla spell truetype xml pdflib plotutils tetex guile perl python libg++ atlas pic berkdb mysql postgres odbc samba gpm zlib java ppds" COMPILER="gcc3" CHOST="i686-pc-linux-gnu" CFLAGS="-march=athlon-mp -O2 -ftracer -ffast-math -pipe -fomit-frame-pointer" CXXFLAGS="-march=athlon-mp -O2 -ftracer -ffast-math -pipe -fomit-frame-pointer" ACCEPT_KEYWORDS="x86" MAKEOPTS="-j4" AUTOCLEAN="yes" SYNC="rsync://rsync.gentoo.org/gentoo-portage" FEATURES="sandbox ccache fixpackages"
Works fine here. Try with gcc-3.3-r1, or 3.3.1. If not, try a vanilla/gentoo kernel.
Well as soon as CVS is up again I will update prelink to 20030811. Test etc ... Also RedHat has decided to make a prelink cron job to reprelink the system preiodically .... we could do soething like that ourselves or do it in portage as in a preiodically spawned process in the background .. Anyway, Stefan
OK I am out of town, I will try gcc-3.3.1 next week when I get back and see what happens.
I just tried with elfutils-0.84 and prelink-20030811 rebuilt with gcc-3.3.1 and the problem persists. I will try a vanilla kernel when I get time.
Never had a problem with this either.
I seem to have got the same problem, but the segfault happens with another library: /usr/X11R6/lib/libXm.so.2.1 I suppose it's from XFree86, but a quick "qpkg -f /usr/X11R6/lib/libXm.so.2.1" gives no results!!! %-/ (The program just quits without output) BTW, I think qpkg should at least say something like: "Dude, no 'build found!". I installed a couple of binary packages which got no ebuild yet (e.g. Maya), and Maya also installs a libXm.so.2, but it's in another directory. I'm using "prelink -aRflv" compiled with gcc-3.2.3 and running under a vanilla kernel v2.6.0-test5. Should you need more info, feel free to ask! ;)
The newest version of prelink seem to have fixed this problem (for me at least).
Ok, closing bug I don't get a crash here also.
Hi guys, it looks like I have the same problem, with latest stable gcc (3.3.2-r5), nvidia-glx (1.0.4496) and prelink (20030811). Which prelink might fix it and is stable enough?
Try prelink-20040317 which I just marked as stable
Hi Stefan, it's working great. Thank you very much for your quick response.