When I try to emerge openoffice-bin, it stops during src_install The output of the emerge-command follows: Calculating dependencies ...done! >>> emerge (1 of 1) app-office/openoffice-bin-1.1.0 to / >>> Downloading http://gentoo.inode.at//distfiles/OOo_1.1.0_LinuxIntel_install. tar.gz --23:14:37-- http://gentoo.inode.at//distfiles/OOo_1.1.0_LinuxIntel_install. tar.gz => `/usr/portage/distfiles/OOo_1.1.0_LinuxIntel_install.tar.gz' Resolving gentoo.inode.at... 81.223.20.162 Connecting to gentoo.inode.at[81.223.20.162]:80... connected. HTTP request sent, awaiting response... 200 OK Length: 78,347,712 [application/x-tar] >>> md5 src_uri ;-) OOo_1.1.0_LinuxIntel_install.tar.gz >>> Unpacking source... >>> Unpacking OOo_1.1.0_LinuxIntel_install.tar.gz to /var/tmp/portage/openoffice-bin-1.1.0/work >>> Source unpacked. >>> Install openoffice-bin-1.1.0 into /var/tmp/portage/openoffice-bin-1.1. 0/image/ category app-office * Installing OpenOffice.org into build root... glibc version: 2.3.2 Initializing installation program.................... !!! ERROR: app-office/openoffice-bin-1.1.0 failed. !!! Function src_install, Line 93, Exitcode 1 !!! The setup program failed Reproducible: Always Steps to Reproduce: 1. emerge openoffice-bin 2. 3. Actual Results: Emerge exits with exitcode 1 in line 93 of src_install laptop richard # emerge info Portage 2.0.49-r15 (default-x86-1.4, gcc-3.3.2, glibc-2.3.2-r9, 2.4. 20-gentoo-r8) ================================================================= System uname: 2.4.20-gentoo-r8 i686 Celeron (Coppermine) Gentoo Base System version 1.4.3.11 ACCEPT_KEYWORDS="x86 ~x86" AUTOCLEAN="yes" CFLAGS=" -O3 -ffast-math -march=pentium3 -funroll-loops -fomit-frame-pointer -fgcse --force-addr -falign-loops=4 -falign-functions=8" CHOST="i686-pc-linux-gnu" COMPILER="gcc3" 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" CXXFLAGS=" -O3 -ffast-math -march=pentium3 -funroll-loops -fomit-frame-pointer -fgcse --force-addr -falign-loops=4 -falign-functions=8" DISTDIR="/usr/portage/distfiles" FEATURES="sandbox ccache autoaddcvs" GENTOO_MIRRORS="http://gentoo.inode.at/ ftp://ftp.easynet.nl/mirror/gentoo/ http://ftp.easynet.nl/mirror/gentoo/ ftp://gentoo.linux.no/pub/gentoo/ ftp: //ftp.tu-clausthal.de/pub/linux/gentoo/ http://gentoo.linux.no/ ftp://ftp. uni-erlangen.de/pub/mirrors/gentoo http://194.83.57.3/sites/www.ibiblio. org/gentoo/ http://mirrors.sec.informatik.tu-darmstadt.de/gentoo http://ftp. uni-erlangen.de/pub/mirrors/gentoo" MAKEOPTS="-j2" PKGDIR="/usr/portage/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="" SYNC="rsync://rsync.europe.gentoo.org/gentoo-portage" USE="x86 oss foomaticdb kde libg++ libwww mikmod motif ncurses xml2 gdbm berkdb slang arts svga guile sdl gpm tcpd python esd imlib qt opengl X acl -afs apm avi bonobo cdr crypt cups dvd encode ethereal evo gif gnome gphoto2 gps gtk gtk2 imap java jpeg ldap mad mbox memlimit mmx mozilla mpeg mpi nls oggvorbis pam pcmcia pda pdflib perl plotutils png pnp quicktime readline samba scanner slp spell sse ssl tiff truetype usb wmf xmms xv zlib -alsa"
Any news? I've hit the same problem. I can add info if it's any help.
I've worked out some more now.... I'm trying to install it in a chroot where it fails. Once I install the rest of the system onto its target, this ebuild installs fine. I haven't tried to run it yet but that's another story if anything's wrong. Back in the chroot.... Trying to run setup manually in the work directory fails as it can't open a display. I ran "xhost +" outside the chroot but still no good. Then I "export DISPLAY="<hostname>:0" inside the chroot and we get further now. The setup program starts and shows the initial progress bar OK but then it just dies before starting up the setup proper. I can't spot any clue why. HTH to solve this.
Well, it seems that things do not work in a chroot. At least thanks to you I have an idea where to look. In this chroot, did you make the /dev and /proc available through: mount -bind /proc /path/to/chroot/proc mount -bind /dev /path/to/chroot/dev If these are not available that could be the cause.
Yes I do mount -bind /dev /proc also just in case it matters also /usr/portage and /usr/local/portage. Happy hunting.
There is now a bug fix in cvs which tries to resolve this problem, could you be so nice to try again and report?
Yes, it emerges fine now, but when I try to start oosetup or any other application I get another error message richard@laptop richard $ oosetup running openoffice.org setup... /usr/bin/oosetup: line 165: 16202 Segmentation fault /opt/OpenOffice.org${PV}/program/setup ${UPDATEFLAG} -R:/etc/openoffice/autoresponse-1.1.0.conf 1>&/dev/null setup failed.. abort
Did you have an install of OpenOffice.org 1.0.x before? If yes, try to remove the .sversionrc and the .openoffice dir in your home directory and start again. Updating from older versions does not really work atm :-(
It builds for me too in my chroot. I haven't tried to install and run it on my target host - I'm using OOo heavily for work at the moment and I daren't chance it but I've got no reason to doubt it will work.
Also running the compiled verion of openoffice in general solves some of these problems if you can get it compiled (which is not trivial).
Removing the ~/.openoffice and .sversionrc does not change anything in behaviour. And I think compiling it myself is not an option as I tried it (with much less aggressive CFLAGS) and after 12 hours it crashes my machine due to the resource intensive compiler runs! (Celeron 800, 192 MB, 512MB Swap)
@ Richard: To be honest: I am quite clueless whats going on here, I've never heard of that problem before, also that you get a Segmentation fault doesn't fit with usual problems. So there has to be something with your setup causing this... Did you ever try to manually install OpenOffice (unpacking and running setup)?
The reporter of this bug asked me to close it, as the original problem is resolved and his new troubles don't seem to be related to our ebuild. So: Doing this.
