As I stated in bug #138545 (when it still was an ebuild request) OpenOffice.org 2.0.3 is supposed to work out-of-the-box on amd64. However, on Gentoo it doesn't. As it reportedly works on non-gentoo systems, I assume the problem has something to do with the gentoo patchset or with a missed (or bad version) dependancy. However, I havn't been able to figure out exactly what. The emerge goes fine for a couple of hours, then it breaks while compiling the canvas (both with and without the cairo USE flag). I'm running stable (amd64, not ~amd64), but with quite an extencive package.keywords, including GCC-4.1.1 and the new java buildsystem. Portage 2.1-r1 (default-linux/amd64/2006.0, gcc-4.1.1, glibc-2.3.6-r4, 2.6.16-gentoo-r7 x86_64) ================================================================= System uname: 2.6.16-gentoo-r7 x86_64 AMD Athlon(tm) 64 X2 Dual Core Processor 4400+ Gentoo Base System version 1.6.15 ccache version 2.3 [enabled] app-admin/eselect-compiler: [Not Present] dev-lang/python: 2.4.2 dev-python/pycrypto: 2.0.1-r5 dev-util/ccache: 2.3 dev-util/confcache: [Not Present] sys-apps/sandbox: 1.2.17 sys-devel/autoconf: 2.13, 2.59-r7 sys-devel/automake: 1.4_p6, 1.5, 1.6.3, 1.7.9-r1, 1.8.5-r3, 1.9.6-r2 sys-devel/binutils: 2.16.1-r3 sys-devel/gcc-config: 1.3.13-r3 sys-devel/libtool: 1.5.22 virtual/os-headers: 2.6.11-r2 ACCEPT_KEYWORDS="amd64" AUTOCLEAN="yes" CBUILD="x86_64-pc-linux-gnu" CFLAGS="-march=k8 -O2 -pipe -fomit-frame-pointer" CHOST="x86_64-pc-linux-gnu" CONFIG_PROTECT="/etc /usr/NX/etc /usr/NX/home /usr/kde/3.5/env /usr/kde/3.5/share/config /usr/kde/3.5/shutdown /usr/lib/X11/xkb /usr/lib64/mozilla/defaults/pref /usr/share/config /usr/share/texmf/dvipdfm/config/ /usr/share/texmf/dvips/config/ /usr/share/texmf/tex/generic/config/ /usr/share/texmf/tex/platex/config/ /usr/share/texmf/xdvi/" CONFIG_PROTECT_MASK="/etc/env.d /etc/env.d/java/ /etc/gconf /etc/java-config/vms/ /etc/revdep-rebuild /etc/terminfo" CXXFLAGS="-march=k8 -O2 -pipe -fomit-frame-pointer" DISTDIR="/usr/portage/distfiles" FEATURES="autoconfig ccache distlocks metadata-transfer parallel-fetch sandbox sfperms strict user-fetch userpriv usersandbox" GENTOO_MIRRORS="http://distfiles.gentoo.org http://www.ibiblio.org/pub/Linux/distributions/gentoo" LANG="sv_SE.utf8" LINGUAS="en sv en_US en_GB sv_SE" MAKEOPTS="-j3" PKGDIR="/usr/portage/packages" PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress --force --whole-file --delete --delete-after --stats --timeout=180 --exclude='/distfiles' --exclude='/local' --exclude='/packages'" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="/usr/local/portage/migration-packages /usr/local/portage/local-overlay" SYNC="rsync://rsync.se.gentoo.org/gentoo-portage" USE="amd64 X a52 aac acl acpi alsa asf audiofile avi bash-completion berkdb bitmap-fonts bzip2 cairo caps cddb cli crypt cups curl dbus dga dlloader dri dvd dvdr dvdread emacs emboss encode fbcon ffmpeg firefox flac foomaticdb fortran ftp gif gmp gpm gtk gtk2 hal iconv idn imagemagick imap imlib innodb isdnlog java javascript jpeg jpeg2k kde kdeenablefinal lcms lzw lzw-tiff mad mjpeg mng mono mozilla mp2 mp3 mp4 mpeg mysql ncurses nls nptl nsplugin ogg opengl oss pam pcre pdf pdflib perl png ppds pppd ps python qt qt3 quicktime readline reflection samba sdl session slp smp speex spell spl sqlite ssl svg tcpd theora tiff timidity truetype truetype-fonts type1-fonts unicode usb v4l v4l2 vcd videos vorbis wmf x264 xcomposite xine xml xml2 xorg xpm xv xvid yahoo zip zlib elibc_glibc input_devices_keyboard input_devices_mouse input_devices_evdev kernel_linux linguas_en linguas_sv linguas_en_US linguas_en_GB linguas_sv_SE userland_GNU video_cards_fglrx video_cards_vesa video_cards_v4l" Unset: CTARGET, EMERGE_DEFAULT_OPTS, INSTALL_MASK, LC_ALL, LDFLAGS, PORTAGE_RSYNC_EXTRA_OPTS
Created attachment 91953 [details] Emerge output when when USE=-cairo
Created attachment 91954 [details] Emerge output when when USE=cairo Emerge output when when USE=cairo In this case I was using x11-libs/cairo-1.2.0 but I've tried with x11-libs/cairo-1.0.4 as well.
Looking at the logs I can't find any errors reported (at least the one with USE=cairo), can you point me at them? Also, can you show me where upstream says that it is expected to work on amd64? I only found various patchsets and a wiki page that says that it is of no use to non-developers..
(In reply to comment #3) > Also, can you show me where upstream says that it is expected to work on amd64? Err, I'd certainly be interested in this as well, 'cos it doesn't work either on Gentoo, or on Ubuntu which I tried. So, if they claim it works, they should fix it. :P
(In reply to comment #4) > Err, I'd certainly be interested in this as well, 'cos it doesn't work either > on Gentoo, or on Ubuntu which I tried. So, if they claim it works, they should > fix it. :P Depends on what you define as "upstream" or "working" I guess ;) To be more precise: AMD64 is still NOT an officially supported architecture for OpenOffice.org. As far as I know, with some patching it should be buildable now, but I'm not sure all the functionality will work correctly. But I've to admit, that I don't have the necessary hardware, so can only speak from "hearsay" But Hanno might know a little bit more about this, as he has been playing with this in his openoffice-overlay
(In reply to comment #5) > Depends on what you define as "upstream" or "working" I guess ;) > > To be more precise: AMD64 is still NOT an officially supported architecture for > OpenOffice.org. As far as I know, with some patching it should be buildable > now, but I'm not sure all the functionality will work correctly. Well, what I have meant it that it certainly does not "work out-of-the-box on amd64", that's what this bug is about. Yeah, it's possible to compile it w/ those patches and some luck, but that's a different story. *g*
i'd prefer to wait until upstream really supports it, other than through distro-specific patchsets. That way we can refer bugs to upstream. I don't think anybody has the time (read: nerves ;)) to support a custom patchset.
My additions to other comments here: - openoffice status on x86_64 is between alpha and beta. If anyone on x86_64 is brave enough to try it they should use my overlay.
I know amd64 isn't oficcialy supported by OO.o, but according to the blog of the main amd64 porter (Pavel Janik: http://blog.janik.cz) it "should work out of the box". However, he is speaking about vanilla OO.o, not ooo-build, without binfilter. So heresay has it that vanilla OO.o, Feudora ooo-build and Debian ooo-build works (there are native amd64 binaries on the net), and thus I conclude that the problem is Gentoo-specific. Regarding the logfiles, it seems they have been truncated. I put the emerge through tee, and only checked the screen output. Should have checked the logfiles as well. Sory about that, I'll redo the emerge and try to attach new logfiles by tomorrow, as soon as my monsterious command "su - -c '( USE=-cairo emerge --verbose --nocolor --nospinner openoffice 2>&1 | tee null_canvas.log ) ; ( USE=cairo emerge --verbose --nocolor --nospinner openoffice 2>&1 | tee cairo_canvas.log )'" is done running...
Created attachment 92051 [details] Emerge output when when USE=-cairo
Created attachment 92052 [details] Emerge output when when USE=cairo
Created attachment 94506 [details, diff] Ebuild patch to get OO.org working on amd64 I found the problem! OO.org comes with boost 1.30.2. However, boost 1.30.x is broken on amd64. By installing a system boost >=1.31.0 and passing --with-system-boost to OO.org configure it works perfectly on amd64. Attached is a small patch that fixes OO.org on amd64 without affecting the other arches. For the next revision you might consider using system boost on other arches as well.
Considering I found the problem, and supplied a fix, I think LATER is about now.
Didn't build for me. Failed with this: Inconsistency detected by ld.so: dl-minimal.c: 137: realloc: Assertion `new == ptr' failed! make[3]: *** [uconvmsg/resources/uconvmsg_root.res] Error 127 make[3]: Leaving directory `/home/portage/portage/openoffice-2.0.3/work/ooo-build-2.0.3.0/build/OOO_2_0_3/icu/unxlngx6.pro/misc/build/icu/source/extra/uconv' So, it's certainly not ready for prime-time.
Created attachment 94658 [details, diff] Patch to migrate OO.org to java-pkg-opt-2 (In reply to comment #14) > Didn't build for me. Failed with this: > > Inconsistency detected by ld.so: dl-minimal.c: 137: realloc: Assertion `new == > ptr' failed! > make[3]: *** [uconvmsg/resources/uconvmsg_root.res] Error 127 > make[3]: Leaving directory > `/home/portage/portage/openoffice-2.0.3/work/ooo-build-2.0.3.0/build/OOO_2_0_3/icu/unxlngx6.pro/misc/build/icu/source/extra/uconv' > > > So, it's certainly not ready for prime-time. This isn't a amd64 bug. I've encountered this particular bug on x86 as well. After much digging I found the problem to be with java: If you are using the new java-build-system (in ~arch) you MUST set the same VM for both generation 1 and generation 2. Another solution (which I've tested earlier) is to port the ebuild to inherit java-pkg-opt-2 instead of java-pkg, but then you need to create a new revision of the package, one that can't be stabilized until the java-heard gets java-config-2 (and related packages) into stable. Some googling shows that old (~2.3.2) glibc can cause the same icu problem as well (at least on x86 and alpha). Attached is a patch migrating OO.org to java-pkg-opt-2. It's not a replacement to the prior patch, but a complement, fixing a different bug. This patch is intended to create new revision of OO.org, thus the KEYWORDS and SRCDIR fixes.
I have no java, and was building with -java.
(In reply to comment #16) > I have no java, and was building with -java. Never tried that. So, either -java don't work on amd64 (will try to build -java on my system tonight), otherwise it's most likely a glibc problems. Are you using a recent version?
Yes, I have a full ~amd64 box. FWIW, I tried it both with and without as-needed, so that's not it. Portage 2.1.1_pre5-r2 (default-linux/amd64/2006.0, gcc-4.1.1/amd64-vanilla, glibc-2.4-r3, 2.6.17-ck1-dfg1 x86_64) ================================================================= System uname: 2.6.17-ck1-dfg1 x86_64 AMD Turion(tm) 64 Mobile Technology ML-28 Gentoo Base System version 1.12.4 Last Sync: Unknown distcc 2.18.3 x86_64-pc-linux-gnu (protocols 1 and 2) (default port 3632) [enabled] ccache version 2.4 [enabled] app-admin/eselect-compiler: 2.0.0_rc2-r1 dev-lang/python: 2.4.3-r1 dev-python/pycrypto: 2.0.1-r5 dev-util/ccache: 2.4-r2 dev-util/confcache: [Not Present] sys-apps/sandbox: 1.2.18.1 sys-devel/autoconf: 2.13, 2.60 sys-devel/automake: 1.4_p6, 1.5, 1.6.3, 1.7.9-r1, 1.8.5-r3, 1.9.6-r2 sys-devel/binutils: 2.17 sys-devel/gcc-config: [Not Present] sys-devel/libtool: 1.5.22 virtual/os-headers: 2.6.16 ACCEPT_KEYWORDS="amd64 ~amd64" AUTOCLEAN="yes" CBUILD="x86_64-pc-linux-gnu" CFLAGS="-O0 -march=athlon64 -pipe -ggdb -Wstrict-aliasing -Wreturn-type -Wimplicit -Wformat" CHOST="x86_64-pc-linux-gnu" CONFIG_PROTECT="/etc /usr/share/X11/xkb" CONFIG_PROTECT_MASK="/etc/env.d /etc/env.d/java/ /etc/eselect/compiler /etc/gconf /etc/java-config/vms/ /etc/revdep-rebuild /etc/terminfo /etc/texmf/web2c" CXXFLAGS="-O0 -march=athlon64 -pipe -ggdb -Wstrict-aliasing -Wreturn-type -Wimplicit -Wformat" DISTDIR="/home/portage/distfiles" FEATURES="autoconfig ccache collision-protect cvs distcc distlocks metadata-transfer multilib-strict nostrip sandbox sfperms sign strict" GENTOO_MIRRORS="http://distfiles.gentoo.org http://distro.ibiblio.org/pub/linux/distributions/gentoo" LC_ALL="en_US.utf8" LDFLAGS="-Wl,--as-needed" LINGUAS="en en_US de" MAKEOPTS="-j5" PKGDIR="/usr/portage/packages" PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress --force --whole-file --delete --delete-after --stats --timeout=180 --exclude='/distfiles' --exclude='/local' --exclude='/packages'" PORTAGE_TMPDIR="/home/portage" PORTDIR="/usr/portage" PORTDIR_OVERLAY="/home/portage/overlays/dang /home/portage/overlays/dev-overlay /home/portage/overlays/gnome /home/portage/overlays/dangtopia /home/portage/overlays/uml-overlay" SYNC="rsync://nobody.nowhere.foo/fuckoff" USE="amd64 X aac aalib acpi alsa apache2 artworkextra avahi avi bash-completion beagle berkdb bitmap-fonts bzip2 cairo cdr cli crypt cscope cups db2 dbus dga directfb divx4linux dlloader dri dvd eds elibc_glibc emboss encode ethereal evo exif fam firefox flac font-server foomaticdb gd gdbm gif gimpprint gnome gnutls gphoto2 gstreamer gtk gtk2 gtkhtml guile hal howl imagemagick imap imlib input_devices_evdev input_devices_keyboard input_devices_mouse input_devices_synaptics ipv6 isdnlog jabber jpeg kernel_linux ldap libg++ libnotify libwww linguas_de linguas_en linguas_en_US logrotate lzo lzw lzw-tiff mad matroska mikmod mime mmap mmx2 mng mono motif mozilla mp3 mpeg mpi ncurses nls nptl nptlonly offensive ogg opengl oscar oss pam pcmcia pcre pda pdflib perl png pppd python qemu-fast quicktime readline reflection rtc samba sdl session sharedmem slang slp smime soap softmmu speex spell spl sqlite ssl svg tcpd theora tiff truetype truetype-fonts type1-fonts unicode usb userland_GNU video_cards_ati video_cards_radeon video_cards_vesa vorbis wifi wmf xface xml xml2 xorg xpm xsl xv xvid zlib" Unset: CTARGET, EMERGE_DEFAULT_OPTS, INSTALL_MASK, LANG, PORTAGE_RSYNC_EXTRA_OPTS
Created attachment 94808 [details, diff] Patch to replace -O0 with -O2 (the only working optimization level) (In reply to comment #18) > Yes, I have a full ~amd64 box. FWIW, I tried it both with and without > as-needed, so that's not it. I've just tested with USE=-java, and it "works for me (tm)". You seems to have more recent versions than me of basicly every package, so I don't think that is a problem. However, I found two things in your emerge --info that might be related. OO.org is known to be broken with all optimization levels but -O2. The ebuild replaces -O1, -O3 and -Os with -O2, but for some reason the ebuild let -O0 pass through (probably becouse noone thought -O0 would be used in any real life situation). OO.org is also known to not work with distcc. The ebuild forces MAKEOPTS=-j1 (which *should* disable distcc) unless you set WANT_DISTCC=true, I hope you didn't? Attached is a patch to fix the forgotten -O0 replacement. This patch should be safe to apply to the stable ebuild. In fact I recommend it, as -O0 is known to be broken with upstream OO.org.
(In reply to comment #14) > Didn't build for me. Failed with this: > > Inconsistency detected by ld.so: dl-minimal.c: 137: realloc: Assertion `new == > ptr' failed! > make[3]: *** [uconvmsg/resources/uconvmsg_root.res] Error 127 > make[3]: Leaving directory > `/home/portage/portage/openoffice-2.0.3/work/ooo-build-2.0.3.0/build/OOO_2_0_3/icu/unxlngx6.pro/misc/build/icu/source/extra/uconv' > > > So, it's certainly not ready for prime-time. > That's a known bug and results from your custom PORTAGE_TMPDIR, see bug #130837, so not specific to this issue
I've added openoffice-2.0.4_rc1 to the tree which has some additions to make AMD64 work. Please test (you'll have to unmask it first) to help in making 2.0.4 ready for primetime (or decide against it, depending on the feedback)
(In reply to comment #21) > I've added openoffice-2.0.4_rc1 to the tree which has some additions to make > AMD64 work. Please test (you'll have to unmask it first) to help in making > 2.0.4 ready for primetime (or decide against it, depending on the feedback) Great. Trying to compile it as I'm writing this. While doing so I'm looking through your ebuilds and found a few things I think is wrong. 1) RDEPEND don't need any specific version of java-config, as openoffice has it's own java version detection and selection system. However, DEPEND should have a dep on ">=dev-java/java-config-2.0.19-r1", which normaly is set by java-pkg-opt-2.eclass, but is overwritten by the ebuild (as DEPEND is redefined, rather than extended). 2) There is no need of a depend on app-shells/bash, as it is a core package specified in /usr/portage/profile/base/packages. 3) The boost dep should be on >=dev-libs/boost-1.31.0, as it will work just as well as 1.33.1. We shouldn't force an unnessesary upgrade on systems with an current install of 1.31.x or 1.32.x. 4) There is no need whatsoever of setting --with-distro="Gentoo64" on amd64. All patches nessesary for amd64 is applied on all arches (as they, since ages, have been considered stable enough). The wiki is outdated in this regard. Then I have some personal oppinions on some posible improvements as well: Firstly I don't see the benefit of using the outdated internal boost on any arch, so imho we should set --with-system-boost on all arches, not only amd64. Secondly I don't think the branding useflag is very usefull. Imho Gentoo should always brand. Thirdly I don't see the use of giving some static configure swiches directly to ./configure and patch some into distro-configs/Gentoo.conf.in. Pick one and stay with that. (In the long run they should preferably be pushed into upstream distro-configs/Gentoo.conf.in). Will report back when the compile is done
(In reply to comment #22) > Great. Trying to compile it as I'm writing this. While doing so I'm looking > through your ebuilds and found a few things I think is wrong. nice > 1) RDEPEND don't need any specific version of java-config, as openoffice has > it's own java version detection and selection system. However, DEPEND should > have a dep on ">=dev-java/java-config-2.0.19-r1", which normaly is set by > java-pkg-opt-2.eclass, but is overwritten by the ebuild (as DEPEND is > redefined, rather than extended). will fix that > 2) There is no need of a depend on app-shells/bash, as it is a core package > specified in /usr/portage/profile/base/packages. also correct > 3) The boost dep should be on >=dev-libs/boost-1.31.0, as it will work just as > well as 1.33.1. We shouldn't force an unnessesary upgrade on systems with an > current install of 1.31.x or 1.32.x. and another one > 4) There is no need whatsoever of setting --with-distro="Gentoo64" on amd64. > All patches nessesary for amd64 is applied on all arches (as they, since ages, > have been considered stable enough). The wiki is outdated in this regard. Actually there is a difference as it still adds the system-java-bits to all the builds, but we are doing that also now (which I missed initially), so will change that > Then I have some personal oppinions on some posible improvements as well: > Firstly I don't see the benefit of using the outdated internal boost on any > arch, so imho we should set --with-system-boost on all arches, not only amd64. And what would be the practical advantage of that other than adding another dependency to an already very dependency-rich ebuild? (just to be clear: I'm not against it and already have considered it, but I think there has to be a good reasoning) > Secondly I don't think the branding useflag is very usefull. Imho Gentoo should > always brand. That was an often requested feature (by both devs and users), setting the use-flag is easy enough so not going to change this > Thirdly I don't see the use of giving some static configure swiches directly to > ./configure and patch some into distro-configs/Gentoo.conf.in. Pick one and > stay with that. (In the long run they should preferably be pushed into upstream > distro-configs/Gentoo.conf.in). Don't know what you are talking about, all the stuff that is not use-flag or arch-specific is already in Gentoo.conf.in. Or are you talking about the removal of the branding-stuff? I've already taken corrected that upstream, it just was after the release we are using.
being curious, I have been bold (or stupid, I haven't decided yet) and unmasked openoffice-2.0.4_rc1 but encountered some problems: first I tried to emerge with USE="-java -gstreamer branding" but in addition to the dependencies of the package emerge died several times, complaining about missing packages. So I installed the following packages, and at least those problems went away: app-shells/tcsh app-arch/rpm dev-java/ant (even though I was emerging with -java set) now I am stuck with the following error, it remains regardless of any combination of +/-java and +/-firefox as useflags: ******************************************************************** * * * Setting up the build environment variables. * * * ******************************************************************** checking solver path... default configure: creating ./config.status config.status: creating set_soenv Setting up the environment for building OpenOffice.org Setting platform independent values... Setting Linux x86-64 specific values... done Died at ./set_soenv line 776. Checking for prebuilt Mozilla libraries ... not found! If you don't use the system mozilla and also not build the needed mozilla libraries yourself you have to provide the needed files LINUXGCCX{inc,lib,runtime}.zip in moz/zipped/ . These files can be found here: <http://tools.openoffice.org/moz_prebuild/680/>. make: *** [stamp/build] Fehler 1 !!! ERROR: app-office/openoffice-2.0.4_rc1 failed. Call stack: ebuild.sh, line 1543: Called dyn_compile ebuild.sh, line 938: Called src_compile openoffice-2.0.4_rc1.ebuild, line 269: Called die !!! Build failed !!! If you need support, post the topmost build error, and the call stack if relevant. !!! When you file a bug report, please include the following information: GENTOO_VM=blackdown-jdk-1.4.2 CLASSPATH="" JAVA_HOME="/opt/blackdown-jdk-1.4.2.03" JAVACFLAGS="-source 1.4 -target 1.4" COMPILER="" and of course, the output of emerge --info Portage 2.1.1_pre5-r3 (default-linux/amd64/2006.0, gcc-4.1.1/amd64-vanilla, glibc-2.4-r3, 2.6.16-gentoo-r9 x86_64) ================================================================= System uname: 2.6.16-gentoo-r9 x86_64 AMD Athlon(tm) 64 Processor 3000+ Gentoo Base System version 1.12.4 Last Sync: Thu, 24 Aug 2006 14:00:01 +0000 app-admin/eselect-compiler: 2.0.0_rc2-r1 dev-lang/python: 2.3.5, 2.4.3-r1 dev-python/pycrypto: 2.0.1-r5 dev-util/ccache: [Not Present] dev-util/confcache: [Not Present] sys-apps/sandbox: 1.2.18.1 sys-devel/autoconf: 2.13, 2.60 sys-devel/automake: 1.4_p6, 1.5, 1.6.3, 1.7.9-r1, 1.8.5-r3, 1.9.6-r2 sys-devel/binutils: 2.17 sys-devel/gcc-config: 2.0.0_rc1 sys-devel/libtool: 1.5.22 virtual/os-headers: 2.6.16 ACCEPT_KEYWORDS="amd64 ~amd64" AUTOCLEAN="yes" CBUILD="x86_64-pc-linux-gnu" CFLAGS="-march=athlon64 -O2 -pipe" CHOST="x86_64-pc-linux-gnu" CONFIG_PROTECT="/etc /usr/NX/etc /usr/NX/home /usr/kde/3.5/env /usr/kde/3.5/share/config /usr/kde/3.5/shutdown /usr/share/X11/xkb /usr/share/config" CONFIG_PROTECT_MASK="/etc/env.d /etc/env.d/java/ /etc/eselect/compiler /etc/gconf /etc/java-config/vms/ /etc/revdep-rebuild /etc/terminfo /etc/texmf/web2c" CXXFLAGS="-march=athlon64 -O2 -pipe" DISTDIR="/usr/portage/distfiles" FEATURES="autoconfig distlocks metadata-transfer parallel-fetch sandbox sfperms strict" GENTOO_MIRRORS="http://distfiles.gentoo.org http://distro.ibiblio.org/pub/linux/distributions/gentoo" LANG="de_DE@euro" LC_ALL="de_DE@euro" LINGUAS="de" MAKEOPTS="-j2" PKGDIR="/usr/portage/packages" PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress --force --whole-file --delete --delete-after --stats --timeout=180 --exclude='/distfiles' --exclude='/local' --exclude='/packages'" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="/usr/local/portage" SYNC="rsync://rsync.gentoo.org/gentoo-portage" USE="amd64 X alsa apache2 arts avi bash-completion berkdb bitmap-fonts bzip2 cairo cli crypt cups curl dbus dlloader dri dvd dvdr eds elibc_glibc emboss encode exif ffmpeg flac foomaticdb fortran gif gphoto2 gpm gstreamer gtk gtk2 hal imlib input_devices_keyboard input_devices_mouse ipv6 isdnlog java jpeg kde kdeenablefinal kdehiddenvisibility kernel_linux linguas_de lzw lzw-tiff mad mp3 mpeg ncurses nls nodrm nptl nptlonly ntpl nvidia ogg opengl pam pcre pdflib perl png postgres pppd python qt qt3 qt4 quicktime readline reflection sdl session speex spell spl ssl svg tcpd tetex theora tiff truetype truetype-fonts type1-fonts unicode usb userland_GNU video_cards_nvidia vorbis xcomposite xine xml xmms xorg xpm xv xvid xvmc zlib" Unset: CTARGET, EMERGE_DEFAULT_OPTS, INSTALL_MASK, LDFLAGS, PORTAGE_RSYNC_EXTRA_OPTS
(In reply to comment #23) > (In reply to comment #22) > > 3) The boost dep should be on >=dev-libs/boost-1.31.0, as it will work just > > as well as 1.33.1. We shouldn't force an unnessesary upgrade on systems > > with an current install of 1.31.x or 1.32.x. > > and another one Seems you forgot to fix this one when you fixed the rest. > > Then I have some personal oppinions on some posible improvements as well: > > Firstly I don't see the benefit of using the outdated internal boost on any > > arch, so imho we should set --with-system-boost on all arches, not only > > amd64. > > And what would be the practical advantage of that other than adding another > dependency to an already very dependency-rich ebuild? (just to be clear: I'm > not against it and already have considered it, but I think there has to be a > good reasoning) You'll compile and install exactly the same amount of code, but you'll get a more recent version, that additionally can be used by other programs (rather than having to install two copies). Gentoo policy > > Secondly I don't think the branding useflag is very usefull. Imho Gentoo > > should always brand. > > That was an often requested feature (by both devs and users), setting the > use-flag is easy enough so not going to change this OK. Then would you consider a "nobranding" useflag instead (thus defaulting to branding). > > Thirdly I don't see the use of giving some static configure swiches directly > > to ./configure and patch some into distro-configs/Gentoo.conf.in. Pick one > > and stay with that. (In the long run they should preferably be pushed into > > upstream distro-configs/Gentoo.conf.in). > > Don't know what you are talking about, all the stuff that is not use-flag or > arch-specific is already in Gentoo.conf.in. --disable-access --disable-post-install-scripts --enable-hunspell --with-system-hunspell --mandir=/usr/share/man is in the ebuild, while --disable-qadevooo --with-dynamic-xinerama --without-myspell-dicts --with-vendor=\"Gentoo Foundation\" is patched into Gentoo.conf.in. Additional some research into java-pkg-opt-2.eclass shows that we don't have to call java-pkg-opt-2_pkg_setup at all, it is automatically called prior to entering any function in the ebuild when USE=java is set. However, this relies on portage 2.1, so we should realy change the dep on ">=dev-java/java-config-2.0.19-r1" to "${JAVA_PKG_E_DEPEND}", which is set by the eclass to include deps on propper versions of java-config and portage.
(In reply to comment #24) > being curious, I have been bold (or stupid, I haven't decided yet) and unmasked > openoffice-2.0.4_rc1 but encountered some problems: emerge --sync and try again, this should be gone now.
(In reply to comment #25) > > Seems you forgot to fix this one when you fixed the rest. Indeed, fixed > --disable-access --disable-post-install-scripts --enable-hunspell > --with-system-hunspell --mandir=/usr/share/man is in the ebuild, while > --disable-qadevooo --with-dynamic-xinerama --without-myspell-dicts > --with-vendor=\"Gentoo Foundation\" is patched into Gentoo.conf.in. > I think there is quite some misunderstanding here: The first ones are configure flags for ooo-build (which is a sort of wrapper around the original openoffice-build-system to make it easier to use), the second ones are the stuff that is pushed directly to the openoffice-configure-run (and they are not "patched into" they are already upstream, see http://cvs.gnome.org/viewcvs/ooo-build/distro-configs/Gentoo.conf.in?rev=1.8&view=markup ). that's the way ooo-build works... > > Additional some research into java-pkg-opt-2.eclass shows that we don't have to > call java-pkg-opt-2_pkg_setup at all, it is automatically called prior to > entering any function in the ebuild when USE=java is set. However, this relies > on portage 2.1, so we should realy change the dep on > ">=dev-java/java-config-2.0.19-r1" to "${JAVA_PKG_E_DEPEND}", which is set by > the eclass to include deps on propper versions of java-config and portage. > fixed, thanks
(In reply to comment #22) > 3) The boost dep should be on >=dev-libs/boost-1.31.0, as it will work just as > well as 1.33.1. We shouldn't force an unnessesary upgrade on systems with an > current install of 1.31.x or 1.32.x. Did you test that? As long as there are no patched versions of <dev-libs/boost-1.33.1 they will break the build on amd64.
Unmasked it, started the emerge and after a whooping 7h compile openoffice-2.0.4_rc1 emerged & works without any problems here. emerge --info Portage 2.1.1_pre5-r3 (default-linux/amd64/2006.0, gcc-4.1.1, glibc-2.4-r3, 2.6.17-suspend2-r4Dudebox-Edition x86_64) ================================================================= System uname: 2.6.17-suspend2-r4Dudebox-Edition x86_64 AMD Athlon(tm) 64 Processor 3200+ Gentoo Base System version 1.12.4 Last Sync: Fri, 25 Aug 2006 03:50:01 +0000 ccache version 2.4 [enabled] app-admin/eselect-compiler: [Not Present] dev-lang/python: 2.4.3-r1 dev-python/pycrypto: 2.0.1-r5 dev-util/ccache: 2.4-r2 dev-util/confcache: [Not Present] sys-apps/sandbox: 1.2.18.1 sys-devel/autoconf: 2.13, 2.60 sys-devel/automake: 1.4_p6, 1.5, 1.6.3, 1.7.9-r1, 1.8.5-r3, 1.9.6-r2 sys-devel/binutils: 2.17 sys-devel/gcc-config: 1.3.13-r3 sys-devel/libtool: 1.5.22 virtual/os-headers: 2.6.16 ACCEPT_KEYWORDS="amd64 ~amd64" AUTOCLEAN="yes" CBUILD="x86_64-pc-linux-gnu" CFLAGS="-march=k8 -msse3 -Os -pipe" CHOST="x86_64-pc-linux-gnu" CONFIG_PROTECT="/etc /usr/kde/3.5/env /usr/kde/3.5/share/config /usr/kde/3.5/shutdown /usr/share/X11/xkb /usr/share/config" CONFIG_PROTECT_MASK="/etc/env.d /etc/env.d/java/ /etc/gconf /etc/java-config/vms/ /etc/revdep-rebuild /etc/terminfo" CXXFLAGS="-march=k8 -msse3 -Os -pipe" DISTDIR="/usr/portage/distfiles" FEATURES="autoconfig buildpkg ccache distlocks metadata-transfer multilib-strict parallel-fetch sandbox sfperms strict userfetch userpriv usersandbox" GENTOO_MIRRORS="ftp://ftp.wh2.tu-dresden.de/pub/mirrors/gentoo ftp://linux.rz.ruhr-uni-bochum.de/gentoo-mirror/ ftp:///ftp-stud.fht-esslingen.de/pub/Mirrors/gentoo/" LDFLAGS="-Wl,-O1" LINGUAS="de" MAKEOPTS="-j2" PKGDIR="/usr/portage/packages" PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress --force --whole-file --delete --delete-after --stats --timeout=180 --exclude='/distfiles' --exclude='/local' --exclude='/packages'" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" SYNC="rsync://server/gentoo-portage" USE="amd64 X a52 aac aalib alsa arts avi berkdb bitmap-fonts bzip2 cairo cli crypt cups dbus directfb dlloader dri dts dvd dvdr dvdread eds elibc_glibc emboss encode firefox foomaticdb fortran gif gnutls gpm gstreamer gtk gtk2 hal imap imlib input_devices_keyboard input_devices_mouse isdnlog java jpeg kde kdeenablefinal kdehiddenvisibility kernel_linux libcaca linguas_de lzw lzw-tiff mad mp3 mp4 mpeg musepack musicbrainz mysql ncurses nls nptl nptlonly nsplugin ogg opengl pam pcre pdf pdflib perl png pppd python qt3 qt4 quicktime readline reflection sasl sdl session spell spl ssl svg tcpd tiff truetype truetype-fonts type1-fonts unicode usb userland_GNU video_cards_radeon visualization vorbis wxwindows x264 xml xorg xpm xv xvid zlib" Unset: CTARGET, EMERGE_DEFAULT_OPTS, INSTALL_MASK, LANG, LC_ALL, PORTAGE_RSYNC_EXTRA_OPTS, PORTDIR_OVERLAY
(In reply to comment #26) > (In reply to comment #24) > > being curious, I have been bold (or stupid, I haven't decided yet) and > > unmasked openoffice-2.0.4_rc1 but encountered some problems: > > emerge --sync and try again, this should be gone now. > OpenOffice.org compiled without error and works just fine.
Compiles and runs on this machine, however, the about dialog shows that it's Openoffice 2.0.3, whereas it's actually 2.0.4. All the other ATs and devs so far have had the same experience. Portage 2.1.1_pre5-r3 (default-linux/amd64/2006.1, gcc-4.1.1/amd64-vanilla, glibc-2.4-r3, 2.6.17-gentoo-r2 x86_64) ================================================================= System uname: 2.6.17-gentoo-r2 x86_64 AMD Sempron(tm) Processor 3300+ Gentoo Base System version 1.12.4 Last Sync: Fri, 25 Aug 2006 07:50:01 +0000 app-admin/eselect-compiler: 2.0.0_rc2-r1 dev-lang/python: 2.4.3-r1 dev-python/pycrypto: 2.0.1-r5 dev-util/ccache: [Not Present] dev-util/confcache: [Not Present] sys-apps/sandbox: 1.2.18.1 sys-devel/autoconf: 2.13, 2.60 sys-devel/automake: 1.4_p6, 1.5, 1.6.3, 1.7.9-r1, 1.8.5-r3, 1.9.6-r2 sys-devel/binutils: 2.17 sys-devel/gcc-config: 1.3.13-r3 sys-devel/libtool: 1.5.22 virtual/os-headers: 2.6.16 ACCEPT_KEYWORDS="amd64 ~amd64" AUTOCLEAN="yes" CBUILD="x86_64-pc-linux-gnu" CFLAGS="-O2 -march=k8 -pipe" CHOST="x86_64-pc-linux-gnu" CONFIG_PROTECT="/etc /usr/kde/3.5/env /usr/kde/3.5/share/config /usr/kde/3.5/shutdown /usr/share/X11/xkb /usr/share/config" CONFIG_PROTECT_MASK="/etc/env.d /etc/env.d/java/ /etc/eselect/compiler /etc/gconf /etc/java-config/vms/ /etc/revdep-rebuild /etc/splash /etc/terminfo" CXXFLAGS="-O2 -march=k8 -pipe" DISTDIR="/usr/distfiles" FEATURES="autoconfig ccache collision-protect distlocks metadata-transfer multilib-strict sandbox sfperms strict" GENTOO_MIRRORS="http://distfiles.gentoo.org http://distro.ibiblio.org/pub/linux/distributions/gentoo" LINGUAS="" MAKEOPTS="-j2" PKGDIR="/usr/portage/packages" PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress --force --whole-file --delete --delete-after --stats --timeout=180 --exclude='/distfiles' --exclude='/local' --exclude='/packages'" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/portage" PORTDIR_OVERLAY="/usr/local/overlays/gentopia /usr/local/overlays/personal /usr/local/overlays/xgl-coffee" SYNC="rsync://rsync5.nl.gentoo.org/gentoo-portage" USE="amd64 X alsa arts berkdb bitmap-fonts cairo cdr cli crypt cups dbus dlloader dri dvd dvdread elibc_glibc ethereal ffmpeg flac flash fortran ftp gdbm gnome gpm gstreamer gtk gtk2 gtkhtml hal imagemagick input_devices_keyboard input_devices_mouse ipv6 isdnlog jabber java javascript jpeg kde kdeenablefinal kernel_linux libg++ mono mp3 mpeg msn ncurses nls nptl nptlonly nsplugin ogg opengl pam pcre pda pdf perl png ppds pppd python qt qt3 qt4 quicktime readline reflection samba session spl ssl tcpd tiff truetype-fonts type1-fonts udev unicode usb userland_GNU userlocales vcd video_cards_nv video_cards_nvidia video_cards_vesa vorbis xine xinerama xml xmms xorg xv xvid zlib" Unset: CTARGET, EMERGE_DEFAULT_OPTS, INSTALL_MASK, LANG, LC_ALL, LDFLAGS, PORTAGE_RSYNC_EXTRA_OPTS
(In reply to comment #31) > Compiles and runs on this machine, however, the about dialog shows that it's > Openoffice 2.0.3, whereas it's actually 2.0.4. Just has not been changed by upstream until now, still a pre-release, so...
It works indeed. It is currently p.masked because it is a rc and not a release, but from amd64's side we're more than happy. Thanks everybody for their efforts.
(In reply to comment #28) > (In reply to comment #22) > > 3) The boost dep should be on >=dev-libs/boost-1.31.0, as it will work just as > > well as 1.33.1. We shouldn't force an unnessesary upgrade on systems with an > > current install of 1.31.x or 1.32.x. > > Did you test that? As long as there are no patched versions of > <dev-libs/boost-1.33.1 they will break the build on amd64. > I'm going to rise the dependency again, thanks for the pointer
just a note... this builds for me in both ~x86 and ~amd64 on a dual-core dual opteron system with MAKEOPTS="-j6" enabled... it would be nice to get others to test this since it drastically reduces the merge time... * app-office/openoffice Sat Sep 23 12:34:06 2006 >>> app-office/openoffice-2.0.4_rc2 merge time: 1 hour, 29 minutes and 32 seconds. with the filter in the ebuild left to set it at -j1 the build takes over 4 hours
Any reason why openoffice-2.0.4 ebuild still contains 'use amd64 && echo "--with-system-boost" >> ${CONFFILE}' and 'echo "--with-system-icu" >> ${CONFFILE}', cosidering that both "--with-system-boost" and "--with-system-icu" are in upstream Gentoo.conf.in now?
Created attachment 99589 [details] bzipped /var/log/portage/app-office:openoffice-2.0.4:20061013-173156.log emerge =openoffice-2.0.4 fails on my ~amd64 :-( root amd64 ~ # emerge --info Portage 2.1.2_pre2-r9 (default-linux/amd64/2006.1/desktop, gcc-4.1.1, glibc-2.5-r0, 2.6.18! x86_64) ================================================================= System uname: 2.6.18! x86_64 AMD Athlon(tm) 64 Processor 3000+ Gentoo Base System version 1.13.0_alpha1 Last Sync: Fri, 13 Oct 2006 15:30:01 +0000 app-admin/eselect-compiler: [Not Present] dev-java/java-config: 1.3.7, 2.0.30 dev-lang/python: 2.5-r1 dev-python/pycrypto: 2.0.1-r5 dev-util/ccache: [Not Present] dev-util/confcache: [Not Present] sys-apps/sandbox: 1.2.20_alpha2 sys-devel/autoconf: 2.13, 2.60 sys-devel/automake: 1.4_p6, 1.5, 1.6.3, 1.7.9-r1, 1.8.5-r3, 1.9.6-r2 sys-devel/binutils: 2.17.50.0.5 sys-devel/gcc-config: 1.3.13-r4 sys-devel/libtool: 1.5.22 virtual/os-headers: 2.6.17-r1 ACCEPT_KEYWORDS="amd64 ~amd64" AUTOCLEAN="yes" CBUILD="x86_64-pc-linux-gnu" CFLAGS="-O2 -march=k8 -fomit-frame-pointer -pipe" CHOST="x86_64-pc-linux-gnu" CONFIG_PROTECT="/etc /usr/kde/3.5/env /usr/kde/3.5/share/config /usr/kde/3.5/shutdown /usr/share/X11/xkb /usr/share/config" CONFIG_PROTECT_MASK="/etc/env.d /etc/env.d/java/ /etc/gconf /etc/java-config/vms/ /etc/revdep-rebuild /etc/sandbox.d /etc/terminfo" CXXFLAGS="-O2 -march=k8 -fomit-frame-pointer -pipe" DISTDIR="/portage/distfiles" FEATURES="autoconfig distlocks metadata-transfer parallel-fetch sandbox sfperms strict" GENTOO_MIRRORS="http://mirror.gentoo.sk/pub http://distfiles.gentoo.org/" LANG="en_US.UTF-8" LC_ALL="en_US.UTF-8" LINGUAS="cs" MAKEOPTS="-j3" PKGDIR="/portage/packages" PORTAGE_RSYNC_EXTRA_OPTS="--exclude-from=/etc/portage/rsync_excludes" PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress --force --whole-file --delete --delete-after --stats --timeout=180 --exclude=/distfiles --exclude=/local --exclude=/packages" PORTAGE_TMPDIR="/20GB/var/tmp" PORTDIR="/portage" PORTDIR_OVERLAY="/overlay /usr/local/portage/xgl-coffee /usr/local/portage/berkano-overlay" SYNC="rsync://rsync.gentoo.org/gentoo-portage" USE="amd64 X a52 aac acl acpi alsa apache2 asf avi bitmap-fonts bzip2 cairo cdparanoia cdr cli css cups curl curlwrappers custom-cflags cvs dbus dga dts dvd dvdr dvdread elibc_glibc emboss enca encode fbcon ffmpeg firefox flash ftp fuse gcc64 gd gdbm gif gpm hal iconv id3 id3v2 imagemagick imlib input_devices_keyboard input_devices_mouse isdnlog jabber javascript jpeg kde kdeenablefinal kdehiddenvisibility kernel_linux lame lcms libg++ linguas_cs mad matroska mime mp3 mpeg mplayer mpm-worker ncurses nls nowebdav nptl nptlonly nsplugin nvidia ogg oggvorbis openal opengl pic png posix ppds pppd print python qt qt3 qt4 quicktime readline reflection sdl session sockets spl sqlite ssl subversion svg symlink theora threads tiff truetype truetype-fonts type1-fonts udev unicode usb userland_GNU userlocales video_cards_nvidia vorbis wma xcomposite xine xml xml2 xorg xpm xrandr xv xvid zlib" Unset: CTARGET, EMERGE_DEFAULT_OPTS, INSTALL_MASK, LDFLAGS
(In reply to comment #36) > Any reason why openoffice-2.0.4 ebuild still contains 'use amd64 && echo > "--with-system-boost" >> ${CONFFILE}' and 'echo "--with-system-icu" >> > ${CONFFILE}', cosidering that both "--with-system-boost" and > "--with-system-icu" are in upstream Gentoo.conf.in now? > Yes, me being stupid ;) Actually I put this upstream some weeks ago but as the RC-period took so long and I'm quite stuffed with real life work atm, I just forgot about that. Thanks for pointing this out, will remove it in a few minutes ;)
(In reply to comment #37) > emerge =openoffice-2.0.4 fails on my ~amd64 :-( amd64 version of sun-jdk-1.5 has a broken libzip.so that conflicts with system libz namespace.
(In reply to comment #39) > amd64 version of sun-jdk-1.5 has a broken libzip.so that conflicts with > system libz namespace. So it's possible to compile it (on amd64) now? Any workaround to this bug?
AFAIK all but >=sun-jdk-1.5 do fine. sun-jdk-1.6 fails on hsqldb because of some sql api changes.
(In reply to comment #41) > AFAIK all but >=sun-jdk-1.5 do fine. sun-jdk-1.6 fails on hsqldb because of > some sql api changes. So I can just run: # java-config -S1 Now using blackdown-jdk-1.4.2 as your generation-2 system JVM and then it should emerge fine?
(In reply to comment #42) > (In reply to comment #41) > > AFAIK all but >=sun-jdk-1.5 do fine. sun-jdk-1.6 fails on hsqldb because of > > some sql api changes. > > So I can just run: > > # java-config -S1 > Now using blackdown-jdk-1.4.2 as your generation-2 system JVM > > and then it should emerge fine? No, it will still use "best" availible VM that matches the virtuals in DEPEND. However, "JAVA_PKG_FORCE_VM=blackdown-jdk-1.4.2 emerge openoffice" will work. (As will setting the vm as you sugested AND emergeing with "JAVA_PKG_ALLOW_VM_CHANGE=no emerge openoffice"). I tried to set DEPEND to "java? ( || ( =virtual/jdk-1.4* !amd64? ( =virtual/jdk-1.5* ) )", however it seams the "best match" algorith ignores useflags, so this didn't work. Best solution I can come up with right now is to force blackdown-jdk-1.4.2 in the ebuild for amd64 by setting DEPEND to "java? ( !amd64? ( || ( =virtual/jdk-1.4* =virtual/jdk-1.5* ) amd64? =dev-java/blackdown-jdk-1.4.2* )" and put "use amd64 && export JAVA_PKG_FORCE_VM=blackdown-jdk-1.4.2" in the ebuild. Not very nice, but unless the root problem is fixed upstream (at sun) or the java team starts honoring useflags in depend-java-query I can't think of anything better.
# JAVA_PKG_FORCE_VM=blackdown-jdk-1.4.2 emerge openoffice fails on make install... it's like *** glibc detected *** /20GB/var/tmp/portage/app-office/openoffice-2.0.4/work/ooo/build/OOO_2_0_4/solver/680/unxlngx6.pro/bin/regcomp.bin: double free or corruption (out): 0x00002b42b7812120 *** ======= Backtrace: ========= /lib/libc.so.6[0x2b42b4b5a5dd] /lib/libc.so.6(cfree+0x76)[0x2b42b4b5bc16] /usr/lib/libpython2.5.so.1.0[0x2b42b72bd8e3] /usr/lib/libpython2.5.so.1.0[0x2b42b72a2ec7] ... and so on... I'm almost sure it's caused by new Python 2.5 :-( I'll try it again with 2.4.3, because 2.5 seems to be very unstable at the moment.
Great, with Python 2.4.3-r4 it finally compiles! :-) Thanks for help.
