| Summary: | Can't emerge x11-base/xorg-x11-6.7.0-r2 because it forces HasMotif to YES | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | Sebastian Bergmann (RETIRED) <sebastian> |
| Component: | New packages | Assignee: | Gentoo X packagers <x11> |
| Status: | RESOLVED FIXED | ||
| Severity: | normal | CC: | askwar, brainysmurf, corsair, gentoo-bugs2, hans.w.wurst, john, leroutier, link.dupont, marduk, pgw99, psmurf, quaggamail, renegart, sebastian_ml, uberlord, vermyndax, weeve, ysgard |
| Priority: | High | ||
| Version: | unspecified | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
| Attachments: |
make log, from where it breaks
Patch to check for motif Simplier patch to delete echo |
||
|
Description
Sebastian Bergmann (RETIRED)
2004-07-16 03:45:08 UTC
same for me, fails with the same error in : making all in lib/dpstk... make[3]: Entering directory `/usr/portage/tmp/portage/xorg-x11-6.7.0-r2/work/xc/lib/dpstk' rm -f FontSample.o unshared/FontSample.o gcc -c -ansi -pedantic -Wno-return-type -w -I/usr/X11R6/include -I../.. -I../../exports/include -Dlinux -D__i386__ -D_POSIX_C_SOURCE=199309L ... use flags : -3dfx +3dnow -cjk -debug +dlloader +doc -ipv6 +mmx +nls +pam -sdk +sse -static i have -r1 installed, working ok, so, it is a new bug in -r2 Created attachment 35563 [details]
make log, from where it breaks
make > xorg-x11-6.7.0-r2.log 2>&1
Same here, relevant info for s and g's:
Portage 2.0.50-r9 (default-x86-2004.0, gcc-3.3.4, glibc-2.3.4.20040619-r0, 2.6.8-rc1)
=================================================================
System uname: 2.6.8-rc1 i686 Intel(R) Pentium(R) M processor 1500MHz
Gentoo Base System version 1.5.1
Autoconf: sys-devel/autoconf-2.59-r4
Automake: sys-devel/automake-1.8.5-r1
ACCEPT_KEYWORDS="x86 ~x86"
AUTOCLEAN="yes"
CFLAGS="-O2 -march=pentium3 -fomit-frame-pointer -pipe"
CHOST="i686-pc-linux-gnu"
COMPILER="gcc3"
{snip}
CXXFLAGS="-O2 -march=pentium3 -fomit-frame-pointer -pipe"
DISTDIR="/usr/portage/distfiles"
FEATURES="autoaddcvs ccache sandbox"
GENTOO_MIRRORS="ftp://ftp.gtlib.cc.gatech.edu/pub/gentoo http://mirrors.tds.net/gentoo ftp://mirrors.tds.net/gentoo"
MAKEOPTS="-j2"
{snip}
SYNC="rsync://192.168.1.3/gentoo-portage"
USE="X alsa apm avi berkdb cdr cdrw crypt cups dvd encode flac foomaticdb gdbm gif gnome gpm gtk gtk2 imlib jpeg libg++ libwww mad mikmod motif mozilla mpeg ncurses oggvorbis opengl oss pam pdflib perl png python quicktime readline sdl slang spell ssl svga tcpd truetype video_cards_radeon x86 xine xml2 xv zlib"
Has anyone managed to get the failure on gcc 3.3.3? This looks a bit like stricter checking in the latest compiler, so it would be interesting to get feedback from someone still on the older version. indeed looks like stricter checking form the looks of it, so i expected gcc-3.4.1-r1 to bail too, it didnt, i have neither gcc-3.3.3 nor gcc-3.3.4 to check this out more, but it merged for me, ill provide my emerge info, i dont have much time to check my and his USE flags, ill do that tonight, if someone finds somehting before then, cool, but for now:
Fri Jul 16 12:28:01 2004 >>> x11-base/xorg-x11-6.7.0-r2
merge time: 19 minutes and 44 seconds.
genlop -t xorg-x11 says it merged fine
X Protocol Version 11, Revision 0, Release 6.7
Build Operating System: Linux 2.6.8-rc1 i686 [ELF]
Current Operating System: Linux nullzone 2.6.8-rc1 #13 Wed Jul 14 11:50:33 CDT 2004 i686
Build Date: 16 July 2004
build date shows it is the one i just build that im on right now
Portage 2.0.51_pre12 (default-x86-2004.0, gcc-3.4.1, glibc-2.3.4.20040605-r0, 2.6.8-rc1)
=================================================================
System uname: 2.6.8-rc1 i686 AMD Athlon(tm) XP 3200+
Gentoo Base System version 1.5.1
ccache version 2.3 [enabled]
Autoconf: sys-devel/autoconf-2.59-r4
Automake: sys-devel/automake-1.8.5-r1
Binutils: sys-devel/binutils-2.14.90.0.8-r1
ACCEPT_KEYWORDS="x86 ~x86"
AUTOCLEAN="yes"
CFLAGS="-O2 -march=athlon-xp -fomit-frame-pointer -pipe -mfpmath=sse"
CHOST="i686-pc-linux-gnu"
COMPILER="gcc3"
CONFIG_PROTECT="/etc /usr/X11R6/lib/X11/xkb /usr/kde/2/share/config /usr/kde/3.2/share/config /usr/kde/3/share/config /usr/lib/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/ /var/qmail/control"
CONFIG_PROTECT_MASK="/etc/gconf /etc/terminfo /etc/env.d"
CXXFLAGS="-O2 -march=athlon-xp -fomit-frame-pointer -pipe -mfpmath=sse"
DISTDIR="/usr/portage/distfiles"
FEATURES="autoaddcvs ccache cvs sandbox"
GENTOO_MIRRORS="ftp://ftp.ndlug.nd.edu/pub/gentoo/ ftp://ftp.ussg.iu.edu/pub/linux/gentoo http://gentoo.ccccom.com"
MAKEOPTS="-j2"
PKGDIR="/usr/portage/packages"
PORTAGE_TMPDIR="/var/tmp"
PORTDIR="/usr/portage"
PORTDIR_OVERLAY=""
SYNC="cvs://morfic@cvs.gentoo.org:/var/cvsroot"
USE="3dnow X alsa apm arts avi berkdb bonobo cdr crypt cups devfs devfs26 dga directfb divx4linux dv dvd dvdr encode escreen esd etwin fbcon fbdev ffmpeg foomaticdb freetype gdbm ggi gif gnome gphoto2 gpm gtk gtk2 gtkhtml ieee1394 imlib java jpeg kde libg++ libwww mad mikmod mmx motif mozilla mpeg mpeg4 mplayer mysql nas ncurses nls nptl oggvorbis opengl pam pdflib perl png python qt quicktime radeon readline sdl silverxp slang spell sqlite sse ssl svga tcpd truetype x86 xine xinerama xml xml2 xmms xv xvid zlib"
if needed ill look into this more later, now ill go with a new blade for my reciprocating saw and chop more trees
same problem: note 3.3.3 Portage 2.0.51_pre13 (!/usr/portage/profiles/default-x86-1.4, gcc-3.3.3, glibc-2.3.4.20040619-r0, 2.6.7 i686 Pentium III (Coppermine)) ================================================================= System uname: 2.6.7 i686 Pentium III (Coppermine) Gentoo Base System version 1.5.1 distcc 2.16 i686-pc-linux-gnu (protocols 1 and 2) (default port 3632) [enabled] ccache version 2.3 [enabled] Autoconf: sys-devel/autoconf-2.59-r4 Automake: sys-devel/automake-1.8.5-r1 Binutils: sys-devel/binutils-2.14.90.0.8-r1 ACCEPT_KEYWORDS="x86 ~x86" AUTOCLEAN="yes" CFLAGS="-march=pentium3 -O2 -pipe -fomit-frame-pointer" CHOST="i686-pc-linux-gnu" COMPILER="gcc3" CONFIG_PROTECT="/etc /usr/kde/2/share/config /usr/kde/3/share/config /usr/share/config /var/qmail/control" CONFIG_PROTECT_MASK="/etc/gconf /etc/terminfo /etc/env.d" CXXFLAGS="-march=pentium3 -O2 -pipe -fomit-frame-pointer" DISTDIR="/storage/distfiles" FEATURES="autoaddcvs buildpkg ccache collision-protect cvs digest distcc fixpackages notitles sandbox strict userpriv usersandbox" GENTOO_MIRRORS="http://ftp.belnet.be/mirror/rsync.gentoo.org/gentoo/ http://ftp.snt.utwente.nl/pub/os/linux/gentoo http://ftp.easynet.nl/mirror/gentoo/ http://ftp.gentoo.skynet.be/pub/gentoo/" MAKEOPTS="-j4" PKGDIR="/usr/portage/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/home/keanu/keep/gentoo-x86" PORTDIR_OVERLAY="/usr/local/portage" SYNC="rsync://bismuth.skynet.be/gentoo-portage" USE="aalib alsa apache2 apm avi berkdb crypt cscope dedicated directfb divx4linux encode fbcon foomaticdb gd gdbm gif gpm gtk2 guile imap imlib ipv6 jabber jpeg libg++ libwww mad maildir mikmod mmx motif mpeg msn mysql ncurses nls oav oggvorbis oscar oss pam pdflib perl pic png postgres python quicktime readline slang slp spell sse ssl svga tcpd tiff usb x86 xml xml2 xvid zlib Too add further confusion to the mix, it compiled perfectly for me Portage 2.0.51_pre13 (!/usr/portage/profiles/default-x86-2004.0, gcc-3.3.4, glibc-2.3.4.20040619-r0, 2.6.7 i686 AMD Athlon(tm) 64 Processor 3200+) ================================================================= System uname: 2.6.7 i686 AMD Athlon(tm) 64 Processor 3200+ Gentoo Base System version 1.5.1 Autoconf: sys-devel/autoconf-2.59-r4 Automake: sys-devel/automake-1.8.5-r1 Binutils: sys-devel/binutils-2.14.90.0.8-r1 ACCEPT_KEYWORDS="x86 ~x86" AUTOCLEAN="yes" CFLAGS="-O2 -pipe -fomit-frame-pointer -march=athlon-xp -msse -msse2 -mmmx" CHOST="i686-pc-linux-gnu" COMPILER="gcc3" CONFIG_PROTECT="/etc /usr/X11R6/lib/X11/xkb /usr/kde/2/share/config /usr/kde/3.2/share/config /usr/kde/3/share/config /usr/lib/mozilla/defaults/pref /usr/share/config /var/qmail/control" CONFIG_PROTECT_MASK="/etc/gconf /etc/terminfo /etc/env.d" CXXFLAGS="-O2 -pipe -fomit-frame-pointer -march=athlon-xp -msse -msse2 -mmmx" DISTDIR="/work/distfiles" FEATURES="ccache cvs fixpackages sandbox userpriv" GENTOO_MIRRORS="http://mirror.aarnet.edu.au/pub/gentoo http://mirror.gentoo.gr.jp http://open-systems.ufl.edu/mirrors/gentoo http://gentoo.seren.com/gentoo" MAKEOPTS="-j2" PKGDIR="/usr/portage/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/work/portage" PORTDIR_OVERLAY="/work/overlay" SYNC="rsync://rsync.gentoo.org/gentoo-portage" USE="X acpi alsa avi bonobo cdr crypt cups dga dvd encode evo fbcon foomaticdb gdbm gif gphoto2 gtk gtk2 gtkhtml guile imagemagick imlib java jpeg kde libg++ libwww mad mikmod motif mozilla mpeg ncurses oggvorbis opengl oss pam pdflib perl png python qt quicktime readline samba sdl slang spell ssl svga tcltk tcpd truetype usb x86 xml xml2 xmms xv zlib" I thought it mightve been portage versions for a minute but no.. Please attach emerge -vpu xorg-x11 From the ebuild: # 9385_all_6.7.0-motif-allow-fhs-and-default-off.patch from upstream # turns off HasMotif YES default on Linux and BSDs, so we need this echo "#define HasMotif YES" >> config/cf/host.def but xorg-x11 doesn't depend on motif and will fail if it's not installed. Interestingly enough openmotif depends on X, damn recursion. I'm on GCC-3.3.3 and I get the exact same crash in my build... For those of you with xorg-x11 allready installed you can just emerge openmotif emerge -u xorg-x11 That will let you upgrade, but im sure thats not helpful if you're doing a fresh install. Created attachment 35618 [details, diff]
Patch to check for motif
This patch checks for the motif use flag. It then checks to see if openmotif
has been installed. If both of these are true then Has Motif gets set to yes.
If not, then it is set to no. This works for me as a -motif person (the ebuild
emerges fine) and I believe it will take into account people who have openmotif
installed. It is some pretty simple if then statements but should still be
tested , especially by people a) who have the motif flag set but openmotif
isn't yet installed and b) people with openmotif.
Its also possible to just take the motif USE flag out of the equation all
together and just see if openmotif is there but I leave that up to the powers
that be.
Jay, your patch makes hella sense -- I'll have a run through on both my boxes (one motif-less, one motif-full) and get it tested (and likely, committed). Thanks :) I merged openmotif now and after that xorg-x11-6.7.0-r2 merged without problems. Well, the patch works, but will break binary package building. If the package is build on a system with USE="motif" and openmotif installed, merging it on a system which has openmotif not installed will break the package. Portage doesn't know that it needs openmotif as a dependency. But currently I don't see a clean way around the circular dependency, except diabling motif support in xorg-x11. *** Bug 57383 has been marked as a duplicate of this bug. *** Good to hear the patch is at least working for some. As to the binary packages issue: wouldn't this have been a problem on -r1 and before when I assume X automatically figured out #HasMotif? I suppose what goes along with that is does #HasMotif make OpenMotif a Runtime dependency - ie, do the packages actually not run if motif isn't there? I generally run -motif systems so I haven't been able to test that.
As for GRP packages, it is probably easy enough to put the patch in something like:
if use !bindist
if use motif ... blah blah
I suppose someone really has to want this flag anyway as they're going to have to build X, build OpenMotif, and rebuild X to get it. Maybe a larger warning einfo is in order for people who have HasMotif turned on so they're warned about using these packages elsewhere.
that patch fixed it for me too *** Bug 57398 has been marked as a duplicate of this bug. *** *** Bug 57409 has been marked as a duplicate of this bug. *** Maybe I am stupid or something, but that patch really do not seem to 'turns off HasMotif YES default on Linux and BSDs', but just add some extra bits _if_ and _only_if_ you already had '#define HasMotif YES' in your host.def. So that comment + change is really totally bogus. While I do admit that Jay's change might be nice for the motif users, you can also really just consider in removing the comment + line. Those who doubt: --- nosferatu work # grep HasMotif patch/9385_all_6.7.0-motif-allow-fhs-and-default-off.patch +#if HasMotif +#endif /* HasMotif */ #if HasMotif +#endif /* HasMotif */ nosferatu work # --- As you should notice - no change in what HasMotif was defined as ... --- nosferatu work # grep HasMotif xc/config/cf/* xc/config/cf/Imake.tmpl:#ifndef HasMotif xc/config/cf/Imake.tmpl:#define HasMotif NO xc/config/cf/Imake.tmpl.orig:#ifndef HasMotif xc/config/cf/Imake.tmpl.orig:#define HasMotif NO xc/config/cf/X11.tmpl:#if HasMotif xc/config/cf/X11.tmpl:#endif /* HasMotif */ xc/config/cf/host.def:#define HasMotif YES xc/config/cf/lynx.cf:/* define HasMotif to YES for LynxOS with Motif xc/config/cf/lynx.cf:#ifndef HasMotif xc/config/cf/lynx.cf:# define HasMotif NO xc/config/cf/lynx.cf:#if HasMotif nosferatu work # --- But here we can notice that it _was_ turned off by default for _all_ archs ... Interesting. Well then I think it is probably better just to subtract those lines and comment out of the -r2 ebuild and not use my check motif patch. I was trying to return things to the way I thought they worked prior to -r2 but if #HasMotif just defaults to off, it probably doesn't make much sense to use the check motif patch. I think that just adds another thing that could possibly break (especially packages as was mentioned). Created attachment 35660 [details, diff]
Simplier patch to delete echo
This is an alternative patch thats probably much simpler. It deletes the
comment about 9385 and the echo into HasMotif. I don't know that it really
obsoletes the former patch (they basically have the same effect for most
people), but this is much less of a headache.
Anyway, people seem to be using/pointing to the other patch just to get
xorg-x11-6.7.0-r2 to build and this would seem to be simplier for them.
There are quite a few circular dependency problems with xlibs needed by some package, but xlibs are only in xfree/xorg, and the said package depends on xfree/xorg ... But deliberately removing motif from the build isnt the right fix, for the time being it should be (if doing a fresh merge of X) merge xorg without motif, then remerge with motif if wanted... If not doing a fresh merge of X well it doesnt matter we could just have a motif? ( x11-libs/openmotif ) depend. The first patch is a lot closer to doing that IMO I don't really see the issue, as AFAIK, you never had a build in Gentoo where XFree or XOrg was built against motif. If there is outstanding bugs with it requested, sure, maybe look at it, but if not, why complicate things? *** Bug 57512 has been marked as a duplicate of this bug. *** O.M.G. Sorry, why does this stay in portage? Can't it be removed. I think this only causes confusion. Sebastian *** Bug 57607 has been marked as a duplicate of this bug. *** Papa Smurf ?? Anyhow, commented it for now until Donnie decides what he wants to do ... I'm back from vacation and looking at this right now. If what appears to be the case actually is, I was deceived by an upstream committer (not purposefully, of course =). I agree with comment #25. I was attempting to fix a problem that didn't exist upstream until shortly AFTER 6.7.0 came out, and I failed to notice that. If someone wants motif support for whatever reason, please file a new enhancement bug and justify it. Thanks for bearing with me, and thanks for helping out, Martin and others who had ideas. |