[gentoo-user] F77 Provided through /etc/portage/env/*
Hi all! I have a little nasty problem. I am using a couple of fortran progs and not all are happy with ifort but some benefit allot of its use. As I am lazy and dont want to change my make.conf all the time I provided the F77, FC and FLAGS for those packages which support ifort through /etc/portage/env/categorie/package. This worked for a long time, but now some thing has changed and I dont know what. The variables are still passed to the emerge but the F77 and FC isnt used by emake any more. See following example: ifort is chosen by the fortran.eclass and configure uses it as well. But in the make part only the ifort spezific FFLAGS are use together with gfortran. Emerging (1 of 1) sci-chemistry/shelx-20060317 to / * shelx-20060317.tgz RMD160 SHA1 SHA256 size ;-) ... [ ok ] * checking ebuild checksums ;-) ... [ ok ] * checking auxfile checksums ;-) ... [ ok ] * checking miscfile checksums ;-) ... [ ok ] * checking shelx-20060317.tgz ;-) ... [ ok ] * You need one of these Fortran Compilers: ifc gfortran * Installed are: ifort gfortran * Using ifort Unpacking source... Unpacking shelx-20060317.tgz to /var/tmp/portage/sci-chemistry/shelx-20060317/work * Applying 20060317-autotool.patch ... [ ok ] * Applying 20060317-gfortran.patch ... [ ok ] * Running eautoreconf in '/var/tmp/portage/sci-chemistry/shelx-20060317/work/unix' ... * Running aclocal ... [ ok ] * Running autoconf ... [ ok ] * Running automake --add-missing --copy --foreign ... [ ok ] Source unpacked. Compiling source in /var/tmp/portage/sci-chemistry/shelx-20060317/work/unix ... ./configure --prefix=/usr --host=i686-pc-linux-gnu --mandir=/usr/share/man --infodir=/usr/share/info --datadir=/usr/share --sysconfdir=/etc --localstatedir=/var/lib FC=ifort --build=i686-pc-linux-gnu checking for a BSD-compatible install... /usr/bin/install -c checking whether build environment is sane... yes checking for a thread-safe mkdir -p... /bin/mkdir -p checking for gawk... gawk checking whether make sets $(MAKE)... yes checking for i686-pc-linux-gnu-gfortran... ifort checking for Fortran compiler default output file name... a.out checking whether the Fortran compiler works... yes checking whether we are cross compiling... no checking for suffix of executables... checking for suffix of object files... o checking whether we are using the GNU Fortran compiler... no checking whether ifort accepts -g... yes checking for i686-pc-linux-gnu-g77... no checking for i686-pc-linux-gnu-xlf... no checking for i686-pc-linux-gnu-f77... no checking for i686-pc-linux-gnu-frt... no checking for i686-pc-linux-gnu-pgf77... no checking for i686-pc-linux-gnu-cf77... no checking for i686-pc-linux-gnu-fort77... no checking for i686-pc-linux-gnu-fl32... no checking for i686-pc-linux-gnu-af77... no checking for i686-pc-linux-gnu-xlf90... no checking for i686-pc-linux-gnu-f90... no checking for i686-pc-linux-gnu-pgf90... no checking for i686-pc-linux-gnu-pghpf... no checking for i686-pc-linux-gnu-epcf90... no checking for i686-pc-linux-gnu-gfortran... i686-pc-linux-gnu-gfortran checking whether we are using the GNU Fortran 77 compiler... no checking whether i686-pc-linux-gnu-gfortran accepts -g... yes configure: creating ./config.status config.status: creating Makefile i686-pc-linux-gnu-gfortran -O3 -march=prescott -openmp -parallel -c -o ciftab.o ciftab.f i686-pc-linux-gnu-gfortran: unrecognized option '-parallel' i686-pc-linux-gnu-gfortran -O3 -march=prescott -openmp -parallel -c -o shelxa.o shelxa.f i686-pc-linux-gnu-gfortran -O3 -march=prescott -openmp -parallel -c -o shelxc.o shelxc.f i686-pc-linux-gnu-gfortran: unrecognized option '-parallel' i686-pc-linux-gnu-gfortran -O3 -march=prescott -openmp -parallel -c -o shelxd.o shelxd.f i686-pc-linux-gnu-gfortran: unrecognized option '-parallel' -- gentoo-user@lists.gentoo.org mailing list
[gentoo-user] Kernel upgrade broke NAT
I upgraded my router's kernel from linux-2.6.18-hardened-r6 to linux-2.6.23-hardened-r7 and now I get errors when starting the firewall: requires NAT which is disabled ERROR: Command /sbin/iptables -A FORWARD -m state --state ESTABLISHED,RELATED -j ACCEPT Failed I used make oldconfig carefully to update the config. I've been back over the current config carefully but I don't see what went wrong. Does anyone have any ideas? - Grant -- gentoo-user@lists.gentoo.org mailing list
Re: [gentoo-user] Kernel upgrade broke NAT
On (22/02/08 06:37) Grant wrote: I upgraded my router's kernel from linux-2.6.18-hardened-r6 to linux-2.6.23-hardened-r7 and now I get errors when starting the firewall: requires NAT which is disabled ERROR: Command /sbin/iptables -A FORWARD -m state --state ESTABLISHED,RELATED -j ACCEPT Failed I used make oldconfig carefully to update the config. I've been back over the current config carefully but I don't see what went wrong. Does anyone have any ideas? - Grant -- gentoo-user@lists.gentoo.org mailing list Hi, Too bad that 'oldconfig' isn't always working :-( Specially from something 2.6.17-18 to 20. Manually check your kernel config (enable NAT etc) IIRC the netwoking sections moved, so much stuff is disabled by default. Also recompile iptables, just to be on the safe side. HTH. Rumen pgprknH3jeN8T.pgp Description: PGP signature
Re: [gentoo-user] Kernel upgrade broke NAT
I upgraded my router's kernel from linux-2.6.18-hardened-r6 to linux-2.6.23-hardened-r7 and now I get errors when starting the firewall: requires NAT which is disabled ERROR: Command /sbin/iptables -A FORWARD -m state --state ESTABLISHED,RELATED -j ACCEPT Failed I used make oldconfig carefully to update the config. I've been back over the current config carefully but I don't see what went wrong. Does anyone have any ideas? - Grant -- gentoo-user@lists.gentoo.org mailing list Hi, Too bad that 'oldconfig' isn't always working :-( Is there a better way to update the config for a new kernel? Specially from something 2.6.17-18 to 20. Manually check your kernel config (enable NAT etc) IIRC the netwoking sections moved, so much stuff is disabled by default. Also recompile iptables, just to be on the safe side. HTH. Rumen Looks like I got it, thank you. - Grant -- gentoo-user@lists.gentoo.org mailing list
Re: [gentoo-user] F77 Provided through /etc/portage/env/*
On Fri, Feb 22, 2008 at 3:56 AM, [EMAIL PROTECTED] wrote: Hi all! I have a little nasty problem. I am using a couple of fortran progs and not all are happy with ifort but some benefit allot of its use. As I am lazy and dont want to change my make.conf all the time I provided the F77, FC and FLAGS for those packages which support ifort through /etc/portage/env/categorie/package. This worked for a long time, but now some thing has changed and I dont know what. The variables are still passed to the emerge but the F77 and FC isnt used by emake any more. See following example: ifort is chosen by the fortran.eclass and configure uses it as well. But in the make part only the ifort spezific FFLAGS are use together with gfortran. What version of portage is this? emerge --info? It is possible that there might be a regression between portage versions. If you've updated portage recently, try downgrading by masking the current version in /etc/portage/package.mask. Emerging (1 of 1) sci-chemistry/shelx-20060317 to / * shelx-20060317.tgz RMD160 SHA1 SHA256 size ;-) ... [ ok ] * checking ebuild checksums ;-) ... [ ok ] * checking auxfile checksums ;-) ... [ ok ] * checking miscfile checksums ;-) ... [ ok ] * checking shelx-20060317.tgz ;-) ... [ ok ] * You need one of these Fortran Compilers: ifc gfortran * Installed are: ifort gfortran * Using ifort Unpacking source... Unpacking shelx-20060317.tgz to /var/tmp/portage/sci-chemistry/shelx-20060317/work * Applying 20060317-autotool.patch ... [ ok ] * Applying 20060317-gfortran.patch ... [ ok ] * Running eautoreconf in '/var/tmp/portage/sci-chemistry/shelx-20060317/work/unix' ... * Running aclocal ... [ ok ] * Running autoconf ... [ ok ] * Running automake --add-missing --copy --foreign ... [ ok ] Source unpacked. Compiling source in /var/tmp/portage/sci-chemistry/shelx-20060317/work/unix ... ./configure --prefix=/usr --host=i686-pc-linux-gnu --mandir=/usr/share/man --infodir=/usr/share/info --datadir=/usr/share --sysconfdir=/etc --localstatedir=/var/lib FC=ifort --build=i686-pc-linux-gnu checking for a BSD-compatible install... /usr/bin/install -c checking whether build environment is sane... yes checking for a thread-safe mkdir -p... /bin/mkdir -p checking for gawk... gawk checking whether make sets $(MAKE)... yes checking for i686-pc-linux-gnu-gfortran... ifort checking for Fortran compiler default output file name... a.out checking whether the Fortran compiler works... yes checking whether we are cross compiling... no checking for suffix of executables... checking for suffix of object files... o checking whether we are using the GNU Fortran compiler... no checking whether ifort accepts -g... yes checking for i686-pc-linux-gnu-g77... no checking for i686-pc-linux-gnu-xlf... no checking for i686-pc-linux-gnu-f77... no checking for i686-pc-linux-gnu-frt... no checking for i686-pc-linux-gnu-pgf77... no checking for i686-pc-linux-gnu-cf77... no checking for i686-pc-linux-gnu-fort77... no checking for i686-pc-linux-gnu-fl32... no checking for i686-pc-linux-gnu-af77... no checking for i686-pc-linux-gnu-xlf90... no checking for i686-pc-linux-gnu-f90... no checking for i686-pc-linux-gnu-pgf90... no checking for i686-pc-linux-gnu-pghpf... no checking for i686-pc-linux-gnu-epcf90... no checking for i686-pc-linux-gnu-gfortran... i686-pc-linux-gnu-gfortran checking whether we are using the GNU Fortran 77 compiler... no checking whether i686-pc-linux-gnu-gfortran accepts -g... yes configure: creating ./config.status config.status: creating Makefile i686-pc-linux-gnu-gfortran -O3 -march=prescott -openmp -parallel -c -o ciftab.o ciftab.f i686-pc-linux-gnu-gfortran: unrecognized option '-parallel' i686-pc-linux-gnu-gfortran -O3 -march=prescott -openmp -parallel -c -o shelxa.o shelxa.f i686-pc-linux-gnu-gfortran -O3 -march=prescott -openmp -parallel -c -o shelxc.o shelxc.f i686-pc-linux-gnu-gfortran: unrecognized option '-parallel' i686-pc-linux-gnu-gfortran -O3 -march=prescott -openmp -parallel -c -o shelxd.o shelxd.f i686-pc-linux-gnu-gfortran: unrecognized option '-parallel' -- gentoo-user@lists.gentoo.org mailing list -- gentoo-user@lists.gentoo.org mailing list
What is LD_LIBRARY_PATH? (Was Re: [gentoo-user] OpenOffice 2.3.1 Won't Start As User)
Philip Webb wrote: 080221 Drew Tomlinson wrote: I emerged OpenOffice 2.3.1 on my amd64 install running gentoo-2.6.23-r5. I can start OpenOffice as root but not as a regular user. How are you starting it ? Do you enter eg 'oocalc' from a CLI or do you click on an icon in a start menu ? If the latter, check that the menu is using the correct command (eg use Kmenuedit). If the former, are there any error messages ? I have started it both ways as a user. As root, I start from the CLI after su to root from my user login. There are no error messages. When starting as a user, the splash graphic just sits there and the CPU usage is basically 100% for both X and oosplash.bin. When starting as root, the splash graphic starts and then a progress bar at the bottom of the splash graphic progresses in about 10 seconds and OpenOffice starts. The progress bar never progresses when starting as a user. I have to kill oosplash.bin with signal 15 after starting as a user to return my system to normal. Some more Googling turned up this post: http://forums.gentoo.org/viewtopic-t-623771-postdays-0-postorder-asc-start-25.html?sid=1514fa8e48f6a12a86c6c66e920161e9 I found that when I su to root, LD_LIBRARY_PATH is not defined. As a user, it is defined as: LD_LIBRARY_PATH=/usr/lib32/xorg:/usr/lib64/xorg If I unset LD_LIBRARY_PATH in a user terminal, then OpenOffice starts. So what is LD_LIBRARY_PATH and why am I seeing this behavior? Thanks, Drew -- Be a Great Magician! Visit The Alchemist's Warehouse http://www.alchemistswarehouse.com -- gentoo-user@lists.gentoo.org mailing list
Re: [gentoo-user] How do I generate libguile-ltdl.so?
Volker Armin Hemmann wrote: On Donnerstag, 21. Februar 2008, Alma J. Wetzker wrote: I just updated portage and tried to update gnucash. The compile errors out trying to find libguile-ltdl.so.1 and libwthreads.so.12. I have tried to re-emerge guile, g-wrap and slib, none of them build the library that I need. (slib also does not link the scm functions that it looks like gnucash looks for, so I may need that too.) What settings or whatever do I need to set to generate libguile-ltdl.*? -- Alma have you followed Jakub's advise from the bug and done a revdep-rebuilt? I have tried. Configuring search environment for revdep-rebuild Checking reverse dependencies... Packages containing binaries and libraries broken by a package update will be emerged. [32;01mCollecting system binaries and libraries...[0m using existing /root/.revdep-rebuild.1_files. [32;01mCollecting complete LD_LIBRARY_PATH...[0m using existing /root/.revdep-rebuild.2_ldpath. [32;01mChecking dynamic linking consistency...[0m using existing /root/.revdep-rebuild.3_rebuild. [32;01mAssigning files to ebuilds...[0m using existing /root/.revdep-rebuild.4_ebuilds. [32;01mEvaluating package order...[0m using existing /root/.revdep-rebuild.5_order. [32;01mAll prepared. Starting rebuild...[0m emerge --oneshot =dev-python/gnome-python-desktop-2.14.0 =sys-fs/cryptsetup-luks-1.0.3-r2 =gnome-base/gnome-panel-2.14.2 =gnome-base/gnome-applets-2.14.2 =app-text/evince-0.5.3-r1 =gnome-extra/deskbar-applet-2.14.2 =gnome-extra/evolution-webcal-2.6.0 =mail-client/evolution-2.6.2-r1 .......... Calculating dependencies . . !!! All ebuilds that could satisfy =dev-python/gnome-python-desktop-2.14.0 have been masked. !!! One of the following masked packages is required to complete your request: - dev-python/gnome-python-desktop-2.14.0 (masked by: ) For more information, see MASKED PACKAGES section in the emerge man page or refer to the Gentoo Handbook. [31;01mrevdep-rebuild failed to emerge all packages[0m [31;01myou have the following choices:[0m - if emerge failed during the build, fix the problems and re-run revdep-rebuild or - use -X or --package-names as first argument (trys to rebuild package, not exact ebuild) or - set ACCEPT_KEYWORDS=~your platform and/or /etc/portage/package.unmask (and remove /root/.revdep-rebuild.5_order to be evaluated again) or - modify the above emerge command and run it manually or - compile or unmerge unsatisfied packages manually, remove temporary files and try again (you can edit package/ebuild list first) [32;01mTo remove temporary files, please run:[0m rm /root/.revdep-rebuild*.?_* the first several libraries are masked, so that seems to be a long and tedious exercise to correct them, one at a time. This is what I did that worked emerge guile(1.8.3) New version, probably made no difference emerge slib (3.1.5-r1) New version, probably made no difference emerge libpcre (7.6-r1) This seems to be the magic step that created 'libguile-ltdl.so.*' and 'libqthreads.so.*' emerge -C gnucash (2.0.5) For some reason, I needed this step in order for the new version of gnucash to compile correctly. It must have been looking for something from the 2.0.5 version. emerge gnucash (2.2.3) This worked. I am not sure what to do about the revdep-build problems, I will work on it later. Thanks for the response! -- Alma -- gentoo-user@lists.gentoo.org mailing list
Re: [gentoo-user] F77 Provided through /etc/portage/env/*
Andrey Falko schrieb: On Fri, Feb 22, 2008 at 3:56 AM, [EMAIL PROTECTED] wrote: Hi all! I have a little nasty problem. I am using a couple of fortran progs and not all are happy with ifort but some benefit allot of its use. As I am lazy and dont want to change my make.conf all the time I provided the F77, FC and FLAGS for those packages which support ifort through /etc/portage/env/categorie/package. This worked for a long time, but now some thing has changed and I dont know what. The variables are still passed to the emerge but the F77 and FC isnt used by emake any more. See following example: ifort is chosen by the fortran.eclass and configure uses it as well. But in the make part only the ifort spezific FFLAGS are use together with gfortran. What version of portage is this? emerge --info? It is possible that there might be a regression between portage versions. If you've updated portage recently, try downgrading by masking the current version in /etc/portage/package.mask. I tryed this first. The second thing I tried was downgrading bash but all that doesnt help. Emerging (1 of 1) sci-chemistry/shelx-20060317 to / * shelx-20060317.tgz RMD160 SHA1 SHA256 size ;-) ... [ ok ] * checking ebuild checksums ;-) ... [ ok ] * checking auxfile checksums ;-) ... [ ok ] * checking miscfile checksums ;-) ... [ ok ] * checking shelx-20060317.tgz ;-) ... [ ok ] * You need one of these Fortran Compilers: ifc gfortran * Installed are: ifort gfortran * Using ifort Unpacking source... Unpacking shelx-20060317.tgz to /var/tmp/portage/sci-chemistry/shelx-20060317/work * Applying 20060317-autotool.patch ... [ ok ] * Applying 20060317-gfortran.patch ... [ ok ] * Running eautoreconf in '/var/tmp/portage/sci-chemistry/shelx-20060317/work/unix' ... * Running aclocal ... [ ok ] * Running autoconf ... [ ok ] * Running automake --add-missing --copy --foreign ... [ ok ] Source unpacked. Compiling source in /var/tmp/portage/sci-chemistry/shelx-20060317/work/unix ... ./configure --prefix=/usr --host=i686-pc-linux-gnu --mandir=/usr/share/man --infodir=/usr/share/info --datadir=/usr/share --sysconfdir=/etc --localstatedir=/var/lib FC=ifort --build=i686-pc-linux-gnu checking for a BSD-compatible install... /usr/bin/install -c checking whether build environment is sane... yes checking for a thread-safe mkdir -p... /bin/mkdir -p checking for gawk... gawk checking whether make sets $(MAKE)... yes checking for i686-pc-linux-gnu-gfortran... ifort checking for Fortran compiler default output file name... a.out checking whether the Fortran compiler works... yes checking whether we are cross compiling... no checking for suffix of executables... checking for suffix of object files... o checking whether we are using the GNU Fortran compiler... no checking whether ifort accepts -g... yes checking for i686-pc-linux-gnu-g77... no checking for i686-pc-linux-gnu-xlf... no checking for i686-pc-linux-gnu-f77... no checking for i686-pc-linux-gnu-frt... no checking for i686-pc-linux-gnu-pgf77... no checking for i686-pc-linux-gnu-cf77... no checking for i686-pc-linux-gnu-fort77... no checking for i686-pc-linux-gnu-fl32... no checking for i686-pc-linux-gnu-af77... no checking for i686-pc-linux-gnu-xlf90... no checking for i686-pc-linux-gnu-f90... no checking for i686-pc-linux-gnu-pgf90... no checking for i686-pc-linux-gnu-pghpf... no checking for i686-pc-linux-gnu-epcf90... no checking for i686-pc-linux-gnu-gfortran... i686-pc-linux-gnu-gfortran checking whether we are using the GNU Fortran 77 compiler... no checking whether i686-pc-linux-gnu-gfortran accepts -g... yes configure: creating ./config.status config.status: creating Makefile i686-pc-linux-gnu-gfortran -O3 -march=prescott -openmp -parallel -c -o ciftab.o ciftab.f i686-pc-linux-gnu-gfortran: unrecognized option '-parallel' i686-pc-linux-gnu-gfortran -O3 -march=prescott -openmp -parallel -c -o shelxa.o shelxa.f i686-pc-linux-gnu-gfortran -O3 -march=prescott -openmp -parallel -c -o shelxc.o shelxc.f i686-pc-linux-gnu-gfortran: unrecognized option '-parallel' i686-pc-linux-gnu-gfortran -O3 -march=prescott -openmp -parallel -c -o shelxd.o shelxd.f i686-pc-linux-gnu-gfortran: unrecognized option '-parallel' -- gentoo-user@lists.gentoo.org mailing list -- gentoo-user@lists.gentoo.org mailing list
Re: [gentoo-user] emerge can not work !
Matthias Guede wrote:
Make sure your working directory is in the path:
PATH=${PATH}:./ ./python /usr/bin/emerge python
!!Big security hole!! ./ is purposely left out of the path so people
can't sneak fake programs in there.
~eric
--
gentoo-user@lists.gentoo.org mailing list
Re: [gentoo-user] How do I generate libguile-ltdl.so?
On Freitag 22 Februar 2008, Alma J. Wetzker wrote: Volker Armin Hemmann wrote: On Donnerstag, 21. Februar 2008, Alma J. Wetzker wrote: I just updated portage and tried to update gnucash. The compile errors out trying to find libguile-ltdl.so.1 and libwthreads.so.12. I have tried to re-emerge guile, g-wrap and slib, none of them build the library that I need. (slib also does not link the scm functions that it looks like gnucash looks for, so I may need that too.) What settings or whatever do I need to set to generate libguile-ltdl.*? -- Alma have you followed Jakub's advise from the bug and done a revdep-rebuilt? I have tried. Configuring search environment for revdep-rebuild Checking reverse dependencies... Packages containing binaries and libraries broken by a package update will be emerged. [32;01mCollecting system binaries and libraries...[0m using existing /root/.revdep-rebuild.1_files. [32;01mCollecting complete LD_LIBRARY_PATH...[0m using existing /root/.revdep-rebuild.2_ldpath. [32;01mChecking dynamic linking consistency...[0m using existing /root/.revdep-rebuild.3_rebuild. [32;01mAssigning files to ebuilds...[0m using existing /root/.revdep-rebuild.4_ebuilds. [32;01mEvaluating package order...[0m using existing /root/.revdep-rebuild.5_order. [32;01mAll prepared. Starting rebuild...[0m emerge --oneshot =dev-python/gnome-python-desktop-2.14.0 =sys-fs/cryptsetup-luks-1.0.3-r2 =gnome-base/gnome-panel-2.14.2 =gnome-base/gnome-applets-2.14.2 =app-text/evince-0.5.3-r1 =gnome-extra/deskbar-applet-2.14.2 =gnome-extra/evolution-webcal-2.6.0 =mail-client/evolution-2.6.2-r1 .......... Calculating dependencies . . !!! All ebuilds that could satisfy =dev-python/gnome-python-desktop-2.14.0 have been masked. !!! One of the following masked packages is required to complete your request: - dev-python/gnome-python-desktop-2.14.0 (masked by: ) For more information, see MASKED PACKAGES section in the emerge man page or refer to the Gentoo Handbook. [31;01mrevdep-rebuild failed to emerge all packages[0m [31;01myou have the following choices:[0m - if emerge failed during the build, fix the problems and re-run revdep-rebuild or - use -X or --package-names as first argument (trys to rebuild package, not exact ebuild) or - set ACCEPT_KEYWORDS=~your platform and/or /etc/portage/package.unmask (and remove /root/.revdep-rebuild.5_order to be evaluated again) or - modify the above emerge command and run it manually or - compile or unmerge unsatisfied packages manually, remove temporary files and try again (you can edit package/ebuild list first) [32;01mTo remove temporary files, please run:[0m rm /root/.revdep-rebuild*.?_* the first several libraries are masked, so that seems to be a long and tedious exercise to correct them, one at a time. This is what I did that worked emerge guile(1.8.3) New version, probably made no difference emerge slib (3.1.5-r1) New version, probably made no difference emerge libpcre (7.6-r1) This seems to be the magic step that created 'libguile-ltdl.so.*' and 'libqthreads.so.*' emerge -C gnucash (2.0.5) For some reason, I needed this step in order for the new version of gnucash to compile correctly. It must have been looking for something from the 2.0.5 version. emerge gnucash (2.2.3) This worked. I am not sure what to do about the revdep-build problems, I will work on it later. Thanks for the response! just put dev-python/gnome-python-desktop-2.14.0 into package.unmask or package.keywords- -- gentoo-user@lists.gentoo.org mailing list
Re: [gentoo-user] F77 Provided through /etc/portage/env/*
On Fri, Feb 22, 2008 at 1:25 PM, Justin [EMAIL PROTECTED] wrote: Andrey Falko schrieb: On Fri, Feb 22, 2008 at 3:56 AM, [EMAIL PROTECTED] wrote: Hi all! I have a little nasty problem. I am using a couple of fortran progs and not all are happy with ifort but some benefit allot of its use. As I am lazy and dont want to change my make.conf all the time I provided the F77, FC and FLAGS for those packages which support ifort through /etc/portage/env/categorie/package. This worked for a long time, but now some thing has changed and I dont know what. The variables are still passed to the emerge but the F77 and FC isnt used by emake any more. See following example: ifort is chosen by the fortran.eclass and configure uses it as well. But in the make part only the ifort spezific FFLAGS are use together with gfortran. What version of portage is this? emerge --info? It is possible that there might be a regression between portage versions. If you've updated portage recently, try downgrading by masking the current version in /etc/portage/package.mask. I tryed this first. The second thing I tried was downgrading bash but all that doesnt help. You can try running emerge --debug, maybe that will hint the problem? Emerging (1 of 1) sci-chemistry/shelx-20060317 to / * shelx-20060317.tgz RMD160 SHA1 SHA256 size ;-) ... [ ok ] * checking ebuild checksums ;-) ... [ ok ] * checking auxfile checksums ;-) ... [ ok ] * checking miscfile checksums ;-) ... [ ok ] * checking shelx-20060317.tgz ;-) ... [ ok ] * You need one of these Fortran Compilers: ifc gfortran * Installed are: ifort gfortran * Using ifort Unpacking source... Unpacking shelx-20060317.tgz to /var/tmp/portage/sci-chemistry/shelx-20060317/work * Applying 20060317-autotool.patch ... [ ok ] * Applying 20060317-gfortran.patch ... [ ok ] * Running eautoreconf in '/var/tmp/portage/sci-chemistry/shelx-20060317/work/unix' ... * Running aclocal ... [ ok ] * Running autoconf ... [ ok ] * Running automake --add-missing --copy --foreign ... [ ok ] Source unpacked. Compiling source in /var/tmp/portage/sci-chemistry/shelx-20060317/work/unix ... ./configure --prefix=/usr --host=i686-pc-linux-gnu --mandir=/usr/share/man --infodir=/usr/share/info --datadir=/usr/share --sysconfdir=/etc --localstatedir=/var/lib FC=ifort --build=i686-pc-linux-gnu checking for a BSD-compatible install... /usr/bin/install -c checking whether build environment is sane... yes checking for a thread-safe mkdir -p... /bin/mkdir -p checking for gawk... gawk checking whether make sets $(MAKE)... yes checking for i686-pc-linux-gnu-gfortran... ifort checking for Fortran compiler default output file name... a.out checking whether the Fortran compiler works... yes checking whether we are cross compiling... no checking for suffix of executables... checking for suffix of object files... o checking whether we are using the GNU Fortran compiler... no checking whether ifort accepts -g... yes checking for i686-pc-linux-gnu-g77... no checking for i686-pc-linux-gnu-xlf... no checking for i686-pc-linux-gnu-f77... no checking for i686-pc-linux-gnu-frt... no checking for i686-pc-linux-gnu-pgf77... no checking for i686-pc-linux-gnu-cf77... no checking for i686-pc-linux-gnu-fort77... no checking for i686-pc-linux-gnu-fl32... no checking for i686-pc-linux-gnu-af77... no checking for i686-pc-linux-gnu-xlf90... no checking for i686-pc-linux-gnu-f90... no checking for i686-pc-linux-gnu-pgf90... no checking for i686-pc-linux-gnu-pghpf... no checking for i686-pc-linux-gnu-epcf90... no checking for i686-pc-linux-gnu-gfortran... i686-pc-linux-gnu-gfortran checking whether we are using the GNU Fortran 77 compiler... no checking whether i686-pc-linux-gnu-gfortran accepts -g... yes configure: creating ./config.status config.status: creating Makefile i686-pc-linux-gnu-gfortran -O3 -march=prescott -openmp -parallel -c -o ciftab.o ciftab.f i686-pc-linux-gnu-gfortran: unrecognized option '-parallel' i686-pc-linux-gnu-gfortran -O3 -march=prescott -openmp -parallel -c -o shelxa.o shelxa.f i686-pc-linux-gnu-gfortran -O3
Re: [gentoo-user] Gnucash 2.2.3 build not finding libguile-ltdl*
Alma J. Wetzker wrote: environment is huge, not sure what you want. I emerged the latest guile (1.8.3) after the first error, then I looked at the build description of fixes and emerged slib (3.1.5-r1), then I reemerged guile. I still do not have libguile-ltdl*, and libqthreads* anywhere that I can find them. Not sure what to do with the libguile.so.[17|12] concern. I also cannot re-emerge gnucash 2.0.5, so nothing is working now. I am not near a guru level, so please be specific when asking for more information. How do I get the missing libraries? -- Alma Have you tried revdep-rebuild yet? -- gentoo-user@lists.gentoo.org mailing list
Re: What is LD_LIBRARY_PATH? (Was Re: [gentoo-user] OpenOffice 2.3.1 Won't Start As User)
On Fri, Feb 22, 2008 at 09:52:39AM -0800, Penguin Lover Drew Tomlinson squawked: There are no error messages. When starting as a user, the splash graphic just sits there and the CPU usage is basically 100% for both X and oosplash.bin. When starting as root, the splash graphic starts and then a progress bar at the bottom of the splash graphic progresses in about 10 seconds and OpenOffice starts. The progress bar never progresses when starting as a user. I have to kill oosplash.bin with signal 15 after starting as a user to return my system to normal. Some more Googling turned up this post: http://forums.gentoo.org/viewtopic-t-623771-postdays-0-postorder-asc-start-25.html?sid=1514fa8e48f6a12a86c6c66e920161e9 I found that when I su to root, LD_LIBRARY_PATH is not defined. As a user, it is defined as: LD_LIBRARY_PATH=/usr/lib32/xorg:/usr/lib64/xorg If I unset LD_LIBRARY_PATH in a user terminal, then OpenOffice starts. So what is LD_LIBRARY_PATH and why am I seeing this behavior? See http://linuxmafia.com/faq/Admin/ld-lib-path.html http://wiki.services.openoffice.org/wiki/LD_LIBRARY_PATH So the question is one of hunting down where this variable is set. Your .bashrc? Perhaps grepping through /etc/env.d? W -- What is the meaning of Life? To search for truth and beauty. To ask questions. To find the answer to this question To rigorously mathematically prove by epislon-delta treatment that this question has no answer. Proof? For all oranges greater than zero, there exists a banana such that for all fruits tastier than the banana, an orange sweeter than a banana implies the orange is yellow. So by the Jedi-Schwartz Inequality, all oranges are not orange. This leads to an obvious contradiction. It's too bad life is a departmental requirement, otherwise I'd P/D/F it. ~Phil Wei Sortir en Pantoufles: up 441 days, 20:06 -- gentoo-user@lists.gentoo.org mailing list
[gentoo-user] emerge --update --deep --newuse world fails to update all packages
I did emerge --update --deep --newuse world and it says: Calculating world dependencies... done! Auto-cleaning packages... No outdated packages were found on your system. But it seems like it is wrong about No outdated packages were found on your system., because when I run emerge -p world it says: These are the packages that would be merged, in order: Calculating world dependencies... done! [ebuild U ] media-gfx/digikam-0.9.2 [0.9.1] LINGUAS=-fa% -nds% [blocks B ] media-plugins/digikamimageplugins (is blocking media-gfx/digikam-0.9.2) So it seems like media-gfx/digikam is outdated despite what emerge --update --deep --newuse world said. I just wonder how much sense this behaviour makes. -- gentoo-user@lists.gentoo.org mailing list
Re: [gentoo-user] emerge can not work !
On Friday 22 February 2008, Eric Martin wrote:
Matthias Guede wrote:
Make sure your working directory is in the path:
PATH=${PATH}:./ ./python /usr/bin/emerge python
!!Big security hole!! ./ is purposely left out of the path so people
can't sneak fake programs in there.
~eric
Very true but in this specific case it is superfluous.
With the sample command given, the PATH will not be searched at all as
all executables are accessed via their directory path.
But you are correct with the warning. People do this solely and only
because Windows does it and some folk are used to it. But WE all know
that Windows is brain-dead :-)
Kids, don't try this at home and dream up false reasons to put . in your
PATH
--
Alan McKinnon
alan dot mckinnon at gmail dot com
--
gentoo-user@lists.gentoo.org mailing list
Re: [gentoo-user] Kernel upgrade broke NAT
On Friday 22 February 2008, Grant wrote: Too bad that 'oldconfig' isn't always working :-( Is there a better way to update the config for a new kernel? Not really. You are dealing with a complex system of configuration settings that react in mysterious ways. AND that seeing as this is a kernel, the hard decisions are best left to humans, not software. So all that oldconfig really does is look for settings that are present in the source tree but not accounted for in your .config For everything else, you get to look yourself. Or, in the usual case, study it for 10 hours to find out what the heck it is, then look for yourself. -- Alan McKinnon alan dot mckinnon at gmail dot com -- gentoo-user@lists.gentoo.org mailing list
[gentoo-user] emerge -ef world and eclean --destructive distfiles do not agree
I am running emerge -ef world eclean --destructive distfiles' emerge -ef world eclean --destructive distfiles ... Each time eclean removes the following files: * Building file list for distfiles cleaning... * Cleaning distfiles... [ 258.9 K ] XML-LibXML-1.65.tar.gz [ 478.6 K ] klibc-1.5.8.tar.bz2 [ 290.2 K ] libassuan-1.0.4.tar.bz2 [ 67.7 K ] mftrace-1.2.9.tar.gz [ 9.2 M ] patch-2.6.24-rc7.bz2 [ 243.4 K ] setuptools-0.6c7.tar.gz * Total space that has been freed in distfiles directory: 10.5 M Then emerge downloads them again. I had expected that those 2 programs should agree on which files belong in the set of distfiles. -- gentoo-user@lists.gentoo.org mailing list
Re: [gentoo-user] emerge -ef world and eclean --destructive distfiles do not agree
On Saturday 23 February 2008, Erik wrote: I am running emerge -ef world eclean --destructive distfiles' emerge -ef world eclean --destructive distfiles ... Each time eclean removes the following files: * Building file list for distfiles cleaning... * Cleaning distfiles... [ 258.9 K ] XML-LibXML-1.65.tar.gz [ 478.6 K ] klibc-1.5.8.tar.bz2 [ 290.2 K ] libassuan-1.0.4.tar.bz2 [ 67.7 K ] mftrace-1.2.9.tar.gz [ 9.2 M ] patch-2.6.24-rc7.bz2 [ 243.4 K ] setuptools-0.6c7.tar.gz * Total space that has been freed in distfiles directory: 10.5 M Then emerge downloads them again. I had expected that those 2 programs should agree on which files belong in the set of distfiles. What happens if you use 'emerge -uNDf world instead of 'emerge -ef world' -e is a special case and does not equate to the same thing as -uND -- Alan McKinnon alan dot mckinnon at gmail dot com -- gentoo-user@lists.gentoo.org mailing list
Re: [gentoo-user] emerge -ef world and eclean --destructive distfiles do not agree
Alan McKinnon wrote: On Saturday 23 February 2008, Erik wrote: I am running emerge -ef world eclean --destructive distfiles' emerge -ef world eclean --destructive distfiles ... Each time eclean removes the following files: * Building file list for distfiles cleaning... * Cleaning distfiles... [ 258.9 K ] XML-LibXML-1.65.tar.gz [ 478.6 K ] klibc-1.5.8.tar.bz2 [ 290.2 K ] libassuan-1.0.4.tar.bz2 [ 67.7 K ] mftrace-1.2.9.tar.gz [ 9.2 M ] patch-2.6.24-rc7.bz2 [ 243.4 K ] setuptools-0.6c7.tar.gz * Total space that has been freed in distfiles directory: 10.5 M Then emerge downloads them again. I had expected that those 2 programs should agree on which files belong in the set of distfiles. What happens if you use 'emerge -uNDf world instead of 'emerge -ef world' -e is a special case and does not equate to the same thing as -uND I think I see it the same way he does. If you do a rm -rf /usr/portage/distfiles/* and then do a emerge -ef world, then do the eclean again, it should not remove anything since everything there is needed to recompile everything in world and their dependencies. It should also not have to redownload anything either. Isn't patch part of system? setuptools too? Dale :-) :-) -- gentoo-user@lists.gentoo.org mailing list
ATI Drivers Sets LD_LIBRARY_PATH (Was Re: What is LD_LIBRARY_PATH? (Was Re: [gentoo-user] OpenOffice 2.3.1 Won't Start As User))
Willie Wong wrote: On Fri, Feb 22, 2008 at 09:52:39AM -0800, Penguin Lover Drew Tomlinson squawked: There are no error messages. When starting as a user, the splash graphic just sits there and the CPU usage is basically 100% for both X and oosplash.bin. When starting as root, the splash graphic starts and then a progress bar at the bottom of the splash graphic progresses in about 10 seconds and OpenOffice starts. The progress bar never progresses when starting as a user. I have to kill oosplash.bin with signal 15 after starting as a user to return my system to normal. Some more Googling turned up this post: http://forums.gentoo.org/viewtopic-t-623771-postdays-0-postorder-asc-start-25.html?sid=1514fa8e48f6a12a86c6c66e920161e9 I found that when I su to root, LD_LIBRARY_PATH is not defined. As a user, it is defined as: LD_LIBRARY_PATH=/usr/lib32/xorg:/usr/lib64/xorg If I unset LD_LIBRARY_PATH in a user terminal, then OpenOffice starts. So what is LD_LIBRARY_PATH and why am I seeing this behavior? See http://linuxmafia.com/faq/Admin/ld-lib-path.html http://wiki.services.openoffice.org/wiki/LD_LIBRARY_PATH So the question is one of hunting down where this variable is set. Your .bashrc? Perhaps grepping through /etc/env.d? Thank you for your reply. The culprit appears to be /etc/profile.d/ati-fglrx.sh. So if setting LD_LIBRARY_PATH is not a good idea, why does the ATI driver do so? And more importantly, how can I workaround this issue? Thanks, Drew -- Be a Great Magician! Visit The Alchemist's Warehouse http://www.alchemistswarehouse.com -- gentoo-user@lists.gentoo.org mailing list
Re: [gentoo-user] emerge -ef world and eclean --destructive distfiles do not agree
Alan McKinnon skrev: On Saturday 23 February 2008, Erik wrote: I am running emerge -ef world eclean --destructive distfiles' emerge -ef world eclean --destructive distfiles ... Each time eclean removes the following files: * Building file list for distfiles cleaning... * Cleaning distfiles... [ 258.9 K ] XML-LibXML-1.65.tar.gz [ 478.6 K ] klibc-1.5.8.tar.bz2 [ 290.2 K ] libassuan-1.0.4.tar.bz2 [ 67.7 K ] mftrace-1.2.9.tar.gz [ 9.2 M ] patch-2.6.24-rc7.bz2 [ 243.4 K ] setuptools-0.6c7.tar.gz * Total space that has been freed in distfiles directory: 10.5 M Then emerge downloads them again. I had expected that those 2 programs should agree on which files belong in the set of distfiles. What happens if you use 'emerge -uNDf world instead of 'emerge -ef world' -e is a special case and does not equate to the same thing as -uND emerge -uNDf world does nothing because the system is up to date. Even if all distfiles are missing, it does nothing (try to move the distfiles directory away while executing it). That command is supposed to download everything that is missing to update the system. And since the system is up to date, nothing is needed to update it and it will therefore not download anything. What I wanted was download everything that is missing to reinstall everything that is currently installed. That is what emerge -ef world should do. Then I wanted to remov -- gentoo-user@lists.gentoo.org mailing list
Re: [gentoo-user] emerge -ef world and eclean --destructive distfiles do not agree
Alan McKinnon skrev: On Saturday 23 February 2008, Erik wrote: I am running emerge -ef world eclean --destructive distfiles' emerge -ef world eclean --destructive distfiles ... Each time eclean removes the following files: * Building file list for distfiles cleaning... * Cleaning distfiles... [ 258.9 K ] XML-LibXML-1.65.tar.gz [ 478.6 K ] klibc-1.5.8.tar.bz2 [ 290.2 K ] libassuan-1.0.4.tar.bz2 [ 67.7 K ] mftrace-1.2.9.tar.gz [ 9.2 M ] patch-2.6.24-rc7.bz2 [ 243.4 K ] setuptools-0.6c7.tar.gz * Total space that has been freed in distfiles directory: 10.5 M Then emerge downloads them again. I had expected that those 2 programs should agree on which files belong in the set of distfiles. What happens if you use 'emerge -uNDf world instead of 'emerge -ef world' -e is a special case and does not equate to the same thing as -uND emerge -uNDf world does nothing because the system is up to date. Even if all distfiles are missing, it does nothing (try to move the distfiles directory away while executing it). That command is supposed to download everything that is missing to update the system. And since the system is up to date, nothing is needed to update it and it will therefore not download anything. What I wanted was download everything that is missing to reinstall everything that is currently installed. That is what emerge -ef world should do. Then I wanted to remove everything that is not needed to reinstall everything that is currently installed. That is what eclean --destructive distfiles should do. Doing both should result in a set of distfiles that is what is needed and only what is needed to reinstall everything that is currently installed (assuming that the system is up to date). But since the 2 commands do not agree, something is broken somewhere. -- gentoo-user@lists.gentoo.org mailing list
Re: [gentoo-user] emerge -ef world and eclean --destructive distfiles do not agree
On 02:41 Sat 23 Feb , Erik wrote: Alan McKinnon skrev: emerge -uNDf world does nothing because the system is up to date. Even if all distfiles are missing, it does nothing (try to move the distfiles directory away while executing it). That command is supposed to download everything that is missing to update the system. And since the system is up to date, nothing is needed to update it and it will therefore not download anything. What I wanted was download everything that is missing to reinstall everything that is currently installed. That is what emerge -ef world should do. Then I wanted to remove everything that is not needed to reinstall everything that is currently installed. That is what eclean --destructive distfiles should do. Doing both should result in a set of distfiles that is what is needed and only what is needed to reinstall everything that is currently installed (assuming that the system is up to date). But since the 2 commands do not agree, something is broken somewhere. -- gentoo-user@lists.gentoo.org mailing list Have you checked to see if the files deleted by eclean are the current versions, or are they old? If they are the most recent you have installed, then there is a problem with eclean. If not, then the problem is with --fetchonly -- gentoo-user@lists.gentoo.org mailing list
Re: ATI Drivers Sets LD_LIBRARY_PATH (Was Re: What is LD_LIBRARY_PATH? (Was Re: [gentoo-user] OpenOffice 2.3.1 Won't Start As User))
On Fri, Feb 22, 2008 at 05:34:54PM -0800, Penguin Lover Drew Tomlinson squawked: LD_LIBRARY_PATH=/usr/lib32/xorg:/usr/lib64/xorg If I unset LD_LIBRARY_PATH in a user terminal, then OpenOffice starts. So what is LD_LIBRARY_PATH and why am I seeing this behavior? See http://linuxmafia.com/faq/Admin/ld-lib-path.html http://wiki.services.openoffice.org/wiki/LD_LIBRARY_PATH So the question is one of hunting down where this variable is set. Your .bashrc? Perhaps grepping through /etc/env.d? Thank you for your reply. The culprit appears to be /etc/profile.d/ati-fglrx.sh. So if setting LD_LIBRARY_PATH is not a good idea, why does the ATI driver do so? And more importantly, how can I workaround this issue? As I don't use fglrx, I can't comment on that behaviour. A workaround is simply to alias oocal in your bashrc to something that unsets the problematic path. W -- W: You see, me and Willetta have been going on for a few weeks now. Phil: Only a few weeks, and she's living in your room? DJP: Will! What's that thing you said earlier about taking things slow! MW rolling on the ground laughing. Sortir en Pantoufles: up 442 days, 5:06 -- gentoo-user@lists.gentoo.org mailing list
