Ok, thank you! I would be glad if you could explain how can we put sage in, for example, /usr/local/ or /opt/, if sage force us to compile without root privileges.
On Thursday, February 2, 2017 at 11:50:14 PM UTC+1, John H Palmieri wrote: > > > > On Thursday, February 2, 2017 at 1:53:43 PM UTC-8, João Alberto Ferreira > wrote: >> >> Well, I read somewhere about this procedure of moving the directory >> before starting sage, but the Installation Manual seems to tell the same >> thing. >> >> "The directory where you built Sage is *NOT* hardcoded. You should be >> able to safely move or rename that directory. (It’s a bug if this is not >> the case.)" >> >> > >> Chapter 2 describes this same procedure of moving the directory, but for >> the precompiled binaries, that for me is the same thing. >> > > We should fix the documentation. See > https://trac.sagemath.org/ticket/22294. > > John > > > >> >> But, in fact, what I experienced was what you said. >> >> On Thursday, February 2, 2017 at 9:28:17 PM UTC+1, HG wrote: >>> >>> I think it has been said on the list : Keep sage where you compiled it ! >>> >>> If you want to change it you have to compile source again ! >>> >>> what I do because I work with git : >>> >>> I make a dir git >>> >>> I compile it there and after I link sage in /usr/bin/sage >>> >>> sudo ln -s /path where is sage/sage /usr/bin/sage >>> >>> >>> >>> Le 02/02/2017 à 20:43, João Alberto Ferreira a écrit : >>> >>> Hi! >>> >>> I just removed "/home/mmsim/tools/lib/64bit" from the LD_LIBRARY_PATH >>> environment variable and sage compiled. >>> >>> I compiled it in my home directory and moved it to /usr/local/ before >>> starting it, and created a symbolic link in /usr/local/bin/, but it shows me >>> >>> python: error while loading shared libraries: libpython2.7.so.1.0: >>> cannot open shared object file: No such file or directory >>> >>> It just works in the place where I compiled it. I will leave the things >>> like that, but I'm curious to know why it do not work. Isn't that correct? >>> Do I Have made something wrong? >>> >>> Thank you! >>> >>> On Thursday, February 2, 2017 at 12:14:51 PM UTC+1, Dima Pasechnik >>> wrote: >>>> >>>> >>>> >>>> On Thursday, February 2, 2017 at 10:36:41 AM UTC, João Alberto Ferreira >>>> wrote: >>>>> >>>>> >>>>> >>>>> On Wednesday, February 1, 2017 at 7:24:56 PM UTC+1, Dima Pasechnik >>>>> wrote: >>>>>> >>>>>> >>>>>> >>>>>> On Wednesday, February 1, 2017 at 6:00:25 PM UTC, João Alberto >>>>>> Ferreira wrote: >>>>>>> >>>>>>> Well, not yet. >>>>>>> >>>>>>> openblas has compiled successfully. The problem now is with R. The R >>>>>>> log follows. >>>>>>> >>>>>>> Apparently it needs libgomp-4 and if I understood correctrly, it >>>>>>> found an older version. But the libgomp-4 is installed >>>>>>> >>>>>>> [defrancaferr_joa@javel sage-7.5.1]$ rpm -q libgomp >>>>>>> libgomp-4.4.7-17.el6.x86_64 >>>>>>> >>>>>> >>>>>> your log says: >>>>>> /home/mmsim/tools/lib/64bit/libgomp.so.1 : version `GOMP_4.0' not >>>>>> found >>>>>> >>>>>> you have some rogue stuff in your PATH or something (LD_LIBRARY_PATH >>>>>> ?) >>>>>> >>>>> >>>>> The environment variables are at the beginning of the whole log file >>>>> >>>> >>>> OK, I overlooked this... >>>> >>>> >>>>> >>>>> >>>>> LD_LIBRARY_PATH=/opt/rh/devtoolset-3/root/usr/lib64:/opt/rh/devtoolset-3/root/usr/lib:/usr/local/lib:/home/mmsim/tools/lib/64bit >>>>> >>>>> >>>>> PATH=/home/defrancaferr_joa/sage-7.5.1/build/bin:/home/defrancaferr_joa/sage-7.5.1/src/bin:/home/defrancaferr_joa/sage-7.5.1/local/bin:/opt/rh/devtoolset-3/root/usr/bin:/home/xfab/xkit/x_all/cadence/xenv:/usr/local/ADS2014_01/bin:/usr/local/texlive/2016/bin/x86_64-linux:/usr/lib64/qt-3.3/bin:/usr/local/bin:/usr/bin:/bin:/usr/local/sbin:/usr/sbin:/sbin:/home/defrancaferr_joa/bin:/home/eclipse:/home/matlab/bin:/home/altera/15.0/quartus/bin:/home/altera/15.0/quartus/sopc_builder/bin:/home/altera/15.0/quartus/bin:/home/ic6.1/tools/bin:/home/ic6.1/tools/dfII/bin:/home/ic6.1/tools/plot/bin:/home/ic6.1/tools/dracula/bin:/home/assura/tools/bin:/home/assura/tools/assura/bin:/home/AMS_4.10/cds/bin:/home/AMS_4.10/programs/bin:/home/mmsim/tools/bin:/home/incisiv/tools/bin:/home/incisiv/tools/dfII/bin:/home/rc/tools/bin:/home/edi/tools/bin:/home/et/tools/bin:/home/ets/tools/bin:/home/ext/tools/bin:/home/confrml/tools/bin:/home/ctos/tools/bin:/home/pve:/home/pve/tools/bin >>>>> >>>>> >>>> yes, this is definitely the problem. >>>> You can just unset these for the particular invocation of make, i.e. >>>> >>>> export MAKE="make -j8" # or whatever numer of cores you have on the box >>>> PATH="/bin:/usr/bin:/sbin:/usr/sbin" LD_LIBRARY_PATH="" $MAKE >>>> >>>> Note that these wrong libraries could have been already selected by >>>> other >>>> parts that are already built, and this may lead to hard to understand >>>> crashes etc. >>>> >>>> That is, I would rebuild from scratch, i.e. do first of all >>>> >>>> make distclean >>>> >>>> HTH >>>> Dima >>>> >>>> >>>>> [defrancaferr_joa@javel sage-7.5.1]$ find >>>>> /opt/rh/devtoolset-3/root/usr/lib64/ -name "libgomp*" -exec ls -ls {} \; >>>>> [defrancaferr_joa@javel sage-7.5.1]$ find >>>>> /opt/rh/devtoolset-3/root/usr/lib/ -name "libgomp*" -exec ls -ls {} \; >>>>> 4 -rw-r--r--. 1 root root 82 3 oct. 2015 >>>>> /opt/rh/devtoolset-3/root/usr/lib/gcc/x86_64-redhat-linux/4.9.2/libgomp.so >>>>> 4 -rw-r--r--. 1 root root 169 3 oct. 2015 >>>>> /opt/rh/devtoolset-3/root/usr/lib/gcc/x86_64-redhat-linux/4.9.2/libgomp.spec >>>>> 176 -rw-r--r--. 1 root root 178766 3 oct. 2015 >>>>> /opt/rh/devtoolset-3/root/usr/lib/gcc/x86_64-redhat-linux/4.9.2/libgomp.a >>>>> 4 -rw-r--r--. 1 root root 78 3 oct. 2015 >>>>> /opt/rh/devtoolset-3/root/usr/lib/gcc/x86_64-redhat-linux/4.9.2/32/libgomp.so >>>>> 140 -rw-r--r--. 1 root root 143328 3 oct. 2015 >>>>> /opt/rh/devtoolset-3/root/usr/lib/gcc/x86_64-redhat-linux/4.9.2/32/libgomp.a >>>>> [defrancaferr_joa@javel sage-7.5.1]$ find /usr/local/lib -name >>>>> "libgomp*" -exec ls -ls {} \; >>>>> [defrancaferr_joa@javel sage-7.5.1]$ find /home/mmsim/tools/lib/64bit >>>>> -name "libgomp*" -exec ls -ls {} \; >>>>> 0 lrwxrwxrwx. 1 cadence softs 16 15 janv. 2016 >>>>> /home/mmsim/tools/lib/64bit/libgomp.so.1 -> libgomp.so.1.0.0 >>>>> 276 -rwxr-xr-x. 1 cadence softs 281302 26 sept. 2011 >>>>> /home/mmsim/tools/lib/64bit/libgomp.so.1.0.0 >>>>> 0 lrwxrwxrwx. 1 cadence softs 16 15 janv. 2016 >>>>> /home/mmsim/tools/lib/64bit/libgomp.so -> libgomp.so.1.0.0 >>>>> >>>>> >>>>>> >>>>>> What is this /home/mmsim and what does it have to do with your >>>>>> account (defrancaferr_joa, I suppose)? >>>>>> >>>>> >>>>> mmsim (multi-mode simulator) is an electronic circuit simulator. I'm >>>>> intended to use this to design an integrated circuit, but this folder >>>>> houses an older version of the simulator (version 14). The newer version >>>>> resides in /home/mmsim15. In fact, this machine is a new one, which the >>>>> system administrator installed by cloning the disk of another machine. >>>>> Now, >>>>> I'm configuring it to my needs through sudo, but things seems to be >>>>> organized in an odd manner (like this simulator in the \home directory) >>>>> >>>>> It seems that the configuration of the integrated circuit design tools >>>>> are messed up and I need, at least, organize the environment variables >>>>> before compile sage. I think I need to edit the .bash_profile (appended). >>>>> What do you think? >>>>> >>>>> >>>>>> (It finds this stuff there first, and tries to use it...) >>>>>> >>>>>> >>>>>> >>>>>>> On Wednesday, February 1, 2017 at 5:56:44 PM UTC+1, João Alberto >>>>>>> Ferreira wrote: >>>>>>>> >>>>>>>> Thank you! >>>>>>>> >>>>>>>> I've done: >>>>>>>> >>>>>>>> [defrancaferr_joa@javel ~]$ sudo yum install centos-release-scl >>>>>>>> [defrancaferr_joa@javel ~]$ sudo yum install devtoolset-3-toolchain >>>>>>>> [defrancaferr_joa@javel ~]$ scl enable devtoolset-3 bash >>>>>>>> [defrancaferr_joa@javel ~]$ as --version >>>>>>>> Assembleur GNU version 2.24 >>>>>>>> Copyright 2013 Free Software Foundation, Inc. >>>>>>>> Ce logiciel est libre; vous pouvez le redistribuer selon les termes de >>>>>>>> la >>>>>>>> version 3 de la licence GNU General Public License ou suivante. >>>>>>>> Ce programme n'est couvert par AUCUNE garantie. >>>>>>>> Cet assembleur a été configuré pour la cible « x86_64-redhat-linux ». >>>>>>>> >>>>>>>> Let's see if it will work. >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> On Wednesday, February 1, 2017 at 5:27:36 PM UTC+1, Dima Pasechnik >>>>>>>> wrote: >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> On Wednesday, February 1, 2017 at 3:52:05 PM UTC, João Alberto >>>>>>>>> Ferreira wrote: >>>>>>>>>> >>>>>>>>>> Hi! >>>>>>>>>> >>>>>>>>>> Here it is the command output. >>>>>>>>>> >>>>>>>>>> [defrancaferr_joa@javel ~]$ as --version >>>>>>>>>> GNU assembler version 2.20.51.0.2-5.44.el6 20100205 >>>>>>>>>> Copyright 2009 Free Software Foundation, Inc. >>>>>>>>>> >>>>>>>>> >>>>>>>>> Thanks, this explains your problem. Your assembler is 7 years old, >>>>>>>>> and your CPU >>>>>>>>> is only 4 years old, if not newer. You are building with >>>>>>>>> (sage-supplied) >>>>>>>>> gcc 4.9, which issues correct assembler commands for your CPU. >>>>>>>>> But your assembler does not understand some of them, as it is too >>>>>>>>> old. >>>>>>>>> >>>>>>>>> Why Centos is so lame in this respect, we've seen many reports >>>>>>>>> like this... >>>>>>>>> You should upgrade your toolchain so that your assembler fully >>>>>>>>> supports you CPU. >>>>>>>>> >>>>>>>>> HTH, >>>>>>>>> Dima >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>>> This program is free software; you may redistribute it under the >>>>>>>>>> terms of >>>>>>>>>> the GNU General Public License version 3 or later. >>>>>>>>>> This program has absolutely no warranty. >>>>>>>>>> This assembler was configured for a target of >>>>>>>>>> `x86_64-redhat-linux'. >>>>>>>>>> >>>>>>>>>> On Wednesday, February 1, 2017 at 12:05:22 PM UTC+1, Dima >>>>>>>>>> Pasechnik wrote: >>>>>>>>>>> >>>>>>>>>>> please post the output of >>>>>>>>>>> >>>>>>>>>>> as --version >>>>>>>>>>> >>>>>>>>>>> on the system. I guess it is too old to understand the whole >>>>>>>>>>> range of assembler commands for your CPU. >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> On Wednesday, February 1, 2017 at 10:09:27 AM UTC, João Alberto >>>>>>>>>>> Ferreira wrote: >>>>>>>>>>>> >>>>>>>>>>>> Hi! >>>>>>>>>>>> >>>>>>>>>>>> I have tried to compile sagemath under a CentOS 6.8 machine, as >>>>>>>>>>>> the binaries for Fedora do not work. The installation halted with >>>>>>>>>>>> an error >>>>>>>>>>>> while building OpenBlas. Made the error is due to the machine CPU, >>>>>>>>>>>> as >>>>>>>>>>>> discussed in https://github.com/JuliaLang/julia/issues/7653, >>>>>>>>>>>> but I have little experience in this. Part of the installation log >>>>>>>>>>>> is >>>>>>>>>>>> attached. I updated the operating system and want to try build >>>>>>>>>>>> Sage again. >>>>>>>>>>>> Does anyone have any sugestion? >>>>>>>>>>>> >>>>>>>>>>>> It seems that all developing tools are installed, as can be >>>>>>>>>>>> seen below >>>>>>>>>>>> >>>>>>>>>>>> [defrancaferr_joa@javel ~]$ rpm -q perl-ExtUtils-MakeMaker >>>>>>>>>>>> perl-ExtUtils-MakeMaker-6.55-141.el6_7.1.x86_64 >>>>>>>>>>>> [defrancaferr_joa@javel ~]$ which perl >>>>>>>>>>>> /usr/bin/perl >>>>>>>>>>>> [defrancaferr_joa@javel ~]$ rpm -q binutils >>>>>>>>>>>> binutils-2.20.51.0.2-5.44.el6.x86_64 >>>>>>>>>>>> [defrancaferr_joa@javel ~]$ rpm -q gcc >>>>>>>>>>>> gcc-4.4.7-17.el6.x86_64 >>>>>>>>>>>> [defrancaferr_joa@javel ~]$ rpm -q make >>>>>>>>>>>> make-3.81-23.el6.x86_64 >>>>>>>>>>>> [defrancaferr_joa@javel ~]$ rpm -q m4 >>>>>>>>>>>> m4-1.4.13-5.el6.x86_64 >>>>>>>>>>>> [defrancaferr_joa@javel ~]$ rpm -q perl >>>>>>>>>>>> perl-5.10.1-141.el6_7.1.x86_64 >>>>>>>>>>>> [defrancaferr_joa@javel ~]$ rpm -q tar >>>>>>>>>>>> tar-1.23-15.el6_8.x86_64 >>>>>>>>>>>> [defrancaferr_joa@javel ~]$ rpm -q git >>>>>>>>>>>> git-1.7.1-4.el6_7.1.x86_64 >>>>>>>>>>>> [defrancaferr_joa@javel ~]$ rpm -q gcc-c++ >>>>>>>>>>>> gcc-c++-4.4.7-17.el6.x86_64 >>>>>>>>>>>> [defrancaferr_joa@javel ~]$ rpm -q gcc-gfortran >>>>>>>>>>>> gcc-gfortran-4.4.7-17.el6.x86_64 >>>>>>>>>>>> [defrancaferr_joa@javel ~]$ rpm -q python >>>>>>>>>>>> python-2.6.6-66.el6_8.x86_64 >>>>>>>>>>>> [defrancaferr_joa@javel ~]$ lscpu >>>>>>>>>>>> Architecture: x86_64 >>>>>>>>>>>> CPU op-mode(s): 32-bit, 64-bit >>>>>>>>>>>> Byte Order: Little Endian >>>>>>>>>>>> CPU(s): 32 >>>>>>>>>>>> On-line CPU(s) list: 0-31 >>>>>>>>>>>> Thread(s) par coeur : 2 >>>>>>>>>>>> Coeur(s) par support CPU :8 >>>>>>>>>>>> Socket(s): 2 >>>>>>>>>>>> Noeud(s) NUMA : 2 >>>>>>>>>>>> ID du vendeur : GenuineIntel >>>>>>>>>>>> Famille CPU : 6 >>>>>>>>>>>> Modèle : 79 >>>>>>>>>>>> Model name: Intel(R) Xeon(R) CPU E5-2620 v4 @ 2.10GHz >>>>>>>>>>>> Version : 1 >>>>>>>>>>>> CPU MHz : 1200.000 >>>>>>>>>>>> BogoMIPS: 4190.00 >>>>>>>>>>>> Virtualisation : VT-x >>>>>>>>>>>> L1d cache : 32K >>>>>>>>>>>> L1i cache : 32K >>>>>>>>>>>> L2 cache : 256K >>>>>>>>>>>> L3 cache : 20480K >>>>>>>>>>>> NUMA node0 CPU(s): 0-7,16-23 >>>>>>>>>>>> NUMA node1 CPU(s): 8-15,24-31 >>>>>>>>>>>> >>>>>>>>>>> -- >>> You received this message because you are subscribed to the Google >>> Groups "sage-support" group. >>> To unsubscribe from this group and stop receiving emails from it, send >>> an email to sage-support...@googlegroups.com. >>> To post to this group, send email to sage-s...@googlegroups.com. >>> Visit this group at https://groups.google.com/group/sage-support. >>> For more options, visit https://groups.google.com/d/optout. >>> >>> >>> -- You received this message because you are subscribed to the Google Groups "sage-support" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-support+unsubscr...@googlegroups.com. To post to this group, send email to sage-support@googlegroups.com. Visit this group at https://groups.google.com/group/sage-support. For more options, visit https://groups.google.com/d/optout.
