Hi OK, removing libatlas3-base again on snapperkob. As I understand is the point is to make SAGE build regardless of whether this is installed.
I still misunderstand something. When you say "link in atlas" do you mean link in the system-wide atlas from libatlas3-base? This should not be necessary to test as it already works on muizenberg. If this was the problem on snapperkob, do we need to depend on system-wude libatlas3-base? I would want then also need to add that to the PPA, right. Also, wouldn't the build need libatlas-dev instead of only libatlas3-base? Excuse my ignorance here. So I can just git pull in a sage-6.4.tar.gz unpacked and get all the latest tickets? Or how do I test just that ticket. Sory, my git workflow is very basic, not very sophisticated yet. Also, how can I build just the R package and not make all of Sage and components? (Also, might it be the quickest way to get 6.4 in a binary is to add libatlas3-base and build, and add a dependency? Then fix this for 6.5?) Regards, Jan On 22 November 2014 14:55, Volker Braun <[email protected]> wrote: > Can you try out http://trac.sagemath.org/17349 instead? This should > always link in atlas, regardless of X11. If anything this should give us > consistent behavior. > > > > > On Saturday, November 22, 2014 12:47:27 PM UTC, pipedream wrote: >> >> Hi >> >> I think I'm getting this the wrong way around. Is it not that >> >> snapperkob build fails where there is NO system ATLAS installed, so it is >> using SAGE ATLAS. >> >> muizenberg works, where libatlas3-base is installed. >> >> How does a fix to NOT use the system ATLAS help, if the system ATLAS was >> not present on snapperkob? >> >> libx11-dev is installed on both. >> >> I am installaing libatlas3-base on snapperkob to try again. >> >> Note there is now load shedding in South Africa, and for the next three >> nights at 8pm South Africa time I expect snapperkob to be down for 2.5 >> hours. I hope this build finishes before that. >> >> Regards, >> Jan >> >> >> On 22 November 2014 01:09, Volker Braun <[email protected]> wrote: >> >>> I've added a commit to always use our ATLAS to http://trac.sagemath.org/ >>> ticket/17349, if nothing else then this will let us debug the issue in >>> a sane way... >>> >>> >>> On Friday, November 21, 2014 10:24:22 PM UTC, Volker Braun wrote: >>>> >>>> R only uses our ATLAS if X development stuff is installed. The whole >>>> logic in the spkg-install is suspicious... >>>> >>>> >>>> >>>> >>>> On Friday, November 21, 2014 7:02:52 PM UTC, pipedream wrote: >>>>> >>>>> Hi >>>>> >>>>> I didn't set SAGE_FAT_BINARY on either. >>>>> >>>>> Actually, I have atlas on muizenberg not snapperkob: >>>>> >>>>> 0 root@muizenberg:~#dpkg -l |grep atlas >>>>> ii libatlas3-base >>>>> 3.10.1-4 amd64 >>>>> Automatically Tuned Linear Algebra Software, generic shared >>>>> 0 root@muizenberg:~# >>>>> >>>>> >>>>> 1 jan@snapperkob:~/sage-6.4$dpkg -l |grep atlas >>>>> 1 jan@snapperkob:~/sage-6.4$ >>>>> >>>>> >>>>> Regards, >>>>> Jan >>>>> >>>>> On 21 November 2014 14:54, Volker Braun <[email protected]> wrote: >>>>> >>>>>> You didn't set SAGE_FAT_BINARY=yes I guess since in that case we >>>>>> configure R --without-ICU >>>>>> >>>>>> I take it you have Ubuntu's version of ATLAS installed on snapperkob >>>>>> and not on muizenberg? >>>>>> >>>>>> >>>>>> >>>>>> On Friday, November 21, 2014 8:21:28 AM UTC, pipedream wrote: >>>>>> >>>>>>> Hi Volker, >>>>>>> >>>>>>> I repeated on snapperkob, not via buildbot just as my own user: >>>>>>> >>>>>>> real 125m41.153s >>>>>>> user 208m24.169s >>>>>>> sys 20m6.421s >>>>>>> *************************************************************** >>>>>>> Error building Sage. >>>>>>> >>>>>>> The following package(s) may have failed to build: >>>>>>> >>>>>>> package: r-3.1.1.p0 >>>>>>> log file: /home/jan/sage-6.4/logs/pkgs/r-3.1.1.p0.log >>>>>>> build directory: /home/jan/sage-6.4/local/var/t >>>>>>> mp/sage/build/r-3.1.1.p0 >>>>>>> >>>>>>> The build directory may contain configuration files and other >>>>>>> potentially >>>>>>> helpful information. WARNING: if you now run 'make' again, the build >>>>>>> directory will, by default, be deleted. Set the environment variable >>>>>>> SAGE_KEEP_BUILT_SPKGS to 'yes' to prevent this. >>>>>>> >>>>>>> make: *** [build] Error 1 >>>>>>> 2 jan@snapperkob:~/sage-6.4$ >>>>>>> >>>>>>> I can try investigate what is different on muizenberg and snapperkob >>>>>>> or my user on muizenberg & snapperkob. >>>>>>> >>>>>>> They are both Dell Optiplex, 64 bit, entirely up to date Ubuntu >>>>>>> 14.04.1, though: >>>>>>> >>>>>>> snapperkob is a text-only server install >>>>>>> muizenberg is a GUI full desktop install >>>>>>> >>>>>>> There is little difference underneath in ubuntu though. It is the >>>>>>> same kernel. (the package linux-server depends on linux-generic which is >>>>>>> also on the desktop, and depends on the actual kernel) >>>>>>> >>>>>>> They both have the Intel E8400 CPU. They both have 8G RAM. >>>>>>> >>>>>>> Comparing the logs for the two R builds, I found these differences >>>>>>> in the configure parts, due to different packages available on the >>>>>>> system: >>>>>>> >>>>>>> muizenberg (built successfully) >>>>>>> >>>>>>> R is now configured for x86_64-unknown-linux-gnu >>>>>>> >>>>>>> Source directory: . >>>>>>> Installation directory: /srv/sysadm/sage/sage-6.4/local >>>>>>> >>>>>>> C compiler: gcc -std=gnu99 -g -O2 >>>>>>> Fortran 77 compiler: gfortran -g -O2 >>>>>>> >>>>>>> C++ compiler: g++ >>>>>>> C++ 11 compiler: g++ -std=c++11 >>>>>>> Fortran 90/95 compiler: gfortran -g -O2 >>>>>>> Obj-C compiler: >>>>>>> >>>>>>> Interfaces supported: >>>>>>> External libraries: readline, ICU, lzma >>>>>>> Additional capabilities: PNG, JPEG, TIFF, NLS, cairo >>>>>>> Options enabled: shared R library, shared BLAS, R >>>>>>> profiling >>>>>>> >>>>>>> Recommended packages: yes >>>>>>> >>>>>>> >>>>>>> And snapperkob (failed): >>>>>>> >>>>>>> R is now configured for x86_64-unknown-linux-gnu >>>>>>> >>>>>>> Source directory: . >>>>>>> Installation directory: /home/jan/sage-6.4/local >>>>>>> >>>>>>> C compiler: gcc -std=gnu99 -g -O2 >>>>>>> Fortran 77 compiler: gfortran -g -O2 >>>>>>> >>>>>>> C++ compiler: g++ >>>>>>> C++ 11 compiler: g++ -std=c++11 >>>>>>> Fortran 90/95 compiler: gfortran -g -O2 >>>>>>> Obj-C compiler: >>>>>>> >>>>>>> Interfaces supported: X11, tcltk >>>>>>> External libraries: readline, BLAS(ATLAS), LAPACK(generic), >>>>>>> ICU >>>>>>> Additional capabilities: PNG, JPEG, NLS, cairo >>>>>>> Options enabled: shared R library, R profiling >>>>>>> >>>>>>> Recommended packages: yes >>>>>>> >>>>>>> >>>>>>> Can you think what I can change on snapperkob to test? BLAS or >>>>>>> LAPACK perhaps? >>>>>>> >>>>>>> Regards, >>>>>>> Jan >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> On 20 November 2014 20:35, Jan Groenewald <[email protected]> wrote: >>>>>>> >>>>>>>> Different machine: >>>>>>>> >>>>>>>> 0 root@muizenberg:/srv/sysadm/sage#dpkg -S >>>>>>>> /usr/lib/x86_64-linux-gnu/libgomp.so.* >>>>>>>> libgomp1:amd64: /usr/lib/x86_64-linux-gnu/libgomp.so.1 >>>>>>>> libgomp1:amd64: /usr/lib/x86_64-linux-gnu/libgomp.so.1.0.0 >>>>>>>> 0 root@muizenberg:/srv/sysadm/sage#dpkg -l libgomp1 >>>>>>>> Desired=Unknown/Install/Remove/Purge/Hold >>>>>>>> | Status=Not/Inst/Conf-files/Unpacked/halF-conf/Half-inst/trig >>>>>>>> -aWait/Trig-pend >>>>>>>> |/ Err?=(none)/Reinst-required (Status,Err: uppercase=bad) >>>>>>>> ||/ Name >>>>>>>> Version Architecture Description >>>>>>>> +++-=======================================-================ >>>>>>>> ========-========================-========================== >>>>>>>> ========================================================= >>>>>>>> ii libgomp1:amd64 >>>>>>>> 4.8.2-19ubuntu1 amd64 GCC OpenMP (GOMP) >>>>>>>> support >>>>>>>> library >>>>>>>> 0 root@muizenberg:/srv/sysadm/sage# >>>>>>>> >>>>>>>> Same version: >>>>>>>> >>>>>>>> 0 jan@snapperkob:~/src$dpkg -S /usr/lib/x86_64-linux-gnu/libg >>>>>>>> omp.so.* >>>>>>>> libgomp1:amd64: /usr/lib/x86_64-linux-gnu/libgomp.so.1 >>>>>>>> libgomp1:amd64: /usr/lib/x86_64-linux-gnu/libgomp.so.1.0.0 >>>>>>>> 0 jan@snapperkob:~/src$dpkg -l libgomp1 >>>>>>>> Desired=Unknown/Install/Remove/Purge/Hold >>>>>>>> | Status=Not/Inst/Conf-files/Unpacked/halF-conf/Half-inst/trig >>>>>>>> -aWait/Trig-pend >>>>>>>> |/ Err?=(none)/Reinst-required (Status,Err: uppercase=bad) >>>>>>>> ||/ Name >>>>>>>> Version Architecture Description >>>>>>>> +++-=======================================-================ >>>>>>>> ========-========================-========================== >>>>>>>> ========================================================= >>>>>>>> ii libgomp1:amd64 >>>>>>>> 4.8.2-19ubuntu1 amd64 GCC OpenMP (GOMP) >>>>>>>> support >>>>>>>> library >>>>>>>> 0 jan@snapperkob:~/src$ >>>>>>>> >>>>>>>> Regards, >>>>>>>> Jan >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> On 20 November 2014 18:32, Volker Braun <[email protected]> wrote: >>>>>>>> >>>>>>>>> Is this on "snapperkob" or on a different machine? Do you >>>>>>>>> have /usr/lib/x86_64-linux-gnu/libgomp.so.* and if yes, which >>>>>>>>> version? >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> On Thursday, November 20, 2014 12:02:18 PM UTC, pipedream wrote: >>>>>>>>>> >>>>>>>>>> Hi >>>>>>>>>> >>>>>>>>>> I rebuilt sage from a cleanly unpacked source tree on Ubuntu >>>>>>>>>> 14.04, and it still did not fail on R >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> export SAGE_INSTALL_GCC=yes >>>>>>>>>> make >>>>>>>>>> ... >>>>>>>>>> Successfully installed r-3.1.1.p0 >>>>>>>>>> >>>>>>>>>> I'd like to assist running any tests to get this fixed. >>>>>>>>>> >>>>>>>>>> Regards, >>>>>>>>>> Jan >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> On 19 November 2014 09:09, Jan Groenewald <[email protected]> >>>>>>>>>> wrote: >>>>>>>>>> >>>>>>>>>>> Hi >>>>>>>>>>> >>>>>>>>>>> I did >>>>>>>>>>> >>>>>>>>>>> $SAGE_INSTALL_GCC=yes make >>>>>>>>>>> >>>>>>>>>>> in the already built source tree, and it seemed to rebuild the >>>>>>>>>>> relevant parts without failing. Is that enough or should I try it >>>>>>>>>>> in a >>>>>>>>>>> cleanly unpacked source tree? >>>>>>>>>>> >>>>>>>>>>> Regards, >>>>>>>>>>> Jan >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> On 18 November 2014 15:56, Jan Groenewald <[email protected]> >>>>>>>>>>> wrote: >>>>>>>>>>> >>>>>>>>>>>> Will do. >>>>>>>>>>>> >>>>>>>>>>>> Regards, >>>>>>>>>>>> Jan >>>>>>>>>>>> >>>>>>>>>>>> >>>>>>>>>>>> On 18 November 2014 15:29, Volker Braun <[email protected]> >>>>>>>>>>>> wrote: >>>>>>>>>>>> >>>>>>>>>>>>> Try building with SAGE_INSTALL_GCC=yes, I'm pretty sure this >>>>>>>>>>>>> is causing the problem. >>>>>>>>>>>>> >>>>>>>>>>>>> On Tuesday, November 18, 2014 11:56:03 AM UTC, pipedream wrote: >>>>>>>>>>>>>> >>>>>>>>>>>>>> Hi >>>>>>>>>>>>>> >>>>>>>>>>>>>> This is strange, the error I see is: >>>>>>>>>>>>>> >>>>>>>>>>>>>> /home/buildbot/build/sage/snapperkob/sage_git/build/local/var/tmp/sage/build/r-3.1.1.p0/src/bin/exec/R: >>>>>>>>>>>>>> /usr/lib/x86_64-linux-gnu/libgomp.so.1: version `GOMP_4.0' not >>>>>>>>>>>>>> found (required by >>>>>>>>>>>>>> /home/buildbot/build/sage/snapperkob/sage_git/build/local/var/tmp/sage/build/r-3.1.1.p0/src/lib/libR.so) >>>>>>>>>>>>>> >>>>>>>>>>>>>> I have the same version of libgomp1 (4.8.2-19ubuntu1) >>>>>>>>>>>>>> installed on my 14.04 64bit desktop as that builder (snapperkob) >>>>>>>>>>>>>> and yet >>>>>>>>>>>>>> sage 6.4 built succesfully on my desktop yesterday: >>>>>>>>>>>>>> >>>>>>>>>>>>>> "Successfully installed r-3.1.1.p0" >>>>>>>>>>>>>> >>>>>>>>>>>>>> Regards, >>>>>>>>>>>>>> Jan >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> On 18 November 2014 13:22, Volker Braun <[email protected]> >>>>>>>>>>>>>> wrote: >>>>>>>>>>>>>> >>>>>>>>>>>>>>> The build failed because R insists on using openmp for some >>>>>>>>>>>>>>> reason, but the openmp libraries are not installed (or of the >>>>>>>>>>>>>>> wrong >>>>>>>>>>>>>>> version): >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> http://build.sagedev.org/release/builders/Zbin%20%20fast%20A >>>>>>>>>>>>>>> IMS%20snapperkob%20%28Ubuntu%2014.04%20x86_64%29%20binary/bu >>>>>>>>>>>>>>> ilds/1/steps/compile_2/logs/r-project >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> On Tuesday, November 18, 2014 5:00:26 AM UTC, pipedream >>>>>>>>>>>>>>> wrote: >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> Hi >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> I do not see the 14.04 Ubuntu binaries here. >>>>>>>>>>>>>>>> http://boxen.math.washington.edu/home/sagemath/sage-mirror/ >>>>>>>>>>>>>>>> linux/64bit/index.html >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> Are they not automatically generated? >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> Regards, >>>>>>>>>>>>>>>> Jan >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> On 17 November 2014 17:17, kcrisman <[email protected]> >>>>>>>>>>>>>>>> wrote: >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> The release notes have a broken link in them: >>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>> http://boxen.math.washington. >>>>>>>>>>>>>>>>>> edu/home/release/sage-6.4/tickets.html >>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> Yes, that has been reported before... >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> By the way, Jeroen, I think I finished all the changes I'm >>>>>>>>>>>>>>>>> going to for now. Long-term, it would be fun to do some >>>>>>>>>>>>>>>>> database thingie >>>>>>>>>>>>>>>>> to standardize names of a few very frequent contributors who >>>>>>>>>>>>>>>>> have them in a >>>>>>>>>>>>>>>>> couple different formats. >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> >>>>>>>>>> -- >>>>>>>>>> .~. >>>>>>>>>> /V\ Jan Groenewald >>>>>>>>>> /( )\ www.aims.ac.za >>>>>>>>>> ^^-^^ >>>>>>>>>> >>>>>>>>> -- >>>>>>>>> You received this message because you are subscribed to the Google >>>>>>>>> Groups "sage-release" group. >>>>>>>>> To unsubscribe from this group and stop receiving emails from it, >>>>>>>>> send an email to [email protected]. >>>>>>>>> To post to this group, send email to [email protected]. >>>>>>>>> Visit this group at http://groups.google.com/group/sage-release. >>>>>>>>> For more options, visit https://groups.google.com/d/optout. >>>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> -- >>>>>>>> .~. >>>>>>>> /V\ Jan Groenewald >>>>>>>> /( )\ www.aims.ac.za >>>>>>>> ^^-^^ >>>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> -- >>>>>>> .~. >>>>>>> /V\ Jan Groenewald >>>>>>> /( )\ www.aims.ac.za >>>>>>> ^^-^^ >>>>>>> >>>>>> -- >>>>>> You received this message because you are subscribed to the Google >>>>>> Groups "sage-release" group. >>>>>> To unsubscribe from this group and stop receiving emails from it, >>>>>> send an email to [email protected]. >>>>>> To post to this group, send email to [email protected]. >>>>>> Visit this group at http://groups.google.com/group/sage-release. >>>>>> For more options, visit https://groups.google.com/d/optout. >>>>>> >>>>> >>>>> >>>>> >>>>> -- >>>>> .~. >>>>> /V\ Jan Groenewald >>>>> /( )\ www.aims.ac.za >>>>> ^^-^^ >>>>> >>>> -- >>> You received this message because you are subscribed to the Google >>> Groups "sage-release" group. >>> To unsubscribe from this group and stop receiving emails from it, send >>> an email to [email protected]. >>> To post to this group, send email to [email protected]. >>> Visit this group at http://groups.google.com/group/sage-release. >>> For more options, visit https://groups.google.com/d/optout. >>> >> >> >> >> -- >> .~. >> /V\ Jan Groenewald >> /( )\ www.aims.ac.za >> ^^-^^ >> > -- > You received this message because you are subscribed to the Google Groups > "sage-release" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To post to this group, send email to [email protected]. > Visit this group at http://groups.google.com/group/sage-release. > For more options, visit https://groups.google.com/d/optout. > -- .~. /V\ Jan Groenewald /( )\ www.aims.ac.za ^^-^^ -- You received this message because you are subscribed to the Google Groups "sage-release" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/sage-release. For more options, visit https://groups.google.com/d/optout.
