> On May 5, 5:44 pm, Francois Bissey <francois.bis...@canterbury.ac.nz> > > wrote: > > > On May 5, 6:35 am, Francois Bissey <francois.bis...@canterbury.ac.nz> > > > > > > wrote: > > > > > On May 4, 8:43 pm, Francois Bissey > > > > > <francois.bis...@canterbury.ac.nz> > > > > > > > > > > wrote: > > > > > > > same problem if I try building a standalone ecl, and if I use > > > > > > > gcc 4.0.1, too. > > > > > > > So this looks like the combination of OSX 10.5 and PPC that is > > > > > > > not working here, nothing Sage-specific. > > > > > > > > > > > > > > Dima > > > > > > > > > > > > > > On May 4, 3:47 pm, Dima Pasechnik <dimp...@gmail.com> wrote: > > > > > > > > On MacOSX 10.5.8 (PPC G4) I cannot proceed past ecl. > > > > > > > > I get crashes (sort or random) after bare.lsp is loaded, > > > > > > > > ie after the log lines > > > > > > > > ;;; About to load cmp/load.lsp > > > > > > > > ;;; > > > > > > > > ;;; Now we are in shape to do something useful. > > > > > > > > ;;; End of bare.lsp > > > > > > > > > > > > Darn, I wish I still had my G4, this is familiar. Can you post > > > > > > the whole log? > > > > > > > > > > please > > > > > seehttp://www1.spms.ntu.edu.sg/~dima/tmp/ppc_ecl_problem_4.7.rc1.gz > > > > > (there are repeated tries to restart the build at the end, > > > > > sorry...) > > > > > > > > > > there are strange messages about sse there, e.g. > > > > > Unknown symbol: ext::sse-pack-as-elt-type > > > > > > > > > > Unknown symbol: ext::sse-pack > > > > > > > > > > Does it misconfigure? (it's has no SSE support, this machine) > > > > > > > > The sse messages are not abnormal. I get them on a amd64 machine > > > > which has sse instructions. I know for a fact that ecl has an > > > > allergy too altivec instructions that results in that kind of stuff > > > > but it isn't in your CFLAGS, unless it has been forced down inside > > > > gcc. > > > > Seehttp://bugs.gentoo.org/show_bug.cgi?id=308909in particular the log > > > > with gcc-4.3.4. > > > > So could you try to resume the build with: > > > > CFLAGS="-mno-altivec -mabi=no-altivec" make > > > > I am hoping that these options are as valid for ppc-macos as they > > > > are for ppc-linux. > > > > > > Indeed, this helps! Although the CFLAGS should be passed to configure. > > > (just 'CFLAGS="-mno-altivec -mabi=no-altivec" make' does not pass > > > CFLAGS anywhere!) > > > This is now trac #11297 > > > > > > > You didn't have any problem with previous ecl did you? > > > > > > No. (Sage 4.6 builds and works on this machine) > > > > Something I should add here. When I had my ppc machine and was > > extending support to Gentoo linux-ppc I also needed to add these > > flags to the compilation of ecl.pyx inside sage itself. Otherwise you hit > > the same wall. > > didn't seem to be needing this. > What were the symptoms at your 2nd wall? > > The build has finished with only the patch from trac #11297 > applied, running tests now. > Apart from the build breaking? If you don't need anything else that's all good. I faintly remember that I had the same kind of error as in http://bugs.gentoo.org/attachment.cgi?id=223091 which may be because when adding -maltivec -mabi=altivec gcc4.4+ doesn't seem to want to do -std=c99 by default or something along these lines. Just adding -std=c99 probably would have enabled the compilation.
Let's see if the test in sage/libs/ecl.pyx pass, if not we may know where there is a miscompilation problem. But with my memory coming back I think it may only be affecting later version of gcc. Francois This email may be confidential and subject to legal privilege, it may not reflect the views of the University of Canterbury, and it is not guaranteed to be virus free. If you are not an intended recipient, please notify the sender immediately and erase all copies of the message and any attachments. Please refer to http://www.canterbury.ac.nz/emaildisclaimer for more information. -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
