On Thu, Mar 5, 2020 at 6:25 PM Vincent Delecroix <20100.delecr...@gmail.com> wrote: > > Le 05/03/2020 à 19:18, Dima Pasechnik a écrit : > > On Thu, Mar 5, 2020 at 6:13 PM Vincent Delecroix > > <20100.delecr...@gmail.com> wrote: > >> > >> I believe it is a problem with pari version. Do you mind making a report > >> directly at > >> > >> https://github.com/sagemath/cypari2 > >> > >> Also, it would help if you provide the following information: > >> > >> 1. whether you have pari installed on your system (independently of > >> Sage)? > >> 2. if yes to 1, whether Sage configure does catch your pari from > >> the system (this information is available in config.log) > > > > from the config.log in the 1st message on the thread: > > > > configure:11939: === checking whether to install the pari SPKG === > > configure:11985: checking installing gmp/mpir or readline? > > configure:11991: result: no > > configure:11993: checking installing PARI/GP packages? > > configure:11997: checking for gp > > configure:12015: found /usr/bin/gp > > configure:12027: result: /usr/bin/gp > > configure:12039: checking is pari_elldata installed? > > configure:12043: result: yes > > configure:12054: checking is pari_galdata installed? > > configure:12058: result: yes > > configure:12069: checking is pari_galpol installed? > > *** at top-level: galoisgetname(12,1)= > > *** ^-------------------- > > *** not a function in function call > > configure:12076: result: no; cannot use system pari/GP without galpol > > package > > configure:12078: Install galpol package and reconfigure. > > configure:12080: Otherwise Sage will build its own pari/GP. > > configure:12084: checking is pari_seadata installed? > > configure:12088: result: yes > > configure:12328: result: using Sage's pari SPKG > > > > That is to say, Pari is installed system-wide, but galpol package is > > missing, and so Sage > > builds its own Pari. > > > > I suspect the error stems from a mix-up of paths, a system-wide Pari > > gets used instead of the Sage's Pari. > > Exactly, it is a problem with cypari2 that has to fight with two pari > installation. Some of the files of cypari2 are generated during the > build. And to do so it calls the gp interpreter (apparently the wrong > one). > > One quick workaround: rename (temporarily) your system gp to > HIDDEN_gp and try again.
I am not sure this will help. Specifically, the failing gcc call gcc -Wno-unused-result -Wsign-compare -DNDEBUG -g -fwrapv -O3 -Wall -Wno-unused -fPIC -I./cypari2 -I/home/alex/build/sage-develop/local/lib/python3.7/site-packages/cysignals -I/home/alex/build/sage-develop/local/include/python3.7m -c cypari2/gen.c -o build/temp.linux-x86_64-3.7/cypari2/gen.o cypari2/gen.c: In function '__pyx_pf_7cypari2_3gen_8Gen_base_40algabsdim': cypari2/gen.c:11949:18: warning: implicit declaration of function 'algabsdim'; did you mean 'algdim'? [-Wimplicit-function-declaration] __pyx_v__ret = algabsdim(__pyx_v__al); ^~~~~~~~~ algdim should have option -I/home/alex/build/sage-develop/local/include to get the correct Sage's Pari headers, but for some reason this does not happen. > > Vincent > > -- > You received this message because you are subscribed to the Google Groups > "sage-devel" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to sage-devel+unsubscr...@googlegroups.com. > To view this discussion on the web visit > https://groups.google.com/d/msgid/sage-devel/f0c94e86-754a-1c1e-37c2-042ee864362b%40gmail.com. -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/CAAWYfq0eonwCW1sc_15OK_%3DxH7cDv8vC27643jG%2BQhb%3Da-Fuqw%40mail.gmail.com.