In fact, ./configure of ECL has an option  --with-libffi-prefix=path
so most probably all we need to do is to point it at the right place.
Try

ECL_CONFIGURE=`pwd`/local  ./sage -i ecl



On Fri, Jan 31, 2020 at 9:33 PM Emmanuel Charpentier
<[email protected]> wrote:
>
>
>
> Le vendredi 31 janvier 2020 22:14:59 UTC+1, Dima Pasechnik a écrit :
>>
>> On Fri, Jan 31, 2020 at 8:58 PM Emmanuel Charpentier
>> <[email protected]> wrote:
>> >
>> >
>> >
>> > Le vendredi 31 janvier 2020 21:50:55 UTC+1, Dima Pasechnik a écrit :
>> >>
>> >> Please send SAGE_ROOT/config.log
>>
>> this looks good, and in fact libffi gets built; indeed, your install.log 
>> says:
>>
>> [libffi-3.2.1] Copying package files from temporary location
>> /usr/local/sage-9/local/var/tmp/sage/build/libffi-3.2.1/inst to
>> /usr/local/sage-9/local
>> [libffi-3.2.1] Successfully installed libffi-3.2.1
>> [libffi-3.2.1] Deleting temporary build directory
>> [libffi-3.2.1] /usr/local/sage-9/local/var/tmp/sage/build/libffi-3.2.1
>> [libffi-3.2.1] Finished installing libffi-3.2.1.spkg
>>
>> You might want to check for presense of libffi.so in local/lib/
>> but I'm positive it's there,
>
>
> You're right. So why didn't I get the compilation log ?
>
>>
>> so it's ECL build that's failing for
>> unrelated reasons, probably cause your  headers are a bit broken
>> somewhere:
>> I guess ECL tries to use system headers, even though it should not.
>>
>> My conjecture is that it was always the case, and it went unnoticed
>> until libffi 3.3 was released, with something broken.
>> Previously, even if libffi-dev (Debian) was installed, its headers
>> were similar enough to the one supplied by Sage to work,
>> so if they were picked up then it still worked.
>>
>> To test this you might try to "apt remove libffi-dev" and then "./sage
>> -i ecl"; if I'm right here then it will work.
>
>
> I can do that but as a very temporary measure: removing libffi-dev break 
> other parts of my system :
>
> charpent@zen-book-flip:/usr/local/sage-9$ sudo apt-get remove -s libffi-dev
> Lecture des listes de paquets... Fait
> Construction de l'arbre des dépendances
> Lecture des informations d'état... Fait
> Les paquets suivants ont été installés automatiquement et ne sont plus 
> nécessaires :
>   gir1.2-poppler-0.18 libblkid-dev libbsd-dev libcairo-script-interpreter2
>   libglib2.0-dev-bin libmount-dev libpixman-1-dev libselinux1-dev
>   libsepol1-dev libxcb-render0-dev libxcb-shm0-dev
> Veuillez utiliser « sudo apt autoremove » pour les supprimer.
> Les paquets suivants seront ENLEVÉS :
>   ghc libcairo2-dev libffi-dev libglib2.0-dev libpoppler-glib-dev
> 0 mis à jour, 0 nouvellement installés, 5 à enlever et 0 non mis à jour.
> Remv ghc [8.8.1+dfsg1+is+8.6.5+dfsg1-2+b1]
> Remv libpoppler-glib-dev [0.71.0-6]
> Remv libcairo2-dev [1.16.0-4]
> Remv libglib2.0-dev [2.62.4-1+b1]
> Remv libffi-dev [3.3-3]
>
> I suippose I have to do the whole cycle (distclean ; configure ; make) 
> again... Stay tuned.
>
>>
>>
>>
>>
>>
>> >
>> >
>> > Attached.
>> >
>> >>
>> >>
>> >> this is the most interesting one, to understand why libffi didn't get 
>> >> built
>> >>
>> >> On Fri, Jan 31, 2020 at 7:58 PM Emmanuel Charpentier
>> >> <[email protected]> wrote:
>> >> >
>> >> > Damn !
>> >> >
>> >> > After make distclean and ./configure --without-system-libffi, make 
>> >> > fails with the same error as before.
>> >> >
>> >> > AND LIBFFI ISN'T COMPILED (no libffi* in logs/pkgs).
>> >> >
>> >> > Relevant logs attached.
>> >> >
>> >> > Le vendredi 31 janvier 2020 20:37:03 UTC+1, Emmanuel Charpentier a 
>> >> > écrit :
>> >> >>
>> >> >>
>> >> >>
>> >> >> Le vendredi 31 janvier 2020 20:30:39 UTC+1, Dima Pasechnik a écrit :
>> >> >>>
>> >> >>> just as a sanity check, please verify that Sage's libffi got built.
>> >> >>
>> >> >>
>> >> >> It did. No joy.
>> >> >>
>> >> >> Currently re-makeing after distclean. Don't hold your breath and have 
>> >> >> a nice walk...
>> >> >>
>> >> >>>
>> >> >>>
>> >> >>> On Fri, Jan 31, 2020 at 7:22 PM Emmanuel Charpentier
>> >> >>> <[email protected]> wrote:
>> >> >>> >
>> >> >>> > Indeed : make has failed with the same failure as before. To 
>> >> >>> > confirm this, I'll distclean and retry, but don't believe it will 
>> >> >>> > be successful. Kep you posted.
>> >> >>> >
>> >> >>> > HTH,
>> >> >>> >
>> >> >>> > Le vendredi 31 janvier 2020 20:17:57 UTC+1, Emmanuel Charpentier a 
>> >> >>> > écrit :
>> >> >>> >>
>> >> >>> >>
>> >> >>> >>
>> >> >>> >> Le vendredi 31 janvier 2020 16:06:12 UTC+1, Dima Pasechnik a écrit 
>> >> >>> >> :
>> >> >>> >>>
>> >> >>> >>> have you updated system's libffi lately?
>> >> >>> >>
>> >> >>> >>
>> >> >>> >> Possibly: I follow Debian testing, but do not look closely each 
>> >> >>> >> and every upgrade. Currently, I have :
>> >> >>> >>
>> >> >>> >> charpent@zen-book-flip:/usr/local/sage-9$ dpkg -l "libffi*" | cat
>> >> >>> >> Souhait=inconnU/Installé/suppRimé/Purgé/H=à garder
>> >> >>> >> | 
>> >> >>> >> État=Non/Installé/fichier-Config/dépaqUeté/échec-conFig/H=semi-installé/W=attend-traitement-déclenchements
>> >> >>> >> |/ Err?=(aucune)/besoin Réinstallation (État,Err: 
>> >> >>> >> majuscule=mauvais)
>> >> >>> >> ||/ Nom              Version      Architecture Description
>> >> >>> >> +++-================-============-============-======================================================
>> >> >>> >> ii  libffi-dev:amd64 3.3-3        amd64        Foreign Function 
>> >> >>> >> Interface library (development files)
>> >> >>> >> un  libffi4-dev      <aucune>     <aucune>     (aucune description 
>> >> >>> >> n'est disponible)
>> >> >>> >> ii  libffi6:amd64    3.2.1-9      amd64        Foreign Function 
>> >> >>> >> Interface library runtime
>> >> >>> >> ii  libffi6:i386     3.2.1-9      i386         Foreign Function 
>> >> >>> >> Interface library runtime
>> >> >>> >> ii  libffi7:amd64    3.3-3        amd64        Foreign Function 
>> >> >>> >> Interface library runtime
>> >> >>> >> ii  libffi7:i386     3.3-3        i386         Foreign Function 
>> >> >>> >> Interface library runtime
>> >> >>> >>
>> >> >>> >>>
>> >> >>> >>> Could you try
>> >> >>> >>>
>> >> >>> >>> ./configure --without-system-libffi
>> >> >>> >>> make
>> >> >>> >>>
>> >> >>> >>> ?
>> >> >>> >>
>> >> >>> >>
>> >> >>> >> Underway. I'm not optimistic: ecl has been recompiled after 
>> >> >>> >> reconfiguration, and the log file still ends with :
>> >> >>> >>
>> >> >>> >> gcc -DECLDIR="\"/usr/local/sage-9/local/lib/ecl-16.1.2\"" -I. 
>> >> >>> >> -I/usr/local/sage-9/local/var/tmp/sage/build/ecl-16.1.2.p5/src/build
>> >> >>> >>  
>> >> >>> >> -I/usr/local/sage-9/local/var/tmp/sage/build/ecl-16.1.2.p5/src/src/c
>> >> >>> >>  -I../ecl/gc -DECL_API -DECL_NO_LEGACY     -D_GNU_SOURCE 
>> >> >>> >> -D_FILE_OFFSET_BITS=64 -g -O2  -fPIC -Dlinux -c -o ffi.o ffi.o.c
>> >> >>> >> /usr/local/sage-9/local/var/tmp/sage/build/ecl-16.1.2.p5/src/src/c/ffi.d:148:34:
>> >> >>> >>  error: 'FFI_SYSV' undeclared here (not in a function)
>> >> >>> >>   148 |         {@':cdecl', FFI_SYSV},
>> >> >>> >>       |                                  ^
>> >> >>> >> make[6]: *** [Makefile:86: ffi.o] Error 1
>> >> >>> >> make[6]: Leaving directory 
>> >> >>> >> '/usr/local/sage-9/local/var/tmp/sage/build/ecl-16.1.2.p5/src/build/c'
>> >> >>> >> make[5]: *** [Makefile:119: libeclmin.a] Error 2
>> >> >>> >> make[5]: Leaving directory 
>> >> >>> >> '/usr/local/sage-9/local/var/tmp/sage/build/ecl-16.1.2.p5/src/build'
>> >> >>> >> make[4]: *** [Makefile:70: all] Error 2
>> >> >>> >> make[4]: Leaving directory 
>> >> >>> >> '/usr/local/sage-9/local/var/tmp/sage/build/ecl-16.1.2.p5/src'
>> >> >>> >> ********************************************************************************
>> >> >>> >> Error building ecl-16.1.2.p5
>> >> >>> >> ********************************************************************************
>> >> >>> >>
>> >> >>> >> I'll keep you posted.
>> >> >>> >>
>> >> >>> >>>
>> >> >>> >>>
>> >> >>> >>> On Fri, Jan 31, 2020 at 2:45 PM Emmanuel Charpentier
>> >> >>> >>> <[email protected]> wrote:
>> >> >>> >>> >
>> >> >>> >>> > On the same machine, I can't rebuild Sage from scratch (in a 
>> >> >>> >>> > different directiory). ./configure + make distclean + make 
>> >> >>> >>> > ptestalllong fails at building ECL, which complaints:
>> >> >>> >>> >
>> >> >>> >>> > gcc -DECLDIR="\"/usr/local/sage-9/local/lib/ecl-16.1.2\"" -I. 
>> >> >>> >>> > -I/usr/local/sage-
>> >> >>> >>> > 9/local/var/tmp/sage/build/ecl-16.1.2.p5/src/build 
>> >> >>> >>> > -I/usr/local/sage-9/local/var
>> >> >>> >>> > /tmp/sage/build/ecl-16.1.2.p5/src/src/c -I../ecl/gc -DECL_API 
>> >> >>> >>> > -DECL_NO_LEGACY
>> >> >>> >>> >   -D_GNU_SOURCE -D_FILE_OFFSET_BITS=64 -g -O2  -fPIC -Dlinux -c 
>> >> >>> >>> > -o ffi.o ffi.o.c
>> >> >>> >>> > /usr/local/sage-9/local/var/tmp/sage/build/ecl-16.1.2.p5/src/src/c/ffi.d:148:34:
>> >> >>> >>> >  error: 'FFI_SYSV' undeclared here (not in a function)
>> >> >>> >>> >   148 |         {@':cdecl', FFI_SYSV},
>> >> >>> >>> >       |                                  ^
>> >> >>> >>> > make[6]: *** [Makefile:86: ffi.o] Error 1
>> >> >>> >>> > make[6]: Leaving directory 
>> >> >>> >>> > '/usr/local/sage-9/local/var/tmp/sage/build/ecl-16.1.
>> >> >>> >>> > 2.p5/src/build/c'
>> >> >>> >>> > make[5]: *** [Makefile:119: libeclmin.a] Error 2
>> >> >>> >>> > make[5]: Leaving directory 
>> >> >>> >>> > '/usr/local/sage-9/local/var/tmp/sage/build/ecl-16.1.
>> >> >>> >>> > 2.p5/src/build'
>> >> >>> >>> > make[4]: *** [Makefile:70: all] Error 2
>> >> >>> >>> > make[4]: Leaving directory 
>> >> >>> >>> > '/usr/local/sage-9/local/var/tmp/sage/build/ecl-16.1.2.p5/src'
>> >> >>> >>> > ********************************************************************************
>> >> >>> >>> > Error building ecl-16.1.2.p5
>> >> >>> >>> > ********************************************************************************
>> >> >>> >>> >
>> >> >>> >>> > Relevant logs attached.
>> >> >>> >>> >
>> >> >>> >>> > Le mardi 28 janvier 2020 16:15:21 UTC+1, Emmanuel Charpentier a 
>> >> >>> >>> > écrit :
>> >> >>> >>> >>
>> >> >>> >>> >> On Debian testing running on core I5 + 8 GB RAM, upgrading 
>> >> >>> >>> >> from Python 3-based 9.1.beta1 to 9.1.beta2:
>> >> >>> >>> >>
>> >> >>> >>> >> passes ptestlong without failure;
>> >> >>> >>> >> does NOT pass ptestalllong, which gets interrupted by multiple 
>> >> >>> >>> >> failures while doctesting findstat.py : see attached 
>> >> >>> >>> >> ptestalllong.log.
>> >> >>> >>> >>
>> >> >>> >>> >> I had similar failures with 9.1.beta1 on another machine, 
>> >> >>> >>> >> which I (probably wrongly) ended up attributing to transient 
>> >> >>> >>> >> networking problems (a repeat attempt succeeded). I'll test 
>> >> >>> >>> >> 9.1.beta2 on this second machine.
>> >> >>> >>> >>
>> >> >>> >>> >> HTH;
>> >> >>> >>> >>
>> >> >>> >>> >> Le mardi 28 janvier 2020 00:04:20 UTC+1, Volker Braun a écrit :
>> >> >>> >>> >>>
>> >> >>> >>> >>> As always, you can get the latest beta version from the 
>> >> >>> >>> >>> "develop" git branch. Alternatively, the self-contained 
>> >> >>> >>> >>> source tarball is at 
>> >> >>> >>> >>> http://www.sagemath.org/download-latest.html
>> >> >>> >>> >>>
>> >> >>> >>> >>>
>> >> >>> >>> >>> 2cbd93e0b7 (tag: 9.1.beta2, trac/develop) Updated SageMath 
>> >> >>> >>> >>> version to 9.1.beta2
>> >> >>> >>> >>> fb9218e13b Trac #29072: unused import cleaning ticket 
>> >> >>> >>> >>> (partial)
>> >> >>> >>> >>> 7cec05e0c7 Trac #28755: "The following package(s) may have 
>> >> >>> >>> >>> failed to build": Improve output
>> >> >>> >>> >>> bb5110ca19 Trac #27315: Minor optimizations to libsingular 
>> >> >>> >>> >>> polynomials
>> >> >>> >>> >>> 18389bdd76 Trac #29070: Queer tableaux for B(-infinity) and 
>> >> >>> >>> >>> B(lambda)
>> >> >>> >>> >>> bdfebfb180 Trac #29069: Fix broken link in crystals catalog
>> >> >>> >>> >>> 96695b33cf Trac #29068: little cleanup of imports and tweak 
>> >> >>> >>> >>> again .lgtm.yml
>> >> >>> >>> >>> 4da45e2169 Trac #29067: Make ./configure --help a bit shorter 
>> >> >>> >>> >>> by combining --with-system-SPKG and --with-system-SPK=force
>> >> >>> >>> >>> 1c5d011b5f Trac #29038: Python package sage_conf: Provides 
>> >> >>> >>> >>> optional configuration information for sagelib
>> >> >>> >>> >>> 9cf951e6ea Trac #29007: faster permutation conversions 
>> >> >>> >>> >>> between GAP and Sage
>> >> >>> >>> >>> aff7f32ac2 Trac #28247: Parametrize the cube/hypercube 
>> >> >>> >>> >>> functions in the library of polytopes
>> >> >>> >>> >>> 0d63d6e276 Trac #21413: A class for ring extensions
>> >> >>> >>> >>> 349a5ab632 Trac #29044: Add default warnings filters in a 
>> >> >>> >>> >>> Python debug build
>> >> >>> >>> >>> 7de6a929a5 Trac #29046: Minor improvements in 
>> >> >>> >>> >>> sage.rings.number_field.galois_group
>> >> >>> >>> >>> 35411d34b1 Trac #29011: speed up number field element 
>> >> >>> >>> >>> conversions
>> >> >>> >>> >>> 10ad03d047 Trac #28787: _add_ method for tensors with indices
>> >> >>> >>> >>> 1a1e1a0935 Trac #4618: Puiseux series
>> >> >>> >>> >>> 3c92cbc37c Trac #29063: spkg-configure.m4 for planarity
>> >> >>> >>> >>> 229d4365d8 Trac #29057: make stack method of Polyhedron use 
>> >> >>> >>> >>> fraction_field
>> >> >>> >>> >>> ed23ee3c7d Trac #29049: partially reverse the logic of py2 / 
>> >> >>> >>> >>> py3 doctest fixing
>> >> >>> >>> >>> 788a2d642f Trac #29048: 1 doctest failing in 
>> >> >>> >>> >>> src/sage/databases/oeis.py with tag internet
>> >> >>> >>> >>> 3dcf55bec6 Trac #29047: 1 internet doctest failing in 
>> >> >>> >>> >>> repl/load.py
>> >> >>> >>> >>> fcc768db49 Trac #28989: Bug in proper cycles of indefinite 
>> >> >>> >>> >>> binary quadratic forms
>> >> >>> >>> >>> f1dc86f78b Trac #29026: update m4ri(e) to Jan 2020 versions
>> >> >>> >>> >>> 3d8399c96a Trac #29035: some details in free pre-Lie algebras
>> >> >>> >>> >>> 3f9ef468a2 Trac #29002: spkg-configure.m4 for sqlite
>> >> >>> >>> >>> 59f6a83618 Trac #28999: qepcad interface is broken with 
>> >> >>> >>> >>> python3
>> >> >>> >>> >>> c10179b2eb Trac #28996: Have `implementation='NTL'` work for 
>> >> >>> >>> >>> series over ZZ
>> >> >>> >>> >>> 44345faf0b Trac #28616: CombinatorialFace: replace Vrepr() 
>> >> >>> >>> >>> and Hrepr() by more consistent names
>> >> >>> >>> >>> 3deaa0a61a Trac #28385: tides does not pass self-tests on 
>> >> >>> >>> >>> 32bit system
>> >> >>> >>> >>> dde78cbd22 (tag: 9.1.beta1) Updated SageMath version to 
>> >> >>> >>> >>> 9.1.beta1
>> >> >>> >>> >>>
>> >> >>> >>> > --
>> >> >>> >>> > 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 view this discussion on the web visit 
>> >> >>> >>> > https://groups.google.com/d/msgid/sage-release/14893264-c655-4932-8b03-be917e5f0892%40googlegroups.com.
>> >> >>> >
>> >> >>> > --
>> >> >>> > 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 view this discussion on the web visit 
>> >> >>> > https://groups.google.com/d/msgid/sage-release/be427777-3869-4e99-82cd-7a8300e91c4f%40googlegroups.com.
>> >> >
>> >> > --
>> >> > 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 view this discussion on the web visit 
>> >> > https://groups.google.com/d/msgid/sage-release/f168a0e4-3ad6-43d0-ada6-58b0744f3768%40googlegroups.com.
>> >
>> > --
>> > 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 view this discussion on the web visit 
>> > https://groups.google.com/d/msgid/sage-release/f5b1c484-fd07-4abe-afc3-8afbd1468f75%40googlegroups.com.
>
> --
> 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 view this discussion on the web visit 
> https://groups.google.com/d/msgid/sage-release/71899b52-1852-46be-9a1b-4ba00bf7c75b%40googlegroups.com.

-- 
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 view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-release/CAAWYfq2inYCjTRp7cwLrQ6akohEwn%2BwDUWE9xcqHd2Y3LdBezg%40mail.gmail.com.

Reply via email to