Missing package?

148 views
Skip to first unread message

brucew...@gmail.com

unread,
Nov 3, 2020, 5:38:09 AM11/3/20
to sage-devel
I have just tried to build sage and got the following error message

[sagelib-9.2.rc0] pkgconfig.pkgconfig.PackageNotFoundError: fflas-ffpack not found

I recently upgraded to Pop!_OS 20.10

Could someone please tell me how to fix this?

Thanks,
Bruce

Dima Pasechnik

unread,
Nov 3, 2020, 7:51:24 AM11/3/20
to sage-devel
On Tue, Nov 3, 2020 at 10:38 AM brucew...@gmail.com
<brucew...@gmail.com> wrote:
>
> I have just tried to build sage and got the following error message
>
> [sagelib-9.2.rc0] pkgconfig.pkgconfig.PackageNotFoundError: fflas-ffpack not found

9.2 is now released - why are you still on rc0?

Most probably this error means that fflas_ffpack was not
successfully installed.
(you can check logs/pkgs/fflas_ffpack-2.4.3.log for errors)
It could also be the case that a buildsystem update broke something,
and you need
to "make distclean"

In any case, it seems that you have a system similar to Ubuntu version
20.10 (released 2020-10-22) - and
probably only very little, if at all, testing, was done for it in 9.2
(released 2020-10-26).

You'd be better switch to 9.3.beta0

Please also post the top-level config.log, and look at the final part
of output of ./configure to see which
system packages you can install.



>
> I recently upgraded to Pop!_OS 20.10
>
> Could someone please tell me how to fix this?
>
> Thanks,
> Bruce
>
> --
> 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+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/39947bb0-0e4f-48ca-9f08-e29bfd7f93f9n%40googlegroups.com.

brucew...@gmail.com

unread,
Nov 4, 2020, 9:49:14 AM11/4/20
to sage-devel
I have tried to switch to 9,3.beta0 (yes, my OS is essentially Ubuntu 20.10)
This failed and I don't know what went wrong. The end of the log file is:

[sagelib-9.3.beta0] gcc -pthread -shared -Wl,-rpath-link,/home/bruce/sage-8.9.beta1/local/lib -L/home/bruce/sage-8.9.beta1/local/lib -Wl,-rpath,/home/bruce/sage-8.9.beta1/local/lib -L. -Wl,-rpath-link,/home/bruce/sage-8.9.beta1/local/lib -L/home/bruce/sage-8.9.beta1/local/lib -Wl,-rpath,/home/bruce/sage-8.9.beta1/local/lib -Wl,-rpath-link,/home/bruce/sage-8.9.beta1/local/lib -L/home/bruce/sage-8.9.beta1/local/lib -Wl,-rpath,/home/bruce/sage-8.9.beta1/local/lib build/temp.linux-x86_64-3.8/build/cythonized/sage/ext/interpreters/wrapper_el.o -L/home/bruce/sage-8.9.beta1/local/lib -o build/lib.linux-x86_64-3.8/sage/ext/interpreters/wrapper_el.cpython-38-x86_64-linux-gnu.so -lpari
[sagelib-9.3.beta0] error: command 'gcc' failed with exit status 1
[sagelib-9.3.beta0]
[sagelib-9.3.beta0] real    42m3.944s
[sagelib-9.3.beta0] user    40m56.660s
[sagelib-9.3.beta0] sys    0m56.600s
make[4]: *** [Makefile:2163: sagelib-no-deps] Error 1
make[3]: *** [Makefile:2163: /home/bruce/sage-8.9.beta1/local/var/lib/sage/installed/sagelib-9.3.beta0] Error 2
make[2]: *** [Makefile:1766: all-start] Error 2
make[2]: Leaving directory '/home/bruce/sage-8.9.beta1/build/make'

real    142m18.202s
user    161m41.339s
sys    9m2.716s
***************************************************************
Error building Sage.

The following package(s) may have failed to build (not necessarily
during this run of 'make all-start'):

It is safe to delete any log files and build directories, but they
contain information that is helpful for debugging build problems.
WARNING: If you now run 'make' again, the build directory of the
same version of the package will, by default, be deleted. Set the
environment variable SAGE_KEEP_BUILT_SPKGS=yes to prevent this.

Dima Pasechnik

unread,
Nov 4, 2020, 10:00:10 AM11/4/20
to sage-devel
sagelib didn't build for some reason that is very hard to see without
seeing more of install.log and config.log.
Please post them here.
(Good news that it went this far, so it's probably something easy to fix)
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/ea466d27-c6ef-415e-b9d8-9d2631703325n%40googlegroups.com.

Bruce Westbury

unread,
Nov 4, 2020, 10:24:05 AM11/4/20
to sage-...@googlegroups.com
The file install.log is over 18MB which is too large to attach.

You received this message because you are subscribed to a topic in the Google Groups "sage-devel" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/sage-devel/gzFzWUdxWwQ/unsubscribe.
To unsubscribe from this group and all its topics, send an email to sage-devel+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/CAAWYfq1%3DEBd%2BfBN7CK_tdWCnSWE%2Bfjw3kjUq5UvSdJgG2amu7g%40mail.gmail.com.
config.log

Dima Pasechnik

unread,
Nov 4, 2020, 10:32:28 AM11/4/20
to sage-devel
well, search for errors in that file, starting at the bottom, and find a relevant error message.
Or put the file on the web somewhere.

Dima Pasechnik

unread,
Nov 4, 2020, 10:57:26 AM11/4/20
to sage-devel
did you follow the advice spelled out in lines 4374-6 of config.log ?

Bruce Westbury

unread,
Nov 4, 2020, 12:19:53 PM11/4/20
to sage-...@googlegroups.com
I have tried following this advice but the build still fails. This advice now appears on lines 2191 to 2195.
The list is shorter now. I have tried again but this is not changing.

I had to take out texlive_generic_extra as it could not find this.

I keep noticing giac coming up.

config.log

Michael Orlitzky

unread,
Nov 4, 2020, 2:07:27 PM11/4/20
to sage-...@googlegroups.com
On 11/4/20 9:49 AM, brucew...@gmail.com wrote:
> I have tried to switch to 9,3.beta0 (yes, my OS is essentially Ubuntu 20.10)
> This failed and I don't know what went wrong. The end of the log file is:

You have to do "make giac" there when it crashes. There's a ticket open
to fix this with a positive review already. After giac installs, another
"make" should finish the sage installation.

brucew...@gmail.com

unread,
Nov 4, 2020, 2:29:17 PM11/4/20
to sage-devel
"make giac" failed. I don't know why.
giac-1.5.0.87p2.p1.log

Matthias Koeppe

unread,
Nov 4, 2020, 4:46:56 PM11/4/20
to sage-devel
On Wednesday, November 4, 2020 at 11:29:17 AM UTC-8, brucew...@gmail.com wrote:
"make giac" failed. I don't know why.


From the log:

/usr/bin/bash ../libtool  --tag=CXX   --mode=link g++ -std=gnu++11  -g -O2  -fno-strict-aliasing -DGIAC_GENERIC_CONSTANTS -no-undefined -Wl,-rpath-link,/home/bruce/sage-8.9.beta1/local/lib -L/home/bruce/sage-8.9.beta1/local/lib -Wl,-rpath,/home/bruce/sage-8.9.beta1/local/lib  -o icas icas.o libxcas.la -lreadline -lncurses -lcurl -lglpk -ldl -lm -lecm -lmpfi -lmpfr -lgmp  -L/home/bruce/sage-8.9.beta1/local/lib -lgsl -lm -lopenblas -lm -L/home/bruce/sage-8.9.beta1/local/lib -lopenblas -lrt -lpthread -lnauty -lcurl -lglpk -ldl -lm -lecm -lmpfi -lmpfr -lgmp 
libtool: link: g++ -std=gnu++11 -g -O2 -fno-strict-aliasing -DGIAC_GENERIC_CONSTANTS -Wl,-rpath-link -Wl,/home/bruce/sage-8.9.beta1/local/lib -Wl,-rpath -Wl,/home/bruce/sage-8.9.beta1/local/lib -o .libs/icas icas.o  -L/home/bruce/sage-8.9.beta1/local/lib ./.libs/libxcas.a /home/bruce/sage-8.9.beta1/local/var/tmp/sage/build/giac-1.5.0.87p2.p1/src/src/.libs/libgiac.so -lntl -lpari -lreadline -lncurses -lgsl -lopenblas -lrt -lpthread -lnauty /usr/lib/x86_64-linux-gnu/libcurl.so -lglpk -ldl -lm -lecm -lmpfi -lmpfr -lgmp -pthread -Wl,-rpath -Wl,/home/bruce/sage-8.9.beta1/local/lib
/usr/bin/ld: /usr/lib/gcc/x86_64-linux-gnu/10/../../../x86_64-linux-gnu/libnauty.so: undefined reference to `graph_free@CLIQUER_1'
/usr/bin/ld: /usr/lib/gcc/x86_64-linux-gnu/10/../../../x86_64-linux-gnu/libnauty.so: undefined reference to `clique_unweighted_find_single@CLIQUER_1'
/usr/bin/ld: /usr/lib/gcc/x86_64-linux-gnu/10/../../../x86_64-linux-gnu/libnauty.so: undefined reference to `graph_new@CLIQUER_1'
collect2: error: ld returned 1 exit status
 
Looks like the system libnauty is broken. What system is this? Try removing the libnauty system package

Dima Pasechnik

unread,
Nov 4, 2020, 5:00:33 PM11/4/20
to sage-devel
it is "mostly" Ubuntu 20.10, so
https://ubuntu.pkgs.org/20.10/ubuntu-universe-arm64/libnauty2_2.7r1+ds-1_arm64.deb.html
- with an interesting dependency, libcliquer
Note that in Sage's configuration there is no such dependency, and
this might be the problem here, I guess


>
> --
> 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+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/2479bef3-22ef-48ec-a96b-7abdbdb53b37o%40googlegroups.com.

Dima Pasechnik

unread,
Nov 4, 2020, 6:07:09 PM11/4/20
to sage-devel
On Wed, Nov 4, 2020 at 10:00 PM Dima Pasechnik <dim...@gmail.com> wrote:
>
> On Wed, Nov 4, 2020 at 9:46 PM Matthias Koeppe <matthia...@gmail.com> wrote:
> >
> > On Wednesday, November 4, 2020 at 11:29:17 AM UTC-8, brucew...@gmail.com wrote:
> >>
> >> "make giac" failed. I don't know why.
> >>>
> >>>
> >
> > From the log:
> >
> > /usr/bin/bash ../libtool --tag=CXX --mode=link g++ -std=gnu++11 -g -O2 -fno-strict-aliasing -DGIAC_GENERIC_CONSTANTS -no-undefined -Wl,-rpath-link,/home/bruce/sage-8.9.beta1/local/lib -L/home/bruce/sage-8.9.beta1/local/lib -Wl,-rpath,/home/bruce/sage-8.9.beta1/local/lib -o icas icas.o libxcas.la -lreadline -lncurses -lcurl -lglpk -ldl -lm -lecm -lmpfi -lmpfr -lgmp -L/home/bruce/sage-8.9.beta1/local/lib -lgsl -lm -lopenblas -lm -L/home/bruce/sage-8.9.beta1/local/lib -lopenblas -lrt -lpthread -lnauty -lcurl -lglpk -ldl -lm -lecm -lmpfi -lmpfr -lgmp
> > libtool: link: g++ -std=gnu++11 -g -O2 -fno-strict-aliasing -DGIAC_GENERIC_CONSTANTS -Wl,-rpath-link -Wl,/home/bruce/sage-8.9.beta1/local/lib -Wl,-rpath -Wl,/home/bruce/sage-8.9.beta1/local/lib -o .libs/icas icas.o -L/home/bruce/sage-8.9.beta1/local/lib ./.libs/libxcas.a /home/bruce/sage-8.9.beta1/local/var/tmp/sage/build/giac-1.5.0.87p2.p1/src/src/.libs/libgiac.so -lntl -lpari -lreadline -lncurses -lgsl -lopenblas -lrt -lpthread -lnauty /usr/lib/x86_64-linux-gnu/libcurl.so -lglpk -ldl -lm -lecm -lmpfi -lmpfr -lgmp -pthread -Wl,-rpath -Wl,/home/bruce/sage-8.9.beta1/local/lib
> > /usr/bin/ld: /usr/lib/gcc/x86_64-linux-gnu/10/../../../x86_64-linux-gnu/libnauty.so: undefined reference to `graph_free@CLIQUER_1'
> > /usr/bin/ld: /usr/lib/gcc/x86_64-linux-gnu/10/../../../x86_64-linux-gnu/libnauty.so: undefined reference to `clique_unweighted_find_single@CLIQUER_1'
> > /usr/bin/ld: /usr/lib/gcc/x86_64-linux-gnu/10/../../../x86_64-linux-gnu/libnauty.so: undefined reference to `graph_new@CLIQUER_1'
> > collect2: error: ld returned 1 exit status
> >
> >
> > Looks like the system libnauty is broken. What system is this? Try removing the libnauty system package
>
> it is "mostly" Ubuntu 20.10, so
> https://ubuntu.pkgs.org/20.10/ubuntu-universe-arm64/libnauty2_2.7r1+ds-1_arm64.deb.html
> - with an interesting dependency, libcliquer
> Note that in Sage's configuration there is no such dependency, and
> this might be the problem here, I guess

yes, that's how it looks like from the config.log posted here. I
suspect that it is a bug in giac, and
one has to link with libcliquer here too (is libnauty underlinked?)

brucew...@gmail.com

unread,
Nov 4, 2020, 6:10:23 PM11/4/20
to sage-devel
I removed libnauty (which was actually libnauty2) did "make libnauty" then "make giac" and finally "make build".
It seems to have worked. That you all for your help and patience.

Matthias Koeppe

unread,
Nov 4, 2020, 6:56:19 PM11/4/20
to sage-devel
On Wednesday, November 4, 2020 at 2:00:33 PM UTC-8, Dima Pasechnik wrote:
On Wed, Nov 4, 2020 at 9:46 PM Matthias Koeppe <matthia...@gmail.com> wrote:
> On Wednesday, November 4, 2020 at 11:29:17 AM UTC-8, brucew...@gmail.com wrote:
>> "make giac" failed. I don't know why.
> From the log:
>
> /usr/bin/bash ../libtool  --tag=CXX   --mode=link g++ -std=gnu++11  -g -O2  -fno-strict-aliasing -DGIAC_GENERIC_CONSTANTS -no-undefined -Wl,-rpath-link,/home/bruce/sage-8.9.beta1/local/lib -L/home/bruce/sage-8.9.beta1/local/lib -Wl,-rpath,/home/bruce/sage-8.9.beta1/local/lib  -o icas icas.o libxcas.la -lreadline -lncurses -lcurl -lglpk -ldl -lm -lecm -lmpfi -lmpfr -lgmp  -L/home/bruce/sage-8.9.beta1/local/lib -lgsl -lm -lopenblas -lm -L/home/bruce/sage-8.9.beta1/local/lib -lopenblas -lrt -lpthread -lnauty -lcurl -lglpk -ldl -lm -lecm -lmpfi -lmpfr -lgmp
> libtool: link: g++ -std=gnu++11 -g -O2 -fno-strict-aliasing -DGIAC_GENERIC_CONSTANTS -Wl,-rpath-link -Wl,/home/bruce/sage-8.9.beta1/local/lib -Wl,-rpath -Wl,/home/bruce/sage-8.9.beta1/local/lib -o .libs/icas icas.o  -L/home/bruce/sage-8.9.beta1/local/lib ./.libs/libxcas.a /home/bruce/sage-8.9.beta1/local/var/tmp/sage/build/giac-1.5.0.87p2.p1/src/src/.libs/libgiac.so -lntl -lpari -lreadline -lncurses -lgsl -lopenblas -lrt -lpthread -lnauty /usr/lib/x86_64-linux-gnu/libcurl.so -lglpk -ldl -lm -lecm -lmpfi -lmpfr -lgmp -pthread -Wl,-rpath -Wl,/home/bruce/sage-8.9.beta1/local/lib
> /usr/bin/ld: /usr/lib/gcc/x86_64-linux-gnu/10/../../../x86_64-linux-gnu/libnauty.so: undefined reference to `graph_free@CLIQUER_1'
> /usr/bin/ld: /usr/lib/gcc/x86_64-linux-gnu/10/../../../x86_64-linux-gnu/libnauty.so: undefined reference to `clique_unweighted_find_single@CLIQUER_1'
> /usr/bin/ld: /usr/lib/gcc/x86_64-linux-gnu/10/../../../x86_64-linux-gnu/libnauty.so: undefined reference to `graph_new@CLIQUER_1'
> collect2: error: ld returned 1 exit status
>
>
> Looks like the system libnauty is broken. What system is this? Try removing the libnauty system package

it is "mostly" Ubuntu 20.10, so
https://ubuntu.pkgs.org/20.10/ubuntu-universe-arm64/libnauty2_2.7r1+ds-1_arm64.deb.html
- with an interesting dependency, libcliquer
Note that in Sage's configuration there is no such dependency, and
this might be the problem here, I guess

nauty packaging is a mess - see https://trac.sagemath.org/ticket/30271

But what's failing here is giac's configure tests for libnauty - we should check whether upstream giac is aware of this



Dima Pasechnik

unread,
Nov 4, 2020, 7:12:30 PM11/4/20
to sage-devel
is this the same as used in Debian/Ubuntu?
 (I think libnauty was brought in there by Macaulay2 people very recently).


But what's failing here is giac's configure tests for libnauty - we should check whether upstream giac is aware of this

could anyone with Ubuntu 20.10 run ldd
on libnauty2.so and post the output here?





--
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+...@googlegroups.com.

Matthias Koeppe

unread,
Nov 4, 2020, 8:03:42 PM11/4/20
to sage-devel
On Wednesday, November 4, 2020 at 4:12:30 PM UTC-8, Dima Pasechnik wrote:

On Wed, 4 Nov 2020, 23:56 Matthias Koeppe, <matthia...@gmail.com> wrote:
On Wednesday, November 4, 2020 at 2:00:33 PM UTC-8, Dima Pasechnik wrote:
On Wed, Nov 4, 2020 at 9:46 PM Matthias Koeppe <matthia...@gmail.com> wrote:
> On Wednesday, November 4, 2020 at 11:29:17 AM UTC-8, brucew...@gmail.com wrote:
>> "make giac" failed. I don't know why.
> From the log:
>
> /usr/bin/bash ../libtool  --tag=CXX   --mode=link g++ -std=gnu++11  -g -O2  -fno-strict-aliasing -DGIAC_GENERIC_CONSTANTS -no-undefined -Wl,-rpath-link,/home/bruce/sage-8.9.beta1/local/lib -L/home/bruce/sage-8.9.beta1/local/lib -Wl,-rpath,/home/bruce/sage-8.9.beta1/local/lib  -o icas icas.o libxcas.la -lreadline -lncurses -lcurl -lglpk -ldl -lm -lecm -lmpfi -lmpfr -lgmp  -L/home/bruce/sage-8.9.beta1/local/lib -lgsl -lm -lopenblas -lm -L/home/bruce/sage-8.9.beta1/local/lib -lopenblas -lrt -lpthread -lnauty -lcurl -lglpk -ldl -lm -lecm -lmpfi -lmpfr -lgmp
> libtool: link: g++ -std=gnu++11 -g -O2 -fno-strict-aliasing -DGIAC_GENERIC_CONSTANTS -Wl,-rpath-link -Wl,/home/bruce/sage-8.9.beta1/local/lib -Wl,-rpath -Wl,/home/bruce/sage-8.9.beta1/local/lib -o .libs/icas icas.o  -L/home/bruce/sage-8.9.beta1/local/lib ./.libs/libxcas.a /home/bruce/sage-8.9.beta1/local/var/tmp/sage/build/giac-1.5.0.87p2.p1/src/src/.libs/libgiac.so -lntl -lpari -lreadline -lncurses -lgsl -lopenblas -lrt -lpthread -lnauty /usr/lib/x86_64-linux-gnu/libcurl.so -lglpk -ldl -lm -lecm -lmpfi -lmpfr -lgmp -pthread -Wl,-rpath -Wl,/home/bruce/sage-8.9.beta1/local/lib
> /usr/bin/ld: /usr/lib/gcc/x86_64-linux-gnu/10/../../../x86_64-linux-gnu/libnauty.so: undefined reference to `graph_free@CLIQUER_1'
> /usr/bin/ld: /usr/lib/gcc/x86_64-linux-gnu/10/../../../x86_64-linux-gnu/libnauty.so: undefined reference to `clique_unweighted_find_single@CLIQUER_1'
> /usr/bin/ld: /usr/lib/gcc/x86_64-linux-gnu/10/../../../x86_64-linux-gnu/libnauty.so: undefined reference to `graph_new@CLIQUER_1'
> collect2: error: ld returned 1 exit status
>
>
> Looks like the system libnauty is broken. What system is this? Try removing the libnauty system package

it is "mostly" Ubuntu 20.10, so
https://ubuntu.pkgs.org/20.10/ubuntu-universe-arm64/libnauty2_2.7r1+ds-1_arm64.deb.html
- with an interesting dependency, libcliquer
Note that in Sage's configuration there is no such dependency, and
this might be the problem here, I guess

nauty packaging is a mess - see https://trac.sagemath.org/ticket/30271

is this the same as used in Debian/Ubuntu?

 
My branch there completes Debian's incomplete autotoolization of nauty.


brucew...@gmail.com

unread,
Nov 5, 2020, 5:08:32 AM11/5/20
to sage-devel
This is off-topic but I don't know the right place to ask this question about using git.

I now have a detached HEAD at 9.3.beta0

I tried "git checkout master" and got some hints and then
fatal: 'master' matched multiple (2) remote tracking branches

How do I get my develop branch back?

Dima Pasechnik

unread,
Nov 5, 2020, 5:32:48 AM11/5/20
to sage-devel
On Thu, Nov 5, 2020 at 10:08 AM brucew...@gmail.com
<brucew...@gmail.com> wrote:
>
> This is off-topic but I don't know the right place to ask this question about using git.
>
> I now have a detached HEAD at 9.3.beta0
>
> I tried "git checkout master" and got some hints and then
> fatal: 'master' matched multiple (2) remote tracking branches
>
> How do I get my develop branch back?

git branch

lists all the local branches you have

Then

git checkout foo # or whatever

will put you back to your (local) branch foo

You probably want to merge 9.3.beta0 into your "develop" branch.

Do

git checkout develop
git fetch trac develop # assuming "trac" is the name of git trac remote
git merge trac/develop

(the latter two commands can be combined into "git pull trac develop")

HTH
Dima
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/46c65437-1729-48b3-9b4b-5305e7bfd20en%40googlegroups.com.

Bruce Westbury

unread,
Nov 5, 2020, 8:20:33 AM11/5/20
to sage-...@googlegroups.com
Thx

You received this message because you are subscribed to a topic in the Google Groups "sage-devel" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/sage-devel/gzFzWUdxWwQ/unsubscribe.
To unsubscribe from this group and all its topics, send an email to sage-devel+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/CAAWYfq1OtUNtVv25DZEOMy5FH5RutPxkYg_FNypFdXxMVM3Bcg%40mail.gmail.com.

parisse

unread,
Nov 5, 2020, 2:22:52 PM11/5/20
to sage-devel
When compiling giac with nauty, I install nauty by hand and statically linked, nothing more than libnauty.a is required.
In configure.ac, the check is done by
AC_CHECK_LIB(nauty,main)
AC_CHECK_HEADERS(nauty/naututil.h)
I can of course add another check if it's required to compile with nauty, just let me know.

Dima Pasechnik

unread,
Nov 5, 2020, 3:59:28 PM11/5/20
to sage-devel


On Thu, 5 Nov 2020, 19:23 parisse, <bernard...@ujf-grenoble.fr> wrote:
When compiling giac with nauty, I install nauty by hand and statically linked, nothing more than libnauty.a is required.
In configure.ac, the check is done by
AC_CHECK_LIB(nauty,main)
AC_CHECK_HEADERS(nauty/naututil.h)
I can of course add another check if it's required to compile with nauty, just let me know.

With latest nauty providing a dynamic library, giac ought to use such a library by default. At least this is what would be expected of giac by various distributions, I think.

What complicates things here is that apparently in some cases this dynamic library is using another dynamic library, lubcliquer, and so it has to be taken care of. (I might be overcomplicating things here)


Dima Pasechnik

unread,
Nov 6, 2020, 5:18:13 AM11/6/20
to sage-devel
On Thu, Nov 5, 2020 at 8:59 PM Dima Pasechnik <dim...@gmail.com> wrote:
>
>
>
> On Thu, 5 Nov 2020, 19:23 parisse, <bernard...@ujf-grenoble.fr> wrote:
>>
>> When compiling giac with nauty, I install nauty by hand and statically linked, nothing more than libnauty.a is required.
>> In configure.ac, the check is done by
>> AC_CHECK_LIB(nauty,main)
>> AC_CHECK_HEADERS(nauty/naututil.h)
>> I can of course add another check if it's required to compile with nauty, just let me know.
>
>
> With latest nauty providing a dynamic library, giac ought to use such a library by default. At least this is what would be expected of giac by various distributions, I think.
>
> What complicates things here is that apparently in some cases this dynamic library is using another dynamic library, lubcliquer, and so it has to be taken care of. (I might be overcomplicating things here)

I've checked that giac (still the latest giac 1.5) in Sage builds fine
on Ubuntu 20.10, and dynamically links libnauty and libcliquer.
So this is all fine. For some reason giac in Ubuntu 20.10 does not do
this, probably Ubuntu/Debian-specific issue.

An improvement, qua robustness, I could suggest, is to use
PKG_CHECK_MODULES in configure.ac
to check (lib)nauty (the package name in pkg-config is nauty), so one would do

PKG_CHECK_MODULES([NAUTY], [nauty >= 2.6], [<stuff to do if found.],
[<stuff to do otherwise])

(this version check is pretty innocent, more as an illustration)

Dima

parisse

unread,
Nov 7, 2020, 3:44:29 AM11/7/20
to sage-devel
It's probably easier just to add
AC_CHECK_LIB(cliquer,main)
Reply all
Reply to author
Forward
0 new messages