Sage 4.6.1 build failure in symmetrica-2.0.p5.spkg

16 views
Skip to first unread message

Paul Leopardi

unread,
Jan 31, 2011, 7:48:06 AM1/31/11
to sage-devel
Hi all,
I'm trying to compile Sage 4.6.1 on openSUSE 11.2 x86_64 and have encountered
the same bug I reported on May 16 last year with Sage 4.4.1.
http://www.mail-archive.com/sage-...@googlegroups.com/msg38879.html
Th fix is to move "-lm" to the right place on the command line. See diff
below. How do I file a bug report for this, so that it is fixed properly?
Best, Paul

diff -rub symmetrica-2.0.p5//patches/makefile.patch
symmetrica-2.0.p5.new//patches/makefile.patch
--- symmetrica-2.0.p5//patches/makefile.patch 2010-01-07 12:25:08.000000000
+1100
+++ symmetrica-2.0.p5.new//patches/makefile.patch 2011-01-31
23:13:46.000000000 +1100
@@ -20,4 +20,4 @@
+ $(CC) -c $(CFLAGS) $<
test: test.c bar.o bi.o boe.o bruch.o classical.o de.o di.o ff.o galois.o
ga.o gra.o hash.o hiccup.o io.o ko.o list.o lo.o ma.o mee.o mem.o mes.o mhe.o
mhh.o mhm.o mhp.o mhs.o mmm.o mms.o mod_dg_sbd.o mo.o mpp.o mps.o mse.o msh.o
msm.o mss.o muir.o na.o nb.o nc.o nu.o part.o pee.o peh.o pem.o perm.o pes.o
phe.o phh.o phm.o phs.o plet.o pme.o pmh.o poly.o ppe.o pph.o ppm.o ppp.o
pps.o pr.o pse.o psh.o psm.o pss.o rest.o rh.o sab.o sb.o sc.o sr.o ta.o teh.o
tem.o tep.o tes.o the.o thm.o thp.o ths.o tme.o tmh.o tmp.o tms.o tpe.o tph.o
tpm.o tps.o tse.o tsh.o tsm.o tsp.o vc.o zo.o zykelind.o zyk.o
- gcc -DALLTRUE -DFAST test.c bar.o bi.o boe.o bruch.o classical.o de.o
di.o ff.o galois.o ga.o gra.o hash.o hiccup.o io.o ko.o list.o lo.o ma.o mee.o
mem.o mes.o mhe.o mhh.o mhm.o mhp.o mhs.o mmm.o mms.o mod_dg_sbd.o mo.o mpp.o
mps.o mse.o msh.o msm.o mss.o muir.o na.o nb.o nc.o nu.o part.o pee.o peh.o
pem.o perm.o pes.o phe.o phh.o phm.o phs.o plet.o pme.o pmh.o poly.o ppe.o
pph.o ppm.o ppp.o pps.o pr.o pse.o psh.o psm.o pss.o rest.o rh.o sab.o sb.o
sc.o sr.o ta.o teh.o tem.o tep.o tes.o the.o thm.o thp.o ths.o tme.o tmh.o
tmp.o tms.o tpe.o tph.o tpm.o tps.o tse.o tsh.o tsm.o tsp.o vc.o zo.o
zykelind.o zyk.o -lm -o test
-+ $(CC) $(CFLAGS) -lm test.c bar.o bi.o boe.o bruch.o classical.o de.o
di.o ff.o galois.o ga.o gra.o hash.o hiccup.o io.o ko.o list.o lo.o ma.o mee.o
mem.o mes.o mhe.o mhh.o mhm.o mhp.o mhs.o mmm.o mms.o mod_dg_sbd.o mo.o mpp.o
mps.o mse.o msh.o msm.o mss.o muir.o na.o nb.o nc.o nu.o part.o pee.o peh.o
pem.o perm.o pes.o phe.o phh.o phm.o phs.o plet.o pme.o pmh.o poly.o ppe.o
pph.o ppm.o ppp.o pps.o pr.o pse.o psh.o psm.o pss.o rest.o rh.o sab.o sb.o
sc.o sr.o ta.o teh.o tem.o tep.o tes.o the.o thm.o thp.o ths.o tme.o tmh.o
tmp.o tms.o tpe.o tph.o tpm.o tps.o tse.o tsh.o tsm.o tsp.o vc.o zo.o
zykelind.o zyk.o -o test
++ $(CC) $(CFLAGS) test.c bar.o bi.o boe.o bruch.o classical.o de.o di.o
ff.o galois.o ga.o gra.o hash.o hiccup.o io.o ko.o list.o lo.o ma.o mee.o
mem.o mes.o mhe.o mhh.o mhm.o mhp.o mhs.o mmm.o mms.o mod_dg_sbd.o mo.o mpp.o
mps.o mse.o msh.o msm.o mss.o muir.o na.o nb.o nc.o nu.o part.o pee.o peh.o
pem.o perm.o pes.o phe.o phh.o phm.o phs.o plet.o pme.o pmh.o poly.o ppe.o
pph.o ppm.o ppp.o pps.o pr.o pse.o psh.o psm.o pss.o rest.o rh.o sab.o sb.o
sc.o sr.o ta.o teh.o tem.o tep.o tes.o the.o thm.o thp.o ths.o tme.o tmh.o
tmp.o tms.o tpe.o tph.o tpm.o tps.o tse.o tsh.o tsm.o tsp.o vc.o zo.o
zykelind.o zyk.o -lm -o test

Volker Braun

unread,
Jan 31, 2011, 10:27:02 AM1/31/11
to sage-...@googlegroups.com
Are you still using gcc 4.4.1? I'm actually surprised that you were able to compile the rest of Sage with that compiler.

I've made a trac ticket


and an updated spkg:


Volker

Paul Leopardi

unread,
Jan 31, 2011, 5:21:07 PM1/31/11
to sage-devel
Thanks. I'm using:
%gcc -v
Using built-in specs.
COLLECT_GCC=gcc
COLLECT_LTO_WRAPPER=/usr/local/libexec/gcc/x86_64-unknown-linux-gnu/
4.5.1/lto-wrapper
Target: x86_64-unknown-linux-gnu
Configured with: ../gcc-4.5.1/configure
Thread model: posix
gcc version 4.5.1 (GCC

Paul Leopardi

unread,
Jan 31, 2011, 5:47:15 PM1/31/11
to sage-devel
Hi Volker

On Feb 1, 2:27 am, Volker Braun <vbraun.n...@gmail.com> wrote:
Thanks again. I tested your spkg. It doesn't work. You also need to
change makefile:

- $(CC) $(CFLAGS) -lm test.c bar.o bi.o boe.o bruch.o

Volker Braun

unread,
Jan 31, 2011, 8:15:10 PM1/31/11
to sage-...@googlegroups.com
Upon a closer look at the symmetrica spkg I noticed a lot of other oddities. Particularly ironic is the fact that the binary that fails to link on your machine is actually not used. I've updated my spkg (http://www.stp.dias.ie/~vbraun/Sage/spkg/symmetrica-2.0.p6.spkg), can you give it another try and let me know if it works for you?

Volker

Paul Leopardi

unread,
Feb 1, 2011, 9:40:28 AM2/1/11
to sage-devel
Hi Volker,
I tried an updated version with just that one change to the position
of -lm in makefile and Sage built OK. I will also try your spkg, but
not now, as it is very late at night here. What to I need to delete so
that only symmetrica is rebuilt?
Best, Paul

Volker Braun

unread,
Feb 1, 2011, 9:48:17 AM2/1/11
to sage-...@googlegroups.com
Delete any cached symmetrica spkg from $SAGE_ROOT/spkg/optional/symmetrica*.spkg if it exists and then

Paul Leopardi

unread,
Feb 1, 2011, 5:14:47 PM2/1/11
to sage-devel
OK. Your revised spkg builds and install correctly now, as far as I
can tell.
Reply all
Reply to author
Forward
0 new messages